{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 28.06361085126286,
  "eval_steps": 500,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "action_query_norm": 156.34859619140624,
      "action_relation_loss": 0.05670318678021431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.009354536950420954,
      "flow_matching_loss": 1.0352736830711364,
      "grad_norm": 10.943965911865234,
      "learning_rate": 6.000000000000001e-07,
      "loss": 1.1721,
      "relation_loss": 0.13681062757968904,
      "selected_pairs": 26709.2,
      "state_query_norm": 119.1670913696289,
      "state_relation_loss": 0.0801074393093586,
      "step": 10,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.13681062757968904,
      "weighted_relation_loss": 0.13681062757968904
    },
    {
      "action_query_norm": 155.7083312988281,
      "action_relation_loss": 0.059563370794057845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.018709073900841908,
      "flow_matching_loss": 1.0236413776874542,
      "grad_norm": 12.195730209350586,
      "learning_rate": 1.2666666666666667e-06,
      "loss": 1.1604,
      "relation_loss": 0.13672868013381959,
      "selected_pairs": 27177.8,
      "state_query_norm": 118.3869514465332,
      "state_relation_loss": 0.07716531008481979,
      "step": 20,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.13672868013381959,
      "weighted_relation_loss": 0.13672868013381959
    },
    {
      "action_query_norm": 155.47974395751953,
      "action_relation_loss": 0.05479946583509445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.02806361085126286,
      "flow_matching_loss": 0.8110633254051208,
      "grad_norm": 7.117291450500488,
      "learning_rate": 1.9333333333333336e-06,
      "loss": 0.9459,
      "relation_loss": 0.13487044498324394,
      "selected_pairs": 26755.1,
      "state_query_norm": 116.18980560302734,
      "state_relation_loss": 0.0800709791481495,
      "step": 30,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.13487044498324394,
      "weighted_relation_loss": 0.13487044498324394
    },
    {
      "action_query_norm": 165.75557708740234,
      "action_relation_loss": 0.05339513085782528,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.037418147801683815,
      "flow_matching_loss": 0.5771536529064178,
      "grad_norm": 2.885411262512207,
      "learning_rate": 2.6e-06,
      "loss": 0.7071,
      "relation_loss": 0.12991497591137885,
      "selected_pairs": 27012.3,
      "state_query_norm": 113.95371322631836,
      "state_relation_loss": 0.07651984393596649,
      "step": 40,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.12991497591137885,
      "weighted_relation_loss": 0.12991497591137885
    },
    {
      "action_query_norm": 173.3781539916992,
      "action_relation_loss": 0.051027079299092296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.04677268475210477,
      "flow_matching_loss": 0.4415735214948654,
      "grad_norm": 1.7212971448898315,
      "learning_rate": 3.2666666666666666e-06,
      "loss": 0.5654,
      "relation_loss": 0.12386175617575645,
      "selected_pairs": 26396.6,
      "state_query_norm": 115.02495498657227,
      "state_relation_loss": 0.0728346761316061,
      "step": 50,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.12386175617575645,
      "weighted_relation_loss": 0.12386175617575645
    },
    {
      "action_query_norm": 180.9956039428711,
      "action_relation_loss": 0.04948233962059021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.05612722170252572,
      "flow_matching_loss": 0.3603812962770462,
      "grad_norm": 1.0965585708618164,
      "learning_rate": 3.9333333333333335e-06,
      "loss": 0.4751,
      "relation_loss": 0.11474426537752151,
      "selected_pairs": 27364.8,
      "state_query_norm": 115.8877082824707,
      "state_relation_loss": 0.06526192575693131,
      "step": 60,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.11474426537752151,
      "weighted_relation_loss": 0.11474426537752151
    },
    {
      "action_query_norm": 180.89956359863282,
      "action_relation_loss": 0.04747598879039287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.06548175865294668,
      "flow_matching_loss": 0.31226853728294374,
      "grad_norm": 0.9854580163955688,
      "learning_rate": 4.6e-06,
      "loss": 0.4147,
      "relation_loss": 0.10244298279285431,
      "selected_pairs": 26454.3,
      "state_query_norm": 119.57897644042968,
      "state_relation_loss": 0.054966995120048524,
      "step": 70,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.10244298279285431,
      "weighted_relation_loss": 0.10244298279285431
    },
    {
      "action_query_norm": 181.21807556152345,
      "action_relation_loss": 0.04429291598498821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.07483629560336763,
      "flow_matching_loss": 0.27374005913734434,
      "grad_norm": 0.6286137700080872,
      "learning_rate": 5.266666666666667e-06,
      "loss": 0.3644,
      "relation_loss": 0.09069925770163537,
      "selected_pairs": 26614.2,
      "state_query_norm": 125.93455810546875,
      "state_relation_loss": 0.04640634171664715,
      "step": 80,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.09069925770163537,
      "weighted_relation_loss": 0.09069925770163537
    },
    {
      "action_query_norm": 176.09258880615235,
      "action_relation_loss": 0.040515653789043427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.08419083255378859,
      "flow_matching_loss": 0.24791160374879836,
      "grad_norm": 0.5397706031799316,
      "learning_rate": 5.933333333333334e-06,
      "loss": 0.33,
      "relation_loss": 0.08210699111223221,
      "selected_pairs": 26897.3,
      "state_query_norm": 129.71632537841796,
      "state_relation_loss": 0.04159133769571781,
      "step": 90,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.08210699111223221,
      "weighted_relation_loss": 0.08210699111223221
    },
    {
      "action_query_norm": 169.29390869140624,
      "action_relation_loss": 0.03785932101309299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.09354536950420954,
      "flow_matching_loss": 0.21942520886659622,
      "grad_norm": 0.4877912998199463,
      "learning_rate": 6.6e-06,
      "loss": 0.2924,
      "relation_loss": 0.07295132055878639,
      "selected_pairs": 26853.0,
      "state_query_norm": 133.27810134887696,
      "state_relation_loss": 0.03509199973195791,
      "step": 100,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.07295132055878639,
      "weighted_relation_loss": 0.07295132055878639
    },
    {
      "action_query_norm": 169.06196136474608,
      "action_relation_loss": 0.0345687747001648,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1028999064546305,
      "flow_matching_loss": 0.2031555414199829,
      "grad_norm": 0.6528446674346924,
      "learning_rate": 7.266666666666668e-06,
      "loss": 0.2681,
      "relation_loss": 0.06497018299996853,
      "selected_pairs": 26207.9,
      "state_query_norm": 140.92886352539062,
      "state_relation_loss": 0.03040140774101019,
      "step": 110,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.06497018299996853,
      "weighted_relation_loss": 0.06497018299996853
    },
    {
      "action_query_norm": 171.64943084716796,
      "action_relation_loss": 0.03360517174005508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.11225444340505145,
      "flow_matching_loss": 0.18917445987462997,
      "grad_norm": 0.7137190699577332,
      "learning_rate": 7.933333333333334e-06,
      "loss": 0.2486,
      "relation_loss": 0.059423967450857165,
      "selected_pairs": 26649.7,
      "state_query_norm": 152.50956115722656,
      "state_relation_loss": 0.025818795897066593,
      "step": 120,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.059423967450857165,
      "weighted_relation_loss": 0.059423967450857165
    },
    {
      "action_query_norm": 171.41458587646486,
      "action_relation_loss": 0.033170217834413054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1216089803554724,
      "flow_matching_loss": 0.18161376863718032,
      "grad_norm": 0.6609107255935669,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.2373,
      "relation_loss": 0.055722708627581594,
      "selected_pairs": 27342.7,
      "state_query_norm": 160.30825805664062,
      "state_relation_loss": 0.022552491165697575,
      "step": 130,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.055722708627581594,
      "weighted_relation_loss": 0.055722708627581594
    },
    {
      "action_query_norm": 168.55827789306642,
      "action_relation_loss": 0.03126271516084671,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.13096351730589337,
      "flow_matching_loss": 0.17236161082983018,
      "grad_norm": 0.6812433004379272,
      "learning_rate": 9.266666666666667e-06,
      "loss": 0.2234,
      "relation_loss": 0.05105141773819923,
      "selected_pairs": 27174.7,
      "state_query_norm": 161.92007446289062,
      "state_relation_loss": 0.01978870201855898,
      "step": 140,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.05105141773819923,
      "weighted_relation_loss": 0.05105141773819923
    },
    {
      "action_query_norm": 171.94832916259764,
      "action_relation_loss": 0.028901061788201333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1403180542563143,
      "flow_matching_loss": 0.1635143369436264,
      "grad_norm": 0.5571377873420715,
      "learning_rate": 9.933333333333334e-06,
      "loss": 0.2104,
      "relation_loss": 0.046893899142742154,
      "selected_pairs": 26826.3,
      "state_query_norm": 177.9437484741211,
      "state_relation_loss": 0.01799283679574728,
      "step": 150,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.046893899142742154,
      "weighted_relation_loss": 0.046893899142742154
    },
    {
      "action_query_norm": 173.61029510498048,
      "action_relation_loss": 0.029193181544542313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.14967259120673526,
      "flow_matching_loss": 0.15216476321220399,
      "grad_norm": 0.991136908531189,
      "learning_rate": 1.06e-05,
      "loss": 0.1977,
      "relation_loss": 0.045566557347774504,
      "selected_pairs": 26622.1,
      "state_query_norm": 181.72623596191406,
      "state_relation_loss": 0.01637337626889348,
      "step": 160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.045566557347774504,
      "weighted_relation_loss": 0.045566557347774504
    },
    {
      "action_query_norm": 174.93113555908204,
      "action_relation_loss": 0.027343202754855155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.15902712815715622,
      "flow_matching_loss": 0.14429329037666322,
      "grad_norm": 0.8376375436782837,
      "learning_rate": 1.1266666666666667e-05,
      "loss": 0.1868,
      "relation_loss": 0.042466671019792554,
      "selected_pairs": 27272.6,
      "state_query_norm": 188.0512680053711,
      "state_relation_loss": 0.01512346900999546,
      "step": 170,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.042466671019792554,
      "weighted_relation_loss": 0.042466671019792554
    },
    {
      "action_query_norm": 173.9644805908203,
      "action_relation_loss": 0.02708689384162426,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.16838166510757718,
      "flow_matching_loss": 0.14010266363620758,
      "grad_norm": 0.4471832513809204,
      "learning_rate": 1.1933333333333333e-05,
      "loss": 0.1807,
      "relation_loss": 0.040558548644185066,
      "selected_pairs": 27079.9,
      "state_query_norm": 194.81756286621095,
      "state_relation_loss": 0.013471654895693064,
      "step": 180,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.040558548644185066,
      "weighted_relation_loss": 0.040558548644185066
    },
    {
      "action_query_norm": 169.7803726196289,
      "action_relation_loss": 0.025112359412014486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.17773620205799812,
      "flow_matching_loss": 0.1349011778831482,
      "grad_norm": 0.5814566016197205,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.1727,
      "relation_loss": 0.037804676592350005,
      "selected_pairs": 26875.0,
      "state_query_norm": 200.27599334716797,
      "state_relation_loss": 0.01269231727346778,
      "step": 190,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.037804676592350005,
      "weighted_relation_loss": 0.037804676592350005
    },
    {
      "action_query_norm": 170.89690551757812,
      "action_relation_loss": 0.024637381173670293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.18709073900841908,
      "flow_matching_loss": 0.13445894420146942,
      "grad_norm": 0.4707700312137604,
      "learning_rate": 1.3266666666666666e-05,
      "loss": 0.1715,
      "relation_loss": 0.03704964183270931,
      "selected_pairs": 26918.3,
      "state_query_norm": 207.0368408203125,
      "state_relation_loss": 0.012412261031568051,
      "step": 200,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.03704964183270931,
      "weighted_relation_loss": 0.03704964183270931
    },
    {
      "action_query_norm": 168.0598342895508,
      "action_relation_loss": 0.023981644585728646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.19644527595884004,
      "flow_matching_loss": 0.12708375379443168,
      "grad_norm": 0.6146612763404846,
      "learning_rate": 1.3933333333333334e-05,
      "loss": 0.1622,
      "relation_loss": 0.03507664278149605,
      "selected_pairs": 27081.4,
      "state_query_norm": 213.82928924560548,
      "state_relation_loss": 0.011094997916370631,
      "step": 210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.03507664278149605,
      "weighted_relation_loss": 0.03507664278149605
    },
    {
      "action_query_norm": 169.00358123779296,
      "action_relation_loss": 0.024483877792954446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.205799812909261,
      "flow_matching_loss": 0.13170734271407128,
      "grad_norm": 0.43268847465515137,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.167,
      "relation_loss": 0.035300766862928865,
      "selected_pairs": 26688.9,
      "state_query_norm": 219.3659637451172,
      "state_relation_loss": 0.010816889535635709,
      "step": 220,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.035300766862928865,
      "weighted_relation_loss": 0.035300766862928865
    },
    {
      "action_query_norm": 173.4494873046875,
      "action_relation_loss": 0.022069139033555986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.21515434985968196,
      "flow_matching_loss": 0.1227548211812973,
      "grad_norm": 0.44474145770072937,
      "learning_rate": 1.5266666666666667e-05,
      "loss": 0.1553,
      "relation_loss": 0.032516518607735634,
      "selected_pairs": 26947.2,
      "state_query_norm": 221.06133880615235,
      "state_relation_loss": 0.010447379853576421,
      "step": 230,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.032516518607735634,
      "weighted_relation_loss": 0.032516518607735634
    },
    {
      "action_query_norm": 179.68383331298827,
      "action_relation_loss": 0.021115554310381413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2245088868101029,
      "flow_matching_loss": 0.11914977431297302,
      "grad_norm": 0.3235006630420685,
      "learning_rate": 1.5933333333333332e-05,
      "loss": 0.1501,
      "relation_loss": 0.030916367284953596,
      "selected_pairs": 26659.7,
      "state_query_norm": 227.0490493774414,
      "state_relation_loss": 0.00980081334710121,
      "step": 240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.030916367284953596,
      "weighted_relation_loss": 0.030916367284953596
    },
    {
      "action_query_norm": 176.90566864013672,
      "action_relation_loss": 0.021502259373664855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.23386342376052385,
      "flow_matching_loss": 0.1212041586637497,
      "grad_norm": 0.5641729831695557,
      "learning_rate": 1.66e-05,
      "loss": 0.1523,
      "relation_loss": 0.031136501207947732,
      "selected_pairs": 26918.7,
      "state_query_norm": 225.28695220947264,
      "state_relation_loss": 0.00963424202054739,
      "step": 250,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.031136501207947732,
      "weighted_relation_loss": 0.031136501207947732
    },
    {
      "action_query_norm": 173.8808151245117,
      "action_relation_loss": 0.02065748143941164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2432179607109448,
      "flow_matching_loss": 0.11762886866927147,
      "grad_norm": 0.571101725101471,
      "learning_rate": 1.726666666666667e-05,
      "loss": 0.1473,
      "relation_loss": 0.029677528142929076,
      "selected_pairs": 26196.7,
      "state_query_norm": 231.3295120239258,
      "state_relation_loss": 0.009020046982914209,
      "step": 260,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.029677528142929076,
      "weighted_relation_loss": 0.029677528142929076
    },
    {
      "action_query_norm": 176.19249725341797,
      "action_relation_loss": 0.01970078404992819,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.25257249766136575,
      "flow_matching_loss": 0.11158700436353683,
      "grad_norm": 0.46756118535995483,
      "learning_rate": 1.7933333333333337e-05,
      "loss": 0.1401,
      "relation_loss": 0.028494757786393165,
      "selected_pairs": 27476.1,
      "state_query_norm": 235.09641876220704,
      "state_relation_loss": 0.008793973736464978,
      "step": 270,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.028494757786393165,
      "weighted_relation_loss": 0.028494757786393165
    },
    {
      "action_query_norm": 184.24525146484376,
      "action_relation_loss": 0.020044147782027722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.26192703461178674,
      "flow_matching_loss": 0.10829925537109375,
      "grad_norm": 0.45089948177337646,
      "learning_rate": 1.86e-05,
      "loss": 0.1366,
      "relation_loss": 0.02830968573689461,
      "selected_pairs": 27263.9,
      "state_query_norm": 239.38871307373046,
      "state_relation_loss": 0.008265537628903984,
      "step": 280,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.02830968573689461,
      "weighted_relation_loss": 0.02830968573689461
    },
    {
      "action_query_norm": 182.7927673339844,
      "action_relation_loss": 0.019819788821041583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.27128157156220767,
      "flow_matching_loss": 0.11182513535022735,
      "grad_norm": 0.42075949907302856,
      "learning_rate": 1.926666666666667e-05,
      "loss": 0.1395,
      "relation_loss": 0.027708981931209565,
      "selected_pairs": 25961.5,
      "state_query_norm": 244.0873062133789,
      "state_relation_loss": 0.00788919343613088,
      "step": 290,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.027708981931209565,
      "weighted_relation_loss": 0.027708981931209565
    },
    {
      "action_query_norm": 188.04600372314454,
      "action_relation_loss": 0.019368887692689896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2806361085126286,
      "flow_matching_loss": 0.11158646047115325,
      "grad_norm": 0.5065628886222839,
      "learning_rate": 1.9933333333333334e-05,
      "loss": 0.1387,
      "relation_loss": 0.0271413903683424,
      "selected_pairs": 26620.2,
      "state_query_norm": 250.0364776611328,
      "state_relation_loss": 0.007772502768784761,
      "step": 300,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0271413903683424,
      "weighted_relation_loss": 0.0271413903683424
    },
    {
      "action_query_norm": 192.4031997680664,
      "action_relation_loss": 0.018724231235682964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2899906454630496,
      "flow_matching_loss": 0.10860741585493087,
      "grad_norm": 0.45178091526031494,
      "learning_rate": 2.06e-05,
      "loss": 0.1348,
      "relation_loss": 0.026146150752902032,
      "selected_pairs": 26466.7,
      "state_query_norm": 253.85086059570312,
      "state_relation_loss": 0.007421919703483581,
      "step": 310,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.026146150752902032,
      "weighted_relation_loss": 0.026146150752902032
    },
    {
      "action_query_norm": 191.3549575805664,
      "action_relation_loss": 0.01914382018148899,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2993451824134705,
      "flow_matching_loss": 0.10721916183829308,
      "grad_norm": 0.4674515724182129,
      "learning_rate": 2.1266666666666667e-05,
      "loss": 0.1335,
      "relation_loss": 0.026293255761265753,
      "selected_pairs": 27105.2,
      "state_query_norm": 255.36990051269532,
      "state_relation_loss": 0.0071494357194751505,
      "step": 320,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.026293255761265753,
      "weighted_relation_loss": 0.026293255761265753
    },
    {
      "action_query_norm": 189.19156646728516,
      "action_relation_loss": 0.018460644967854024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3086997193638915,
      "flow_matching_loss": 0.10727651491761207,
      "grad_norm": 0.43764981627464294,
      "learning_rate": 2.1933333333333332e-05,
      "loss": 0.1326,
      "relation_loss": 0.025339188426733016,
      "selected_pairs": 26965.0,
      "state_query_norm": 259.3135055541992,
      "state_relation_loss": 0.00687854359857738,
      "step": 330,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.025339188426733016,
      "weighted_relation_loss": 0.025339188426733016
    },
    {
      "action_query_norm": 190.91317749023438,
      "action_relation_loss": 0.017972008138895035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.31805425631431244,
      "flow_matching_loss": 0.10627304092049598,
      "grad_norm": 0.44696635007858276,
      "learning_rate": 2.26e-05,
      "loss": 0.131,
      "relation_loss": 0.024741335585713387,
      "selected_pairs": 26916.4,
      "state_query_norm": 263.61985778808594,
      "state_relation_loss": 0.00676932749338448,
      "step": 340,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.024741335585713387,
      "weighted_relation_loss": 0.024741335585713387
    },
    {
      "action_query_norm": 188.02275848388672,
      "action_relation_loss": 0.017416397854685784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3274087932647334,
      "flow_matching_loss": 0.105187775939703,
      "grad_norm": 0.48083198070526123,
      "learning_rate": 2.326666666666667e-05,
      "loss": 0.1295,
      "relation_loss": 0.02427672054618597,
      "selected_pairs": 26783.9,
      "state_query_norm": 255.0870101928711,
      "state_relation_loss": 0.006860322831198573,
      "step": 350,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.02427672054618597,
      "weighted_relation_loss": 0.02427672054618597
    },
    {
      "action_query_norm": 194.61307373046876,
      "action_relation_loss": 0.017900021374225618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.33676333021515437,
      "flow_matching_loss": 0.10755072087049485,
      "grad_norm": 0.6397559642791748,
      "learning_rate": 2.3933333333333337e-05,
      "loss": 0.1318,
      "relation_loss": 0.0242563184350729,
      "selected_pairs": 26576.7,
      "state_query_norm": 275.3360870361328,
      "state_relation_loss": 0.0063562972471117975,
      "step": 360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0242563184350729,
      "weighted_relation_loss": 0.0242563184350729
    },
    {
      "action_query_norm": 196.5666702270508,
      "action_relation_loss": 0.017876990884542466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3461178671655753,
      "flow_matching_loss": 0.10511581748723983,
      "grad_norm": 0.3782017230987549,
      "learning_rate": 2.46e-05,
      "loss": 0.1293,
      "relation_loss": 0.024229776114225388,
      "selected_pairs": 26965.0,
      "state_query_norm": 264.994873046875,
      "state_relation_loss": 0.006352785462513566,
      "step": 370,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.024229776114225388,
      "weighted_relation_loss": 0.024229776114225388
    },
    {
      "action_query_norm": 196.13507385253905,
      "action_relation_loss": 0.01694877715781331,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.35547240411599623,
      "flow_matching_loss": 0.10143234580755234,
      "grad_norm": 0.47637519240379333,
      "learning_rate": 2.5266666666666666e-05,
      "loss": 0.1248,
      "relation_loss": 0.023339928314089774,
      "selected_pairs": 26673.6,
      "state_query_norm": 268.204931640625,
      "state_relation_loss": 0.006391150783747435,
      "step": 380,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.023339928314089774,
      "weighted_relation_loss": 0.023339928314089774
    },
    {
      "action_query_norm": 202.13196868896483,
      "action_relation_loss": 0.01694293674081564,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3648269410664172,
      "flow_matching_loss": 0.10565074384212494,
      "grad_norm": 0.6303591132164001,
      "learning_rate": 2.5933333333333338e-05,
      "loss": 0.1289,
      "relation_loss": 0.023246569745242595,
      "selected_pairs": 26691.8,
      "state_query_norm": 269.8357666015625,
      "state_relation_loss": 0.006303633283823729,
      "step": 390,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.023246569745242595,
      "weighted_relation_loss": 0.023246569745242595
    },
    {
      "action_query_norm": 199.57242584228516,
      "action_relation_loss": 0.017384649161249398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.37418147801683815,
      "flow_matching_loss": 0.10397771745920181,
      "grad_norm": 0.4011478126049042,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.1272,
      "relation_loss": 0.02325275242328644,
      "selected_pairs": 26829.5,
      "state_query_norm": 277.0225463867188,
      "state_relation_loss": 0.005868103541433811,
      "step": 400,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.02325275242328644,
      "weighted_relation_loss": 0.02325275242328644
    },
    {
      "action_query_norm": 201.31470947265626,
      "action_relation_loss": 0.016834885720163582,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.38353601496725914,
      "flow_matching_loss": 0.09908855706453323,
      "grad_norm": 0.4193701148033142,
      "learning_rate": 2.7266666666666668e-05,
      "loss": 0.1219,
      "relation_loss": 0.022821297869086266,
      "selected_pairs": 27249.4,
      "state_query_norm": 284.1251586914062,
      "state_relation_loss": 0.005986412288621068,
      "step": 410,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.022821297869086266,
      "weighted_relation_loss": 0.022821297869086266
    },
    {
      "action_query_norm": 204.30000762939454,
      "action_relation_loss": 0.016125809960067274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3928905519176801,
      "flow_matching_loss": 0.09765045419335365,
      "grad_norm": 0.3880195617675781,
      "learning_rate": 2.7933333333333332e-05,
      "loss": 0.1198,
      "relation_loss": 0.022162182070314885,
      "selected_pairs": 26803.8,
      "state_query_norm": 278.12110900878906,
      "state_relation_loss": 0.006036372063681484,
      "step": 420,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.022162182070314885,
      "weighted_relation_loss": 0.022162182070314885
    },
    {
      "action_query_norm": 201.9203338623047,
      "action_relation_loss": 0.017199635691940784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.402245088868101,
      "flow_matching_loss": 0.10407502055168152,
      "grad_norm": 0.33431318402290344,
      "learning_rate": 2.86e-05,
      "loss": 0.1269,
      "relation_loss": 0.022864147275686263,
      "selected_pairs": 26985.8,
      "state_query_norm": 287.71343383789065,
      "state_relation_loss": 0.005664511257782579,
      "step": 430,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.022864147275686263,
      "weighted_relation_loss": 0.022864147275686263
    },
    {
      "action_query_norm": 203.89520111083985,
      "action_relation_loss": 0.01650876775383949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.411599625818522,
      "flow_matching_loss": 0.09851630851626396,
      "grad_norm": 0.41798070073127747,
      "learning_rate": 2.926666666666667e-05,
      "loss": 0.1207,
      "relation_loss": 0.02214676607400179,
      "selected_pairs": 26354.5,
      "state_query_norm": 288.05559692382815,
      "state_relation_loss": 0.0056379984598606825,
      "step": 440,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.02214676607400179,
      "weighted_relation_loss": 0.02214676607400179
    },
    {
      "action_query_norm": 207.59705352783203,
      "action_relation_loss": 0.01615890525281429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.42095416276894293,
      "flow_matching_loss": 0.100409497320652,
      "grad_norm": 0.400863379240036,
      "learning_rate": 2.9933333333333337e-05,
      "loss": 0.1219,
      "relation_loss": 0.02152617983520031,
      "selected_pairs": 27014.6,
      "state_query_norm": 291.11940612792966,
      "state_relation_loss": 0.0053672745358198885,
      "step": 450,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.02152617983520031,
      "weighted_relation_loss": 0.02152617983520031
    },
    {
      "action_query_norm": 209.32013244628905,
      "action_relation_loss": 0.016596207022666933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4303086997193639,
      "flow_matching_loss": 0.1002658873796463,
      "grad_norm": 0.541985273361206,
      "learning_rate": 3.06e-05,
      "loss": 0.122,
      "relation_loss": 0.021702773123979568,
      "selected_pairs": 26617.4,
      "state_query_norm": 291.30585021972655,
      "state_relation_loss": 0.005106566101312637,
      "step": 460,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.021702773123979568,
      "weighted_relation_loss": 0.021702773123979568
    },
    {
      "action_query_norm": 211.4908477783203,
      "action_relation_loss": 0.016127614583820105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.43966323666978485,
      "flow_matching_loss": 0.09618836790323257,
      "grad_norm": 0.4196713864803314,
      "learning_rate": 3.126666666666666e-05,
      "loss": 0.1173,
      "relation_loss": 0.0210930859670043,
      "selected_pairs": 25988.2,
      "state_query_norm": 298.14178466796875,
      "state_relation_loss": 0.004965471290051937,
      "step": 470,
      "valid_action_samples": 228.4,
      "weighted_auxiliary_loss": 0.0210930859670043,
      "weighted_relation_loss": 0.0210930859670043
    },
    {
      "action_query_norm": 210.32933197021484,
      "action_relation_loss": 0.016049108002334833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4490177736202058,
      "flow_matching_loss": 0.09737321808934211,
      "grad_norm": 0.4941789209842682,
      "learning_rate": 3.1933333333333335e-05,
      "loss": 0.1183,
      "relation_loss": 0.02094301749020815,
      "selected_pairs": 26416.2,
      "state_query_norm": 297.79090270996096,
      "state_relation_loss": 0.004893909441307187,
      "step": 480,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.02094301749020815,
      "weighted_relation_loss": 0.02094301749020815
    },
    {
      "action_query_norm": 210.15566864013672,
      "action_relation_loss": 0.015483498387038708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4583723105706268,
      "flow_matching_loss": 0.09782718196511268,
      "grad_norm": 0.4438914954662323,
      "learning_rate": 3.26e-05,
      "loss": 0.1182,
      "relation_loss": 0.0203521266579628,
      "selected_pairs": 26572.2,
      "state_query_norm": 299.9333984375,
      "state_relation_loss": 0.004868628410622478,
      "step": 490,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0203521266579628,
      "weighted_relation_loss": 0.0203521266579628
    },
    {
      "action_query_norm": 207.11070556640624,
      "action_relation_loss": 0.015314573794603348,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4677268475210477,
      "flow_matching_loss": 0.09876130148768425,
      "grad_norm": 0.44449955224990845,
      "learning_rate": 3.326666666666667e-05,
      "loss": 0.1189,
      "relation_loss": 0.020168454758822917,
      "selected_pairs": 26546.6,
      "state_query_norm": 299.75938110351564,
      "state_relation_loss": 0.004853881057351827,
      "step": 500,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.020168454758822917,
      "weighted_relation_loss": 0.020168454758822917
    },
    {
      "action_query_norm": 211.4548553466797,
      "action_relation_loss": 0.01622598385438323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.47708138447146864,
      "flow_matching_loss": 0.09440036565065384,
      "grad_norm": 0.4907554090023041,
      "learning_rate": 3.3933333333333336e-05,
      "loss": 0.1153,
      "relation_loss": 0.020891417004168033,
      "selected_pairs": 27058.2,
      "state_query_norm": 307.6534729003906,
      "state_relation_loss": 0.004665433429181576,
      "step": 510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.020891417004168033,
      "weighted_relation_loss": 0.020891417004168033
    },
    {
      "action_query_norm": 211.91691131591796,
      "action_relation_loss": 0.014888055622577667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4864359214218896,
      "flow_matching_loss": 0.09368254020810127,
      "grad_norm": 0.37119364738464355,
      "learning_rate": 3.46e-05,
      "loss": 0.1135,
      "relation_loss": 0.01983028221875429,
      "selected_pairs": 27478.1,
      "state_query_norm": 303.629638671875,
      "state_relation_loss": 0.004942226922139525,
      "step": 520,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.01983028221875429,
      "weighted_relation_loss": 0.01983028221875429
    },
    {
      "action_query_norm": 211.26343841552733,
      "action_relation_loss": 0.015448653511703015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.49579045837231056,
      "flow_matching_loss": 0.09340699389576912,
      "grad_norm": 0.3198866844177246,
      "learning_rate": 3.526666666666667e-05,
      "loss": 0.1135,
      "relation_loss": 0.020095591247081757,
      "selected_pairs": 26298.5,
      "state_query_norm": 311.665283203125,
      "state_relation_loss": 0.004646937921643257,
      "step": 530,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.020095591247081757,
      "weighted_relation_loss": 0.020095591247081757
    },
    {
      "action_query_norm": 214.90834350585936,
      "action_relation_loss": 0.014818134997040034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5051449953227315,
      "flow_matching_loss": 0.09619156867265702,
      "grad_norm": 0.3592257499694824,
      "learning_rate": 3.593333333333334e-05,
      "loss": 0.1156,
      "relation_loss": 0.01940017957240343,
      "selected_pairs": 27060.5,
      "state_query_norm": 310.57477416992185,
      "state_relation_loss": 0.004582044761627912,
      "step": 540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.01940017957240343,
      "weighted_relation_loss": 0.01940017957240343
    },
    {
      "action_query_norm": 219.9906753540039,
      "action_relation_loss": 0.014732781238853932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5144995322731525,
      "flow_matching_loss": 0.09322524964809417,
      "grad_norm": 0.34182506799697876,
      "learning_rate": 3.66e-05,
      "loss": 0.1126,
      "relation_loss": 0.019330866262316705,
      "selected_pairs": 26225.1,
      "state_query_norm": 312.1492065429687,
      "state_relation_loss": 0.004598085442557931,
      "step": 550,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.019330866262316705,
      "weighted_relation_loss": 0.019330866262316705
    },
    {
      "action_query_norm": 217.5544906616211,
      "action_relation_loss": 0.015570538118481636,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5238540692235735,
      "flow_matching_loss": 0.09298321604728699,
      "grad_norm": 0.4336444139480591,
      "learning_rate": 3.726666666666667e-05,
      "loss": 0.1131,
      "relation_loss": 0.02013465706259012,
      "selected_pairs": 26645.2,
      "state_query_norm": 313.70924072265626,
      "state_relation_loss": 0.004564119083806872,
      "step": 560,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.02013465706259012,
      "weighted_relation_loss": 0.02013465706259012
    },
    {
      "action_query_norm": 219.93272094726564,
      "action_relation_loss": 0.015219246409833432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5332086061739943,
      "flow_matching_loss": 0.09131983369588852,
      "grad_norm": 0.3457176089286804,
      "learning_rate": 3.793333333333334e-05,
      "loss": 0.111,
      "relation_loss": 0.019691070914268492,
      "selected_pairs": 26852.9,
      "state_query_norm": 317.4314392089844,
      "state_relation_loss": 0.004471824248321354,
      "step": 570,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.019691070914268492,
      "weighted_relation_loss": 0.019691070914268492
    },
    {
      "action_query_norm": 216.9746856689453,
      "action_relation_loss": 0.014844344556331634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5425631431244153,
      "flow_matching_loss": 0.0922582909464836,
      "grad_norm": 0.3067861497402191,
      "learning_rate": 3.86e-05,
      "loss": 0.1113,
      "relation_loss": 0.019054364040493967,
      "selected_pairs": 27201.6,
      "state_query_norm": 317.26533203125,
      "state_relation_loss": 0.004210019414313137,
      "step": 580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.019054364040493967,
      "weighted_relation_loss": 0.019054364040493967
    },
    {
      "action_query_norm": 215.0322052001953,
      "action_relation_loss": 0.014724243711680175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5519176800748363,
      "flow_matching_loss": 0.09568068534135818,
      "grad_norm": 0.32470229268074036,
      "learning_rate": 3.926666666666667e-05,
      "loss": 0.1146,
      "relation_loss": 0.018914848379790784,
      "selected_pairs": 26848.9,
      "state_query_norm": 322.1008239746094,
      "state_relation_loss": 0.004190604761242866,
      "step": 590,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.018914848379790784,
      "weighted_relation_loss": 0.018914848379790784
    },
    {
      "action_query_norm": 217.8053726196289,
      "action_relation_loss": 0.014375157095491886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5612722170252572,
      "flow_matching_loss": 0.09065111353993416,
      "grad_norm": 0.40412378311157227,
      "learning_rate": 3.993333333333333e-05,
      "loss": 0.109,
      "relation_loss": 0.01839221138507128,
      "selected_pairs": 26365.9,
      "state_query_norm": 330.23615112304685,
      "state_relation_loss": 0.004017054196447134,
      "step": 600,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.01839221138507128,
      "weighted_relation_loss": 0.01839221138507128
    },
    {
      "action_query_norm": 217.8391372680664,
      "action_relation_loss": 0.014796881098300218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5706267539756782,
      "flow_matching_loss": 0.09127567633986473,
      "grad_norm": 0.39654552936553955,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.1101,
      "relation_loss": 0.018844937719404698,
      "selected_pairs": 26413.6,
      "state_query_norm": 322.48821716308595,
      "state_relation_loss": 0.004048056830652058,
      "step": 610,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.018844937719404698,
      "weighted_relation_loss": 0.018844937719404698
    },
    {
      "action_query_norm": 221.2395217895508,
      "action_relation_loss": 0.014119661785662175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5799812909260992,
      "flow_matching_loss": 0.0922117292881012,
      "grad_norm": 0.4937744438648224,
      "learning_rate": 4.126666666666667e-05,
      "loss": 0.1103,
      "relation_loss": 0.01813014466315508,
      "selected_pairs": 27270.6,
      "state_query_norm": 330.6699249267578,
      "state_relation_loss": 0.004010483087040484,
      "step": 620,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.01813014466315508,
      "weighted_relation_loss": 0.01813014466315508
    },
    {
      "action_query_norm": 222.9442169189453,
      "action_relation_loss": 0.014167955983430148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5893358278765201,
      "flow_matching_loss": 0.09349744766950607,
      "grad_norm": 0.49693506956100464,
      "learning_rate": 4.1933333333333334e-05,
      "loss": 0.1115,
      "relation_loss": 0.01805143840610981,
      "selected_pairs": 26306.5,
      "state_query_norm": 331.32606506347656,
      "state_relation_loss": 0.0038834821432828905,
      "step": 630,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.01805143840610981,
      "weighted_relation_loss": 0.01805143840610981
    },
    {
      "action_query_norm": 221.0087158203125,
      "action_relation_loss": 0.014279259275645018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.598690364826941,
      "flow_matching_loss": 0.09312881901860237,
      "grad_norm": 0.5436987280845642,
      "learning_rate": 4.26e-05,
      "loss": 0.1113,
      "relation_loss": 0.01813252791762352,
      "selected_pairs": 27222.9,
      "state_query_norm": 325.2220520019531,
      "state_relation_loss": 0.0038532684789970516,
      "step": 640,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.01813252791762352,
      "weighted_relation_loss": 0.01813252791762352
    },
    {
      "action_query_norm": 220.63835754394532,
      "action_relation_loss": 0.012888561189174651,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.608044901777362,
      "flow_matching_loss": 0.08936340585350991,
      "grad_norm": 0.3858391344547272,
      "learning_rate": 4.3266666666666664e-05,
      "loss": 0.1061,
      "relation_loss": 0.016724778432399036,
      "selected_pairs": 26754.4,
      "state_query_norm": 335.53118896484375,
      "state_relation_loss": 0.00383621733635664,
      "step": 650,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.016724778432399036,
      "weighted_relation_loss": 0.016724778432399036
    },
    {
      "action_query_norm": 223.3227081298828,
      "action_relation_loss": 0.014017087593674659,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.617399438727783,
      "flow_matching_loss": 0.09128605872392655,
      "grad_norm": 0.26014721393585205,
      "learning_rate": 4.3933333333333335e-05,
      "loss": 0.1091,
      "relation_loss": 0.01776522574946284,
      "selected_pairs": 26672.1,
      "state_query_norm": 334.91395263671876,
      "state_relation_loss": 0.0037481382256373762,
      "step": 660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.01776522574946284,
      "weighted_relation_loss": 0.01776522574946284
    },
    {
      "action_query_norm": 226.03655242919922,
      "action_relation_loss": 0.014803342148661613,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6267539756782039,
      "flow_matching_loss": 0.09005392193794251,
      "grad_norm": 0.2950807511806488,
      "learning_rate": 4.46e-05,
      "loss": 0.1085,
      "relation_loss": 0.01842021569609642,
      "selected_pairs": 27316.5,
      "state_query_norm": 334.5619171142578,
      "state_relation_loss": 0.003616873570717871,
      "step": 670,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.01842021569609642,
      "weighted_relation_loss": 0.01842021569609642
    },
    {
      "action_query_norm": 221.92382354736327,
      "action_relation_loss": 0.013909899815917014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6361085126286249,
      "flow_matching_loss": 0.09028064906597137,
      "grad_norm": 0.4200601875782013,
      "learning_rate": 4.526666666666667e-05,
      "loss": 0.1079,
      "relation_loss": 0.017588314600288868,
      "selected_pairs": 26925.6,
      "state_query_norm": 339.3471343994141,
      "state_relation_loss": 0.003678414551541209,
      "step": 680,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.017588314600288868,
      "weighted_relation_loss": 0.017588314600288868
    },
    {
      "action_query_norm": 217.9522506713867,
      "action_relation_loss": 0.01342523330822587,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6454630495790459,
      "flow_matching_loss": 0.08956214413046837,
      "grad_norm": 0.3484449088573456,
      "learning_rate": 4.5933333333333336e-05,
      "loss": 0.1065,
      "relation_loss": 0.016971748508512973,
      "selected_pairs": 27382.9,
      "state_query_norm": 339.6793975830078,
      "state_relation_loss": 0.0035465151071548463,
      "step": 690,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.016971748508512973,
      "weighted_relation_loss": 0.016971748508512973
    },
    {
      "action_query_norm": 228.83826141357423,
      "action_relation_loss": 0.01386057659983635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6548175865294668,
      "flow_matching_loss": 0.09336926266551018,
      "grad_norm": 0.38785600662231445,
      "learning_rate": 4.660000000000001e-05,
      "loss": 0.1107,
      "relation_loss": 0.017310648411512374,
      "selected_pairs": 26574.2,
      "state_query_norm": 347.0936706542969,
      "state_relation_loss": 0.0034500718582421543,
      "step": 700,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.017310648411512374,
      "weighted_relation_loss": 0.017310648411512374
    },
    {
      "action_query_norm": 222.18727111816406,
      "action_relation_loss": 0.01339414892718196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6641721234798877,
      "flow_matching_loss": 0.08985045775771142,
      "grad_norm": 0.27836355566978455,
      "learning_rate": 4.726666666666667e-05,
      "loss": 0.1069,
      "relation_loss": 0.01707143560051918,
      "selected_pairs": 26826.8,
      "state_query_norm": 347.63486938476564,
      "state_relation_loss": 0.0036772867199033497,
      "step": 710,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.01707143560051918,
      "weighted_relation_loss": 0.01707143560051918
    },
    {
      "action_query_norm": 228.45032958984376,
      "action_relation_loss": 0.014463832881301642,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6735266604303087,
      "flow_matching_loss": 0.09195146486163139,
      "grad_norm": 0.4453020393848419,
      "learning_rate": 4.793333333333334e-05,
      "loss": 0.11,
      "relation_loss": 0.018003985472023486,
      "selected_pairs": 27413.5,
      "state_query_norm": 352.8913635253906,
      "state_relation_loss": 0.003540152660571039,
      "step": 720,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.018003985472023486,
      "weighted_relation_loss": 0.018003985472023486
    },
    {
      "action_query_norm": 230.90093688964845,
      "action_relation_loss": 0.013585979491472245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6828811973807296,
      "flow_matching_loss": 0.08971146047115326,
      "grad_norm": 0.3518011271953583,
      "learning_rate": 4.86e-05,
      "loss": 0.1069,
      "relation_loss": 0.01715909717604518,
      "selected_pairs": 26964.0,
      "state_query_norm": 353.3800323486328,
      "state_relation_loss": 0.003573117428459227,
      "step": 730,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01715909717604518,
      "weighted_relation_loss": 0.01715909717604518
    },
    {
      "action_query_norm": 224.4369857788086,
      "action_relation_loss": 0.013308028876781463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6922357343311506,
      "flow_matching_loss": 0.09197193384170532,
      "grad_norm": 0.2849276661872864,
      "learning_rate": 4.926666666666667e-05,
      "loss": 0.1087,
      "relation_loss": 0.016681540943682193,
      "selected_pairs": 27245.2,
      "state_query_norm": 347.0234039306641,
      "state_relation_loss": 0.003373512183316052,
      "step": 740,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.016681540943682193,
      "weighted_relation_loss": 0.016681540943682193
    },
    {
      "action_query_norm": 224.2882095336914,
      "action_relation_loss": 0.013281203527003526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7015902712815716,
      "flow_matching_loss": 0.08503602892160415,
      "grad_norm": 0.37960514426231384,
      "learning_rate": 4.993333333333334e-05,
      "loss": 0.1016,
      "relation_loss": 0.016558773070573806,
      "selected_pairs": 27059.8,
      "state_query_norm": 352.4554107666016,
      "state_relation_loss": 0.003277569520287216,
      "step": 750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.016558773070573806,
      "weighted_relation_loss": 0.016558773070573806
    },
    {
      "action_query_norm": 225.71568756103517,
      "action_relation_loss": 0.01314536239951849,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7109448082319925,
      "flow_matching_loss": 0.08880419880151749,
      "grad_norm": 0.4837154150009155,
      "learning_rate": 5.0600000000000003e-05,
      "loss": 0.1054,
      "relation_loss": 0.016637859307229518,
      "selected_pairs": 26211.2,
      "state_query_norm": 363.5725067138672,
      "state_relation_loss": 0.0034924971405416727,
      "step": 760,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.016637859307229518,
      "weighted_relation_loss": 0.016637859307229518
    },
    {
      "action_query_norm": 233.72270965576172,
      "action_relation_loss": 0.013146562967449427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7202993451824135,
      "flow_matching_loss": 0.0853109009563923,
      "grad_norm": 0.47389280796051025,
      "learning_rate": 5.1266666666666675e-05,
      "loss": 0.102,
      "relation_loss": 0.016654969565570353,
      "selected_pairs": 27034.8,
      "state_query_norm": 352.07353515625,
      "state_relation_loss": 0.003508406411856413,
      "step": 770,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.016654969565570353,
      "weighted_relation_loss": 0.016654969565570353
    },
    {
      "action_query_norm": 229.8061309814453,
      "action_relation_loss": 0.013817547634243966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7296538821328344,
      "flow_matching_loss": 0.08950724676251412,
      "grad_norm": 0.30203914642333984,
      "learning_rate": 5.193333333333333e-05,
      "loss": 0.1068,
      "relation_loss": 0.017318281903862955,
      "selected_pairs": 27082.2,
      "state_query_norm": 362.45533752441406,
      "state_relation_loss": 0.003500734153203666,
      "step": 780,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.017318281903862955,
      "weighted_relation_loss": 0.017318281903862955
    },
    {
      "action_query_norm": 229.54263000488282,
      "action_relation_loss": 0.01308958763256669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7390084190832554,
      "flow_matching_loss": 0.08771313950419427,
      "grad_norm": 0.3314289450645447,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 0.1042,
      "relation_loss": 0.016441250499337912,
      "selected_pairs": 27058.4,
      "state_query_norm": 364.64829406738284,
      "state_relation_loss": 0.0033516627503558994,
      "step": 790,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.016441250499337912,
      "weighted_relation_loss": 0.016441250499337912
    },
    {
      "action_query_norm": 233.9319595336914,
      "action_relation_loss": 0.013499524723738433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7483629560336763,
      "flow_matching_loss": 0.08809012249112129,
      "grad_norm": 0.34830406308174133,
      "learning_rate": 5.326666666666666e-05,
      "loss": 0.1048,
      "relation_loss": 0.016750586312264205,
      "selected_pairs": 27107.8,
      "state_query_norm": 371.0688781738281,
      "state_relation_loss": 0.0032510612858459354,
      "step": 800,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.016750586312264205,
      "weighted_relation_loss": 0.016750586312264205
    },
    {
      "action_query_norm": 235.97161102294922,
      "action_relation_loss": 0.012647254113107919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7577174929840973,
      "flow_matching_loss": 0.0866236500442028,
      "grad_norm": 0.38417452573776245,
      "learning_rate": 5.3933333333333334e-05,
      "loss": 0.1024,
      "relation_loss": 0.015810718573629857,
      "selected_pairs": 27045.8,
      "state_query_norm": 368.72792663574216,
      "state_relation_loss": 0.003163464250974357,
      "step": 810,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.015810718573629857,
      "weighted_relation_loss": 0.015810718573629857
    },
    {
      "action_query_norm": 234.35484313964844,
      "action_relation_loss": 0.01299060070887208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7670720299345183,
      "flow_matching_loss": 0.08620683029294014,
      "grad_norm": 0.35274171829223633,
      "learning_rate": 5.4600000000000006e-05,
      "loss": 0.1024,
      "relation_loss": 0.016171031259000303,
      "selected_pairs": 27290.8,
      "state_query_norm": 371.44394836425784,
      "state_relation_loss": 0.0031804305501282213,
      "step": 820,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.016171031259000303,
      "weighted_relation_loss": 0.016171031259000303
    },
    {
      "action_query_norm": 234.6689422607422,
      "action_relation_loss": 0.01256947973743081,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7764265668849392,
      "flow_matching_loss": 0.08926319107413291,
      "grad_norm": 0.3702718913555145,
      "learning_rate": 5.5266666666666664e-05,
      "loss": 0.105,
      "relation_loss": 0.015696106664836406,
      "selected_pairs": 26780.2,
      "state_query_norm": 368.6138946533203,
      "state_relation_loss": 0.003126626927405596,
      "step": 830,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.015696106664836406,
      "weighted_relation_loss": 0.015696106664836406
    },
    {
      "action_query_norm": 233.0064666748047,
      "action_relation_loss": 0.012305836286395788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7857811038353602,
      "flow_matching_loss": 0.08344559073448181,
      "grad_norm": 0.34673985838890076,
      "learning_rate": 5.5933333333333335e-05,
      "loss": 0.0989,
      "relation_loss": 0.01543336408212781,
      "selected_pairs": 27173.0,
      "state_query_norm": 369.6529998779297,
      "state_relation_loss": 0.0031275274232029915,
      "step": 840,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.01543336408212781,
      "weighted_relation_loss": 0.01543336408212781
    },
    {
      "action_query_norm": 237.9439910888672,
      "action_relation_loss": 0.012690324056893586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7951356407857811,
      "flow_matching_loss": 0.0846167892217636,
      "grad_norm": 0.3351242244243622,
      "learning_rate": 5.66e-05,
      "loss": 0.1003,
      "relation_loss": 0.015727206878364087,
      "selected_pairs": 26987.0,
      "state_query_norm": 380.4346435546875,
      "state_relation_loss": 0.0030368829844519495,
      "step": 850,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.015727206878364087,
      "weighted_relation_loss": 0.015727206878364087
    },
    {
      "action_query_norm": 234.6016403198242,
      "action_relation_loss": 0.012873700354248285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.804490177736202,
      "flow_matching_loss": 0.08486724570393563,
      "grad_norm": 0.30925190448760986,
      "learning_rate": 5.726666666666667e-05,
      "loss": 0.1009,
      "relation_loss": 0.01598596340045333,
      "selected_pairs": 26989.0,
      "state_query_norm": 381.24256896972656,
      "state_relation_loss": 0.0031122632790356874,
      "step": 860,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.01598596340045333,
      "weighted_relation_loss": 0.01598596340045333
    },
    {
      "action_query_norm": 231.49498596191407,
      "action_relation_loss": 0.012349178548902273,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.813844714686623,
      "flow_matching_loss": 0.08860827684402466,
      "grad_norm": 0.457789808511734,
      "learning_rate": 5.7933333333333337e-05,
      "loss": 0.1039,
      "relation_loss": 0.015313336066901684,
      "selected_pairs": 26689.3,
      "state_query_norm": 380.47706909179686,
      "state_relation_loss": 0.002964157541282475,
      "step": 870,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.015313336066901684,
      "weighted_relation_loss": 0.015313336066901684
    },
    {
      "action_query_norm": 235.61329193115233,
      "action_relation_loss": 0.012856569513678551,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.823199251637044,
      "flow_matching_loss": 0.08298399522900582,
      "grad_norm": 0.460379421710968,
      "learning_rate": 5.86e-05,
      "loss": 0.0988,
      "relation_loss": 0.015826940536499023,
      "selected_pairs": 26760.4,
      "state_query_norm": 380.5898101806641,
      "state_relation_loss": 0.002970371022820473,
      "step": 880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.015826940536499023,
      "weighted_relation_loss": 0.015826940536499023
    },
    {
      "action_query_norm": 236.3978271484375,
      "action_relation_loss": 0.012560487259179354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8325537885874649,
      "flow_matching_loss": 0.08506830334663391,
      "grad_norm": 0.31901058554649353,
      "learning_rate": 5.926666666666667e-05,
      "loss": 0.1006,
      "relation_loss": 0.015488325525075198,
      "selected_pairs": 26894.0,
      "state_query_norm": 379.30198974609374,
      "state_relation_loss": 0.002927838172763586,
      "step": 890,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.015488325525075198,
      "weighted_relation_loss": 0.015488325525075198
    },
    {
      "action_query_norm": 231.87041015625,
      "action_relation_loss": 0.012933508213609456,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8419083255378859,
      "flow_matching_loss": 0.08619587048888207,
      "grad_norm": 0.3820139467716217,
      "learning_rate": 5.9933333333333345e-05,
      "loss": 0.102,
      "relation_loss": 0.015836820378899573,
      "selected_pairs": 27289.2,
      "state_query_norm": 381.2234344482422,
      "state_relation_loss": 0.0029033121885731817,
      "step": 900,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.015836820378899573,
      "weighted_relation_loss": 0.015836820378899573
    },
    {
      "action_query_norm": 231.6465606689453,
      "action_relation_loss": 0.012945462483912706,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8512628624883068,
      "flow_matching_loss": 0.08438000679016114,
      "grad_norm": 0.2748344838619232,
      "learning_rate": 6.06e-05,
      "loss": 0.1002,
      "relation_loss": 0.015842645335942505,
      "selected_pairs": 26898.3,
      "state_query_norm": 388.04327392578125,
      "state_relation_loss": 0.002897182945162058,
      "step": 910,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.015842645335942505,
      "weighted_relation_loss": 0.015842645335942505
    },
    {
      "action_query_norm": 235.36663055419922,
      "action_relation_loss": 0.012005392462015152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8606173994387278,
      "flow_matching_loss": 0.08423610851168632,
      "grad_norm": 0.3656565248966217,
      "learning_rate": 6.126666666666667e-05,
      "loss": 0.0992,
      "relation_loss": 0.01496089855208993,
      "selected_pairs": 26660.5,
      "state_query_norm": 384.0993957519531,
      "state_relation_loss": 0.0029555060900747776,
      "step": 920,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.01496089855208993,
      "weighted_relation_loss": 0.01496089855208993
    },
    {
      "action_query_norm": 239.78863830566405,
      "action_relation_loss": 0.012172649521380663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8699719363891487,
      "flow_matching_loss": 0.08329113125801087,
      "grad_norm": 0.3069477081298828,
      "learning_rate": 6.193333333333333e-05,
      "loss": 0.0983,
      "relation_loss": 0.015006606746464968,
      "selected_pairs": 27086.8,
      "state_query_norm": 390.50787048339845,
      "state_relation_loss": 0.0028339572250843047,
      "step": 930,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.015006606746464968,
      "weighted_relation_loss": 0.015006606746464968
    },
    {
      "action_query_norm": 243.5544891357422,
      "action_relation_loss": 0.012416546884924173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8793264733395697,
      "flow_matching_loss": 0.08282178714871406,
      "grad_norm": 0.26281872391700745,
      "learning_rate": 6.26e-05,
      "loss": 0.098,
      "relation_loss": 0.015143201034516097,
      "selected_pairs": 27623.5,
      "state_query_norm": 394.6821563720703,
      "state_relation_loss": 0.002726654103025794,
      "step": 940,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.015143201034516097,
      "weighted_relation_loss": 0.015143201034516097
    },
    {
      "action_query_norm": 249.79690704345703,
      "action_relation_loss": 0.012272367440164089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8886810102899907,
      "flow_matching_loss": 0.08282658383250237,
      "grad_norm": 0.32420557737350464,
      "learning_rate": 6.326666666666667e-05,
      "loss": 0.098,
      "relation_loss": 0.015132930222898721,
      "selected_pairs": 27269.9,
      "state_query_norm": 394.7885467529297,
      "state_relation_loss": 0.0028605628525838257,
      "step": 950,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.015132930222898721,
      "weighted_relation_loss": 0.015132930222898721
    },
    {
      "action_query_norm": 249.05016632080077,
      "action_relation_loss": 0.011855636816471815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8980355472404116,
      "flow_matching_loss": 0.0822502240538597,
      "grad_norm": 0.24315956234931946,
      "learning_rate": 6.393333333333333e-05,
      "loss": 0.0968,
      "relation_loss": 0.014544754661619664,
      "selected_pairs": 26485.8,
      "state_query_norm": 393.7887023925781,
      "state_relation_loss": 0.002689117891713977,
      "step": 960,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.014544754661619664,
      "weighted_relation_loss": 0.014544754661619664
    },
    {
      "action_query_norm": 240.27865447998047,
      "action_relation_loss": 0.012054040934890509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9073900841908326,
      "flow_matching_loss": 0.08129123002290725,
      "grad_norm": 0.31697922945022583,
      "learning_rate": 6.460000000000001e-05,
      "loss": 0.0961,
      "relation_loss": 0.014778620377182961,
      "selected_pairs": 26531.5,
      "state_query_norm": 401.0913818359375,
      "state_relation_loss": 0.0027245793491601942,
      "step": 970,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.014778620377182961,
      "weighted_relation_loss": 0.014778620377182961
    },
    {
      "action_query_norm": 242.85347595214844,
      "action_relation_loss": 0.011554727051407098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9167446211412535,
      "flow_matching_loss": 0.0837326280772686,
      "grad_norm": 0.33094367384910583,
      "learning_rate": 6.526666666666666e-05,
      "loss": 0.0981,
      "relation_loss": 0.014361275918781757,
      "selected_pairs": 26920.1,
      "state_query_norm": 401.80116882324216,
      "state_relation_loss": 0.0028065489372238518,
      "step": 980,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.014361275918781757,
      "weighted_relation_loss": 0.014361275918781757
    },
    {
      "action_query_norm": 239.4491195678711,
      "action_relation_loss": 0.012033148016780616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9260991580916744,
      "flow_matching_loss": 0.0843633159995079,
      "grad_norm": 0.2750435173511505,
      "learning_rate": 6.593333333333334e-05,
      "loss": 0.0992,
      "relation_loss": 0.01479917224496603,
      "selected_pairs": 27012.6,
      "state_query_norm": 404.5852111816406,
      "state_relation_loss": 0.002766024274751544,
      "step": 990,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.01479917224496603,
      "weighted_relation_loss": 0.01479917224496603
    },
    {
      "action_query_norm": 243.8745361328125,
      "action_relation_loss": 0.011759748868644237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9354536950420954,
      "flow_matching_loss": 0.08662145659327507,
      "grad_norm": 0.3341834843158722,
      "learning_rate": 6.66e-05,
      "loss": 0.1011,
      "relation_loss": 0.014482045359909535,
      "selected_pairs": 26686.1,
      "state_query_norm": 398.69081726074216,
      "state_relation_loss": 0.0027222965145483613,
      "step": 1000,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.014482045359909535,
      "weighted_relation_loss": 0.014482045359909535
    },
    {
      "action_query_norm": 248.8827697753906,
      "action_relation_loss": 0.011886944156140089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9448082319925164,
      "flow_matching_loss": 0.08551373183727265,
      "grad_norm": 0.399579793214798,
      "learning_rate": 6.726666666666667e-05,
      "loss": 0.1,
      "relation_loss": 0.014501920808106661,
      "selected_pairs": 26782.0,
      "state_query_norm": 397.0442657470703,
      "state_relation_loss": 0.0026149764889851213,
      "step": 1010,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.014501920808106661,
      "weighted_relation_loss": 0.014501920808106661
    },
    {
      "action_query_norm": 246.36113891601562,
      "action_relation_loss": 0.0116085110232234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9541627689429373,
      "flow_matching_loss": 0.08171378225088119,
      "grad_norm": 0.5260331034660339,
      "learning_rate": 6.793333333333334e-05,
      "loss": 0.096,
      "relation_loss": 0.014271720126271249,
      "selected_pairs": 27155.1,
      "state_query_norm": 405.73089904785155,
      "state_relation_loss": 0.0026632092194631696,
      "step": 1020,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.014271720126271249,
      "weighted_relation_loss": 0.014271720126271249
    },
    {
      "action_query_norm": 242.53585662841797,
      "action_relation_loss": 0.012282603699713946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9635173058933583,
      "flow_matching_loss": 0.08449356481432915,
      "grad_norm": 0.41593465209007263,
      "learning_rate": 6.860000000000001e-05,
      "loss": 0.0996,
      "relation_loss": 0.015058992337435484,
      "selected_pairs": 26726.0,
      "state_query_norm": 399.18565979003904,
      "state_relation_loss": 0.0027763884514570235,
      "step": 1030,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.015058992337435484,
      "weighted_relation_loss": 0.015058992337435484
    },
    {
      "action_query_norm": 243.46815032958983,
      "action_relation_loss": 0.011875123344361782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9728718428437793,
      "flow_matching_loss": 0.08415975868701935,
      "grad_norm": 0.3066416382789612,
      "learning_rate": 6.926666666666667e-05,
      "loss": 0.0988,
      "relation_loss": 0.014615075662732124,
      "selected_pairs": 27193.9,
      "state_query_norm": 408.0427276611328,
      "state_relation_loss": 0.002739952434785664,
      "step": 1040,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.014615075662732124,
      "weighted_relation_loss": 0.014615075662732124
    },
    {
      "action_query_norm": 243.2266387939453,
      "action_relation_loss": 0.012004354503005743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9822263797942001,
      "flow_matching_loss": 0.08215547427535057,
      "grad_norm": 0.2399974763393402,
      "learning_rate": 6.993333333333334e-05,
      "loss": 0.0967,
      "relation_loss": 0.014548908453434706,
      "selected_pairs": 26000.8,
      "state_query_norm": 413.2910858154297,
      "state_relation_loss": 0.002544554159976542,
      "step": 1050,
      "valid_action_samples": 228.5,
      "weighted_auxiliary_loss": 0.014548908453434706,
      "weighted_relation_loss": 0.014548908453434706
    },
    {
      "action_query_norm": 249.8976043701172,
      "action_relation_loss": 0.01167795853689313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9915809167446211,
      "flow_matching_loss": 0.08118498921394349,
      "grad_norm": 0.25143295526504517,
      "learning_rate": 7.06e-05,
      "loss": 0.0955,
      "relation_loss": 0.014319652691483498,
      "selected_pairs": 27615.0,
      "state_query_norm": 414.6811920166016,
      "state_relation_loss": 0.002641694084741175,
      "step": 1060,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.014319652691483498,
      "weighted_relation_loss": 0.014319652691483498
    },
    {
      "action_query_norm": 245.91777954101562,
      "action_relation_loss": 0.011073994170874357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.000935453695042,
      "flow_matching_loss": 0.08153709098696708,
      "grad_norm": 0.26455000042915344,
      "learning_rate": 7.126666666666667e-05,
      "loss": 0.0951,
      "relation_loss": 0.013535943441092968,
      "selected_pairs": 24331.3,
      "state_query_norm": 413.57008056640626,
      "state_relation_loss": 0.002461949177086353,
      "step": 1070,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.013535943441092968,
      "weighted_relation_loss": 0.013535943441092968
    },
    {
      "action_query_norm": 249.8277359008789,
      "action_relation_loss": 0.011377304792404175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.010289990645463,
      "flow_matching_loss": 0.08413148522377015,
      "grad_norm": 0.30802926421165466,
      "learning_rate": 7.193333333333334e-05,
      "loss": 0.098,
      "relation_loss": 0.01383595671504736,
      "selected_pairs": 26854.3,
      "state_query_norm": 419.6433349609375,
      "state_relation_loss": 0.002458651829510927,
      "step": 1080,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.01383595671504736,
      "weighted_relation_loss": 0.01383595671504736
    },
    {
      "action_query_norm": 256.7717819213867,
      "action_relation_loss": 0.01157594071701169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.019644527595884,
      "flow_matching_loss": 0.08136507421731949,
      "grad_norm": 0.35841310024261475,
      "learning_rate": 7.26e-05,
      "loss": 0.0955,
      "relation_loss": 0.0140877490863204,
      "selected_pairs": 26685.2,
      "state_query_norm": 421.9861236572266,
      "state_relation_loss": 0.0025118083693087103,
      "step": 1090,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0140877490863204,
      "weighted_relation_loss": 0.0140877490863204
    },
    {
      "action_query_norm": 260.6109588623047,
      "action_relation_loss": 0.012063075602054597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.028999064546305,
      "flow_matching_loss": 0.08540282174944877,
      "grad_norm": 0.39008331298828125,
      "learning_rate": 7.326666666666667e-05,
      "loss": 0.1001,
      "relation_loss": 0.014665587432682514,
      "selected_pairs": 27292.0,
      "state_query_norm": 412.6672027587891,
      "state_relation_loss": 0.002602511830627918,
      "step": 1100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.014665587432682514,
      "weighted_relation_loss": 0.014665587432682514
    },
    {
      "action_query_norm": 260.6937713623047,
      "action_relation_loss": 0.01166062792763114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0383536014967258,
      "flow_matching_loss": 0.0824704073369503,
      "grad_norm": 0.40451472997665405,
      "learning_rate": 7.393333333333333e-05,
      "loss": 0.0966,
      "relation_loss": 0.014169315993785857,
      "selected_pairs": 26446.3,
      "state_query_norm": 417.71253967285156,
      "state_relation_loss": 0.002508687926456332,
      "step": 1110,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.014169315993785857,
      "weighted_relation_loss": 0.014169315993785857
    },
    {
      "action_query_norm": 251.38077697753906,
      "action_relation_loss": 0.011767330579459667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.047708138447147,
      "flow_matching_loss": 0.0823388360440731,
      "grad_norm": 0.24011710286140442,
      "learning_rate": 7.46e-05,
      "loss": 0.0967,
      "relation_loss": 0.014368886966258288,
      "selected_pairs": 27480.5,
      "state_query_norm": 423.03505249023436,
      "state_relation_loss": 0.00260155638679862,
      "step": 1120,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.014368886966258288,
      "weighted_relation_loss": 0.014368886966258288
    },
    {
      "action_query_norm": 261.8115997314453,
      "action_relation_loss": 0.011866064183413983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0570626753975678,
      "flow_matching_loss": 0.08268067091703415,
      "grad_norm": 0.2806282341480255,
      "learning_rate": 7.526666666666668e-05,
      "loss": 0.097,
      "relation_loss": 0.014306250121444463,
      "selected_pairs": 27033.8,
      "state_query_norm": 424.10560607910156,
      "state_relation_loss": 0.002440185961313546,
      "step": 1130,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.014306250121444463,
      "weighted_relation_loss": 0.014306250121444463
    },
    {
      "action_query_norm": 258.0733947753906,
      "action_relation_loss": 0.011651199217885733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0664172123479887,
      "flow_matching_loss": 0.08084065690636635,
      "grad_norm": 0.3404388725757599,
      "learning_rate": 7.593333333333334e-05,
      "loss": 0.095,
      "relation_loss": 0.014117935858666896,
      "selected_pairs": 26409.6,
      "state_query_norm": 432.7680389404297,
      "state_relation_loss": 0.0024667367804795505,
      "step": 1140,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.014117935858666896,
      "weighted_relation_loss": 0.014117935858666896
    },
    {
      "action_query_norm": 261.20819091796875,
      "action_relation_loss": 0.011920175142586231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0757717492984098,
      "flow_matching_loss": 0.08031724393367767,
      "grad_norm": 0.23492512106895447,
      "learning_rate": 7.66e-05,
      "loss": 0.0945,
      "relation_loss": 0.014196499437093734,
      "selected_pairs": 27662.5,
      "state_query_norm": 428.80010375976565,
      "state_relation_loss": 0.0022763241082429885,
      "step": 1150,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.014196499437093734,
      "weighted_relation_loss": 0.014196499437093734
    },
    {
      "action_query_norm": 262.0367431640625,
      "action_relation_loss": 0.012001251801848412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0851262862488307,
      "flow_matching_loss": 0.08272170946002007,
      "grad_norm": 0.34754613041877747,
      "learning_rate": 7.726666666666667e-05,
      "loss": 0.097,
      "relation_loss": 0.014325746987015008,
      "selected_pairs": 26627.6,
      "state_query_norm": 423.7959747314453,
      "state_relation_loss": 0.0023244951851665975,
      "step": 1160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.014325746987015008,
      "weighted_relation_loss": 0.014325746987015008
    },
    {
      "action_query_norm": 260.43076477050784,
      "action_relation_loss": 0.011618376523256303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0944808231992516,
      "flow_matching_loss": 0.08288098350167275,
      "grad_norm": 0.23399938642978668,
      "learning_rate": 7.793333333333333e-05,
      "loss": 0.0968,
      "relation_loss": 0.013903234805911779,
      "selected_pairs": 26926.1,
      "state_query_norm": 434.76788024902345,
      "state_relation_loss": 0.002284858259372413,
      "step": 1170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.013903234805911779,
      "weighted_relation_loss": 0.013903234805911779
    },
    {
      "action_query_norm": 268.86260681152345,
      "action_relation_loss": 0.011757200956344605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1038353601496727,
      "flow_matching_loss": 0.08299504369497299,
      "grad_norm": 0.2899233400821686,
      "learning_rate": 7.860000000000001e-05,
      "loss": 0.0971,
      "relation_loss": 0.01407624753192067,
      "selected_pairs": 26729.7,
      "state_query_norm": 433.57032470703126,
      "state_relation_loss": 0.002319046459160745,
      "step": 1180,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.01407624753192067,
      "weighted_relation_loss": 0.01407624753192067
    },
    {
      "action_query_norm": 265.9371643066406,
      "action_relation_loss": 0.01201260657981038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1131898971000935,
      "flow_matching_loss": 0.08353727757930755,
      "grad_norm": 0.21713463962078094,
      "learning_rate": 7.926666666666666e-05,
      "loss": 0.0979,
      "relation_loss": 0.014362269546836615,
      "selected_pairs": 27038.9,
      "state_query_norm": 437.4046264648438,
      "state_relation_loss": 0.0023496629437431693,
      "step": 1190,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.014362269546836615,
      "weighted_relation_loss": 0.014362269546836615
    },
    {
      "action_query_norm": 263.3661590576172,
      "action_relation_loss": 0.011347485519945622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1225444340505144,
      "flow_matching_loss": 0.08218451887369156,
      "grad_norm": 0.2965366244316101,
      "learning_rate": 7.993333333333334e-05,
      "loss": 0.0959,
      "relation_loss": 0.013729773368686438,
      "selected_pairs": 26204.3,
      "state_query_norm": 439.1734893798828,
      "state_relation_loss": 0.002382287802174687,
      "step": 1200,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.013729773368686438,
      "weighted_relation_loss": 0.013729773368686438
    },
    {
      "action_query_norm": 263.7786804199219,
      "action_relation_loss": 0.010681074671447277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1318989710009355,
      "flow_matching_loss": 0.08374074473977089,
      "grad_norm": 0.26677241921424866,
      "learning_rate": 8.060000000000001e-05,
      "loss": 0.0968,
      "relation_loss": 0.01302667586132884,
      "selected_pairs": 26853.9,
      "state_query_norm": 441.2104431152344,
      "state_relation_loss": 0.00234560112003237,
      "step": 1210,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.01302667586132884,
      "weighted_relation_loss": 0.01302667586132884
    },
    {
      "action_query_norm": 263.9011627197266,
      "action_relation_loss": 0.011373414099216462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1412535079513564,
      "flow_matching_loss": 0.08142412975430488,
      "grad_norm": 0.39701247215270996,
      "learning_rate": 8.126666666666667e-05,
      "loss": 0.0951,
      "relation_loss": 0.01362802293151617,
      "selected_pairs": 26970.1,
      "state_query_norm": 437.99493408203125,
      "state_relation_loss": 0.002254608948715031,
      "step": 1220,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01362802293151617,
      "weighted_relation_loss": 0.01362802293151617
    },
    {
      "action_query_norm": 257.1745803833008,
      "action_relation_loss": 0.011663947906345128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1506080449017775,
      "flow_matching_loss": 0.08072505965828895,
      "grad_norm": 0.29396888613700867,
      "learning_rate": 8.193333333333334e-05,
      "loss": 0.0946,
      "relation_loss": 0.013894722051918506,
      "selected_pairs": 26903.1,
      "state_query_norm": 444.0351226806641,
      "state_relation_loss": 0.002230774238705635,
      "step": 1230,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.013894722051918506,
      "weighted_relation_loss": 0.013894722051918506
    },
    {
      "action_query_norm": 260.5042785644531,
      "action_relation_loss": 0.010988788679242134,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1599625818521984,
      "flow_matching_loss": 0.08388394415378571,
      "grad_norm": 0.3370097279548645,
      "learning_rate": 8.26e-05,
      "loss": 0.0971,
      "relation_loss": 0.013192425295710564,
      "selected_pairs": 26621.1,
      "state_query_norm": 440.5410491943359,
      "state_relation_loss": 0.002203636453486979,
      "step": 1240,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.013192425295710564,
      "weighted_relation_loss": 0.013192425295710564
    },
    {
      "action_query_norm": 264.6748901367188,
      "action_relation_loss": 0.01158053707331419,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1693171188026192,
      "flow_matching_loss": 0.08268251195549965,
      "grad_norm": 0.2750389277935028,
      "learning_rate": 8.326666666666667e-05,
      "loss": 0.0965,
      "relation_loss": 0.013838985934853554,
      "selected_pairs": 27023.8,
      "state_query_norm": 446.3229217529297,
      "state_relation_loss": 0.002258448931388557,
      "step": 1250,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.013838985934853554,
      "weighted_relation_loss": 0.013838985934853554
    },
    {
      "action_query_norm": 266.4870361328125,
      "action_relation_loss": 0.011114886961877346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1786716557530403,
      "flow_matching_loss": 0.08370445221662522,
      "grad_norm": 0.2966930866241455,
      "learning_rate": 8.393333333333335e-05,
      "loss": 0.0971,
      "relation_loss": 0.01335816429927945,
      "selected_pairs": 27037.5,
      "state_query_norm": 449.0581573486328,
      "state_relation_loss": 0.0022432774072512983,
      "step": 1260,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.01335816429927945,
      "weighted_relation_loss": 0.01335816429927945
    },
    {
      "action_query_norm": 266.22263793945314,
      "action_relation_loss": 0.011273269727826118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1880261927034612,
      "flow_matching_loss": 0.08068537563085557,
      "grad_norm": 0.23128046095371246,
      "learning_rate": 8.46e-05,
      "loss": 0.0941,
      "relation_loss": 0.013389275595545768,
      "selected_pairs": 27059.5,
      "state_query_norm": 459.3358856201172,
      "state_relation_loss": 0.002116005856078118,
      "step": 1270,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.013389275595545768,
      "weighted_relation_loss": 0.013389275595545768
    },
    {
      "action_query_norm": 264.9741729736328,
      "action_relation_loss": 0.011372811999171973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.197380729653882,
      "flow_matching_loss": 0.0839454673230648,
      "grad_norm": 0.2795444428920746,
      "learning_rate": 8.526666666666667e-05,
      "loss": 0.0974,
      "relation_loss": 0.013475274853408336,
      "selected_pairs": 26659.5,
      "state_query_norm": 448.52244873046874,
      "state_relation_loss": 0.0021024626679718495,
      "step": 1280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.013475274853408336,
      "weighted_relation_loss": 0.013475274853408336
    },
    {
      "action_query_norm": 259.95801849365233,
      "action_relation_loss": 0.011713621020317078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2067352666043032,
      "flow_matching_loss": 0.08361533954739571,
      "grad_norm": 0.2942962646484375,
      "learning_rate": 8.593333333333333e-05,
      "loss": 0.0975,
      "relation_loss": 0.013857221510261298,
      "selected_pairs": 27108.8,
      "state_query_norm": 445.5007385253906,
      "state_relation_loss": 0.0021436003735288978,
      "step": 1290,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.013857221510261298,
      "weighted_relation_loss": 0.013857221510261298
    },
    {
      "action_query_norm": 266.32752685546876,
      "action_relation_loss": 0.011289734859019518,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.216089803554724,
      "flow_matching_loss": 0.08079455196857452,
      "grad_norm": 0.33781981468200684,
      "learning_rate": 8.66e-05,
      "loss": 0.0942,
      "relation_loss": 0.01341935619711876,
      "selected_pairs": 26697.4,
      "state_query_norm": 447.5376037597656,
      "state_relation_loss": 0.0021296213380992413,
      "step": 1300,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.01341935619711876,
      "weighted_relation_loss": 0.01341935619711876
    },
    {
      "action_query_norm": 271.74082946777344,
      "action_relation_loss": 0.011379566695541143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.225444340505145,
      "flow_matching_loss": 0.08057250455021858,
      "grad_norm": 0.2596014738082886,
      "learning_rate": 8.726666666666667e-05,
      "loss": 0.0942,
      "relation_loss": 0.013619746174663305,
      "selected_pairs": 26462.9,
      "state_query_norm": 456.1616149902344,
      "state_relation_loss": 0.002240179548971355,
      "step": 1310,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.013619746174663305,
      "weighted_relation_loss": 0.013619746174663305
    },
    {
      "action_query_norm": 276.0668914794922,
      "action_relation_loss": 0.011093659605830909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.234798877455566,
      "flow_matching_loss": 0.07797239646315575,
      "grad_norm": 0.3054620623588562,
      "learning_rate": 8.793333333333333e-05,
      "loss": 0.0912,
      "relation_loss": 0.013266978599131108,
      "selected_pairs": 27136.2,
      "state_query_norm": 458.8184539794922,
      "state_relation_loss": 0.002173319004941732,
      "step": 1320,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.013266978599131108,
      "weighted_relation_loss": 0.013266978599131108
    },
    {
      "action_query_norm": 279.28221435546874,
      "action_relation_loss": 0.01119423247873783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.244153414405987,
      "flow_matching_loss": 0.08281337693333626,
      "grad_norm": 0.19885088503360748,
      "learning_rate": 8.86e-05,
      "loss": 0.0962,
      "relation_loss": 0.013392242416739465,
      "selected_pairs": 26976.3,
      "state_query_norm": 469.48223876953125,
      "state_relation_loss": 0.0021980098448693753,
      "step": 1330,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.013392242416739465,
      "weighted_relation_loss": 0.013392242416739465
    },
    {
      "action_query_norm": 275.85328369140626,
      "action_relation_loss": 0.01144208712503314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2535079513564078,
      "flow_matching_loss": 0.07885327637195587,
      "grad_norm": 0.1737913191318512,
      "learning_rate": 8.926666666666668e-05,
      "loss": 0.0925,
      "relation_loss": 0.013625126797705889,
      "selected_pairs": 27592.7,
      "state_query_norm": 458.15456237792966,
      "state_relation_loss": 0.002183039684314281,
      "step": 1340,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.013625126797705889,
      "weighted_relation_loss": 0.013625126797705889
    },
    {
      "action_query_norm": 277.73646850585936,
      "action_relation_loss": 0.011062880605459213,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.262862488306829,
      "flow_matching_loss": 0.07889830395579338,
      "grad_norm": 0.3117070496082306,
      "learning_rate": 8.993333333333334e-05,
      "loss": 0.0922,
      "relation_loss": 0.013267109356820583,
      "selected_pairs": 26967.6,
      "state_query_norm": 461.5815124511719,
      "state_relation_loss": 0.002204228681512177,
      "step": 1350,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.013267109356820583,
      "weighted_relation_loss": 0.013267109356820583
    },
    {
      "action_query_norm": 284.7958740234375,
      "action_relation_loss": 0.011105439439415932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2722170252572498,
      "flow_matching_loss": 0.07854876667261124,
      "grad_norm": 0.29297831654548645,
      "learning_rate": 9.06e-05,
      "loss": 0.0918,
      "relation_loss": 0.013265404105186462,
      "selected_pairs": 26574.5,
      "state_query_norm": 470.0400085449219,
      "state_relation_loss": 0.002159964735619724,
      "step": 1360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.013265404105186462,
      "weighted_relation_loss": 0.013265404105186462
    },
    {
      "action_query_norm": 282.0662567138672,
      "action_relation_loss": 0.010824947990477085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2815715622076707,
      "flow_matching_loss": 0.07903732657432556,
      "grad_norm": 0.35402414202690125,
      "learning_rate": 9.126666666666667e-05,
      "loss": 0.0918,
      "relation_loss": 0.012797798402607441,
      "selected_pairs": 26501.1,
      "state_query_norm": 459.6134490966797,
      "state_relation_loss": 0.001972850481979549,
      "step": 1370,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.012797798402607441,
      "weighted_relation_loss": 0.012797798402607441
    },
    {
      "action_query_norm": 283.6295532226562,
      "action_relation_loss": 0.010352396126836538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2909260991580918,
      "flow_matching_loss": 0.08203864991664886,
      "grad_norm": 0.3254261016845703,
      "learning_rate": 9.193333333333334e-05,
      "loss": 0.0944,
      "relation_loss": 0.012410507164895534,
      "selected_pairs": 27263.0,
      "state_query_norm": 475.52655029296875,
      "state_relation_loss": 0.0020581111079081894,
      "step": 1380,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.012410507164895534,
      "weighted_relation_loss": 0.012410507164895534
    },
    {
      "action_query_norm": 275.9576385498047,
      "action_relation_loss": 0.010759715177118778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3002806361085126,
      "flow_matching_loss": 0.08212254866957665,
      "grad_norm": 0.24337846040725708,
      "learning_rate": 9.260000000000001e-05,
      "loss": 0.095,
      "relation_loss": 0.01288755126297474,
      "selected_pairs": 26693.4,
      "state_query_norm": 470.95176391601564,
      "state_relation_loss": 0.0021278361789882185,
      "step": 1390,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.01288755126297474,
      "weighted_relation_loss": 0.01288755126297474
    },
    {
      "action_query_norm": 271.5451690673828,
      "action_relation_loss": 0.010763261932879686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3096351730589335,
      "flow_matching_loss": 0.08345879167318344,
      "grad_norm": 0.23792342841625214,
      "learning_rate": 9.326666666666667e-05,
      "loss": 0.0964,
      "relation_loss": 0.01291835866868496,
      "selected_pairs": 26919.3,
      "state_query_norm": 464.310009765625,
      "state_relation_loss": 0.002155096735805273,
      "step": 1400,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.01291835866868496,
      "weighted_relation_loss": 0.01291835866868496
    },
    {
      "action_query_norm": 274.6008605957031,
      "action_relation_loss": 0.010536031890660524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3189897100093546,
      "flow_matching_loss": 0.08002631589770318,
      "grad_norm": 0.28046268224716187,
      "learning_rate": 9.393333333333334e-05,
      "loss": 0.0927,
      "relation_loss": 0.012653046939522027,
      "selected_pairs": 27339.2,
      "state_query_norm": 475.7815734863281,
      "state_relation_loss": 0.0021170149091631176,
      "step": 1410,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.012653046939522027,
      "weighted_relation_loss": 0.012653046939522027
    },
    {
      "action_query_norm": 277.5609924316406,
      "action_relation_loss": 0.010843990556895732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3283442469597755,
      "flow_matching_loss": 0.08116699159145355,
      "grad_norm": 0.332953542470932,
      "learning_rate": 9.46e-05,
      "loss": 0.0941,
      "relation_loss": 0.0129211344756186,
      "selected_pairs": 26665.1,
      "state_query_norm": 480.4478363037109,
      "state_relation_loss": 0.0020771439303644,
      "step": 1420,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0129211344756186,
      "weighted_relation_loss": 0.0129211344756186
    },
    {
      "action_query_norm": 273.21979064941405,
      "action_relation_loss": 0.01069302847608924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3376987839101964,
      "flow_matching_loss": 0.07783697843551636,
      "grad_norm": 0.26040446758270264,
      "learning_rate": 9.526666666666667e-05,
      "loss": 0.0905,
      "relation_loss": 0.012676681205630303,
      "selected_pairs": 26660.6,
      "state_query_norm": 473.11793518066406,
      "state_relation_loss": 0.001983652892522514,
      "step": 1430,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.012676681205630303,
      "weighted_relation_loss": 0.012676681205630303
    },
    {
      "action_query_norm": 275.72548828125,
      "action_relation_loss": 0.010340601019561291,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3470533208606175,
      "flow_matching_loss": 0.0806191086769104,
      "grad_norm": 0.32334208488464355,
      "learning_rate": 9.593333333333334e-05,
      "loss": 0.0929,
      "relation_loss": 0.012273192871361971,
      "selected_pairs": 26951.2,
      "state_query_norm": 479.1010772705078,
      "state_relation_loss": 0.0019325919565744698,
      "step": 1440,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.012273192871361971,
      "weighted_relation_loss": 0.012273192871361971
    },
    {
      "action_query_norm": 283.63291931152344,
      "action_relation_loss": 0.010498819313943386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3564078578110383,
      "flow_matching_loss": 0.07909120246767998,
      "grad_norm": 0.1767537146806717,
      "learning_rate": 9.66e-05,
      "loss": 0.0916,
      "relation_loss": 0.012461675330996513,
      "selected_pairs": 26723.7,
      "state_query_norm": 488.84166564941404,
      "state_relation_loss": 0.0019628559704869984,
      "step": 1450,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.012461675330996513,
      "weighted_relation_loss": 0.012461675330996513
    },
    {
      "action_query_norm": 284.08611450195315,
      "action_relation_loss": 0.010527454689145089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3657623947614592,
      "flow_matching_loss": 0.07894821465015411,
      "grad_norm": 0.29302841424942017,
      "learning_rate": 9.726666666666667e-05,
      "loss": 0.0914,
      "relation_loss": 0.012469009682536124,
      "selected_pairs": 27150.1,
      "state_query_norm": 480.61761779785155,
      "state_relation_loss": 0.001941554865334183,
      "step": 1460,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.012469009682536124,
      "weighted_relation_loss": 0.012469009682536124
    },
    {
      "action_query_norm": 288.36452026367186,
      "action_relation_loss": 0.009931312594562768,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3751169317118803,
      "flow_matching_loss": 0.0786671593785286,
      "grad_norm": 0.21998761594295502,
      "learning_rate": 9.793333333333333e-05,
      "loss": 0.0905,
      "relation_loss": 0.01183113269507885,
      "selected_pairs": 27109.4,
      "state_query_norm": 486.68898010253906,
      "state_relation_loss": 0.0018998201470822097,
      "step": 1470,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.01183113269507885,
      "weighted_relation_loss": 0.01183113269507885
    },
    {
      "action_query_norm": 290.42010192871095,
      "action_relation_loss": 0.01075031403452158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3844714686623012,
      "flow_matching_loss": 0.0783044770359993,
      "grad_norm": 0.26925143599510193,
      "learning_rate": 9.86e-05,
      "loss": 0.091,
      "relation_loss": 0.012646086607128381,
      "selected_pairs": 26978.5,
      "state_query_norm": 484.5998077392578,
      "state_relation_loss": 0.0018957726773805917,
      "step": 1480,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.012646086607128381,
      "weighted_relation_loss": 0.012646086607128381
    },
    {
      "action_query_norm": 286.93030700683596,
      "action_relation_loss": 0.010528853442519904,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.393826005612722,
      "flow_matching_loss": 0.08103190883994102,
      "grad_norm": 0.26271966099739075,
      "learning_rate": 9.926666666666668e-05,
      "loss": 0.0935,
      "relation_loss": 0.012481431197375058,
      "selected_pairs": 26717.2,
      "state_query_norm": 487.15635070800784,
      "state_relation_loss": 0.0019525777082890273,
      "step": 1490,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.012481431197375058,
      "weighted_relation_loss": 0.012481431197375058
    },
    {
      "action_query_norm": 288.1251159667969,
      "action_relation_loss": 0.010530505049973726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4031805425631432,
      "flow_matching_loss": 0.07888642251491547,
      "grad_norm": 0.21228653192520142,
      "learning_rate": 9.993333333333334e-05,
      "loss": 0.0913,
      "relation_loss": 0.012426033057272435,
      "selected_pairs": 26481.1,
      "state_query_norm": 495.32353515625,
      "state_relation_loss": 0.001895527949091047,
      "step": 1500,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.012426033057272435,
      "weighted_relation_loss": 0.012426033057272435
    },
    {
      "action_query_norm": 287.8960906982422,
      "action_relation_loss": 0.011256755143404008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.412535079513564,
      "flow_matching_loss": 0.0807700976729393,
      "grad_norm": 0.2821492850780487,
      "learning_rate": 9.999997539434007e-05,
      "loss": 0.094,
      "relation_loss": 0.013217385858297348,
      "selected_pairs": 26966.8,
      "state_query_norm": 486.27240295410155,
      "state_relation_loss": 0.0019606306566856803,
      "step": 1510,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.013217385858297348,
      "weighted_relation_loss": 0.013217385858297348
    },
    {
      "action_query_norm": 296.3557373046875,
      "action_relation_loss": 0.010539949499070644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.421889616463985,
      "flow_matching_loss": 0.07876260504126549,
      "grad_norm": 0.2927161157131195,
      "learning_rate": 9.999989033776898e-05,
      "loss": 0.0912,
      "relation_loss": 0.012487151008099318,
      "selected_pairs": 27545.6,
      "state_query_norm": 479.2643798828125,
      "state_relation_loss": 0.0019472014857456088,
      "step": 1520,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.012487151008099318,
      "weighted_relation_loss": 0.012487151008099318
    },
    {
      "action_query_norm": 295.3998321533203,
      "action_relation_loss": 0.010526895429939032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.431244153414406,
      "flow_matching_loss": 0.08033071458339691,
      "grad_norm": 0.2350798398256302,
      "learning_rate": 9.999974452661641e-05,
      "loss": 0.0928,
      "relation_loss": 0.012429316435009241,
      "selected_pairs": 26709.9,
      "state_query_norm": 490.7585083007813,
      "state_relation_loss": 0.0019024210167117416,
      "step": 1530,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.012429316435009241,
      "weighted_relation_loss": 0.012429316435009241
    },
    {
      "action_query_norm": 291.2217620849609,
      "action_relation_loss": 0.010511873383074999,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.440598690364827,
      "flow_matching_loss": 0.07766651064157486,
      "grad_norm": 0.2869342267513275,
      "learning_rate": 9.999953796105959e-05,
      "loss": 0.0901,
      "relation_loss": 0.012467924132943153,
      "selected_pairs": 26944.6,
      "state_query_norm": 479.54847106933596,
      "state_relation_loss": 0.001956050773151219,
      "step": 1540,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.012467924132943153,
      "weighted_relation_loss": 0.012467924132943153
    },
    {
      "action_query_norm": 285.52098388671874,
      "action_relation_loss": 0.011157539393752813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4499532273152478,
      "flow_matching_loss": 0.07603230327367783,
      "grad_norm": 0.291239857673645,
      "learning_rate": 9.999927064134949e-05,
      "loss": 0.0891,
      "relation_loss": 0.013052567932754755,
      "selected_pairs": 26316.8,
      "state_query_norm": 480.57911071777346,
      "state_relation_loss": 0.001895028550643474,
      "step": 1550,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.013052567932754755,
      "weighted_relation_loss": 0.013052567932754755
    },
    {
      "action_query_norm": 300.2772491455078,
      "action_relation_loss": 0.009833483211696147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4593077642656689,
      "flow_matching_loss": 0.07922354489564895,
      "grad_norm": 0.21879512071609497,
      "learning_rate": 9.999894256781095e-05,
      "loss": 0.0909,
      "relation_loss": 0.011718196887522936,
      "selected_pairs": 26826.8,
      "state_query_norm": 494.8099426269531,
      "state_relation_loss": 0.0018847136758267879,
      "step": 1560,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.011718196887522936,
      "weighted_relation_loss": 0.011718196887522936
    },
    {
      "action_query_norm": 302.8892578125,
      "action_relation_loss": 0.010666837263852359,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4686623012160898,
      "flow_matching_loss": 0.07817206010222436,
      "grad_norm": 0.20387497544288635,
      "learning_rate": 9.99985537408426e-05,
      "loss": 0.0907,
      "relation_loss": 0.012497013993561268,
      "selected_pairs": 27079.3,
      "state_query_norm": 500.6191711425781,
      "state_relation_loss": 0.0018301768344826997,
      "step": 1570,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.012497013993561268,
      "weighted_relation_loss": 0.012497013993561268
    },
    {
      "action_query_norm": 302.0485504150391,
      "action_relation_loss": 0.010679745767265558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4780168381665106,
      "flow_matching_loss": 0.07664949223399162,
      "grad_norm": 0.2678481936454773,
      "learning_rate": 9.999810416091688e-05,
      "loss": 0.0893,
      "relation_loss": 0.012648854311555624,
      "selected_pairs": 27011.6,
      "state_query_norm": 503.4222137451172,
      "state_relation_loss": 0.001969108625780791,
      "step": 1580,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.012648854311555624,
      "weighted_relation_loss": 0.012648854311555624
    },
    {
      "action_query_norm": 298.17633361816405,
      "action_relation_loss": 0.010005980264395475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4873713751169317,
      "flow_matching_loss": 0.0783376544713974,
      "grad_norm": 0.1907239407300949,
      "learning_rate": 9.99975938285801e-05,
      "loss": 0.0902,
      "relation_loss": 0.011875806469470262,
      "selected_pairs": 26389.4,
      "state_query_norm": 503.72585144042966,
      "state_relation_loss": 0.0018698261003009975,
      "step": 1590,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.011875806469470262,
      "weighted_relation_loss": 0.011875806469470262
    },
    {
      "action_query_norm": 295.3674377441406,
      "action_relation_loss": 0.01061379574239254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4967259120673526,
      "flow_matching_loss": 0.07888604179024697,
      "grad_norm": 0.20223917067050934,
      "learning_rate": 9.999702274445236e-05,
      "loss": 0.0914,
      "relation_loss": 0.012504427134990692,
      "selected_pairs": 27599.6,
      "state_query_norm": 495.2479248046875,
      "state_relation_loss": 0.0018906313693150879,
      "step": 1600,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.012504427134990692,
      "weighted_relation_loss": 0.012504427134990692
    },
    {
      "action_query_norm": 300.24432678222655,
      "action_relation_loss": 0.009680497832596302,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5060804490177735,
      "flow_matching_loss": 0.07817223817110061,
      "grad_norm": 0.23305438458919525,
      "learning_rate": 9.999639090922756e-05,
      "loss": 0.0897,
      "relation_loss": 0.011517758760601283,
      "selected_pairs": 26897.9,
      "state_query_norm": 509.5516082763672,
      "state_relation_loss": 0.0018372609396465124,
      "step": 1610,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.011517758760601283,
      "weighted_relation_loss": 0.011517758760601283
    },
    {
      "action_query_norm": 300.33778076171876,
      "action_relation_loss": 0.00977215487509966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5154349859681946,
      "flow_matching_loss": 0.0730219803750515,
      "grad_norm": 0.33594322204589844,
      "learning_rate": 9.999569832367346e-05,
      "loss": 0.0846,
      "relation_loss": 0.011575306020677089,
      "selected_pairs": 27243.8,
      "state_query_norm": 523.8225219726562,
      "state_relation_loss": 0.0018031510640867054,
      "step": 1620,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.011575306020677089,
      "weighted_relation_loss": 0.011575306020677089
    },
    {
      "action_query_norm": 292.4248962402344,
      "action_relation_loss": 0.009830664005130529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5247895229186157,
      "flow_matching_loss": 0.07778273299336433,
      "grad_norm": 0.22464494407176971,
      "learning_rate": 9.999494498863162e-05,
      "loss": 0.0894,
      "relation_loss": 0.011645096447318792,
      "selected_pairs": 26732.0,
      "state_query_norm": 501.1803375244141,
      "state_relation_loss": 0.0018144323490560054,
      "step": 1630,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.011645096447318792,
      "weighted_relation_loss": 0.011645096447318792
    },
    {
      "action_query_norm": 292.64882507324216,
      "action_relation_loss": 0.01004371214658022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5341440598690363,
      "flow_matching_loss": 0.07513096854090691,
      "grad_norm": 0.3097492456436157,
      "learning_rate": 9.99941309050174e-05,
      "loss": 0.0869,
      "relation_loss": 0.011814218014478683,
      "selected_pairs": 27085.0,
      "state_query_norm": 496.7731109619141,
      "state_relation_loss": 0.001770505856256932,
      "step": 1640,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.011814218014478683,
      "weighted_relation_loss": 0.011814218014478683
    },
    {
      "action_query_norm": 300.5580322265625,
      "action_relation_loss": 0.010503777023404836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5434985968194574,
      "flow_matching_loss": 0.07926465049386025,
      "grad_norm": 0.2722613513469696,
      "learning_rate": 9.999325607381999e-05,
      "loss": 0.0916,
      "relation_loss": 0.012329481542110443,
      "selected_pairs": 26320.7,
      "state_query_norm": 514.9336730957032,
      "state_relation_loss": 0.0018257046467624605,
      "step": 1650,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.012329481542110443,
      "weighted_relation_loss": 0.012329481542110443
    },
    {
      "action_query_norm": 300.4426788330078,
      "action_relation_loss": 0.009882618393748998,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5528531337698785,
      "flow_matching_loss": 0.08025520145893097,
      "grad_norm": 0.22783870995044708,
      "learning_rate": 9.999232049610238e-05,
      "loss": 0.092,
      "relation_loss": 0.011701426841318608,
      "selected_pairs": 26759.7,
      "state_query_norm": 508.7567657470703,
      "state_relation_loss": 0.0018188083195127547,
      "step": 1660,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.011701426841318608,
      "weighted_relation_loss": 0.011701426841318608
    },
    {
      "action_query_norm": 300.8139678955078,
      "action_relation_loss": 0.009993266686797142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5622076707202992,
      "flow_matching_loss": 0.07671134695410728,
      "grad_norm": 0.23254115879535675,
      "learning_rate": 9.999132417300142e-05,
      "loss": 0.0885,
      "relation_loss": 0.011819570697844029,
      "selected_pairs": 26814.8,
      "state_query_norm": 509.09418334960935,
      "state_relation_loss": 0.0018263039411976934,
      "step": 1670,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.011819570697844029,
      "weighted_relation_loss": 0.011819570697844029
    },
    {
      "action_query_norm": 297.91973266601565,
      "action_relation_loss": 0.010233238246291876,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5715622076707203,
      "flow_matching_loss": 0.07618497684597969,
      "grad_norm": 0.2832983136177063,
      "learning_rate": 9.99902671057277e-05,
      "loss": 0.0882,
      "relation_loss": 0.01199220959097147,
      "selected_pairs": 26572.3,
      "state_query_norm": 510.9808349609375,
      "state_relation_loss": 0.0017589712399058044,
      "step": 1680,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.01199220959097147,
      "weighted_relation_loss": 0.01199220959097147
    },
    {
      "action_query_norm": 297.2290283203125,
      "action_relation_loss": 0.009895204938948155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5809167446211414,
      "flow_matching_loss": 0.07742695659399032,
      "grad_norm": 0.3185674250125885,
      "learning_rate": 9.998914929556569e-05,
      "loss": 0.089,
      "relation_loss": 0.011592027265578508,
      "selected_pairs": 26998.2,
      "state_query_norm": 506.1217864990234,
      "state_relation_loss": 0.001696822326630354,
      "step": 1690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.011592027265578508,
      "weighted_relation_loss": 0.011592027265578508
    },
    {
      "action_query_norm": 300.87950439453124,
      "action_relation_loss": 0.009484822582453489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.590271281571562,
      "flow_matching_loss": 0.07846665158867835,
      "grad_norm": 0.24853064119815826,
      "learning_rate": 9.998797074387361e-05,
      "loss": 0.0897,
      "relation_loss": 0.01126782912760973,
      "selected_pairs": 26685.9,
      "state_query_norm": 515.3996765136719,
      "state_relation_loss": 0.0017830066732130944,
      "step": 1700,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.01126782912760973,
      "weighted_relation_loss": 0.01126782912760973
    },
    {
      "action_query_norm": 301.0535583496094,
      "action_relation_loss": 0.010357186757028103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5996258185219832,
      "flow_matching_loss": 0.07953323423862457,
      "grad_norm": 0.2104903906583786,
      "learning_rate": 9.99867314520835e-05,
      "loss": 0.0916,
      "relation_loss": 0.012090703006833792,
      "selected_pairs": 26668.3,
      "state_query_norm": 530.3461090087891,
      "state_relation_loss": 0.0017335162498056888,
      "step": 1710,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.012090703006833792,
      "weighted_relation_loss": 0.012090703006833792
    },
    {
      "action_query_norm": 303.7970794677734,
      "action_relation_loss": 0.009973815269768238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6089803554724043,
      "flow_matching_loss": 0.07733592092990875,
      "grad_norm": 0.24150843918323517,
      "learning_rate": 9.998543142170126e-05,
      "loss": 0.0891,
      "relation_loss": 0.011751953233033418,
      "selected_pairs": 26461.7,
      "state_query_norm": 520.6509552001953,
      "state_relation_loss": 0.0017781378468498588,
      "step": 1720,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.011751953233033418,
      "weighted_relation_loss": 0.011751953233033418
    },
    {
      "action_query_norm": 301.8589263916016,
      "action_relation_loss": 0.009548309911042451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.618334892422825,
      "flow_matching_loss": 0.07716958820819855,
      "grad_norm": 0.20401398837566376,
      "learning_rate": 9.99840706543065e-05,
      "loss": 0.0885,
      "relation_loss": 0.011316597554832697,
      "selected_pairs": 27149.7,
      "state_query_norm": 529.7447082519532,
      "state_relation_loss": 0.0017682877602055668,
      "step": 1730,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.011316597554832697,
      "weighted_relation_loss": 0.011316597554832697
    },
    {
      "action_query_norm": 308.66968688964846,
      "action_relation_loss": 0.010337577760219574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.627689429373246,
      "flow_matching_loss": 0.07568058595061303,
      "grad_norm": 0.2556317448616028,
      "learning_rate": 9.998264915155274e-05,
      "loss": 0.0878,
      "relation_loss": 0.012118035089224577,
      "selected_pairs": 26826.4,
      "state_query_norm": 527.469369506836,
      "state_relation_loss": 0.0017804573173634708,
      "step": 1740,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.012118035089224577,
      "weighted_relation_loss": 0.012118035089224577
    },
    {
      "action_query_norm": 313.0710815429687,
      "action_relation_loss": 0.009316805191338063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.637043966323667,
      "flow_matching_loss": 0.07659081444144249,
      "grad_norm": 0.2575783431529999,
      "learning_rate": 9.998116691516718e-05,
      "loss": 0.0877,
      "relation_loss": 0.011091559007763863,
      "selected_pairs": 26640.6,
      "state_query_norm": 542.1224426269531,
      "state_relation_loss": 0.0017747538979165256,
      "step": 1750,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.011091559007763863,
      "weighted_relation_loss": 0.011091559007763863
    },
    {
      "action_query_norm": 309.8066009521484,
      "action_relation_loss": 0.010196651890873909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.646398503274088,
      "flow_matching_loss": 0.0772792048752308,
      "grad_norm": 0.27210909128189087,
      "learning_rate": 9.997962394695091e-05,
      "loss": 0.0893,
      "relation_loss": 0.01199893020093441,
      "selected_pairs": 26644.2,
      "state_query_norm": 541.1378601074218,
      "state_relation_loss": 0.0018022782285697758,
      "step": 1760,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.01199893020093441,
      "weighted_relation_loss": 0.01199893020093441
    },
    {
      "action_query_norm": 304.96779479980466,
      "action_relation_loss": 0.00953561570495367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6557530402245089,
      "flow_matching_loss": 0.0720097865909338,
      "grad_norm": 0.17290040850639343,
      "learning_rate": 9.997802024877875e-05,
      "loss": 0.0833,
      "relation_loss": 0.011260084062814712,
      "selected_pairs": 26640.7,
      "state_query_norm": 526.5202819824219,
      "state_relation_loss": 0.0017244682763703167,
      "step": 1770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.011260084062814712,
      "weighted_relation_loss": 0.011260084062814712
    },
    {
      "action_query_norm": 307.7735931396484,
      "action_relation_loss": 0.010100467689335346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.66510757717493,
      "flow_matching_loss": 0.07565473690629006,
      "grad_norm": 0.258993536233902,
      "learning_rate": 9.99763558225994e-05,
      "loss": 0.0874,
      "relation_loss": 0.011793691664934158,
      "selected_pairs": 26638.9,
      "state_query_norm": 525.7177825927735,
      "state_relation_loss": 0.0016932240221649408,
      "step": 1780,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.011793691664934158,
      "weighted_relation_loss": 0.011793691664934158
    },
    {
      "action_query_norm": 314.5115478515625,
      "action_relation_loss": 0.009665680397301913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6744621141253508,
      "flow_matching_loss": 0.07350990027189255,
      "grad_norm": 0.2473670244216919,
      "learning_rate": 9.997463067043526e-05,
      "loss": 0.0848,
      "relation_loss": 0.011313025373965502,
      "selected_pairs": 27342.1,
      "state_query_norm": 523.0127777099609,
      "state_relation_loss": 0.0016473448602482676,
      "step": 1790,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.011313025373965502,
      "weighted_relation_loss": 0.011313025373965502
    },
    {
      "action_query_norm": 315.47767639160156,
      "action_relation_loss": 0.009825000865384936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6838166510757717,
      "flow_matching_loss": 0.0740852627903223,
      "grad_norm": 0.21468450129032135,
      "learning_rate": 9.997284479438253e-05,
      "loss": 0.0855,
      "relation_loss": 0.011413146369159221,
      "selected_pairs": 27130.7,
      "state_query_norm": 521.4224304199219,
      "state_relation_loss": 0.0015881454921327532,
      "step": 1800,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.011413146369159221,
      "weighted_relation_loss": 0.011413146369159221
    },
    {
      "action_query_norm": 316.05882263183594,
      "action_relation_loss": 0.00960097024217248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6931711880261928,
      "flow_matching_loss": 0.07471544221043587,
      "grad_norm": 0.28254377841949463,
      "learning_rate": 9.997099819661127e-05,
      "loss": 0.0859,
      "relation_loss": 0.011139926500618458,
      "selected_pairs": 26480.5,
      "state_query_norm": 533.2640502929687,
      "state_relation_loss": 0.0015389563632197678,
      "step": 1810,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.011139926500618458,
      "weighted_relation_loss": 0.011139926500618458
    },
    {
      "action_query_norm": 315.5699462890625,
      "action_relation_loss": 0.009789381735026836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7025257249766137,
      "flow_matching_loss": 0.07766021937131881,
      "grad_norm": 0.2479434311389923,
      "learning_rate": 9.996909087936524e-05,
      "loss": 0.0891,
      "relation_loss": 0.01140918182209134,
      "selected_pairs": 27339.8,
      "state_query_norm": 538.5714721679688,
      "state_relation_loss": 0.0016198000637814402,
      "step": 1820,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.01140918182209134,
      "weighted_relation_loss": 0.01140918182209134
    },
    {
      "action_query_norm": 312.3800750732422,
      "action_relation_loss": 0.009945463156327605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7118802619270346,
      "flow_matching_loss": 0.08009628430008889,
      "grad_norm": 0.22002077102661133,
      "learning_rate": 9.996712284496201e-05,
      "loss": 0.0916,
      "relation_loss": 0.011553068459033967,
      "selected_pairs": 27126.5,
      "state_query_norm": 539.1115905761719,
      "state_relation_loss": 0.0016076052677817642,
      "step": 1830,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.011553068459033967,
      "weighted_relation_loss": 0.011553068459033967
    },
    {
      "action_query_norm": 311.6152862548828,
      "action_relation_loss": 0.00982297509908676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7212347988774557,
      "flow_matching_loss": 0.07689912691712379,
      "grad_norm": 0.2618357837200165,
      "learning_rate": 9.996509409579293e-05,
      "loss": 0.0884,
      "relation_loss": 0.011529655288904905,
      "selected_pairs": 26743.6,
      "state_query_norm": 541.8806213378906,
      "state_relation_loss": 0.0017066802829504013,
      "step": 1840,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.011529655288904905,
      "weighted_relation_loss": 0.011529655288904905
    },
    {
      "action_query_norm": 305.13718872070314,
      "action_relation_loss": 0.009380652103573084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7305893358278766,
      "flow_matching_loss": 0.07955270782113075,
      "grad_norm": 0.22231051325798035,
      "learning_rate": 9.996300463432312e-05,
      "loss": 0.0907,
      "relation_loss": 0.011110715940594674,
      "selected_pairs": 27202.9,
      "state_query_norm": 540.6952575683594,
      "state_relation_loss": 0.0017300638370215892,
      "step": 1850,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.011110715940594674,
      "weighted_relation_loss": 0.011110715940594674
    },
    {
      "action_query_norm": 312.00623474121096,
      "action_relation_loss": 0.00977797769010067,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7399438727782974,
      "flow_matching_loss": 0.0718812182545662,
      "grad_norm": 0.22110536694526672,
      "learning_rate": 9.996085446309148e-05,
      "loss": 0.0833,
      "relation_loss": 0.011369862593710423,
      "selected_pairs": 26477.1,
      "state_query_norm": 540.2490783691406,
      "state_relation_loss": 0.0015918849618174135,
      "step": 1860,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.011369862593710423,
      "weighted_relation_loss": 0.011369862593710423
    },
    {
      "action_query_norm": 313.84104309082034,
      "action_relation_loss": 0.009197423979640008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7492984097287185,
      "flow_matching_loss": 0.0744894377887249,
      "grad_norm": 0.21278278529644012,
      "learning_rate": 9.995864358471066e-05,
      "loss": 0.0853,
      "relation_loss": 0.010829896759241819,
      "selected_pairs": 26878.5,
      "state_query_norm": 533.6172302246093,
      "state_relation_loss": 0.0016324728028848767,
      "step": 1870,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.010829896759241819,
      "weighted_relation_loss": 0.010829896759241819
    },
    {
      "action_query_norm": 315.3880340576172,
      "action_relation_loss": 0.00965677686035633,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7586529466791394,
      "flow_matching_loss": 0.0740131065249443,
      "grad_norm": 0.19421692192554474,
      "learning_rate": 9.99563720018671e-05,
      "loss": 0.0853,
      "relation_loss": 0.011278317030519246,
      "selected_pairs": 27152.7,
      "state_query_norm": 551.6971435546875,
      "state_relation_loss": 0.001621540228370577,
      "step": 1880,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.011278317030519246,
      "weighted_relation_loss": 0.011278317030519246
    },
    {
      "action_query_norm": 315.35313720703124,
      "action_relation_loss": 0.009429930523037911,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7680074836295603,
      "flow_matching_loss": 0.07579517066478729,
      "grad_norm": 0.20513617992401123,
      "learning_rate": 9.995403971732098e-05,
      "loss": 0.0868,
      "relation_loss": 0.010980577766895294,
      "selected_pairs": 27152.4,
      "state_query_norm": 545.0426513671875,
      "state_relation_loss": 0.0015506471856497229,
      "step": 1890,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.010980577766895294,
      "weighted_relation_loss": 0.010980577766895294
    },
    {
      "action_query_norm": 311.0481292724609,
      "action_relation_loss": 0.009864544868469239,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7773620205799814,
      "flow_matching_loss": 0.0710784450173378,
      "grad_norm": 0.29409536719322205,
      "learning_rate": 9.995164673390625e-05,
      "loss": 0.0825,
      "relation_loss": 0.011457893159240484,
      "selected_pairs": 27033.6,
      "state_query_norm": 549.5893981933593,
      "state_relation_loss": 0.001593348302412778,
      "step": 1900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.011457893159240484,
      "weighted_relation_loss": 0.011457893159240484
    },
    {
      "action_query_norm": 312.1933319091797,
      "action_relation_loss": 0.009647067729383707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7867165575304023,
      "flow_matching_loss": 0.07492240741848946,
      "grad_norm": 0.26136595010757446,
      "learning_rate": 9.994919305453059e-05,
      "loss": 0.0861,
      "relation_loss": 0.011180013231933118,
      "selected_pairs": 26486.7,
      "state_query_norm": 546.342041015625,
      "state_relation_loss": 0.0015329454792663455,
      "step": 1910,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.011180013231933118,
      "weighted_relation_loss": 0.011180013231933118
    },
    {
      "action_query_norm": 316.36683349609376,
      "action_relation_loss": 0.009249035269021988,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7960710944808231,
      "flow_matching_loss": 0.07685123085975647,
      "grad_norm": 0.2705959379673004,
      "learning_rate": 9.994667868217548e-05,
      "loss": 0.0877,
      "relation_loss": 0.010818608477711678,
      "selected_pairs": 27858.6,
      "state_query_norm": 551.472705078125,
      "state_relation_loss": 0.0015695732552558183,
      "step": 1920,
      "valid_action_samples": 236.5,
      "weighted_auxiliary_loss": 0.010818608477711678,
      "weighted_relation_loss": 0.010818608477711678
    },
    {
      "action_query_norm": 316.63890686035154,
      "action_relation_loss": 0.01020288011059165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8054256314312442,
      "flow_matching_loss": 0.07833636328577995,
      "grad_norm": 0.30072271823883057,
      "learning_rate": 9.99441036198961e-05,
      "loss": 0.0901,
      "relation_loss": 0.011751140374690295,
      "selected_pairs": 27152.7,
      "state_query_norm": 552.90400390625,
      "state_relation_loss": 0.0015482602640986443,
      "step": 1930,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.011751140374690295,
      "weighted_relation_loss": 0.011751140374690295
    },
    {
      "action_query_norm": 321.79747619628904,
      "action_relation_loss": 0.00972325075417757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8147801683816651,
      "flow_matching_loss": 0.0748776525259018,
      "grad_norm": 0.35343071818351746,
      "learning_rate": 9.99414678708214e-05,
      "loss": 0.0861,
      "relation_loss": 0.011257641483098269,
      "selected_pairs": 26075.9,
      "state_query_norm": 565.4462036132812,
      "state_relation_loss": 0.0015343907405622303,
      "step": 1940,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.011257641483098269,
      "weighted_relation_loss": 0.011257641483098269
    },
    {
      "action_query_norm": 321.52976989746094,
      "action_relation_loss": 0.008906353311613203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.824134705332086,
      "flow_matching_loss": 0.07548805773258209,
      "grad_norm": 0.3878430426120758,
      "learning_rate": 9.993877143815407e-05,
      "loss": 0.086,
      "relation_loss": 0.010488234832882882,
      "selected_pairs": 26204.5,
      "state_query_norm": 556.9283935546875,
      "state_relation_loss": 0.0015818814514204859,
      "step": 1950,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.010488234832882882,
      "weighted_relation_loss": 0.010488234832882882
    },
    {
      "action_query_norm": 318.87359008789065,
      "action_relation_loss": 0.009789906907826662,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.833489242282507,
      "flow_matching_loss": 0.0762317143380642,
      "grad_norm": 0.37762710452079773,
      "learning_rate": 9.993601432517053e-05,
      "loss": 0.0876,
      "relation_loss": 0.01139867315068841,
      "selected_pairs": 26369.8,
      "state_query_norm": 556.746533203125,
      "state_relation_loss": 0.0016087661613710224,
      "step": 1960,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.01139867315068841,
      "weighted_relation_loss": 0.01139867315068841
    },
    {
      "action_query_norm": 317.1628082275391,
      "action_relation_loss": 0.00963465142995119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.842843779232928,
      "flow_matching_loss": 0.07389320507645607,
      "grad_norm": 0.30499914288520813,
      "learning_rate": 9.993319653522091e-05,
      "loss": 0.0851,
      "relation_loss": 0.011191727127879858,
      "selected_pairs": 26216.1,
      "state_query_norm": 552.0266845703125,
      "state_relation_loss": 0.001557075756136328,
      "step": 1970,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.011191727127879858,
      "weighted_relation_loss": 0.011191727127879858
    },
    {
      "action_query_norm": 325.1887268066406,
      "action_relation_loss": 0.009636826068162917,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8521983161833488,
      "flow_matching_loss": 0.07548631951212884,
      "grad_norm": 0.25221389532089233,
      "learning_rate": 9.993031807172911e-05,
      "loss": 0.0867,
      "relation_loss": 0.011231561377644538,
      "selected_pairs": 26673.4,
      "state_query_norm": 560.6740600585938,
      "state_relation_loss": 0.0015947354142554105,
      "step": 1980,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.011231561377644538,
      "weighted_relation_loss": 0.011231561377644538
    },
    {
      "action_query_norm": 321.71189880371094,
      "action_relation_loss": 0.009254517406225205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.86155285313377,
      "flow_matching_loss": 0.07608823254704475,
      "grad_norm": 0.19590510427951813,
      "learning_rate": 9.992737893819273e-05,
      "loss": 0.0869,
      "relation_loss": 0.010764640849083662,
      "selected_pairs": 26311.4,
      "state_query_norm": 564.8483093261718,
      "state_relation_loss": 0.0015101234195753932,
      "step": 1990,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.010764640849083662,
      "weighted_relation_loss": 0.010764640849083662
    },
    {
      "action_query_norm": 323.7912200927734,
      "action_relation_loss": 0.00946514131501317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8709073900841908,
      "flow_matching_loss": 0.07691389173269272,
      "grad_norm": 0.2030879706144333,
      "learning_rate": 9.992437913818312e-05,
      "loss": 0.088,
      "relation_loss": 0.01104846941307187,
      "selected_pairs": 26806.5,
      "state_query_norm": 554.1645080566407,
      "state_relation_loss": 0.0015833280398510397,
      "step": 2000,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.01104846941307187,
      "weighted_relation_loss": 0.01104846941307187
    },
    {
      "action_query_norm": 323.82774658203124,
      "action_relation_loss": 0.00917677436955273,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8802619270346117,
      "flow_matching_loss": 0.07474645227193832,
      "grad_norm": 0.2879800498485565,
      "learning_rate": 9.992131867534526e-05,
      "loss": 0.0854,
      "relation_loss": 0.010612810589373111,
      "selected_pairs": 27383.8,
      "state_query_norm": 559.1664794921875,
      "state_relation_loss": 0.0014360362431034446,
      "step": 2010,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.010612810589373111,
      "weighted_relation_loss": 0.010612810589373111
    },
    {
      "action_query_norm": 319.2225677490234,
      "action_relation_loss": 0.009767386689782142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8896164639850328,
      "flow_matching_loss": 0.07419850900769234,
      "grad_norm": 0.24198180437088013,
      "learning_rate": 9.991819755339796e-05,
      "loss": 0.0854,
      "relation_loss": 0.011183600407093763,
      "selected_pairs": 26844.6,
      "state_query_norm": 554.0669372558593,
      "state_relation_loss": 0.001416213728953153,
      "step": 2020,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.011183600407093763,
      "weighted_relation_loss": 0.011183600407093763
    },
    {
      "action_query_norm": 323.53090515136716,
      "action_relation_loss": 0.009663437120616436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8989710009354537,
      "flow_matching_loss": 0.07315900847315789,
      "grad_norm": 0.23553121089935303,
      "learning_rate": 9.991501577613365e-05,
      "loss": 0.0843,
      "relation_loss": 0.011112257838249207,
      "selected_pairs": 26619.7,
      "state_query_norm": 556.1265014648437,
      "state_relation_loss": 0.001448820810765028,
      "step": 2030,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.011112257838249207,
      "weighted_relation_loss": 0.011112257838249207
    },
    {
      "action_query_norm": 322.810693359375,
      "action_relation_loss": 0.009282930195331574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9083255378858746,
      "flow_matching_loss": 0.07349469438195229,
      "grad_norm": 0.22108136117458344,
      "learning_rate": 9.99117733474185e-05,
      "loss": 0.0843,
      "relation_loss": 0.010755771864205598,
      "selected_pairs": 26871.5,
      "state_query_norm": 556.0154479980469,
      "state_relation_loss": 0.001472841831855476,
      "step": 2040,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.010755771864205598,
      "weighted_relation_loss": 0.010755771864205598
    },
    {
      "action_query_norm": 329.3446960449219,
      "action_relation_loss": 0.009533802699297666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9176800748362957,
      "flow_matching_loss": 0.07488640770316124,
      "grad_norm": 0.2641567885875702,
      "learning_rate": 9.990847027119234e-05,
      "loss": 0.086,
      "relation_loss": 0.011065517272800206,
      "selected_pairs": 26341.8,
      "state_query_norm": 569.9385620117188,
      "state_relation_loss": 0.0015317145967856049,
      "step": 2050,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.011065517272800206,
      "weighted_relation_loss": 0.011065517272800206
    },
    {
      "action_query_norm": 336.4641082763672,
      "action_relation_loss": 0.009182014223188161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9270346117867165,
      "flow_matching_loss": 0.07124858424067497,
      "grad_norm": 0.21372216939926147,
      "learning_rate": 9.990510655146877e-05,
      "loss": 0.0819,
      "relation_loss": 0.010675872024148703,
      "selected_pairs": 26916.8,
      "state_query_norm": 587.1649291992187,
      "state_relation_loss": 0.0014938578009605409,
      "step": 2060,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.010675872024148703,
      "weighted_relation_loss": 0.010675872024148703
    },
    {
      "action_query_norm": 332.58291015625,
      "action_relation_loss": 0.008732755575329066,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9363891487371374,
      "flow_matching_loss": 0.07459966614842414,
      "grad_norm": 0.24324022233486176,
      "learning_rate": 9.990168219233496e-05,
      "loss": 0.0848,
      "relation_loss": 0.01023390470072627,
      "selected_pairs": 27174.8,
      "state_query_norm": 569.4153076171875,
      "state_relation_loss": 0.0015011491370387376,
      "step": 2070,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.01023390470072627,
      "weighted_relation_loss": 0.01023390470072627
    },
    {
      "action_query_norm": 325.02532653808595,
      "action_relation_loss": 0.0094491146504879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9457436856875585,
      "flow_matching_loss": 0.07317892014980316,
      "grad_norm": 0.2200683206319809,
      "learning_rate": 9.989819719795188e-05,
      "loss": 0.0841,
      "relation_loss": 0.010922251921147107,
      "selected_pairs": 27056.8,
      "state_query_norm": 577.358154296875,
      "state_relation_loss": 0.0014731371775269508,
      "step": 2080,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.010922251921147107,
      "weighted_relation_loss": 0.010922251921147107
    },
    {
      "action_query_norm": 322.29990234375,
      "action_relation_loss": 0.0090999742038548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9550982226379794,
      "flow_matching_loss": 0.07073545083403587,
      "grad_norm": 0.1405741572380066,
      "learning_rate": 9.989465157255412e-05,
      "loss": 0.0812,
      "relation_loss": 0.010465367790311575,
      "selected_pairs": 26812.1,
      "state_query_norm": 575.2248901367187,
      "state_relation_loss": 0.0013653936563059687,
      "step": 2090,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010465367790311575,
      "weighted_relation_loss": 0.010465367790311575
    },
    {
      "action_query_norm": 325.5595397949219,
      "action_relation_loss": 0.009181003831326962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9644527595884003,
      "flow_matching_loss": 0.07460486069321633,
      "grad_norm": 0.20943784713745117,
      "learning_rate": 9.989104532044994e-05,
      "loss": 0.0852,
      "relation_loss": 0.010598688200116158,
      "selected_pairs": 26047.8,
      "state_query_norm": 574.9546264648437,
      "state_relation_loss": 0.001417684426996857,
      "step": 2100,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.010598688200116158,
      "weighted_relation_loss": 0.010598688200116158
    },
    {
      "action_query_norm": 330.2423889160156,
      "action_relation_loss": 0.009087205119431018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9738072965388214,
      "flow_matching_loss": 0.07211251556873322,
      "grad_norm": 0.25949254631996155,
      "learning_rate": 9.988737844602128e-05,
      "loss": 0.0826,
      "relation_loss": 0.010504894703626633,
      "selected_pairs": 27162.7,
      "state_query_norm": 566.8181335449219,
      "state_relation_loss": 0.0014176895725540816,
      "step": 2110,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.010504894703626633,
      "weighted_relation_loss": 0.010504894703626633
    },
    {
      "action_query_norm": 331.7891845703125,
      "action_relation_loss": 0.008520531794056296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9831618334892422,
      "flow_matching_loss": 0.07143871001899242,
      "grad_norm": 0.22387227416038513,
      "learning_rate": 9.988365095372372e-05,
      "loss": 0.0814,
      "relation_loss": 0.009915887191891671,
      "selected_pairs": 27179.4,
      "state_query_norm": 562.5632019042969,
      "state_relation_loss": 0.0013953556073829532,
      "step": 2120,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.009915887191891671,
      "weighted_relation_loss": 0.009915887191891671
    },
    {
      "action_query_norm": 326.28285217285156,
      "action_relation_loss": 0.00918989609926939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9925163704396631,
      "flow_matching_loss": 0.07284893095493317,
      "grad_norm": 0.23562274873256683,
      "learning_rate": 9.987986284808654e-05,
      "loss": 0.0834,
      "relation_loss": 0.010582503397017717,
      "selected_pairs": 26916.4,
      "state_query_norm": 579.2980651855469,
      "state_relation_loss": 0.001392607286106795,
      "step": 2130,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.010582503397017717,
      "weighted_relation_loss": 0.010582503397017717
    },
    {
      "action_query_norm": 332.2514373779297,
      "action_relation_loss": 0.008572004735469818,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.001870907390084,
      "flow_matching_loss": 0.07366078346967697,
      "grad_norm": 0.2364606112241745,
      "learning_rate": 9.987601413371264e-05,
      "loss": 0.0836,
      "relation_loss": 0.00994697269052267,
      "selected_pairs": 24556.9,
      "state_query_norm": 569.4292236328125,
      "state_relation_loss": 0.001374968036543578,
      "step": 2140,
      "valid_action_samples": 215.2,
      "weighted_auxiliary_loss": 0.00994697269052267,
      "weighted_relation_loss": 0.00994697269052267
    },
    {
      "action_query_norm": 340.01105346679685,
      "action_relation_loss": 0.009166507190093398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.0112254443405053,
      "flow_matching_loss": 0.07748696208000183,
      "grad_norm": 0.22263631224632263,
      "learning_rate": 9.987210481527855e-05,
      "loss": 0.0881,
      "relation_loss": 0.010624999925494195,
      "selected_pairs": 26400.0,
      "state_query_norm": 573.4179443359375,
      "state_relation_loss": 0.0014584927470423282,
      "step": 2150,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.010624999925494195,
      "weighted_relation_loss": 0.010624999925494195
    },
    {
      "action_query_norm": 335.6146240234375,
      "action_relation_loss": 0.009134072437882423,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.020579981290926,
      "flow_matching_loss": 0.07345826551318169,
      "grad_norm": 0.48433127999305725,
      "learning_rate": 9.98681348975345e-05,
      "loss": 0.084,
      "relation_loss": 0.010573856439441443,
      "selected_pairs": 26648.1,
      "state_query_norm": 577.6876220703125,
      "state_relation_loss": 0.0014397840946912765,
      "step": 2160,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.010573856439441443,
      "weighted_relation_loss": 0.010573856439441443
    },
    {
      "action_query_norm": 323.8755767822266,
      "action_relation_loss": 0.009117335267364978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.029934518241347,
      "flow_matching_loss": 0.07622007951140404,
      "grad_norm": 0.2998368740081787,
      "learning_rate": 9.986410438530427e-05,
      "loss": 0.0868,
      "relation_loss": 0.01058588856831193,
      "selected_pairs": 27128.4,
      "state_query_norm": 565.7700805664062,
      "state_relation_loss": 0.001468553359154612,
      "step": 2170,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.01058588856831193,
      "weighted_relation_loss": 0.01058588856831193
    },
    {
      "action_query_norm": 333.0249908447266,
      "action_relation_loss": 0.009196165110915899,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.039289055191768,
      "flow_matching_loss": 0.07114869728684425,
      "grad_norm": 0.1882820427417755,
      "learning_rate": 9.986001328348534e-05,
      "loss": 0.0818,
      "relation_loss": 0.010654117073863744,
      "selected_pairs": 27291.6,
      "state_query_norm": 584.2804382324218,
      "state_relation_loss": 0.0014579519163817167,
      "step": 2180,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.010654117073863744,
      "weighted_relation_loss": 0.010654117073863744
    },
    {
      "action_query_norm": 326.5463012695312,
      "action_relation_loss": 0.0090038544498384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.048643592142189,
      "flow_matching_loss": 0.07083504498004914,
      "grad_norm": 0.18174166977405548,
      "learning_rate": 9.985586159704878e-05,
      "loss": 0.0812,
      "relation_loss": 0.010379403177648783,
      "selected_pairs": 26761.1,
      "state_query_norm": 584.0802673339844,
      "state_relation_loss": 0.0013755487860180437,
      "step": 2190,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.010379403177648783,
      "weighted_relation_loss": 0.010379403177648783
    },
    {
      "action_query_norm": 329.83460998535156,
      "action_relation_loss": 0.008841428533196449,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.05799812909261,
      "flow_matching_loss": 0.07370148971676826,
      "grad_norm": 0.18971653282642365,
      "learning_rate": 9.985164933103929e-05,
      "loss": 0.0839,
      "relation_loss": 0.010203849617391824,
      "selected_pairs": 26971.8,
      "state_query_norm": 586.9882141113281,
      "state_relation_loss": 0.0013624210376292467,
      "step": 2200,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.010203849617391824,
      "weighted_relation_loss": 0.010203849617391824
    },
    {
      "action_query_norm": 334.9351348876953,
      "action_relation_loss": 0.009011577442288398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.067352666043031,
      "flow_matching_loss": 0.07366714626550674,
      "grad_norm": 0.16014939546585083,
      "learning_rate": 9.984737649057513e-05,
      "loss": 0.0841,
      "relation_loss": 0.010414230730384588,
      "selected_pairs": 26928.6,
      "state_query_norm": 584.5063232421875,
      "state_relation_loss": 0.0014026532531715929,
      "step": 2210,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.010414230730384588,
      "weighted_relation_loss": 0.010414230730384588
    },
    {
      "action_query_norm": 335.7343719482422,
      "action_relation_loss": 0.009675111901015044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.0767072029934517,
      "flow_matching_loss": 0.07504901960492134,
      "grad_norm": 0.23817282915115356,
      "learning_rate": 9.984304308084827e-05,
      "loss": 0.0861,
      "relation_loss": 0.011056935694068671,
      "selected_pairs": 26822.9,
      "state_query_norm": 584.7154418945313,
      "state_relation_loss": 0.0013818238279782236,
      "step": 2220,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.011056935694068671,
      "weighted_relation_loss": 0.011056935694068671
    },
    {
      "action_query_norm": 334.1658081054687,
      "action_relation_loss": 0.008807945484295487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.086061739943873,
      "flow_matching_loss": 0.07281148657202721,
      "grad_norm": 0.20512127876281738,
      "learning_rate": 9.983864910712416e-05,
      "loss": 0.083,
      "relation_loss": 0.01022504298016429,
      "selected_pairs": 27156.8,
      "state_query_norm": 589.5847106933594,
      "state_relation_loss": 0.0014170974958688022,
      "step": 2230,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.01022504298016429,
      "weighted_relation_loss": 0.01022504298016429
    },
    {
      "action_query_norm": 335.6060333251953,
      "action_relation_loss": 0.008694261219352484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.095416276894294,
      "flow_matching_loss": 0.07223614156246186,
      "grad_norm": 0.20532603561878204,
      "learning_rate": 9.98341945747419e-05,
      "loss": 0.0823,
      "relation_loss": 0.010083797946572303,
      "selected_pairs": 27174.6,
      "state_query_norm": 577.965087890625,
      "state_relation_loss": 0.0013895367737859488,
      "step": 2240,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010083797946572303,
      "weighted_relation_loss": 0.010083797946572303
    },
    {
      "action_query_norm": 340.62095642089844,
      "action_relation_loss": 0.008488766802474856,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1047708138447145,
      "flow_matching_loss": 0.0728467047214508,
      "grad_norm": 0.24193590879440308,
      "learning_rate": 9.98296794891142e-05,
      "loss": 0.0827,
      "relation_loss": 0.009846445452421904,
      "selected_pairs": 27131.5,
      "state_query_norm": 587.1596862792969,
      "state_relation_loss": 0.0013576787547208368,
      "step": 2250,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.009846445452421904,
      "weighted_relation_loss": 0.009846445452421904
    },
    {
      "action_query_norm": 334.7041015625,
      "action_relation_loss": 0.009008313901722432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1141253507951356,
      "flow_matching_loss": 0.07480405271053314,
      "grad_norm": 0.2396889626979828,
      "learning_rate": 9.982510385572725e-05,
      "loss": 0.0853,
      "relation_loss": 0.01052397359162569,
      "selected_pairs": 26824.4,
      "state_query_norm": 586.2537414550782,
      "state_relation_loss": 0.0015156597713939846,
      "step": 2260,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.01052397359162569,
      "weighted_relation_loss": 0.01052397359162569
    },
    {
      "action_query_norm": 345.488037109375,
      "action_relation_loss": 0.008641348453238606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1234798877455567,
      "flow_matching_loss": 0.0728083796799183,
      "grad_norm": 0.1865755170583725,
      "learning_rate": 9.982046768014094e-05,
      "loss": 0.0829,
      "relation_loss": 0.01012468421831727,
      "selected_pairs": 26792.6,
      "state_query_norm": 594.0317016601563,
      "state_relation_loss": 0.0014833358582109213,
      "step": 2270,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.01012468421831727,
      "weighted_relation_loss": 0.01012468421831727
    },
    {
      "action_query_norm": 340.9667236328125,
      "action_relation_loss": 0.008570838999003173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1328344246959774,
      "flow_matching_loss": 0.06894328556954861,
      "grad_norm": 0.20377217233181,
      "learning_rate": 9.981577096798863e-05,
      "loss": 0.079,
      "relation_loss": 0.010020730644464492,
      "selected_pairs": 26942.0,
      "state_query_norm": 602.3439575195313,
      "state_relation_loss": 0.0014498916803859173,
      "step": 2280,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.010020730644464492,
      "weighted_relation_loss": 0.010020730644464492
    },
    {
      "action_query_norm": 342.3474914550781,
      "action_relation_loss": 0.008581650536507368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1421889616463985,
      "flow_matching_loss": 0.06883103996515275,
      "grad_norm": 0.2104639858007431,
      "learning_rate": 9.981101372497727e-05,
      "loss": 0.0789,
      "relation_loss": 0.010029701422899962,
      "selected_pairs": 26789.0,
      "state_query_norm": 595.5771240234375,
      "state_relation_loss": 0.0014480507816188037,
      "step": 2290,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.010029701422899962,
      "weighted_relation_loss": 0.010029701422899962
    },
    {
      "action_query_norm": 332.89032592773435,
      "action_relation_loss": 0.008652942348271608,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1515434985968196,
      "flow_matching_loss": 0.06928434744477271,
      "grad_norm": 0.2209376096725464,
      "learning_rate": 9.980619595688737e-05,
      "loss": 0.0794,
      "relation_loss": 0.010067323222756386,
      "selected_pairs": 26996.1,
      "state_query_norm": 597.6714599609375,
      "state_relation_loss": 0.001414380851201713,
      "step": 2300,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.010067323222756386,
      "weighted_relation_loss": 0.010067323222756386
    },
    {
      "action_query_norm": 337.7707458496094,
      "action_relation_loss": 0.00863246787339449,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1608980355472402,
      "flow_matching_loss": 0.07244512513279915,
      "grad_norm": 0.27100327610969543,
      "learning_rate": 9.980131766957295e-05,
      "loss": 0.0824,
      "relation_loss": 0.009970434848219157,
      "selected_pairs": 26411.7,
      "state_query_norm": 595.9198486328125,
      "state_relation_loss": 0.0013379670446738601,
      "step": 2310,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.009970434848219157,
      "weighted_relation_loss": 0.009970434848219157
    },
    {
      "action_query_norm": 339.8186340332031,
      "action_relation_loss": 0.008706530928611756,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1702525724976613,
      "flow_matching_loss": 0.07333957254886628,
      "grad_norm": 0.20782005786895752,
      "learning_rate": 9.979637886896163e-05,
      "loss": 0.0834,
      "relation_loss": 0.010065642930567265,
      "selected_pairs": 26457.8,
      "state_query_norm": 607.1313659667969,
      "state_relation_loss": 0.0013591119437478482,
      "step": 2320,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.010065642930567265,
      "weighted_relation_loss": 0.010065642930567265
    },
    {
      "action_query_norm": 338.82781982421875,
      "action_relation_loss": 0.00898115085437894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1796071094480824,
      "flow_matching_loss": 0.0711911354213953,
      "grad_norm": 0.19874659180641174,
      "learning_rate": 9.979137956105447e-05,
      "loss": 0.0815,
      "relation_loss": 0.010347925219684839,
      "selected_pairs": 26272.8,
      "state_query_norm": 588.8064697265625,
      "state_relation_loss": 0.0013667743187397718,
      "step": 2330,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.010347925219684839,
      "weighted_relation_loss": 0.010347925219684839
    },
    {
      "action_query_norm": 343.3072021484375,
      "action_relation_loss": 0.009028808772563934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.188961646398503,
      "flow_matching_loss": 0.074659825861454,
      "grad_norm": 0.14821317791938782,
      "learning_rate": 9.978631975192613e-05,
      "loss": 0.0851,
      "relation_loss": 0.010400382336229086,
      "selected_pairs": 27174.2,
      "state_query_norm": 608.7400939941406,
      "state_relation_loss": 0.001371573528740555,
      "step": 2340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010400382336229086,
      "weighted_relation_loss": 0.010400382336229086
    },
    {
      "action_query_norm": 341.9522186279297,
      "action_relation_loss": 0.009007824864238501,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.198316183348924,
      "flow_matching_loss": 0.06854491271078586,
      "grad_norm": 0.21034689247608185,
      "learning_rate": 9.978119944772475e-05,
      "loss": 0.0789,
      "relation_loss": 0.0103441814891994,
      "selected_pairs": 27102.6,
      "state_query_norm": 601.7193908691406,
      "state_relation_loss": 0.0013363566366024315,
      "step": 2350,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0103441814891994,
      "weighted_relation_loss": 0.0103441814891994
    },
    {
      "action_query_norm": 335.55926513671875,
      "action_relation_loss": 0.009095208253711462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2076707202993453,
      "flow_matching_loss": 0.0735008880496025,
      "grad_norm": 0.21092960238456726,
      "learning_rate": 9.977601865467197e-05,
      "loss": 0.0839,
      "relation_loss": 0.010447981115430594,
      "selected_pairs": 26735.8,
      "state_query_norm": 600.1146728515625,
      "state_relation_loss": 0.0013527728617191315,
      "step": 2360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.010447981115430594,
      "weighted_relation_loss": 0.010447981115430594
    },
    {
      "action_query_norm": 336.6060028076172,
      "action_relation_loss": 0.008746938314288855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.217025257249766,
      "flow_matching_loss": 0.07164075151085854,
      "grad_norm": 0.16852466762065887,
      "learning_rate": 9.977077737906297e-05,
      "loss": 0.0817,
      "relation_loss": 0.010067252349108458,
      "selected_pairs": 27174.6,
      "state_query_norm": 597.0492553710938,
      "state_relation_loss": 0.0013203139416873455,
      "step": 2370,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010067252349108458,
      "weighted_relation_loss": 0.010067252349108458
    },
    {
      "action_query_norm": 339.4058135986328,
      "action_relation_loss": 0.008319697575643659,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.226379794200187,
      "flow_matching_loss": 0.07264334708452225,
      "grad_norm": 0.2112162858247757,
      "learning_rate": 9.976547562726636e-05,
      "loss": 0.0822,
      "relation_loss": 0.009584833215922118,
      "selected_pairs": 26759.0,
      "state_query_norm": 601.9111877441406,
      "state_relation_loss": 0.0012651356169953941,
      "step": 2380,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.009584833215922118,
      "weighted_relation_loss": 0.009584833215922118
    },
    {
      "action_query_norm": 340.1018341064453,
      "action_relation_loss": 0.008783800713717937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.235734331150608,
      "flow_matching_loss": 0.0683930203318596,
      "grad_norm": 0.21452516317367554,
      "learning_rate": 9.976011340572429e-05,
      "loss": 0.0785,
      "relation_loss": 0.01005885163322091,
      "selected_pairs": 27153.0,
      "state_query_norm": 594.245166015625,
      "state_relation_loss": 0.0012750509195029736,
      "step": 2390,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.01005885163322091,
      "weighted_relation_loss": 0.01005885163322091
    },
    {
      "action_query_norm": 343.95375061035156,
      "action_relation_loss": 0.007876318600028753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.245088868101029,
      "flow_matching_loss": 0.07144081741571426,
      "grad_norm": 0.21591748297214508,
      "learning_rate": 9.975469072095237e-05,
      "loss": 0.0806,
      "relation_loss": 0.009124743100255728,
      "selected_pairs": 26781.3,
      "state_query_norm": 605.8194519042969,
      "state_relation_loss": 0.0012484242906793952,
      "step": 2400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.009124743100255728,
      "weighted_relation_loss": 0.009124743100255728
    },
    {
      "action_query_norm": 347.0225463867188,
      "action_relation_loss": 0.008290671138092875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.25444340505145,
      "flow_matching_loss": 0.06930705755949021,
      "grad_norm": 0.3113096058368683,
      "learning_rate": 9.974920757953965e-05,
      "loss": 0.0788,
      "relation_loss": 0.009520746069028974,
      "selected_pairs": 26433.6,
      "state_query_norm": 594.8046142578125,
      "state_relation_loss": 0.0012300748843699695,
      "step": 2410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.009520746069028974,
      "weighted_relation_loss": 0.009520746069028974
    },
    {
      "action_query_norm": 339.6500244140625,
      "action_relation_loss": 0.008308030478656293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.263797942001871,
      "flow_matching_loss": 0.07066471949219703,
      "grad_norm": 0.2540135979652405,
      "learning_rate": 9.97436639881487e-05,
      "loss": 0.0802,
      "relation_loss": 0.009550410602241755,
      "selected_pairs": 27059.9,
      "state_query_norm": 600.9197570800782,
      "state_relation_loss": 0.001242380088660866,
      "step": 2420,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.009550410602241755,
      "weighted_relation_loss": 0.009550410602241755
    },
    {
      "action_query_norm": 329.9006408691406,
      "action_relation_loss": 0.008587311813607812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2731524789522917,
      "flow_matching_loss": 0.0708794467151165,
      "grad_norm": 0.2575356364250183,
      "learning_rate": 9.973805995351545e-05,
      "loss": 0.0808,
      "relation_loss": 0.009889953583478928,
      "selected_pairs": 26532.1,
      "state_query_norm": 603.8726806640625,
      "state_relation_loss": 0.0013026417233049869,
      "step": 2430,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.009889953583478928,
      "weighted_relation_loss": 0.009889953583478928
    },
    {
      "action_query_norm": 333.2282470703125,
      "action_relation_loss": 0.008606671635061503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2825070159027128,
      "flow_matching_loss": 0.07209787592291832,
      "grad_norm": 0.2142474353313446,
      "learning_rate": 9.973239548244939e-05,
      "loss": 0.082,
      "relation_loss": 0.009897368401288987,
      "selected_pairs": 26501.9,
      "state_query_norm": 602.6123413085937,
      "state_relation_loss": 0.0012906968710012733,
      "step": 2440,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.009897368401288987,
      "weighted_relation_loss": 0.009897368401288987
    },
    {
      "action_query_norm": 337.9939849853516,
      "action_relation_loss": 0.008615677896887063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.291861552853134,
      "flow_matching_loss": 0.07223771959543228,
      "grad_norm": 0.26907315850257874,
      "learning_rate": 9.972667058183333e-05,
      "loss": 0.082,
      "relation_loss": 0.00979532515630126,
      "selected_pairs": 26342.6,
      "state_query_norm": 616.061572265625,
      "state_relation_loss": 0.0011796471895650028,
      "step": 2450,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.00979532515630126,
      "weighted_relation_loss": 0.00979532515630126
    },
    {
      "action_query_norm": 345.37193603515624,
      "action_relation_loss": 0.008583726920187473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.301216089803555,
      "flow_matching_loss": 0.06962973587214946,
      "grad_norm": 0.23551060259342194,
      "learning_rate": 9.972088525862362e-05,
      "loss": 0.0795,
      "relation_loss": 0.009830453433096409,
      "selected_pairs": 26660.3,
      "state_query_norm": 615.0199035644531,
      "state_relation_loss": 0.0012467264197766781,
      "step": 2460,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.009830453433096409,
      "weighted_relation_loss": 0.009830453433096409
    },
    {
      "action_query_norm": 339.8311309814453,
      "action_relation_loss": 0.00883093667216599,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3105706267539756,
      "flow_matching_loss": 0.07192888520658017,
      "grad_norm": 0.33799538016319275,
      "learning_rate": 9.971503951984995e-05,
      "loss": 0.082,
      "relation_loss": 0.010063547268509864,
      "selected_pairs": 26997.4,
      "state_query_norm": 611.2715209960937,
      "state_relation_loss": 0.0012326105497777462,
      "step": 2470,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.010063547268509864,
      "weighted_relation_loss": 0.010063547268509864
    },
    {
      "action_query_norm": 342.1459197998047,
      "action_relation_loss": 0.008866642462089657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3199251637043967,
      "flow_matching_loss": 0.07146613895893097,
      "grad_norm": 0.2937348484992981,
      "learning_rate": 9.970913337261543e-05,
      "loss": 0.0816,
      "relation_loss": 0.010158620495349168,
      "selected_pairs": 26327.1,
      "state_query_norm": 617.7766906738282,
      "state_relation_loss": 0.0012919781263917684,
      "step": 2480,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.010158620495349168,
      "weighted_relation_loss": 0.010158620495349168
    },
    {
      "action_query_norm": 342.57410888671876,
      "action_relation_loss": 0.008949638903141021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3292797006548174,
      "flow_matching_loss": 0.06974714398384094,
      "grad_norm": 0.2574823200702667,
      "learning_rate": 9.97031668240966e-05,
      "loss": 0.08,
      "relation_loss": 0.010239785630255937,
      "selected_pairs": 26544.0,
      "state_query_norm": 608.1541625976563,
      "state_relation_loss": 0.0012901466921903192,
      "step": 2490,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.010239785630255937,
      "weighted_relation_loss": 0.010239785630255937
    },
    {
      "action_query_norm": 343.7083404541016,
      "action_relation_loss": 0.008664720365777612,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3386342376052385,
      "flow_matching_loss": 0.06824349090456963,
      "grad_norm": 0.2136549949645996,
      "learning_rate": 9.969713988154339e-05,
      "loss": 0.0782,
      "relation_loss": 0.009948336984962225,
      "selected_pairs": 26850.5,
      "state_query_norm": 618.1813293457031,
      "state_relation_loss": 0.0012836167472414673,
      "step": 2500,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009948336984962225,
      "weighted_relation_loss": 0.009948336984962225
    },
    {
      "action_query_norm": 349.68564453125,
      "action_relation_loss": 0.008840922033414245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3479887745556596,
      "flow_matching_loss": 0.06924765855073929,
      "grad_norm": 0.2498205602169037,
      "learning_rate": 9.969105255227906e-05,
      "loss": 0.0794,
      "relation_loss": 0.010134530812501907,
      "selected_pairs": 27292.8,
      "state_query_norm": 625.7083862304687,
      "state_relation_loss": 0.00129360860446468,
      "step": 2510,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.010134530812501907,
      "weighted_relation_loss": 0.010134530812501907
    },
    {
      "action_query_norm": 349.10691833496094,
      "action_relation_loss": 0.008654259936884045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3573433115060807,
      "flow_matching_loss": 0.06786833368241788,
      "grad_norm": 0.3132263422012329,
      "learning_rate": 9.968490484370035e-05,
      "loss": 0.0779,
      "relation_loss": 0.009990497026592493,
      "selected_pairs": 26249.9,
      "state_query_norm": 615.186279296875,
      "state_relation_loss": 0.001336237194482237,
      "step": 2520,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.009990497026592493,
      "weighted_relation_loss": 0.009990497026592493
    },
    {
      "action_query_norm": 344.45789489746096,
      "action_relation_loss": 0.009076655376702547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3666978484565013,
      "flow_matching_loss": 0.07272676452994346,
      "grad_norm": 0.30793297290802,
      "learning_rate": 9.967869676327726e-05,
      "loss": 0.0831,
      "relation_loss": 0.010346346534788609,
      "selected_pairs": 27034.6,
      "state_query_norm": 609.2814514160157,
      "state_relation_loss": 0.0012696910882368684,
      "step": 2530,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.010346346534788609,
      "weighted_relation_loss": 0.010346346534788609
    },
    {
      "action_query_norm": 349.0975769042969,
      "action_relation_loss": 0.00883959881030023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3760523854069224,
      "flow_matching_loss": 0.06814943104982377,
      "grad_norm": 0.2705758512020111,
      "learning_rate": 9.96724283185532e-05,
      "loss": 0.0782,
      "relation_loss": 0.010052875895053149,
      "selected_pairs": 27012.2,
      "state_query_norm": 610.7695495605469,
      "state_relation_loss": 0.0012132771778851747,
      "step": 2540,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.010052875895053149,
      "weighted_relation_loss": 0.010052875895053149
    },
    {
      "action_query_norm": 347.1614044189453,
      "action_relation_loss": 0.008605641080066562,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.385406922357343,
      "flow_matching_loss": 0.07171859219670296,
      "grad_norm": 0.2514707148075104,
      "learning_rate": 9.966609951714494e-05,
      "loss": 0.0816,
      "relation_loss": 0.00992283308878541,
      "selected_pairs": 27196.4,
      "state_query_norm": 620.8810668945313,
      "state_relation_loss": 0.0013171920320019126,
      "step": 2550,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00992283308878541,
      "weighted_relation_loss": 0.00992283308878541
    },
    {
      "action_query_norm": 341.9038330078125,
      "action_relation_loss": 0.008787093078717589,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.394761459307764,
      "flow_matching_loss": 0.0674242302775383,
      "grad_norm": 0.15056993067264557,
      "learning_rate": 9.965971036674255e-05,
      "loss": 0.0775,
      "relation_loss": 0.010084462072700261,
      "selected_pairs": 27084.3,
      "state_query_norm": 619.1083435058594,
      "state_relation_loss": 0.0012973689357750119,
      "step": 2560,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.010084462072700261,
      "weighted_relation_loss": 0.010084462072700261
    },
    {
      "action_query_norm": 350.02003479003906,
      "action_relation_loss": 0.008747863490134478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4041159962581853,
      "flow_matching_loss": 0.06862119659781456,
      "grad_norm": 0.1828744113445282,
      "learning_rate": 9.965326087510947e-05,
      "loss": 0.0786,
      "relation_loss": 0.010020100325345994,
      "selected_pairs": 26484.4,
      "state_query_norm": 623.4389038085938,
      "state_relation_loss": 0.0012722369632683695,
      "step": 2570,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.010020100325345994,
      "weighted_relation_loss": 0.010020100325345994
    },
    {
      "action_query_norm": 351.2462585449219,
      "action_relation_loss": 0.008326546056196094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4134705332086064,
      "flow_matching_loss": 0.06909056901931762,
      "grad_norm": 0.20167666673660278,
      "learning_rate": 9.964675105008244e-05,
      "loss": 0.0787,
      "relation_loss": 0.009595135226845741,
      "selected_pairs": 27292.2,
      "state_query_norm": 617.4175598144532,
      "state_relation_loss": 0.0012685892055742443,
      "step": 2580,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.009595135226845741,
      "weighted_relation_loss": 0.009595135226845741
    },
    {
      "action_query_norm": 344.2136657714844,
      "action_relation_loss": 0.008456465229392052,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.422825070159027,
      "flow_matching_loss": 0.06937461160123348,
      "grad_norm": 0.2102804034948349,
      "learning_rate": 9.964018089957147e-05,
      "loss": 0.0791,
      "relation_loss": 0.009720206260681152,
      "selected_pairs": 26940.8,
      "state_query_norm": 615.5442993164063,
      "state_relation_loss": 0.001263740984722972,
      "step": 2590,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.009720206260681152,
      "weighted_relation_loss": 0.009720206260681152
    },
    {
      "action_query_norm": 353.4313568115234,
      "action_relation_loss": 0.008098434237763286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.432179607109448,
      "flow_matching_loss": 0.06912503987550736,
      "grad_norm": 0.27858492732048035,
      "learning_rate": 9.963355043155996e-05,
      "loss": 0.0785,
      "relation_loss": 0.009354585595428944,
      "selected_pairs": 27510.0,
      "state_query_norm": 613.0406799316406,
      "state_relation_loss": 0.0012561513460241259,
      "step": 2600,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.009354585595428944,
      "weighted_relation_loss": 0.009354585595428944
    },
    {
      "action_query_norm": 349.1479217529297,
      "action_relation_loss": 0.008367313677445054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.441534144059869,
      "flow_matching_loss": 0.06763085350394249,
      "grad_norm": 0.24594822525978088,
      "learning_rate": 9.962685965410455e-05,
      "loss": 0.0773,
      "relation_loss": 0.00968241849914193,
      "selected_pairs": 27034.6,
      "state_query_norm": 620.0779113769531,
      "state_relation_loss": 0.0013151048799045384,
      "step": 2610,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00968241849914193,
      "weighted_relation_loss": 0.00968241849914193
    },
    {
      "action_query_norm": 347.73292541503906,
      "action_relation_loss": 0.008531262446194888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.45088868101029,
      "flow_matching_loss": 0.06969553828239441,
      "grad_norm": 0.23033173382282257,
      "learning_rate": 9.962010857533514e-05,
      "loss": 0.0795,
      "relation_loss": 0.00978461978957057,
      "selected_pairs": 27198.1,
      "state_query_norm": 617.4984924316407,
      "state_relation_loss": 0.0012533573084510864,
      "step": 2620,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00978461978957057,
      "weighted_relation_loss": 0.00978461978957057
    },
    {
      "action_query_norm": 339.03828125,
      "action_relation_loss": 0.008756248652935028,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.460243217960711,
      "flow_matching_loss": 0.07220882549881935,
      "grad_norm": 0.1897290199995041,
      "learning_rate": 9.961329720345493e-05,
      "loss": 0.0822,
      "relation_loss": 0.010010148584842681,
      "selected_pairs": 26640.0,
      "state_query_norm": 613.995849609375,
      "state_relation_loss": 0.001253899873699993,
      "step": 2630,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.010010148584842681,
      "weighted_relation_loss": 0.010010148584842681
    },
    {
      "action_query_norm": 344.3998046875,
      "action_relation_loss": 0.00833776816725731,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.469597754911132,
      "flow_matching_loss": 0.07021749615669251,
      "grad_norm": 0.1938534528017044,
      "learning_rate": 9.96064255467404e-05,
      "loss": 0.0798,
      "relation_loss": 0.009559532906860114,
      "selected_pairs": 27201.6,
      "state_query_norm": 618.89560546875,
      "state_relation_loss": 0.0012217648210935295,
      "step": 2640,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.009559532906860114,
      "weighted_relation_loss": 0.009559532906860114
    },
    {
      "action_query_norm": 347.2222961425781,
      "action_relation_loss": 0.008183017233386635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4789522918615527,
      "flow_matching_loss": 0.06763231679797173,
      "grad_norm": 0.18630658090114594,
      "learning_rate": 9.959949361354126e-05,
      "loss": 0.0771,
      "relation_loss": 0.009434562642127275,
      "selected_pairs": 27217.7,
      "state_query_norm": 620.4400024414062,
      "state_relation_loss": 0.0012515452923253178,
      "step": 2650,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.009434562642127275,
      "weighted_relation_loss": 0.009434562642127275
    },
    {
      "action_query_norm": 362.11134948730466,
      "action_relation_loss": 0.008660709438845515,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.488306828811974,
      "flow_matching_loss": 0.069956673681736,
      "grad_norm": 0.21469679474830627,
      "learning_rate": 9.959250141228045e-05,
      "loss": 0.0799,
      "relation_loss": 0.009939785953611136,
      "selected_pairs": 26270.9,
      "state_query_norm": 632.0920288085938,
      "state_relation_loss": 0.001279076363425702,
      "step": 2660,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.009939785953611136,
      "weighted_relation_loss": 0.009939785953611136
    },
    {
      "action_query_norm": 360.1907592773438,
      "action_relation_loss": 0.00860459585674107,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.497661365762395,
      "flow_matching_loss": 0.06940729320049285,
      "grad_norm": 0.20034879446029663,
      "learning_rate": 9.958544895145414e-05,
      "loss": 0.0793,
      "relation_loss": 0.009879876766353846,
      "selected_pairs": 26920.5,
      "state_query_norm": 625.6768127441406,
      "state_relation_loss": 0.0012752809445373714,
      "step": 2670,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.009879876766353846,
      "weighted_relation_loss": 0.009879876766353846
    },
    {
      "action_query_norm": 354.33485107421876,
      "action_relation_loss": 0.008475269144400954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5070159027128156,
      "flow_matching_loss": 0.07103624418377877,
      "grad_norm": 0.16459505259990692,
      "learning_rate": 9.957833623963177e-05,
      "loss": 0.0808,
      "relation_loss": 0.009732349403202533,
      "selected_pairs": 27243.8,
      "state_query_norm": 633.3385192871094,
      "state_relation_loss": 0.0012570802588015794,
      "step": 2680,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.009732349403202533,
      "weighted_relation_loss": 0.009732349403202533
    },
    {
      "action_query_norm": 360.33935546875,
      "action_relation_loss": 0.008400630857795476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5163704396632367,
      "flow_matching_loss": 0.06897400841116905,
      "grad_norm": 0.20679886639118195,
      "learning_rate": 9.957116328545593e-05,
      "loss": 0.0786,
      "relation_loss": 0.009625749289989471,
      "selected_pairs": 27427.9,
      "state_query_norm": 632.0200988769532,
      "state_relation_loss": 0.0012251185369677841,
      "step": 2690,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.009625749289989471,
      "weighted_relation_loss": 0.009625749289989471
    },
    {
      "action_query_norm": 358.67979736328124,
      "action_relation_loss": 0.008247475093230604,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.525724976613658,
      "flow_matching_loss": 0.06749854497611522,
      "grad_norm": 0.19516398012638092,
      "learning_rate": 9.956393009764244e-05,
      "loss": 0.077,
      "relation_loss": 0.009457746334373952,
      "selected_pairs": 26687.0,
      "state_query_norm": 633.4752075195313,
      "state_relation_loss": 0.0012102711130864917,
      "step": 2700,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.009457746334373952,
      "weighted_relation_loss": 0.009457746334373952
    },
    {
      "action_query_norm": 363.2331909179687,
      "action_relation_loss": 0.00834380192682147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5350795135640785,
      "flow_matching_loss": 0.06905141845345497,
      "grad_norm": 0.20016014575958252,
      "learning_rate": 9.955663668498032e-05,
      "loss": 0.0786,
      "relation_loss": 0.009558094013482333,
      "selected_pairs": 27298.0,
      "state_query_norm": 641.2092102050781,
      "state_relation_loss": 0.0012142922496423125,
      "step": 2710,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.009558094013482333,
      "weighted_relation_loss": 0.009558094013482333
    },
    {
      "action_query_norm": 357.59478759765625,
      "action_relation_loss": 0.008403062680736184,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5444340505144996,
      "flow_matching_loss": 0.07007838264107705,
      "grad_norm": 0.17560994625091553,
      "learning_rate": 9.954928305633173e-05,
      "loss": 0.0797,
      "relation_loss": 0.00962993698194623,
      "selected_pairs": 26898.4,
      "state_query_norm": 634.7375854492187,
      "state_relation_loss": 0.00122687432449311,
      "step": 2720,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00962993698194623,
      "weighted_relation_loss": 0.00962993698194623
    },
    {
      "action_query_norm": 352.9806823730469,
      "action_relation_loss": 0.007998283440247177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.55378858746492,
      "flow_matching_loss": 0.06668653637170792,
      "grad_norm": 0.19343522191047668,
      "learning_rate": 9.954186922063204e-05,
      "loss": 0.0759,
      "relation_loss": 0.00926145063713193,
      "selected_pairs": 26890.7,
      "state_query_norm": 630.1834899902344,
      "state_relation_loss": 0.0012631671852432191,
      "step": 2730,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00926145063713193,
      "weighted_relation_loss": 0.00926145063713193
    },
    {
      "action_query_norm": 359.7926727294922,
      "action_relation_loss": 0.007932637864723802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5631431244153413,
      "flow_matching_loss": 0.06754603683948517,
      "grad_norm": 0.23756450414657593,
      "learning_rate": 9.953439518688974e-05,
      "loss": 0.0767,
      "relation_loss": 0.009177821967750788,
      "selected_pairs": 26826.6,
      "state_query_norm": 644.6063720703125,
      "state_relation_loss": 0.0012451839982531964,
      "step": 2740,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.009177821967750788,
      "weighted_relation_loss": 0.009177821967750788
    },
    {
      "action_query_norm": 365.7309997558594,
      "action_relation_loss": 0.008161461027339102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5724976613657624,
      "flow_matching_loss": 0.06807134449481964,
      "grad_norm": 0.18374815583229065,
      "learning_rate": 9.952686096418652e-05,
      "loss": 0.0774,
      "relation_loss": 0.009304484352469444,
      "selected_pairs": 26439.8,
      "state_query_norm": 643.0809204101563,
      "state_relation_loss": 0.0011430232436396182,
      "step": 2750,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.009304484352469444,
      "weighted_relation_loss": 0.009304484352469444
    },
    {
      "action_query_norm": 366.4694030761719,
      "action_relation_loss": 0.008085809368640184,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5818521983161835,
      "flow_matching_loss": 0.06922352835536003,
      "grad_norm": 0.17012427747249603,
      "learning_rate": 9.951926656167714e-05,
      "loss": 0.0785,
      "relation_loss": 0.009307865332812072,
      "selected_pairs": 26209.9,
      "state_query_norm": 639.2936828613281,
      "state_relation_loss": 0.0012220560456626117,
      "step": 2760,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.009307865332812072,
      "weighted_relation_loss": 0.009307865332812072
    },
    {
      "action_query_norm": 364.3841888427734,
      "action_relation_loss": 0.00815849918872118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.591206735266604,
      "flow_matching_loss": 0.06755220219492912,
      "grad_norm": 0.1984652727842331,
      "learning_rate": 9.951161198858953e-05,
      "loss": 0.0769,
      "relation_loss": 0.009342560730874538,
      "selected_pairs": 27171.9,
      "state_query_norm": 643.0988708496094,
      "state_relation_loss": 0.0011840615305118262,
      "step": 2770,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.009342560730874538,
      "weighted_relation_loss": 0.009342560730874538
    },
    {
      "action_query_norm": 360.08973388671876,
      "action_relation_loss": 0.00804750295355916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6005612722170253,
      "flow_matching_loss": 0.06729800291359425,
      "grad_norm": 0.15562666952610016,
      "learning_rate": 9.950389725422471e-05,
      "loss": 0.0765,
      "relation_loss": 0.009245685022324323,
      "selected_pairs": 26412.3,
      "state_query_norm": 633.9723449707031,
      "state_relation_loss": 0.0011981820920482279,
      "step": 2780,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.009245685022324323,
      "weighted_relation_loss": 0.009245685022324323
    },
    {
      "action_query_norm": 356.4265441894531,
      "action_relation_loss": 0.007707556802779436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6099158091674464,
      "flow_matching_loss": 0.0668487049639225,
      "grad_norm": 0.16888293623924255,
      "learning_rate": 9.949612236795682e-05,
      "loss": 0.0758,
      "relation_loss": 0.00893228454515338,
      "selected_pairs": 26475.2,
      "state_query_norm": 635.2164306640625,
      "state_relation_loss": 0.001224727730732411,
      "step": 2790,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00893228454515338,
      "weighted_relation_loss": 0.00893228454515338
    },
    {
      "action_query_norm": 358.64093322753905,
      "action_relation_loss": 0.008367453375831247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.619270346117867,
      "flow_matching_loss": 0.06811228394508362,
      "grad_norm": 0.20554019510746002,
      "learning_rate": 9.948828733923305e-05,
      "loss": 0.0776,
      "relation_loss": 0.009515162650495767,
      "selected_pairs": 26981.5,
      "state_query_norm": 644.4478210449219,
      "state_relation_loss": 0.0011477092397399248,
      "step": 2800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.009515162650495767,
      "weighted_relation_loss": 0.009515162650495767
    },
    {
      "action_query_norm": 358.6804138183594,
      "action_relation_loss": 0.00831278320401907,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.628624883068288,
      "flow_matching_loss": 0.06946376115083694,
      "grad_norm": 0.24309515953063965,
      "learning_rate": 9.948039217757374e-05,
      "loss": 0.0789,
      "relation_loss": 0.009477431792765856,
      "selected_pairs": 26777.2,
      "state_query_norm": 643.4040466308594,
      "state_relation_loss": 0.0011646485072560608,
      "step": 2810,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.009477431792765856,
      "weighted_relation_loss": 0.009477431792765856
    },
    {
      "action_query_norm": 350.9569854736328,
      "action_relation_loss": 0.007870867289602757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.637979420018709,
      "flow_matching_loss": 0.06750498414039612,
      "grad_norm": 0.19894631206989288,
      "learning_rate": 9.947243689257225e-05,
      "loss": 0.0765,
      "relation_loss": 0.009041175339370966,
      "selected_pairs": 26851.5,
      "state_query_norm": 647.3078369140625,
      "state_relation_loss": 0.0011703080032020806,
      "step": 2820,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009041175339370966,
      "weighted_relation_loss": 0.009041175339370966
    },
    {
      "action_query_norm": 357.73349914550784,
      "action_relation_loss": 0.007833652943372727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.64733395696913,
      "flow_matching_loss": 0.06429466232657433,
      "grad_norm": 0.18637549877166748,
      "learning_rate": 9.946442149389497e-05,
      "loss": 0.0733,
      "relation_loss": 0.008984773233532906,
      "selected_pairs": 27103.5,
      "state_query_norm": 648.3685791015625,
      "state_relation_loss": 0.001151120220310986,
      "step": 2830,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.008984773233532906,
      "weighted_relation_loss": 0.008984773233532906
    },
    {
      "action_query_norm": 364.7059051513672,
      "action_relation_loss": 0.0075699059758335355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.656688493919551,
      "flow_matching_loss": 0.06754940524697303,
      "grad_norm": 0.17662754654884338,
      "learning_rate": 9.945634599128139e-05,
      "loss": 0.0763,
      "relation_loss": 0.008764341613277793,
      "selected_pairs": 26966.8,
      "state_query_norm": 649.4652404785156,
      "state_relation_loss": 0.0011944355792365968,
      "step": 2840,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.008764341613277793,
      "weighted_relation_loss": 0.008764341613277793
    },
    {
      "action_query_norm": 353.78644104003905,
      "action_relation_loss": 0.007809357438236475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.666043030869972,
      "flow_matching_loss": 0.07050311788916588,
      "grad_norm": 0.17787186801433563,
      "learning_rate": 9.944821039454402e-05,
      "loss": 0.0795,
      "relation_loss": 0.009021586831659078,
      "selected_pairs": 26851.4,
      "state_query_norm": 645.5938598632813,
      "state_relation_loss": 0.0012122293119318783,
      "step": 2850,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009021586831659078,
      "weighted_relation_loss": 0.009021586831659078
    },
    {
      "action_query_norm": 358.1372589111328,
      "action_relation_loss": 0.008109597070142627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6753975678203927,
      "flow_matching_loss": 0.06797061413526535,
      "grad_norm": 0.15066230297088623,
      "learning_rate": 9.944001471356835e-05,
      "loss": 0.0773,
      "relation_loss": 0.00930261444300413,
      "selected_pairs": 26412.8,
      "state_query_norm": 649.9956115722656,
      "state_relation_loss": 0.001193017326295376,
      "step": 2860,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.00930261444300413,
      "weighted_relation_loss": 0.00930261444300413
    },
    {
      "action_query_norm": 366.0387786865234,
      "action_relation_loss": 0.007888393010944128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.684752104770814,
      "flow_matching_loss": 0.06760683096945286,
      "grad_norm": 0.3047311007976532,
      "learning_rate": 9.94317589583129e-05,
      "loss": 0.0766,
      "relation_loss": 0.009034117124974728,
      "selected_pairs": 26855.5,
      "state_query_norm": 637.9294494628906,
      "state_relation_loss": 0.001145724067464471,
      "step": 2870,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009034117124974728,
      "weighted_relation_loss": 0.009034117124974728
    },
    {
      "action_query_norm": 363.4650817871094,
      "action_relation_loss": 0.008089482598006726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.694106641721235,
      "flow_matching_loss": 0.06784008666872979,
      "grad_norm": 0.24274751543998718,
      "learning_rate": 9.942344313880922e-05,
      "loss": 0.0772,
      "relation_loss": 0.009311707410961389,
      "selected_pairs": 26990.4,
      "state_query_norm": 637.7208068847656,
      "state_relation_loss": 0.001222224871162325,
      "step": 2880,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.009311707410961389,
      "weighted_relation_loss": 0.009311707410961389
    },
    {
      "action_query_norm": 363.32189025878904,
      "action_relation_loss": 0.008585609681904315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.703461178671656,
      "flow_matching_loss": 0.06967544481158257,
      "grad_norm": 0.21514566242694855,
      "learning_rate": 9.941506726516179e-05,
      "loss": 0.0794,
      "relation_loss": 0.009760561492294073,
      "selected_pairs": 26854.9,
      "state_query_norm": 651.254736328125,
      "state_relation_loss": 0.0011749518336728216,
      "step": 2890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009760561492294073,
      "weighted_relation_loss": 0.009760561492294073
    },
    {
      "action_query_norm": 362.7378387451172,
      "action_relation_loss": 0.008319542091339827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7128157156220767,
      "flow_matching_loss": 0.068732338398695,
      "grad_norm": 0.1741291880607605,
      "learning_rate": 9.94066313475481e-05,
      "loss": 0.0782,
      "relation_loss": 0.009467284381389617,
      "selected_pairs": 27153.3,
      "state_query_norm": 651.1895385742188,
      "state_relation_loss": 0.0011477424413897097,
      "step": 2900,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.009467284381389617,
      "weighted_relation_loss": 0.009467284381389617
    },
    {
      "action_query_norm": 361.73793334960936,
      "action_relation_loss": 0.00814419104717672,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.722170252572498,
      "flow_matching_loss": 0.07232236042618752,
      "grad_norm": 0.17087174952030182,
      "learning_rate": 9.939813539621857e-05,
      "loss": 0.0817,
      "relation_loss": 0.00934835122898221,
      "selected_pairs": 27087.3,
      "state_query_norm": 644.35400390625,
      "state_relation_loss": 0.0012041603331454098,
      "step": 2910,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00934835122898221,
      "weighted_relation_loss": 0.00934835122898221
    },
    {
      "action_query_norm": 365.55576782226564,
      "action_relation_loss": 0.008152588829398155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7315247895229184,
      "flow_matching_loss": 0.06868748143315315,
      "grad_norm": 0.2062525749206543,
      "learning_rate": 9.93895794214966e-05,
      "loss": 0.078,
      "relation_loss": 0.009333096630871295,
      "selected_pairs": 27243.4,
      "state_query_norm": 633.1876098632813,
      "state_relation_loss": 0.0011805077432654799,
      "step": 2920,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.009333096630871295,
      "weighted_relation_loss": 0.009333096630871295
    },
    {
      "action_query_norm": 363.95762023925784,
      "action_relation_loss": 0.007934445329010487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7408793264733395,
      "flow_matching_loss": 0.06757656820118427,
      "grad_norm": 0.2973158359527588,
      "learning_rate": 9.938096343377852e-05,
      "loss": 0.0767,
      "relation_loss": 0.009120921697467565,
      "selected_pairs": 27533.1,
      "state_query_norm": 644.3457580566406,
      "state_relation_loss": 0.0011864764150232078,
      "step": 2930,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.009120921697467565,
      "weighted_relation_loss": 0.009120921697467565
    },
    {
      "action_query_norm": 367.0849090576172,
      "action_relation_loss": 0.008023749385029078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7502338634237606,
      "flow_matching_loss": 0.06591596528887748,
      "grad_norm": 0.19262436032295227,
      "learning_rate": 9.937228744353353e-05,
      "loss": 0.0751,
      "relation_loss": 0.009151682257652283,
      "selected_pairs": 27085.0,
      "state_query_norm": 646.78466796875,
      "state_relation_loss": 0.001127932930830866,
      "step": 2940,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.009151682257652283,
      "weighted_relation_loss": 0.009151682257652283
    },
    {
      "action_query_norm": 362.49671630859376,
      "action_relation_loss": 0.008260230580344797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7595884003741817,
      "flow_matching_loss": 0.06629604622721672,
      "grad_norm": 0.1973838210105896,
      "learning_rate": 9.936355146130379e-05,
      "loss": 0.0757,
      "relation_loss": 0.009417899232357741,
      "selected_pairs": 26713.4,
      "state_query_norm": 652.2145812988281,
      "state_relation_loss": 0.0011576684541068972,
      "step": 2950,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.009417899232357741,
      "weighted_relation_loss": 0.009417899232357741
    },
    {
      "action_query_norm": 365.1019592285156,
      "action_relation_loss": 0.007661481015384197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7689429373246024,
      "flow_matching_loss": 0.07188760563731193,
      "grad_norm": 0.19410890340805054,
      "learning_rate": 9.935475549770435e-05,
      "loss": 0.0807,
      "relation_loss": 0.008850066224113107,
      "selected_pairs": 26845.7,
      "state_query_norm": 650.9405029296875,
      "state_relation_loss": 0.0011885851388797164,
      "step": 2960,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008850066224113107,
      "weighted_relation_loss": 0.008850066224113107
    },
    {
      "action_query_norm": 360.8576232910156,
      "action_relation_loss": 0.008216638350859284,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7782974742750235,
      "flow_matching_loss": 0.06652151420712471,
      "grad_norm": 0.28737926483154297,
      "learning_rate": 9.934589956342315e-05,
      "loss": 0.0758,
      "relation_loss": 0.009327090252190829,
      "selected_pairs": 26941.4,
      "state_query_norm": 667.0247619628906,
      "state_relation_loss": 0.0011104519478976726,
      "step": 2970,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.009327090252190829,
      "weighted_relation_loss": 0.009327090252190829
    },
    {
      "action_query_norm": 362.22362365722654,
      "action_relation_loss": 0.007875457778573037,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.787652011225444,
      "flow_matching_loss": 0.0676540419459343,
      "grad_norm": 0.22484219074249268,
      "learning_rate": 9.933698366922093e-05,
      "loss": 0.0766,
      "relation_loss": 0.008987088873982429,
      "selected_pairs": 26626.7,
      "state_query_norm": 645.6607238769532,
      "state_relation_loss": 0.0011116311303339898,
      "step": 2980,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.008987088873982429,
      "weighted_relation_loss": 0.008987088873982429
    },
    {
      "action_query_norm": 366.56712036132814,
      "action_relation_loss": 0.007694002892822027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7970065481758652,
      "flow_matching_loss": 0.07007924690842629,
      "grad_norm": 0.19393758475780487,
      "learning_rate": 9.93280078259314e-05,
      "loss": 0.0789,
      "relation_loss": 0.008821860421448946,
      "selected_pairs": 27068.2,
      "state_query_norm": 646.4672546386719,
      "state_relation_loss": 0.0011278576916083694,
      "step": 2990,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.008821860421448946,
      "weighted_relation_loss": 0.008821860421448946
    },
    {
      "action_query_norm": 372.0503326416016,
      "action_relation_loss": 0.007945046247914433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8063610851262863,
      "flow_matching_loss": 0.06835817769169808,
      "grad_norm": 0.2102833241224289,
      "learning_rate": 9.931897204446104e-05,
      "loss": 0.0774,
      "relation_loss": 0.009027891978621482,
      "selected_pairs": 27126.1,
      "state_query_norm": 652.4312133789062,
      "state_relation_loss": 0.001082845765631646,
      "step": 3000,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.009027891978621482,
      "weighted_relation_loss": 0.009027891978621482
    },
    {
      "action_query_norm": 376.2794647216797,
      "action_relation_loss": 0.008070154394954444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8157156220767074,
      "flow_matching_loss": 0.06779616810381413,
      "grad_norm": 0.2176474779844284,
      "learning_rate": 9.930987633578915e-05,
      "loss": 0.0769,
      "relation_loss": 0.00914982296526432,
      "selected_pairs": 26575.7,
      "state_query_norm": 665.4732727050781,
      "state_relation_loss": 0.0010796686634421348,
      "step": 3010,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.00914982296526432,
      "weighted_relation_loss": 0.00914982296526432
    },
    {
      "action_query_norm": 374.23923950195314,
      "action_relation_loss": 0.007886103400960565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.825070159027128,
      "flow_matching_loss": 0.0683508075773716,
      "grad_norm": 0.15549378097057343,
      "learning_rate": 9.93007207109679e-05,
      "loss": 0.0774,
      "relation_loss": 0.009026744216680527,
      "selected_pairs": 26312.2,
      "state_query_norm": 661.0112121582031,
      "state_relation_loss": 0.001140640932135284,
      "step": 3020,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.009026744216680527,
      "weighted_relation_loss": 0.009026744216680527
    },
    {
      "action_query_norm": 370.2215179443359,
      "action_relation_loss": 0.008009162778034807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.834424695977549,
      "flow_matching_loss": 0.06878905221819878,
      "grad_norm": 0.16277144849300385,
      "learning_rate": 9.929150518112224e-05,
      "loss": 0.0779,
      "relation_loss": 0.009150848444551229,
      "selected_pairs": 26597.0,
      "state_query_norm": 656.9447387695312,
      "state_relation_loss": 0.0011416854686103761,
      "step": 3030,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.009150848444551229,
      "weighted_relation_loss": 0.009150848444551229
    },
    {
      "action_query_norm": 370.8384979248047,
      "action_relation_loss": 0.008220879035070539,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.84377923292797,
      "flow_matching_loss": 0.06517311632633209,
      "grad_norm": 0.16556373238563538,
      "learning_rate": 9.928222975744991e-05,
      "loss": 0.0745,
      "relation_loss": 0.009319860953837633,
      "selected_pairs": 27404.9,
      "state_query_norm": 657.9594116210938,
      "state_relation_loss": 0.0010989820351824165,
      "step": 3040,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.009319860953837633,
      "weighted_relation_loss": 0.009319860953837633
    },
    {
      "action_query_norm": 359.24280700683596,
      "action_relation_loss": 0.008065673569217325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.853133769878391,
      "flow_matching_loss": 0.06836543045938015,
      "grad_norm": 0.16282252967357635,
      "learning_rate": 9.92728944512214e-05,
      "loss": 0.0775,
      "relation_loss": 0.009176712017506361,
      "selected_pairs": 26829.9,
      "state_query_norm": 648.5251586914062,
      "state_relation_loss": 0.001111038401722908,
      "step": 3050,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.009176712017506361,
      "weighted_relation_loss": 0.009176712017506361
    },
    {
      "action_query_norm": 365.2908966064453,
      "action_relation_loss": 0.008207297883927822,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.862488306828812,
      "flow_matching_loss": 0.06915698572993279,
      "grad_norm": 0.19269749522209167,
      "learning_rate": 9.926349927378001e-05,
      "loss": 0.0785,
      "relation_loss": 0.009308390133082867,
      "selected_pairs": 26735.0,
      "state_query_norm": 651.9323547363281,
      "state_relation_loss": 0.001101092144381255,
      "step": 3060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.009308390133082867,
      "weighted_relation_loss": 0.009308390133082867
    },
    {
      "action_query_norm": 366.29368286132814,
      "action_relation_loss": 0.007422742992639541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.871842843779233,
      "flow_matching_loss": 0.06555492244660854,
      "grad_norm": 0.14276106655597687,
      "learning_rate": 9.925404423654174e-05,
      "loss": 0.0741,
      "relation_loss": 0.008513035206124186,
      "selected_pairs": 26500.2,
      "state_query_norm": 670.5847106933594,
      "state_relation_loss": 0.0010902921669185162,
      "step": 3070,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.008513035206124186,
      "weighted_relation_loss": 0.008513035206124186
    },
    {
      "action_query_norm": 365.9617645263672,
      "action_relation_loss": 0.007850180007517337,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.881197380729654,
      "flow_matching_loss": 0.06881036274135113,
      "grad_norm": 0.20220614969730377,
      "learning_rate": 9.924452935099537e-05,
      "loss": 0.0778,
      "relation_loss": 0.00895205968990922,
      "selected_pairs": 26760.8,
      "state_query_norm": 665.1907897949219,
      "state_relation_loss": 0.0011018797056749463,
      "step": 3080,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.00895205968990922,
      "weighted_relation_loss": 0.00895205968990922
    },
    {
      "action_query_norm": 371.9969909667969,
      "action_relation_loss": 0.007761413604021073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.890551917680075,
      "flow_matching_loss": 0.06540590189397336,
      "grad_norm": 0.19710440933704376,
      "learning_rate": 9.92349546287024e-05,
      "loss": 0.0743,
      "relation_loss": 0.008860632963478565,
      "selected_pairs": 26761.6,
      "state_query_norm": 666.8272521972656,
      "state_relation_loss": 0.0010992193361744284,
      "step": 3090,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.008860632963478565,
      "weighted_relation_loss": 0.008860632963478565
    },
    {
      "action_query_norm": 374.76500549316404,
      "action_relation_loss": 0.00756876370869577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8999064546304956,
      "flow_matching_loss": 0.06323905922472477,
      "grad_norm": 0.16615889966487885,
      "learning_rate": 9.9225320081297e-05,
      "loss": 0.0719,
      "relation_loss": 0.008643615245819091,
      "selected_pairs": 26596.9,
      "state_query_norm": 662.7346374511719,
      "state_relation_loss": 0.0010748514789156616,
      "step": 3100,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.008643615245819091,
      "weighted_relation_loss": 0.008643615245819091
    },
    {
      "action_query_norm": 375.8738250732422,
      "action_relation_loss": 0.008116508089005946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9092609915809167,
      "flow_matching_loss": 0.06495024599134921,
      "grad_norm": 0.19087593257427216,
      "learning_rate": 9.921562572048606e-05,
      "loss": 0.0741,
      "relation_loss": 0.009185128286480904,
      "selected_pairs": 26825.8,
      "state_query_norm": 659.5930541992187,
      "state_relation_loss": 0.0010686201916541905,
      "step": 3110,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.009185128286480904,
      "weighted_relation_loss": 0.009185128286480904
    },
    {
      "action_query_norm": 375.49513549804686,
      "action_relation_loss": 0.007564858766272664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9186155285313378,
      "flow_matching_loss": 0.06518805138766766,
      "grad_norm": 0.1737152338027954,
      "learning_rate": 9.920587155804913e-05,
      "loss": 0.0738,
      "relation_loss": 0.008648435957729817,
      "selected_pairs": 26923.6,
      "state_query_norm": 666.0405883789062,
      "state_relation_loss": 0.0010835773195140065,
      "step": 3120,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.008648435957729817,
      "weighted_relation_loss": 0.008648435957729817
    },
    {
      "action_query_norm": 365.8643035888672,
      "action_relation_loss": 0.00790767096914351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.927970065481759,
      "flow_matching_loss": 0.06706517711281776,
      "grad_norm": 0.1650819182395935,
      "learning_rate": 9.919605760583845e-05,
      "loss": 0.0761,
      "relation_loss": 0.009046125132590533,
      "selected_pairs": 27149.6,
      "state_query_norm": 659.1567993164062,
      "state_relation_loss": 0.0011384542565792799,
      "step": 3130,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.009046125132590533,
      "weighted_relation_loss": 0.009046125132590533
    },
    {
      "action_query_norm": 370.1566162109375,
      "action_relation_loss": 0.007976640434935688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9373246024321795,
      "flow_matching_loss": 0.06331698782742023,
      "grad_norm": 0.17333868145942688,
      "learning_rate": 9.91861838757789e-05,
      "loss": 0.0724,
      "relation_loss": 0.009074420807883144,
      "selected_pairs": 26799.0,
      "state_query_norm": 666.565869140625,
      "state_relation_loss": 0.0010977802681736647,
      "step": 3140,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.009074420807883144,
      "weighted_relation_loss": 0.009074420807883144
    },
    {
      "action_query_norm": 377.0458465576172,
      "action_relation_loss": 0.0074509632308036085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9466791393826006,
      "flow_matching_loss": 0.06602950133383274,
      "grad_norm": 0.1618310511112213,
      "learning_rate": 9.917625037986798e-05,
      "loss": 0.0746,
      "relation_loss": 0.008534600026905537,
      "selected_pairs": 26776.1,
      "state_query_norm": 650.9912841796875,
      "state_relation_loss": 0.0010836368019226938,
      "step": 3150,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008534600026905537,
      "weighted_relation_loss": 0.008534600026905537
    },
    {
      "action_query_norm": 376.5486297607422,
      "action_relation_loss": 0.00747634400613606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9560336763330213,
      "flow_matching_loss": 0.06328612044453621,
      "grad_norm": 0.18152153491973877,
      "learning_rate": 9.916625713017583e-05,
      "loss": 0.0718,
      "relation_loss": 0.008531892392784356,
      "selected_pairs": 26776.4,
      "state_query_norm": 647.8156188964844,
      "state_relation_loss": 0.001055548369185999,
      "step": 3160,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008531892392784356,
      "weighted_relation_loss": 0.008531892392784356
    },
    {
      "action_query_norm": 380.50587463378906,
      "action_relation_loss": 0.00756259704940021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9653882132834424,
      "flow_matching_loss": 0.06534818187355995,
      "grad_norm": 0.15045855939388275,
      "learning_rate": 9.915620413884519e-05,
      "loss": 0.074,
      "relation_loss": 0.008611133508384227,
      "selected_pairs": 26784.3,
      "state_query_norm": 669.0486022949219,
      "state_relation_loss": 0.001048536424059421,
      "step": 3170,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008611133508384227,
      "weighted_relation_loss": 0.008611133508384227
    },
    {
      "action_query_norm": 378.05244140625,
      "action_relation_loss": 0.007422317983582616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9747427502338635,
      "flow_matching_loss": 0.06739708706736565,
      "grad_norm": 0.14675918221473694,
      "learning_rate": 9.914609141809139e-05,
      "loss": 0.076,
      "relation_loss": 0.008567404607310891,
      "selected_pairs": 26921.9,
      "state_query_norm": 667.9309204101562,
      "state_relation_loss": 0.001145086600445211,
      "step": 3180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.008567404607310891,
      "weighted_relation_loss": 0.008567404607310891
    },
    {
      "action_query_norm": 373.1508514404297,
      "action_relation_loss": 0.007981549389660359,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9840972871842846,
      "flow_matching_loss": 0.06869405433535576,
      "grad_norm": 0.1377776861190796,
      "learning_rate": 9.913591898020235e-05,
      "loss": 0.0777,
      "relation_loss": 0.008995716646313667,
      "selected_pairs": 27247.1,
      "state_query_norm": 676.0834289550781,
      "state_relation_loss": 0.0010141672624740749,
      "step": 3190,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.008995716646313667,
      "weighted_relation_loss": 0.008995716646313667
    },
    {
      "action_query_norm": 369.3069732666016,
      "action_relation_loss": 0.008159521268680692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9934518241347052,
      "flow_matching_loss": 0.07004631385207176,
      "grad_norm": 0.18439671397209167,
      "learning_rate": 9.912568683753853e-05,
      "loss": 0.0793,
      "relation_loss": 0.009250304941087962,
      "selected_pairs": 26684.8,
      "state_query_norm": 672.0885375976562,
      "state_relation_loss": 0.0010907836956903338,
      "step": 3200,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.009250304941087962,
      "weighted_relation_loss": 0.009250304941087962
    },
    {
      "action_query_norm": 380.80939025878905,
      "action_relation_loss": 0.007557179685682059,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0028063610851263,
      "flow_matching_loss": 0.06636974439024926,
      "grad_norm": 0.2499120980501175,
      "learning_rate": 9.911539500253295e-05,
      "loss": 0.075,
      "relation_loss": 0.008650158438831568,
      "selected_pairs": 23666.0,
      "state_query_norm": 685.99267578125,
      "state_relation_loss": 0.0010929788171779365,
      "step": 3210,
      "valid_action_samples": 211.5,
      "weighted_auxiliary_loss": 0.008650158438831568,
      "weighted_relation_loss": 0.008650158438831568
    },
    {
      "action_query_norm": 367.21005859375,
      "action_relation_loss": 0.008226330857723951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0121608980355474,
      "flow_matching_loss": 0.0657748132944107,
      "grad_norm": 0.1967742145061493,
      "learning_rate": 9.910504348769118e-05,
      "loss": 0.0751,
      "relation_loss": 0.009339430462568998,
      "selected_pairs": 26881.7,
      "state_query_norm": 666.5167541503906,
      "state_relation_loss": 0.0011130996281281114,
      "step": 3220,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.009339430462568998,
      "weighted_relation_loss": 0.009339430462568998
    },
    {
      "action_query_norm": 375.39056091308595,
      "action_relation_loss": 0.008042486710473895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.021515434985968,
      "flow_matching_loss": 0.0655230350792408,
      "grad_norm": 0.2625417113304138,
      "learning_rate": 9.909463230559127e-05,
      "loss": 0.0747,
      "relation_loss": 0.00918938615359366,
      "selected_pairs": 27205.0,
      "state_query_norm": 669.8966918945313,
      "state_relation_loss": 0.0011468994664028287,
      "step": 3230,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00918938615359366,
      "weighted_relation_loss": 0.00918938615359366
    },
    {
      "action_query_norm": 375.19935913085936,
      "action_relation_loss": 0.007851759204640985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.030869971936389,
      "flow_matching_loss": 0.06700734905898571,
      "grad_norm": 0.21019881963729858,
      "learning_rate": 9.908416146888376e-05,
      "loss": 0.0759,
      "relation_loss": 0.008938202960416674,
      "selected_pairs": 27251.6,
      "state_query_norm": 668.3295654296875,
      "state_relation_loss": 0.0010864437441341578,
      "step": 3240,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.008938202960416674,
      "weighted_relation_loss": 0.008938202960416674
    },
    {
      "action_query_norm": 381.223486328125,
      "action_relation_loss": 0.008039573393762111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0402245088868103,
      "flow_matching_loss": 0.0656412836164236,
      "grad_norm": 0.24299287796020508,
      "learning_rate": 9.907363099029175e-05,
      "loss": 0.0748,
      "relation_loss": 0.009133014641702176,
      "selected_pairs": 26344.2,
      "state_query_norm": 663.6451354980469,
      "state_relation_loss": 0.00109344128286466,
      "step": 3250,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.009133014641702176,
      "weighted_relation_loss": 0.009133014641702176
    },
    {
      "action_query_norm": 386.7778015136719,
      "action_relation_loss": 0.007567906565964222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.049579045837231,
      "flow_matching_loss": 0.06560971550643444,
      "grad_norm": 0.16380061209201813,
      "learning_rate": 9.906304088261073e-05,
      "loss": 0.0743,
      "relation_loss": 0.008681748062372208,
      "selected_pairs": 27233.9,
      "state_query_norm": 671.1055603027344,
      "state_relation_loss": 0.0011138414614833892,
      "step": 3260,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008681748062372208,
      "weighted_relation_loss": 0.008681748062372208
    },
    {
      "action_query_norm": 382.0475311279297,
      "action_relation_loss": 0.007776444731280208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.058933582787652,
      "flow_matching_loss": 0.06614168584346772,
      "grad_norm": 0.15198415517807007,
      "learning_rate": 9.905239115870872e-05,
      "loss": 0.075,
      "relation_loss": 0.008903922606259584,
      "selected_pairs": 26854.1,
      "state_query_norm": 677.4609924316406,
      "state_relation_loss": 0.00112747794482857,
      "step": 3270,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008903922606259584,
      "weighted_relation_loss": 0.008903922606259584
    },
    {
      "action_query_norm": 382.02813720703125,
      "action_relation_loss": 0.007451476249843836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.068288119738073,
      "flow_matching_loss": 0.06808988153934478,
      "grad_norm": 0.16179916262626648,
      "learning_rate": 9.90416818315261e-05,
      "loss": 0.0767,
      "relation_loss": 0.008605526015162467,
      "selected_pairs": 26877.9,
      "state_query_norm": 670.2976745605469,
      "state_relation_loss": 0.001154049683827907,
      "step": 3280,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.008605526015162467,
      "weighted_relation_loss": 0.008605526015162467
    },
    {
      "action_query_norm": 381.56500244140625,
      "action_relation_loss": 0.007380283111706376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.077642656688494,
      "flow_matching_loss": 0.06534657739102841,
      "grad_norm": 0.14778923988342285,
      "learning_rate": 9.903091291407573e-05,
      "loss": 0.0739,
      "relation_loss": 0.008551307860761881,
      "selected_pairs": 27170.2,
      "state_query_norm": 680.2337463378906,
      "state_relation_loss": 0.001171024702489376,
      "step": 3290,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.008551307860761881,
      "weighted_relation_loss": 0.008551307860761881
    },
    {
      "action_query_norm": 383.20037841796875,
      "action_relation_loss": 0.007725099893286824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.086997193638915,
      "flow_matching_loss": 0.06704280748963357,
      "grad_norm": 0.12592445313930511,
      "learning_rate": 9.902008441944286e-05,
      "loss": 0.0759,
      "relation_loss": 0.008895818144083023,
      "selected_pairs": 26972.9,
      "state_query_norm": 672.6708740234375,
      "state_relation_loss": 0.001170718134380877,
      "step": 3300,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.008895818144083023,
      "weighted_relation_loss": 0.008895818144083023
    },
    {
      "action_query_norm": 377.7701019287109,
      "action_relation_loss": 0.007547882525250316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.096351730589336,
      "flow_matching_loss": 0.06400039084255696,
      "grad_norm": 0.18786752223968506,
      "learning_rate": 9.900919636078512e-05,
      "loss": 0.0727,
      "relation_loss": 0.008684848342090845,
      "selected_pairs": 26237.4,
      "state_query_norm": 666.78232421875,
      "state_relation_loss": 0.0011369658284820616,
      "step": 3310,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.008684848342090845,
      "weighted_relation_loss": 0.008684848342090845
    },
    {
      "action_query_norm": 387.6018859863281,
      "action_relation_loss": 0.007753303507342935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1057062675397566,
      "flow_matching_loss": 0.06537817567586898,
      "grad_norm": 0.19137868285179138,
      "learning_rate": 9.899824875133255e-05,
      "loss": 0.0743,
      "relation_loss": 0.008872062340378762,
      "selected_pairs": 26895.5,
      "state_query_norm": 681.8591918945312,
      "state_relation_loss": 0.0011187587748281657,
      "step": 3320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.008872062340378762,
      "weighted_relation_loss": 0.008872062340378762
    },
    {
      "action_query_norm": 387.02625427246096,
      "action_relation_loss": 0.007200382184237242,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1150608044901777,
      "flow_matching_loss": 0.0644128829240799,
      "grad_norm": 0.14456234872341156,
      "learning_rate": 9.898724160438749e-05,
      "loss": 0.0727,
      "relation_loss": 0.008260565111413598,
      "selected_pairs": 26834.6,
      "state_query_norm": 676.8868225097656,
      "state_relation_loss": 0.0010601829271763563,
      "step": 3330,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.008260565111413598,
      "weighted_relation_loss": 0.008260565111413598
    },
    {
      "action_query_norm": 385.14824829101565,
      "action_relation_loss": 0.007376530161127448,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.124415341440599,
      "flow_matching_loss": 0.06536822952330112,
      "grad_norm": 0.19837957620620728,
      "learning_rate": 9.89761749333247e-05,
      "loss": 0.0738,
      "relation_loss": 0.008444945560768247,
      "selected_pairs": 27175.2,
      "state_query_norm": 666.5469604492188,
      "state_relation_loss": 0.0010684153356123715,
      "step": 3340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.008444945560768247,
      "weighted_relation_loss": 0.008444945560768247
    },
    {
      "action_query_norm": 388.0183044433594,
      "action_relation_loss": 0.00731935054063797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1337698783910195,
      "flow_matching_loss": 0.06433978900313378,
      "grad_norm": 0.1629357635974884,
      "learning_rate": 9.896504875159122e-05,
      "loss": 0.0728,
      "relation_loss": 0.008425172697752715,
      "selected_pairs": 27089.8,
      "state_query_norm": 677.7584167480469,
      "state_relation_loss": 0.0011058221687562763,
      "step": 3350,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.008425172697752715,
      "weighted_relation_loss": 0.008425172697752715
    },
    {
      "action_query_norm": 387.5427673339844,
      "action_relation_loss": 0.00699802148155868,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1431244153414406,
      "flow_matching_loss": 0.0679480068385601,
      "grad_norm": 0.18526959419250488,
      "learning_rate": 9.89538630727064e-05,
      "loss": 0.076,
      "relation_loss": 0.008079102402552963,
      "selected_pairs": 27230.4,
      "state_query_norm": 682.8958679199219,
      "state_relation_loss": 0.0010810809792019427,
      "step": 3360,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008079102402552963,
      "weighted_relation_loss": 0.008079102402552963
    },
    {
      "action_query_norm": 385.42681274414065,
      "action_relation_loss": 0.007315668882802129,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1524789522918617,
      "flow_matching_loss": 0.06682976484298705,
      "grad_norm": 0.2229853719472885,
      "learning_rate": 9.894261791026189e-05,
      "loss": 0.0752,
      "relation_loss": 0.008333831373602153,
      "selected_pairs": 26273.9,
      "state_query_norm": 688.6982788085937,
      "state_relation_loss": 0.0010181625606492163,
      "step": 3370,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.008333831373602153,
      "weighted_relation_loss": 0.008333831373602153
    },
    {
      "action_query_norm": 388.46644897460936,
      "action_relation_loss": 0.00721491458825767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1618334892422824,
      "flow_matching_loss": 0.06589227765798569,
      "grad_norm": 0.201499342918396,
      "learning_rate": 9.893131327792165e-05,
      "loss": 0.0742,
      "relation_loss": 0.008297401992604137,
      "selected_pairs": 27385.8,
      "state_query_norm": 688.3526611328125,
      "state_relation_loss": 0.001082487392704934,
      "step": 3380,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.008297401992604137,
      "weighted_relation_loss": 0.008297401992604137
    },
    {
      "action_query_norm": 385.22461853027346,
      "action_relation_loss": 0.00749524924904108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1711880261927035,
      "flow_matching_loss": 0.06398913934826851,
      "grad_norm": 0.22957758605480194,
      "learning_rate": 9.891994918942182e-05,
      "loss": 0.0726,
      "relation_loss": 0.008576849987730385,
      "selected_pairs": 27008.0,
      "state_query_norm": 696.7300598144532,
      "state_relation_loss": 0.0010816007968969643,
      "step": 3390,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008576849987730385,
      "weighted_relation_loss": 0.008576849987730385
    },
    {
      "action_query_norm": 386.9494934082031,
      "action_relation_loss": 0.007603006483986974,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1805425631431246,
      "flow_matching_loss": 0.06543483883142472,
      "grad_norm": 0.17236211895942688,
      "learning_rate": 9.890852565857092e-05,
      "loss": 0.0741,
      "relation_loss": 0.008681826805695892,
      "selected_pairs": 27017.3,
      "state_query_norm": 714.6140502929687,
      "state_relation_loss": 0.001078820344991982,
      "step": 3400,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008681826805695892,
      "weighted_relation_loss": 0.008681826805695892
    },
    {
      "action_query_norm": 379.1946746826172,
      "action_relation_loss": 0.007543927244842053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.189897100093545,
      "flow_matching_loss": 0.0627247542142868,
      "grad_norm": 0.14553703367710114,
      "learning_rate": 9.889704269924954e-05,
      "loss": 0.0714,
      "relation_loss": 0.008627641526982189,
      "selected_pairs": 27201.0,
      "state_query_norm": 704.6485717773437,
      "state_relation_loss": 0.0010837142181117088,
      "step": 3410,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.008627641526982189,
      "weighted_relation_loss": 0.008627641526982189
    },
    {
      "action_query_norm": 385.01787719726565,
      "action_relation_loss": 0.00735282483510673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1992516370439663,
      "flow_matching_loss": 0.06652221791446208,
      "grad_norm": 0.1546635776758194,
      "learning_rate": 9.888550032541059e-05,
      "loss": 0.075,
      "relation_loss": 0.008434021938592195,
      "selected_pairs": 26783.7,
      "state_query_norm": 688.4713684082031,
      "state_relation_loss": 0.0010811970685608685,
      "step": 3420,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008434021938592195,
      "weighted_relation_loss": 0.008434021938592195
    },
    {
      "action_query_norm": 383.4348999023438,
      "action_relation_loss": 0.007468912797048688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2086061739943874,
      "flow_matching_loss": 0.06848076544702053,
      "grad_norm": 0.1660563051700592,
      "learning_rate": 9.887389855107916e-05,
      "loss": 0.077,
      "relation_loss": 0.008482929132878781,
      "selected_pairs": 26970.5,
      "state_query_norm": 692.2619873046875,
      "state_relation_loss": 0.0010140163183677941,
      "step": 3430,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.008482929132878781,
      "weighted_relation_loss": 0.008482929132878781
    },
    {
      "action_query_norm": 389.23582763671874,
      "action_relation_loss": 0.007409192668274045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.217960710944808,
      "flow_matching_loss": 0.06678997725248337,
      "grad_norm": 0.15473699569702148,
      "learning_rate": 9.886223739035248e-05,
      "loss": 0.0753,
      "relation_loss": 0.008469022158533335,
      "selected_pairs": 26622.7,
      "state_query_norm": 695.4076843261719,
      "state_relation_loss": 0.001059829513542354,
      "step": 3440,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.008469022158533335,
      "weighted_relation_loss": 0.008469022158533335
    },
    {
      "action_query_norm": 390.1435119628906,
      "action_relation_loss": 0.007445723423734308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.227315247895229,
      "flow_matching_loss": 0.06430377997457981,
      "grad_norm": 0.13456223905086517,
      "learning_rate": 9.885051685739997e-05,
      "loss": 0.0727,
      "relation_loss": 0.008438999066129327,
      "selected_pairs": 26367.4,
      "state_query_norm": 690.1060363769532,
      "state_relation_loss": 0.00099327567149885,
      "step": 3450,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.008438999066129327,
      "weighted_relation_loss": 0.008438999066129327
    },
    {
      "action_query_norm": 387.0421203613281,
      "action_relation_loss": 0.007576883491128683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2366697848456503,
      "flow_matching_loss": 0.06298882812261582,
      "grad_norm": 0.20047225058078766,
      "learning_rate": 9.883873696646316e-05,
      "loss": 0.0716,
      "relation_loss": 0.008628946030512452,
      "selected_pairs": 27223.7,
      "state_query_norm": 696.5748046875,
      "state_relation_loss": 0.0010520625917706638,
      "step": 3460,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008628946030512452,
      "weighted_relation_loss": 0.008628946030512452
    },
    {
      "action_query_norm": 390.11827392578124,
      "action_relation_loss": 0.007333340123295784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.246024321796071,
      "flow_matching_loss": 0.06371869854629039,
      "grad_norm": 0.28478536009788513,
      "learning_rate": 9.882689773185575e-05,
      "loss": 0.0721,
      "relation_loss": 0.008334648096933961,
      "selected_pairs": 26931.8,
      "state_query_norm": 691.6960388183594,
      "state_relation_loss": 0.001001307845581323,
      "step": 3470,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.008334648096933961,
      "weighted_relation_loss": 0.008334648096933961
    },
    {
      "action_query_norm": 393.1161315917969,
      "action_relation_loss": 0.007536414358764887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.255378858746492,
      "flow_matching_loss": 0.06500666067004204,
      "grad_norm": 0.18527081608772278,
      "learning_rate": 9.881499916796353e-05,
      "loss": 0.0736,
      "relation_loss": 0.00855405731126666,
      "selected_pairs": 27037.1,
      "state_query_norm": 690.6622131347656,
      "state_relation_loss": 0.0010176429990679027,
      "step": 3480,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00855405731126666,
      "weighted_relation_loss": 0.00855405731126666
    },
    {
      "action_query_norm": 389.0035064697266,
      "action_relation_loss": 0.007166387001052499,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.264733395696913,
      "flow_matching_loss": 0.06279184222221375,
      "grad_norm": 0.1987791806459427,
      "learning_rate": 9.880304128924434e-05,
      "loss": 0.071,
      "relation_loss": 0.00815856410190463,
      "selected_pairs": 26569.1,
      "state_query_norm": 686.9383239746094,
      "state_relation_loss": 0.0009921770193614065,
      "step": 3490,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.00815856410190463,
      "weighted_relation_loss": 0.00815856410190463
    },
    {
      "action_query_norm": 381.6559539794922,
      "action_relation_loss": 0.0070255252998322245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2740879326473338,
      "flow_matching_loss": 0.06680831238627434,
      "grad_norm": 0.2447911649942398,
      "learning_rate": 9.879102411022817e-05,
      "loss": 0.0749,
      "relation_loss": 0.008047591242939234,
      "selected_pairs": 26096.5,
      "state_query_norm": 690.9340637207031,
      "state_relation_loss": 0.001022065884899348,
      "step": 3500,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.008047591242939234,
      "weighted_relation_loss": 0.008047591242939234
    },
    {
      "action_query_norm": 380.65390930175784,
      "action_relation_loss": 0.006949157733470201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.283442469597755,
      "flow_matching_loss": 0.063317034766078,
      "grad_norm": 0.18127375841140747,
      "learning_rate": 9.877894764551703e-05,
      "loss": 0.0713,
      "relation_loss": 0.007978282263502478,
      "selected_pairs": 27153.3,
      "state_query_norm": 694.7358093261719,
      "state_relation_loss": 0.0010291245067492127,
      "step": 3510,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.007978282263502478,
      "weighted_relation_loss": 0.007978282263502478
    },
    {
      "action_query_norm": 389.5143127441406,
      "action_relation_loss": 0.007585029723122716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.292797006548176,
      "flow_matching_loss": 0.06424644477665424,
      "grad_norm": 0.18380607664585114,
      "learning_rate": 9.876681190978494e-05,
      "loss": 0.0729,
      "relation_loss": 0.008607773529365658,
      "selected_pairs": 26804.7,
      "state_query_norm": 691.8899536132812,
      "state_relation_loss": 0.0010227439051959664,
      "step": 3520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.008607773529365658,
      "weighted_relation_loss": 0.008607773529365658
    },
    {
      "action_query_norm": 388.0685638427734,
      "action_relation_loss": 0.007167704496532678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3021515434985966,
      "flow_matching_loss": 0.06360607780516148,
      "grad_norm": 0.19496427476406097,
      "learning_rate": 9.875461691777797e-05,
      "loss": 0.0718,
      "relation_loss": 0.008220573514699936,
      "selected_pairs": 26123.5,
      "state_query_norm": 701.6595764160156,
      "state_relation_loss": 0.0010528689657803626,
      "step": 3530,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.008220573514699936,
      "weighted_relation_loss": 0.008220573514699936
    },
    {
      "action_query_norm": 395.80902404785155,
      "action_relation_loss": 0.007140406966209411,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3115060804490177,
      "flow_matching_loss": 0.06441025771200656,
      "grad_norm": 0.24690492451190948,
      "learning_rate": 9.874236268431417e-05,
      "loss": 0.0726,
      "relation_loss": 0.008198738191276788,
      "selected_pairs": 26579.6,
      "state_query_norm": 704.8451049804687,
      "state_relation_loss": 0.0010583312716335059,
      "step": 3540,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.008198738191276788,
      "weighted_relation_loss": 0.008198738191276788
    },
    {
      "action_query_norm": 392.51897888183595,
      "action_relation_loss": 0.007443704782053829,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.320860617399439,
      "flow_matching_loss": 0.064454310759902,
      "grad_norm": 0.1934770792722702,
      "learning_rate": 9.873004922428361e-05,
      "loss": 0.073,
      "relation_loss": 0.008516400540247559,
      "selected_pairs": 26803.9,
      "state_query_norm": 700.3651062011719,
      "state_relation_loss": 0.0010726957232691347,
      "step": 3550,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.008516400540247559,
      "weighted_relation_loss": 0.008516400540247559
    },
    {
      "action_query_norm": 394.08460998535156,
      "action_relation_loss": 0.007057646661996842,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.33021515434986,
      "flow_matching_loss": 0.06265891194343567,
      "grad_norm": 0.22564993798732758,
      "learning_rate": 9.871767655264829e-05,
      "loss": 0.0708,
      "relation_loss": 0.008120117569342255,
      "selected_pairs": 26502.2,
      "state_query_norm": 698.5000305175781,
      "state_relation_loss": 0.0010624708142131568,
      "step": 3560,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.008120117569342255,
      "weighted_relation_loss": 0.008120117569342255
    },
    {
      "action_query_norm": 393.5711334228516,
      "action_relation_loss": 0.006727721123024821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3395696913002806,
      "flow_matching_loss": 0.06348718442022801,
      "grad_norm": 0.15088006854057312,
      "learning_rate": 9.87052446844422e-05,
      "loss": 0.0712,
      "relation_loss": 0.007686922745779157,
      "selected_pairs": 27664.0,
      "state_query_norm": 692.794287109375,
      "state_relation_loss": 0.0009592015761882067,
      "step": 3570,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.007686922745779157,
      "weighted_relation_loss": 0.007686922745779157
    },
    {
      "action_query_norm": 396.0874420166016,
      "action_relation_loss": 0.007034708140417933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3489242282507017,
      "flow_matching_loss": 0.06554033495485782,
      "grad_norm": 0.17770810425281525,
      "learning_rate": 9.869275363477122e-05,
      "loss": 0.0736,
      "relation_loss": 0.008052971865981817,
      "selected_pairs": 27594.8,
      "state_query_norm": 705.1978149414062,
      "state_relation_loss": 0.0010182636207900942,
      "step": 3580,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.008052971865981817,
      "weighted_relation_loss": 0.008052971865981817
    },
    {
      "action_query_norm": 395.78504638671876,
      "action_relation_loss": 0.007149422448128462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3582787652011223,
      "flow_matching_loss": 0.0639942653477192,
      "grad_norm": 0.185628280043602,
      "learning_rate": 9.868020341881312e-05,
      "loss": 0.0722,
      "relation_loss": 0.008205501874908806,
      "selected_pairs": 27530.4,
      "state_query_norm": 701.3209533691406,
      "state_relation_loss": 0.001056079368572682,
      "step": 3590,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.008205501874908806,
      "weighted_relation_loss": 0.008205501874908806
    },
    {
      "action_query_norm": 392.1460235595703,
      "action_relation_loss": 0.007038707798346877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3676333021515434,
      "flow_matching_loss": 0.06578187458217144,
      "grad_norm": 0.2340817153453827,
      "learning_rate": 9.866759405181765e-05,
      "loss": 0.0739,
      "relation_loss": 0.00812147236429155,
      "selected_pairs": 27323.8,
      "state_query_norm": 708.6688293457031,
      "state_relation_loss": 0.0010827645135577768,
      "step": 3600,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.00812147236429155,
      "weighted_relation_loss": 0.00812147236429155
    },
    {
      "action_query_norm": 388.8676483154297,
      "action_relation_loss": 0.007175638386979699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3769878391019645,
      "flow_matching_loss": 0.06392326690256596,
      "grad_norm": 0.1724216192960739,
      "learning_rate": 9.865492554910633e-05,
      "loss": 0.0721,
      "relation_loss": 0.00819052462466061,
      "selected_pairs": 26366.8,
      "state_query_norm": 701.6062683105469,
      "state_relation_loss": 0.001014886237680912,
      "step": 3610,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.00819052462466061,
      "weighted_relation_loss": 0.00819052462466061
    },
    {
      "action_query_norm": 393.0197814941406,
      "action_relation_loss": 0.007214157888665795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3863423760523856,
      "flow_matching_loss": 0.06409392058849335,
      "grad_norm": 0.18048283457756042,
      "learning_rate": 9.864219792607262e-05,
      "loss": 0.0723,
      "relation_loss": 0.00824389448389411,
      "selected_pairs": 26994.0,
      "state_query_norm": 699.7971008300781,
      "state_relation_loss": 0.0010297365544829518,
      "step": 3620,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00824389448389411,
      "weighted_relation_loss": 0.00824389448389411
    },
    {
      "action_query_norm": 390.9386260986328,
      "action_relation_loss": 0.007168002240359783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3956969130028063,
      "flow_matching_loss": 0.06131569184362888,
      "grad_norm": 0.21553319692611694,
      "learning_rate": 9.862941119818177e-05,
      "loss": 0.0695,
      "relation_loss": 0.008186232531443238,
      "selected_pairs": 27164.5,
      "state_query_norm": 691.0631408691406,
      "state_relation_loss": 0.0010182302154134958,
      "step": 3630,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.008186232531443238,
      "weighted_relation_loss": 0.008186232531443238
    },
    {
      "action_query_norm": 396.81976013183595,
      "action_relation_loss": 0.0068404755089432005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4050514499532274,
      "flow_matching_loss": 0.06360835917294025,
      "grad_norm": 0.18953709304332733,
      "learning_rate": 9.861656538097086e-05,
      "loss": 0.0714,
      "relation_loss": 0.007841564808040857,
      "selected_pairs": 27365.8,
      "state_query_norm": 701.5041198730469,
      "state_relation_loss": 0.0010010892467107624,
      "step": 3640,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.007841564808040857,
      "weighted_relation_loss": 0.007841564808040857
    },
    {
      "action_query_norm": 393.43743896484375,
      "action_relation_loss": 0.007134722266346216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.414405986903648,
      "flow_matching_loss": 0.06188592277467251,
      "grad_norm": 0.18207088112831116,
      "learning_rate": 9.860366049004877e-05,
      "loss": 0.07,
      "relation_loss": 0.008151249866932631,
      "selected_pairs": 26963.9,
      "state_query_norm": 713.824609375,
      "state_relation_loss": 0.0010165274899918586,
      "step": 3650,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.008151249866932631,
      "weighted_relation_loss": 0.008151249866932631
    },
    {
      "action_query_norm": 394.0212799072266,
      "action_relation_loss": 0.006951711373403668,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.423760523854069,
      "flow_matching_loss": 0.0631231177598238,
      "grad_norm": 0.1264079064130783,
      "learning_rate": 9.859069654109615e-05,
      "loss": 0.071,
      "relation_loss": 0.007911152951419354,
      "selected_pairs": 26942.7,
      "state_query_norm": 716.8044799804687,
      "state_relation_loss": 0.0009594416711479426,
      "step": 3660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007911152951419354,
      "weighted_relation_loss": 0.007911152951419354
    },
    {
      "action_query_norm": 399.38300170898435,
      "action_relation_loss": 0.006864533154293894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4331150608044902,
      "flow_matching_loss": 0.06169675774872303,
      "grad_norm": 0.20912295579910278,
      "learning_rate": 9.857767354986545e-05,
      "loss": 0.0695,
      "relation_loss": 0.007846436323598027,
      "selected_pairs": 26559.0,
      "state_query_norm": 700.1602233886719,
      "state_relation_loss": 0.0009819031751248986,
      "step": 3670,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.007846436323598027,
      "weighted_relation_loss": 0.007846436323598027
    },
    {
      "action_query_norm": 392.86358032226565,
      "action_relation_loss": 0.0067328006960451605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4424695977549113,
      "flow_matching_loss": 0.05958600752055645,
      "grad_norm": 0.2004556655883789,
      "learning_rate": 9.856459153218078e-05,
      "loss": 0.0673,
      "relation_loss": 0.007686669193208218,
      "selected_pairs": 26618.7,
      "state_query_norm": 708.5995483398438,
      "state_relation_loss": 0.0009538686135783792,
      "step": 3680,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.007686669193208218,
      "weighted_relation_loss": 0.007686669193208218
    },
    {
      "action_query_norm": 389.8677490234375,
      "action_relation_loss": 0.0070375566836446525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.451824134705332,
      "flow_matching_loss": 0.061306466907262804,
      "grad_norm": 0.19637097418308258,
      "learning_rate": 9.855145050393808e-05,
      "loss": 0.0693,
      "relation_loss": 0.00803128331899643,
      "selected_pairs": 26457.6,
      "state_query_norm": 709.3021423339844,
      "state_relation_loss": 0.0009937266586348414,
      "step": 3690,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.00803128331899643,
      "weighted_relation_loss": 0.00803128331899643
    },
    {
      "action_query_norm": 395.53215942382815,
      "action_relation_loss": 0.006842458201572299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.461178671655753,
      "flow_matching_loss": 0.06358380578458309,
      "grad_norm": 0.16527411341667175,
      "learning_rate": 9.85382504811049e-05,
      "loss": 0.0714,
      "relation_loss": 0.007823867164552212,
      "selected_pairs": 26831.6,
      "state_query_norm": 717.795263671875,
      "state_relation_loss": 0.0009814088232815265,
      "step": 3700,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.007823867164552212,
      "weighted_relation_loss": 0.007823867164552212
    },
    {
      "action_query_norm": 397.46519775390624,
      "action_relation_loss": 0.006782476138323545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4705332086061738,
      "flow_matching_loss": 0.06459562331438065,
      "grad_norm": 0.21602940559387207,
      "learning_rate": 9.852499147972054e-05,
      "loss": 0.0723,
      "relation_loss": 0.007752854144200683,
      "selected_pairs": 27253.7,
      "state_query_norm": 717.4208312988281,
      "state_relation_loss": 0.0009703779884148389,
      "step": 3710,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.007752854144200683,
      "weighted_relation_loss": 0.007752854144200683
    },
    {
      "action_query_norm": 398.31434936523436,
      "action_relation_loss": 0.006766056967899203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.479887745556595,
      "flow_matching_loss": 0.06200545206665993,
      "grad_norm": 0.20199188590049744,
      "learning_rate": 9.851167351589592e-05,
      "loss": 0.0697,
      "relation_loss": 0.007723573362454772,
      "selected_pairs": 26944.6,
      "state_query_norm": 705.8310546875,
      "state_relation_loss": 0.0009575165109708905,
      "step": 3720,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007723573362454772,
      "weighted_relation_loss": 0.007723573362454772
    },
    {
      "action_query_norm": 400.79892883300784,
      "action_relation_loss": 0.006720272544771433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.489242282507016,
      "flow_matching_loss": 0.06544768027961254,
      "grad_norm": 0.21365688741207123,
      "learning_rate": 9.849829660581363e-05,
      "loss": 0.0731,
      "relation_loss": 0.007648543082177639,
      "selected_pairs": 26756.8,
      "state_query_norm": 711.9827880859375,
      "state_relation_loss": 0.0009282705199439078,
      "step": 3730,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.007648543082177639,
      "weighted_relation_loss": 0.007648543082177639
    },
    {
      "action_query_norm": 398.0320251464844,
      "action_relation_loss": 0.0070838346146047115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.498596819457437,
      "flow_matching_loss": 0.06570137552917003,
      "grad_norm": 0.14177371561527252,
      "learning_rate": 9.848486076572787e-05,
      "loss": 0.0738,
      "relation_loss": 0.008101822575554252,
      "selected_pairs": 26853.2,
      "state_query_norm": 713.7001159667968,
      "state_relation_loss": 0.0010179879318457097,
      "step": 3740,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008101822575554252,
      "weighted_relation_loss": 0.008101822575554252
    },
    {
      "action_query_norm": 401.5370208740234,
      "action_relation_loss": 0.00717120012268424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5079513564078577,
      "flow_matching_loss": 0.06346829235553741,
      "grad_norm": 0.16301000118255615,
      "learning_rate": 9.847136601196446e-05,
      "loss": 0.0716,
      "relation_loss": 0.008145078038796782,
      "selected_pairs": 27221.8,
      "state_query_norm": 699.6335815429687,
      "state_relation_loss": 0.0009738780616316944,
      "step": 3750,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008145078038796782,
      "weighted_relation_loss": 0.008145078038796782
    },
    {
      "action_query_norm": 399.1304046630859,
      "action_relation_loss": 0.006809721468016505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.517305893358279,
      "flow_matching_loss": 0.0629256971180439,
      "grad_norm": 0.12274125963449478,
      "learning_rate": 9.845781236092078e-05,
      "loss": 0.0707,
      "relation_loss": 0.007742219697684049,
      "selected_pairs": 26802.9,
      "state_query_norm": 712.2187866210937,
      "state_relation_loss": 0.000932498148176819,
      "step": 3760,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007742219697684049,
      "weighted_relation_loss": 0.007742219697684049
    },
    {
      "action_query_norm": 399.6181701660156,
      "action_relation_loss": 0.006856130016967654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5266604303086995,
      "flow_matching_loss": 0.0659097209572792,
      "grad_norm": 0.16825531423091888,
      "learning_rate": 9.844419982906583e-05,
      "loss": 0.0738,
      "relation_loss": 0.00785936377942562,
      "selected_pairs": 26096.2,
      "state_query_norm": 710.8408386230469,
      "state_relation_loss": 0.0010032337449956685,
      "step": 3770,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.00785936377942562,
      "weighted_relation_loss": 0.00785936377942562
    },
    {
      "action_query_norm": 401.5776885986328,
      "action_relation_loss": 0.006932115461677313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5360149672591206,
      "flow_matching_loss": 0.06457687951624394,
      "grad_norm": 0.1393176168203354,
      "learning_rate": 9.843052843294008e-05,
      "loss": 0.0725,
      "relation_loss": 0.007947116484865546,
      "selected_pairs": 26648.6,
      "state_query_norm": 698.9640380859375,
      "state_relation_loss": 0.0010150009766221046,
      "step": 3780,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.007947116484865546,
      "weighted_relation_loss": 0.007947116484865546
    },
    {
      "action_query_norm": 390.631103515625,
      "action_relation_loss": 0.007091419259086251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5453695042095417,
      "flow_matching_loss": 0.06309323161840438,
      "grad_norm": 0.1921764761209488,
      "learning_rate": 9.841679818915559e-05,
      "loss": 0.0712,
      "relation_loss": 0.00807256563566625,
      "selected_pairs": 26686.9,
      "state_query_norm": 711.8557006835938,
      "state_relation_loss": 0.0009811463882215322,
      "step": 3790,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00807256563566625,
      "weighted_relation_loss": 0.00807256563566625
    },
    {
      "action_query_norm": 401.26146240234374,
      "action_relation_loss": 0.007192698493599892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5547240411599628,
      "flow_matching_loss": 0.0639937624335289,
      "grad_norm": 0.15908809006214142,
      "learning_rate": 9.840300911439591e-05,
      "loss": 0.0721,
      "relation_loss": 0.008153311582282186,
      "selected_pairs": 26385.9,
      "state_query_norm": 718.4446960449219,
      "state_relation_loss": 0.0009606130886822938,
      "step": 3800,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.008153311582282186,
      "weighted_relation_loss": 0.008153311582282186
    },
    {
      "action_query_norm": 407.10276794433594,
      "action_relation_loss": 0.00710004628635943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5640785781103834,
      "flow_matching_loss": 0.06443818248808383,
      "grad_norm": 0.1927213817834854,
      "learning_rate": 9.838916122541603e-05,
      "loss": 0.0726,
      "relation_loss": 0.00811538458801806,
      "selected_pairs": 27129.3,
      "state_query_norm": 717.51767578125,
      "state_relation_loss": 0.0010153383191209286,
      "step": 3810,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.00811538458801806,
      "weighted_relation_loss": 0.00811538458801806
    },
    {
      "action_query_norm": 391.95400085449216,
      "action_relation_loss": 0.007305774278938771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5734331150608045,
      "flow_matching_loss": 0.06291351765394211,
      "grad_norm": 0.16460593044757843,
      "learning_rate": 9.837525453904246e-05,
      "loss": 0.0712,
      "relation_loss": 0.008317936304956674,
      "selected_pairs": 26584.5,
      "state_query_norm": 711.4651916503906,
      "state_relation_loss": 0.0010121620900463313,
      "step": 3820,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.008317936304956674,
      "weighted_relation_loss": 0.008317936304956674
    },
    {
      "action_query_norm": 397.83263244628904,
      "action_relation_loss": 0.00712260757572949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.582787652011225,
      "flow_matching_loss": 0.06310501247644425,
      "grad_norm": 0.15409880876541138,
      "learning_rate": 9.836128907217314e-05,
      "loss": 0.0712,
      "relation_loss": 0.008121377369388939,
      "selected_pairs": 26873.1,
      "state_query_norm": 707.8771362304688,
      "state_relation_loss": 0.000998769816942513,
      "step": 3830,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.008121377369388939,
      "weighted_relation_loss": 0.008121377369388939
    },
    {
      "action_query_norm": 398.1146301269531,
      "action_relation_loss": 0.007055708253756165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5921421889616463,
      "flow_matching_loss": 0.06169956661760807,
      "grad_norm": 0.22117817401885986,
      "learning_rate": 9.834726484177743e-05,
      "loss": 0.0698,
      "relation_loss": 0.008073999686166644,
      "selected_pairs": 27064.6,
      "state_query_norm": 713.1999755859375,
      "state_relation_loss": 0.0010182914615143091,
      "step": 3840,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.008073999686166644,
      "weighted_relation_loss": 0.008073999686166644
    },
    {
      "action_query_norm": 404.90271911621096,
      "action_relation_loss": 0.0072540690191090105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6014967259120674,
      "flow_matching_loss": 0.067806226760149,
      "grad_norm": 0.2033134549856186,
      "learning_rate": 9.833318186489609e-05,
      "loss": 0.0761,
      "relation_loss": 0.00826564566232264,
      "selected_pairs": 26455.8,
      "state_query_norm": 719.3313659667969,
      "state_relation_loss": 0.0010115767014212905,
      "step": 3850,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.00826564566232264,
      "weighted_relation_loss": 0.00826564566232264
    },
    {
      "action_query_norm": 410.69834289550784,
      "action_relation_loss": 0.007243971526622772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6108512628624885,
      "flow_matching_loss": 0.06179692298173904,
      "grad_norm": 0.1419924944639206,
      "learning_rate": 9.831904015864126e-05,
      "loss": 0.07,
      "relation_loss": 0.008226971188560129,
      "selected_pairs": 27270.3,
      "state_query_norm": 719.4438781738281,
      "state_relation_loss": 0.000982999720145017,
      "step": 3860,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.008226971188560129,
      "weighted_relation_loss": 0.008226971188560129
    },
    {
      "action_query_norm": 406.06895446777344,
      "action_relation_loss": 0.0074370648246258496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.620205799812909,
      "flow_matching_loss": 0.06166799776256084,
      "grad_norm": 0.1716959923505783,
      "learning_rate": 9.830483974019645e-05,
      "loss": 0.07,
      "relation_loss": 0.00837714271619916,
      "selected_pairs": 26848.1,
      "state_query_norm": 735.7979187011719,
      "state_relation_loss": 0.000940077897394076,
      "step": 3870,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.00837714271619916,
      "weighted_relation_loss": 0.00837714271619916
    },
    {
      "action_query_norm": 410.30523986816405,
      "action_relation_loss": 0.006885201530531048,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6295603367633302,
      "flow_matching_loss": 0.06470456719398499,
      "grad_norm": 0.17699843645095825,
      "learning_rate": 9.82905806268165e-05,
      "loss": 0.0726,
      "relation_loss": 0.007857287535443902,
      "selected_pairs": 27089.5,
      "state_query_norm": 722.5205810546875,
      "state_relation_loss": 0.0009720860223751515,
      "step": 3880,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.007857287535443902,
      "weighted_relation_loss": 0.007857287535443902
    },
    {
      "action_query_norm": 406.5542419433594,
      "action_relation_loss": 0.007294748071581125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6389148737137513,
      "flow_matching_loss": 0.06561991274356842,
      "grad_norm": 0.2048773467540741,
      "learning_rate": 9.82762628358276e-05,
      "loss": 0.0739,
      "relation_loss": 0.008257572120055557,
      "selected_pairs": 27197.4,
      "state_query_norm": 732.016259765625,
      "state_relation_loss": 0.0009628240019083023,
      "step": 3890,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.008257572120055557,
      "weighted_relation_loss": 0.008257572120055557
    },
    {
      "action_query_norm": 407.2564727783203,
      "action_relation_loss": 0.006910635391250253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.648269410664172,
      "flow_matching_loss": 0.06458414904773235,
      "grad_norm": 0.14978767931461334,
      "learning_rate": 9.826188638462718e-05,
      "loss": 0.0725,
      "relation_loss": 0.007889380073174834,
      "selected_pairs": 27010.7,
      "state_query_norm": 717.4031188964843,
      "state_relation_loss": 0.000978744646999985,
      "step": 3900,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.007889380073174834,
      "weighted_relation_loss": 0.007889380073174834
    },
    {
      "action_query_norm": 403.2829193115234,
      "action_relation_loss": 0.00704929674975574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.657623947614593,
      "flow_matching_loss": 0.06334667392075062,
      "grad_norm": 0.2295282781124115,
      "learning_rate": 9.824745129068402e-05,
      "loss": 0.0714,
      "relation_loss": 0.008064008504152297,
      "selected_pairs": 27371.1,
      "state_query_norm": 715.3221130371094,
      "state_relation_loss": 0.0010147117835003882,
      "step": 3910,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.008064008504152297,
      "weighted_relation_loss": 0.008064008504152297
    },
    {
      "action_query_norm": 399.61887512207034,
      "action_relation_loss": 0.006781174149364233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.666978484565014,
      "flow_matching_loss": 0.060763391107320784,
      "grad_norm": 0.18646010756492615,
      "learning_rate": 9.82329575715381e-05,
      "loss": 0.0686,
      "relation_loss": 0.007808456430211663,
      "selected_pairs": 27181.1,
      "state_query_norm": 715.5792663574218,
      "state_relation_loss": 0.0010272822750266642,
      "step": 3920,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.007808456430211663,
      "weighted_relation_loss": 0.007808456430211663
    },
    {
      "action_query_norm": 398.89281311035154,
      "action_relation_loss": 0.006974774086847902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.676333021515435,
      "flow_matching_loss": 0.06342085972428321,
      "grad_norm": 0.21648405492305756,
      "learning_rate": 9.821840524480066e-05,
      "loss": 0.0714,
      "relation_loss": 0.00797724504955113,
      "selected_pairs": 26801.4,
      "state_query_norm": 723.5109313964844,
      "state_relation_loss": 0.0010024711198639124,
      "step": 3930,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.00797724504955113,
      "weighted_relation_loss": 0.00797724504955113
    },
    {
      "action_query_norm": 401.4898284912109,
      "action_relation_loss": 0.007001541554927826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.685687558465856,
      "flow_matching_loss": 0.061547928303480146,
      "grad_norm": 0.1677580177783966,
      "learning_rate": 9.820379432815414e-05,
      "loss": 0.0696,
      "relation_loss": 0.008006267994642258,
      "selected_pairs": 26735.2,
      "state_query_norm": 735.6474487304688,
      "state_relation_loss": 0.00100472632329911,
      "step": 3940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.008006267994642258,
      "weighted_relation_loss": 0.008006267994642258
    },
    {
      "action_query_norm": 411.0325500488281,
      "action_relation_loss": 0.007185287494212389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.695042095416277,
      "flow_matching_loss": 0.0653906013816595,
      "grad_norm": 0.15148232877254486,
      "learning_rate": 9.81891248393522e-05,
      "loss": 0.0735,
      "relation_loss": 0.008145091962069273,
      "selected_pairs": 26802.2,
      "state_query_norm": 743.4899169921875,
      "state_relation_loss": 0.0009598043572623283,
      "step": 3950,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.008145091962069273,
      "weighted_relation_loss": 0.008145091962069273
    },
    {
      "action_query_norm": 402.0284057617188,
      "action_relation_loss": 0.006799586955457926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7043966323666977,
      "flow_matching_loss": 0.06263516582548619,
      "grad_norm": 0.17495912313461304,
      "learning_rate": 9.817439679621963e-05,
      "loss": 0.0704,
      "relation_loss": 0.00776052507571876,
      "selected_pairs": 27225.0,
      "state_query_norm": 728.9520751953125,
      "state_relation_loss": 0.0009609381959307939,
      "step": 3960,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00776052507571876,
      "weighted_relation_loss": 0.00776052507571876
    },
    {
      "action_query_norm": 405.2349792480469,
      "action_relation_loss": 0.006886648014187813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.713751169317119,
      "flow_matching_loss": 0.06541098430752754,
      "grad_norm": 0.18629275262355804,
      "learning_rate": 9.815961021665243e-05,
      "loss": 0.0732,
      "relation_loss": 0.007822630880400538,
      "selected_pairs": 26802.7,
      "state_query_norm": 728.8755493164062,
      "state_relation_loss": 0.0009359827905427665,
      "step": 3970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007822630880400538,
      "weighted_relation_loss": 0.007822630880400538
    },
    {
      "action_query_norm": 410.6949096679688,
      "action_relation_loss": 0.006831803452223539,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.72310570626754,
      "flow_matching_loss": 0.06135713681578636,
      "grad_norm": 0.14856980741024017,
      "learning_rate": 9.814476511861763e-05,
      "loss": 0.0692,
      "relation_loss": 0.00781518965959549,
      "selected_pairs": 26387.8,
      "state_query_norm": 730.091845703125,
      "state_relation_loss": 0.0009833861549850553,
      "step": 3980,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.00781518965959549,
      "weighted_relation_loss": 0.00781518965959549
    },
    {
      "action_query_norm": 400.2174102783203,
      "action_relation_loss": 0.0069323906674981116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7324602432179605,
      "flow_matching_loss": 0.060787854716181755,
      "grad_norm": 0.18008841574192047,
      "learning_rate": 9.812986152015348e-05,
      "loss": 0.0687,
      "relation_loss": 0.007882720930501819,
      "selected_pairs": 26509.3,
      "state_query_norm": 738.5783874511719,
      "state_relation_loss": 0.0009503303794190287,
      "step": 3990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.007882720930501819,
      "weighted_relation_loss": 0.007882720930501819
    },
    {
      "action_query_norm": 402.3547393798828,
      "action_relation_loss": 0.006815038854256272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7418147801683816,
      "flow_matching_loss": 0.06327481046319008,
      "grad_norm": 0.20727945864200592,
      "learning_rate": 9.811489943936922e-05,
      "loss": 0.071,
      "relation_loss": 0.007755542639642954,
      "selected_pairs": 26943.1,
      "state_query_norm": 729.8058715820313,
      "state_relation_loss": 0.000940503814490512,
      "step": 4000,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007755542639642954,
      "weighted_relation_loss": 0.007755542639642954
    },
    {
      "action_query_norm": 407.2535888671875,
      "action_relation_loss": 0.006866884976625442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7511693171188027,
      "flow_matching_loss": 0.06152021512389183,
      "grad_norm": 0.16843678057193756,
      "learning_rate": 9.809987889444522e-05,
      "loss": 0.0694,
      "relation_loss": 0.007862733351066709,
      "selected_pairs": 27343.5,
      "state_query_norm": 716.6135925292969,
      "state_relation_loss": 0.0009958484442904592,
      "step": 4010,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.007862733351066709,
      "weighted_relation_loss": 0.007862733351066709
    },
    {
      "action_query_norm": 409.97719116210936,
      "action_relation_loss": 0.007056138943880796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7605238540692234,
      "flow_matching_loss": 0.06206820979714393,
      "grad_norm": 0.18004651367664337,
      "learning_rate": 9.808479990363282e-05,
      "loss": 0.0701,
      "relation_loss": 0.00802780305966735,
      "selected_pairs": 26659.5,
      "state_query_norm": 737.3279296875,
      "state_relation_loss": 0.0009716641041450202,
      "step": 4020,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.00802780305966735,
      "weighted_relation_loss": 0.00802780305966735
    },
    {
      "action_query_norm": 408.18443908691404,
      "action_relation_loss": 0.006817694567143917,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7698783910196445,
      "flow_matching_loss": 0.06351795420050621,
      "grad_norm": 0.22551605105400085,
      "learning_rate": 9.806966248525445e-05,
      "loss": 0.0713,
      "relation_loss": 0.007782294834032655,
      "selected_pairs": 26875.0,
      "state_query_norm": 735.6773193359375,
      "state_relation_loss": 0.0009646002901718021,
      "step": 4030,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.007782294834032655,
      "weighted_relation_loss": 0.007782294834032655
    },
    {
      "action_query_norm": 405.4850158691406,
      "action_relation_loss": 0.006706809718161821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7792329279700656,
      "flow_matching_loss": 0.06174325905740261,
      "grad_norm": 0.15731172263622284,
      "learning_rate": 9.805446665770348e-05,
      "loss": 0.0695,
      "relation_loss": 0.0077083644457161425,
      "selected_pairs": 26783.4,
      "state_query_norm": 723.7242431640625,
      "state_relation_loss": 0.0010015547857619822,
      "step": 4040,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0077083644457161425,
      "weighted_relation_loss": 0.0077083644457161425
    },
    {
      "action_query_norm": 413.14450988769534,
      "action_relation_loss": 0.006861191568896174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7885874649204867,
      "flow_matching_loss": 0.06314216703176498,
      "grad_norm": 0.14457368850708008,
      "learning_rate": 9.803921243944429e-05,
      "loss": 0.071,
      "relation_loss": 0.007844635332003237,
      "selected_pairs": 27195.7,
      "state_query_norm": 738.5393188476562,
      "state_relation_loss": 0.0009834437223616987,
      "step": 4050,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.007844635332003237,
      "weighted_relation_loss": 0.007844635332003237
    },
    {
      "action_query_norm": 419.11762084960935,
      "action_relation_loss": 0.00703148078173399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7979420018709074,
      "flow_matching_loss": 0.06353770270943641,
      "grad_norm": 0.15141558647155762,
      "learning_rate": 9.802389984901218e-05,
      "loss": 0.0715,
      "relation_loss": 0.007993189617991447,
      "selected_pairs": 26687.4,
      "state_query_norm": 742.4719177246094,
      "state_relation_loss": 0.0009617088246159256,
      "step": 4060,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007993189617991447,
      "weighted_relation_loss": 0.007993189617991447
    },
    {
      "action_query_norm": 409.07545776367186,
      "action_relation_loss": 0.006687619211152196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8072965388213285,
      "flow_matching_loss": 0.06306162849068642,
      "grad_norm": 0.17146161198616028,
      "learning_rate": 9.80085289050134e-05,
      "loss": 0.0707,
      "relation_loss": 0.007625689217820764,
      "selected_pairs": 26848.1,
      "state_query_norm": 735.4686767578125,
      "state_relation_loss": 0.0009380699950270355,
      "step": 4070,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.007625689217820764,
      "weighted_relation_loss": 0.007625689217820764
    },
    {
      "action_query_norm": 404.391796875,
      "action_relation_loss": 0.007080101082101465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.816651075771749,
      "flow_matching_loss": 0.06393289640545845,
      "grad_norm": 0.1385727822780609,
      "learning_rate": 9.799309962612508e-05,
      "loss": 0.072,
      "relation_loss": 0.0080667810048908,
      "selected_pairs": 27242.0,
      "state_query_norm": 724.6154235839844,
      "state_relation_loss": 0.0009866799809969962,
      "step": 4080,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0080667810048908,
      "weighted_relation_loss": 0.0080667810048908
    },
    {
      "action_query_norm": 416.5584747314453,
      "action_relation_loss": 0.006878374749794603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.82600561272217,
      "flow_matching_loss": 0.06065562851727009,
      "grad_norm": 0.15363693237304688,
      "learning_rate": 9.797761203109527e-05,
      "loss": 0.0684,
      "relation_loss": 0.007780324947088957,
      "selected_pairs": 26869.9,
      "state_query_norm": 739.1142150878907,
      "state_relation_loss": 0.0009019502380397171,
      "step": 4090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.007780324947088957,
      "weighted_relation_loss": 0.007780324947088957
    },
    {
      "action_query_norm": 412.0905395507813,
      "action_relation_loss": 0.006507491786032915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8353601496725913,
      "flow_matching_loss": 0.0601975005120039,
      "grad_norm": 0.1617167443037033,
      "learning_rate": 9.796206613874283e-05,
      "loss": 0.0676,
      "relation_loss": 0.007450906885787844,
      "selected_pairs": 27269.9,
      "state_query_norm": 732.6114440917969,
      "state_relation_loss": 0.00094341512885876,
      "step": 4100,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.007450906885787844,
      "weighted_relation_loss": 0.007450906885787844
    },
    {
      "action_query_norm": 405.776806640625,
      "action_relation_loss": 0.007495176326483488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8447146866230124,
      "flow_matching_loss": 0.06122880764305592,
      "grad_norm": 0.16708283126354218,
      "learning_rate": 9.794646196795754e-05,
      "loss": 0.0697,
      "relation_loss": 0.008442451898008584,
      "selected_pairs": 27082.6,
      "state_query_norm": 733.6357482910156,
      "state_relation_loss": 0.0009472756122704595,
      "step": 4110,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.008442451898008584,
      "weighted_relation_loss": 0.008442451898008584
    },
    {
      "action_query_norm": 405.8884307861328,
      "action_relation_loss": 0.00691370964050293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.854069223573433,
      "flow_matching_loss": 0.061759402975440025,
      "grad_norm": 0.16660064458847046,
      "learning_rate": 9.793079953769987e-05,
      "loss": 0.0696,
      "relation_loss": 0.007842263253405691,
      "selected_pairs": 26350.1,
      "state_query_norm": 725.8053588867188,
      "state_relation_loss": 0.0009285535255912691,
      "step": 4120,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.007842263253405691,
      "weighted_relation_loss": 0.007842263253405691
    },
    {
      "action_query_norm": 410.07146606445315,
      "action_relation_loss": 0.006846889760345221,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.863423760523854,
      "flow_matching_loss": 0.06410034708678722,
      "grad_norm": 0.13482019305229187,
      "learning_rate": 9.79150788670012e-05,
      "loss": 0.0718,
      "relation_loss": 0.007742182072252035,
      "selected_pairs": 26756.2,
      "state_query_norm": 747.5435424804688,
      "state_relation_loss": 0.0008952923410106451,
      "step": 4130,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.007742182072252035,
      "weighted_relation_loss": 0.007742182072252035
    },
    {
      "action_query_norm": 411.2005615234375,
      "action_relation_loss": 0.007231373246759177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.872778297474275,
      "flow_matching_loss": 0.060393215715885164,
      "grad_norm": 0.17092566192150116,
      "learning_rate": 9.78992999749636e-05,
      "loss": 0.0686,
      "relation_loss": 0.008171527413651347,
      "selected_pairs": 27103.1,
      "state_query_norm": 746.1945373535157,
      "state_relation_loss": 0.0009401542192790657,
      "step": 4140,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.008171527413651347,
      "weighted_relation_loss": 0.008171527413651347
    },
    {
      "action_query_norm": 411.0408416748047,
      "action_relation_loss": 0.006560101872310042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.882132834424696,
      "flow_matching_loss": 0.06120159924030304,
      "grad_norm": 0.16510392725467682,
      "learning_rate": 9.788346288075994e-05,
      "loss": 0.0687,
      "relation_loss": 0.007485431712120771,
      "selected_pairs": 26664.0,
      "state_query_norm": 747.1611145019531,
      "state_relation_loss": 0.0009253297816030681,
      "step": 4150,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.007485431712120771,
      "weighted_relation_loss": 0.007485431712120771
    },
    {
      "action_query_norm": 410.5645263671875,
      "action_relation_loss": 0.007185590080916881,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.891487371375117,
      "flow_matching_loss": 0.06385989934206009,
      "grad_norm": 0.13075274229049683,
      "learning_rate": 9.786756760363373e-05,
      "loss": 0.072,
      "relation_loss": 0.008146026032045483,
      "selected_pairs": 26664.3,
      "state_query_norm": 750.06416015625,
      "state_relation_loss": 0.0009604359918739646,
      "step": 4160,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.008146026032045483,
      "weighted_relation_loss": 0.008146026032045483
    },
    {
      "action_query_norm": 408.3644744873047,
      "action_relation_loss": 0.006261381041258574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.900841908325538,
      "flow_matching_loss": 0.06079896055161953,
      "grad_norm": 0.18632850050926208,
      "learning_rate": 9.78516141628993e-05,
      "loss": 0.068,
      "relation_loss": 0.007193627534434199,
      "selected_pairs": 26811.7,
      "state_query_norm": 751.864111328125,
      "state_relation_loss": 0.0009322465339209884,
      "step": 4170,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007193627534434199,
      "weighted_relation_loss": 0.007193627534434199
    },
    {
      "action_query_norm": 406.14104614257815,
      "action_relation_loss": 0.006654303800314665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9101964452759588,
      "flow_matching_loss": 0.06200545206665993,
      "grad_norm": 0.18757519125938416,
      "learning_rate": 9.783560257794154e-05,
      "loss": 0.0696,
      "relation_loss": 0.007575698802247644,
      "selected_pairs": 26062.4,
      "state_query_norm": 739.2834045410157,
      "state_relation_loss": 0.0009213951241690665,
      "step": 4180,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.007575698802247644,
      "weighted_relation_loss": 0.007575698802247644
    },
    {
      "action_query_norm": 407.58290100097656,
      "action_relation_loss": 0.006512936344370246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.91955098222638,
      "flow_matching_loss": 0.06398464441299438,
      "grad_norm": 0.2224249243736267,
      "learning_rate": 9.781953286821603e-05,
      "loss": 0.0714,
      "relation_loss": 0.007452275138348341,
      "selected_pairs": 26019.4,
      "state_query_norm": 744.78447265625,
      "state_relation_loss": 0.0009393387997988611,
      "step": 4190,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.007452275138348341,
      "weighted_relation_loss": 0.007452275138348341
    },
    {
      "action_query_norm": 414.55419921875,
      "action_relation_loss": 0.00653504990041256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9289055191768005,
      "flow_matching_loss": 0.06097790710628033,
      "grad_norm": 0.20392285287380219,
      "learning_rate": 9.780340505324901e-05,
      "loss": 0.0684,
      "relation_loss": 0.0074529294855892655,
      "selected_pairs": 27131.4,
      "state_query_norm": 752.3321105957032,
      "state_relation_loss": 0.0009178795851767063,
      "step": 4200,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0074529294855892655,
      "weighted_relation_loss": 0.0074529294855892655
    },
    {
      "action_query_norm": 409.19182739257815,
      "action_relation_loss": 0.006534115457907319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9382600561272216,
      "flow_matching_loss": 0.06502302959561349,
      "grad_norm": 0.1584748476743698,
      "learning_rate": 9.778721915263727e-05,
      "loss": 0.0725,
      "relation_loss": 0.007481257384642959,
      "selected_pairs": 26735.3,
      "state_query_norm": 740.4098083496094,
      "state_relation_loss": 0.0009471418044995517,
      "step": 4210,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.007481257384642959,
      "weighted_relation_loss": 0.007481257384642959
    },
    {
      "action_query_norm": 413.0910278320313,
      "action_relation_loss": 0.006323695089668036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9476145930776427,
      "flow_matching_loss": 0.06011633910238743,
      "grad_norm": 0.23171301186084747,
      "learning_rate": 9.777097518604824e-05,
      "loss": 0.0674,
      "relation_loss": 0.007259923173114658,
      "selected_pairs": 26348.0,
      "state_query_norm": 745.3683288574218,
      "state_relation_loss": 0.0009362280427012593,
      "step": 4220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.007259923173114658,
      "weighted_relation_loss": 0.007259923173114658
    },
    {
      "action_query_norm": 410.3033020019531,
      "action_relation_loss": 0.006490144040435552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.956969130028064,
      "flow_matching_loss": 0.06054063588380813,
      "grad_norm": 0.14603407680988312,
      "learning_rate": 9.775467317321984e-05,
      "loss": 0.0679,
      "relation_loss": 0.00735287768766284,
      "selected_pairs": 26295.2,
      "state_query_norm": 756.0780822753907,
      "state_relation_loss": 0.0008627336646895856,
      "step": 4230,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.00735287768766284,
      "weighted_relation_loss": 0.00735287768766284
    },
    {
      "action_query_norm": 414.24570922851564,
      "action_relation_loss": 0.006921310862526298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9663236669784845,
      "flow_matching_loss": 0.05921631269156933,
      "grad_norm": 0.1581653505563736,
      "learning_rate": 9.773831313396055e-05,
      "loss": 0.067,
      "relation_loss": 0.007830829126760364,
      "selected_pairs": 26737.3,
      "state_query_norm": 754.7788024902344,
      "state_relation_loss": 0.0009095183049794286,
      "step": 4240,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.007830829126760364,
      "weighted_relation_loss": 0.007830829126760364
    },
    {
      "action_query_norm": 415.1636444091797,
      "action_relation_loss": 0.006644904287531972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9756782039289056,
      "flow_matching_loss": 0.061271768808364865,
      "grad_norm": 0.21688716113567352,
      "learning_rate": 9.77218950881494e-05,
      "loss": 0.0688,
      "relation_loss": 0.007558651268482208,
      "selected_pairs": 26757.6,
      "state_query_norm": 770.8243347167969,
      "state_relation_loss": 0.0009137470275163651,
      "step": 4250,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.007558651268482208,
      "weighted_relation_loss": 0.007558651268482208
    },
    {
      "action_query_norm": 409.7297607421875,
      "action_relation_loss": 0.006746068969368935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9850327408793262,
      "flow_matching_loss": 0.061087895557284354,
      "grad_norm": 0.19110378623008728,
      "learning_rate": 9.770541905573583e-05,
      "loss": 0.0688,
      "relation_loss": 0.00766938142478466,
      "selected_pairs": 27221.4,
      "state_query_norm": 751.5141479492188,
      "state_relation_loss": 0.0009233125194441528,
      "step": 4260,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00766938142478466,
      "weighted_relation_loss": 0.00766938142478466
    },
    {
      "action_query_norm": 414.4842254638672,
      "action_relation_loss": 0.006687554437667132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9943872778297473,
      "flow_matching_loss": 0.061255845800042154,
      "grad_norm": 0.14876070618629456,
      "learning_rate": 9.768888505673976e-05,
      "loss": 0.0689,
      "relation_loss": 0.007599745923653245,
      "selected_pairs": 26926.1,
      "state_query_norm": 762.1838500976562,
      "state_relation_loss": 0.0009121914044953883,
      "step": 4270,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.007599745923653245,
      "weighted_relation_loss": 0.007599745923653245
    },
    {
      "action_query_norm": 418.52330322265624,
      "action_relation_loss": 0.006374837271869183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.003741814780168,
      "flow_matching_loss": 0.0591424509882927,
      "grad_norm": 0.1296294629573822,
      "learning_rate": 9.767229311125162e-05,
      "loss": 0.0664,
      "relation_loss": 0.00729694040492177,
      "selected_pairs": 24219.5,
      "state_query_norm": 753.5568359375,
      "state_relation_loss": 0.0009221030923072249,
      "step": 4280,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.00729694040492177,
      "weighted_relation_loss": 0.00729694040492177
    },
    {
      "action_query_norm": 411.7259765625,
      "action_relation_loss": 0.006339479563757777,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.0130963517305895,
      "flow_matching_loss": 0.05698495097458363,
      "grad_norm": 0.18128591775894165,
      "learning_rate": 9.765564323943211e-05,
      "loss": 0.0642,
      "relation_loss": 0.0072376866359263655,
      "selected_pairs": 27033.7,
      "state_query_norm": 751.4887451171875,
      "state_relation_loss": 0.0008982070139609277,
      "step": 4290,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0072376866359263655,
      "weighted_relation_loss": 0.0072376866359263655
    },
    {
      "action_query_norm": 416.0613098144531,
      "action_relation_loss": 0.00623908881098032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.022450888681011,
      "flow_matching_loss": 0.05887463688850403,
      "grad_norm": 0.22091516852378845,
      "learning_rate": 9.763893546151244e-05,
      "loss": 0.066,
      "relation_loss": 0.007134787971153855,
      "selected_pairs": 27106.7,
      "state_query_norm": 754.7933776855468,
      "state_relation_loss": 0.0008956991718150675,
      "step": 4300,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.007134787971153855,
      "weighted_relation_loss": 0.007134787971153855
    },
    {
      "action_query_norm": 411.88306579589846,
      "action_relation_loss": 0.006256158975884318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.031805425631431,
      "flow_matching_loss": 0.061232464388012886,
      "grad_norm": 0.1954587697982788,
      "learning_rate": 9.762216979779412e-05,
      "loss": 0.0684,
      "relation_loss": 0.007161057973280549,
      "selected_pairs": 26526.0,
      "state_query_norm": 767.1010864257812,
      "state_relation_loss": 0.0009048990847077221,
      "step": 4310,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.007161057973280549,
      "weighted_relation_loss": 0.007161057973280549
    },
    {
      "action_query_norm": 415.0070861816406,
      "action_relation_loss": 0.006307155964896083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.041159962581852,
      "flow_matching_loss": 0.05827256441116333,
      "grad_norm": 0.15484343469142914,
      "learning_rate": 9.760534626864902e-05,
      "loss": 0.0655,
      "relation_loss": 0.007242387812584639,
      "selected_pairs": 26904.8,
      "state_query_norm": 755.9444946289062,
      "state_relation_loss": 0.0009352318535093218,
      "step": 4320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007242387812584639,
      "weighted_relation_loss": 0.007242387812584639
    },
    {
      "action_query_norm": 409.5476440429687,
      "action_relation_loss": 0.006428384687751532,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.050514499532273,
      "flow_matching_loss": 0.06252833344042301,
      "grad_norm": 0.15658441185951233,
      "learning_rate": 9.758846489451931e-05,
      "loss": 0.0699,
      "relation_loss": 0.007329765008762479,
      "selected_pairs": 27129.6,
      "state_query_norm": 753.6307434082031,
      "state_relation_loss": 0.0009013803501147777,
      "step": 4330,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.007329765008762479,
      "weighted_relation_loss": 0.007329765008762479
    },
    {
      "action_query_norm": 415.3850982666016,
      "action_relation_loss": 0.006219606194645166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.059869036482694,
      "flow_matching_loss": 0.059351757168769836,
      "grad_norm": 0.17559921741485596,
      "learning_rate": 9.757152569591748e-05,
      "loss": 0.0665,
      "relation_loss": 0.007142071146517992,
      "selected_pairs": 26892.2,
      "state_query_norm": 762.2675964355469,
      "state_relation_loss": 0.0009224649518728256,
      "step": 4340,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007142071146517992,
      "weighted_relation_loss": 0.007142071146517992
    },
    {
      "action_query_norm": 421.23060302734376,
      "action_relation_loss": 0.006392243131995201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.069223573433115,
      "flow_matching_loss": 0.060314896330237386,
      "grad_norm": 0.20080304145812988,
      "learning_rate": 9.75545286934262e-05,
      "loss": 0.0676,
      "relation_loss": 0.007306073233485222,
      "selected_pairs": 26662.3,
      "state_query_norm": 765.4833679199219,
      "state_relation_loss": 0.0009138300549238921,
      "step": 4350,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.007306073233485222,
      "weighted_relation_loss": 0.007306073233485222
    },
    {
      "action_query_norm": 421.1108673095703,
      "action_relation_loss": 0.006688068015500903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.078578110383536,
      "flow_matching_loss": 0.060809670761227605,
      "grad_norm": 0.17574094235897064,
      "learning_rate": 9.753747390769847e-05,
      "loss": 0.0684,
      "relation_loss": 0.007592476904392242,
      "selected_pairs": 26919.9,
      "state_query_norm": 752.8660522460938,
      "state_relation_loss": 0.0009044088947121054,
      "step": 4360,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.007592476904392242,
      "weighted_relation_loss": 0.007592476904392242
    },
    {
      "action_query_norm": 422.0478088378906,
      "action_relation_loss": 0.006212955852970481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.087932647333957,
      "flow_matching_loss": 0.05761498361825943,
      "grad_norm": 0.14438508450984955,
      "learning_rate": 9.752036135945744e-05,
      "loss": 0.0647,
      "relation_loss": 0.007095285505056381,
      "selected_pairs": 26532.4,
      "state_query_norm": 752.9587768554687,
      "state_relation_loss": 0.0008823296113405377,
      "step": 4370,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.007095285505056381,
      "weighted_relation_loss": 0.007095285505056381
    },
    {
      "action_query_norm": 430.4370635986328,
      "action_relation_loss": 0.0064035261049866675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.097287184284378,
      "flow_matching_loss": 0.058077339082956314,
      "grad_norm": 0.18849828839302063,
      "learning_rate": 9.750319106949649e-05,
      "loss": 0.0653,
      "relation_loss": 0.007266627065837384,
      "selected_pairs": 26993.3,
      "state_query_norm": 754.3430480957031,
      "state_relation_loss": 0.000863100914284587,
      "step": 4380,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.007266627065837384,
      "weighted_relation_loss": 0.007266627065837384
    },
    {
      "action_query_norm": 428.0318664550781,
      "action_relation_loss": 0.006691429717466235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.106641721234799,
      "flow_matching_loss": 0.06205615922808647,
      "grad_norm": 0.15661261975765228,
      "learning_rate": 9.748596305867913e-05,
      "loss": 0.0697,
      "relation_loss": 0.007616225769743324,
      "selected_pairs": 27151.0,
      "state_query_norm": 758.1889587402344,
      "state_relation_loss": 0.0009247960755601525,
      "step": 4390,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.007616225769743324,
      "weighted_relation_loss": 0.007616225769743324
    },
    {
      "action_query_norm": 418.5235321044922,
      "action_relation_loss": 0.006284131482243538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.11599625818522,
      "flow_matching_loss": 0.061679143086075784,
      "grad_norm": 0.15457957983016968,
      "learning_rate": 9.746867734793903e-05,
      "loss": 0.0689,
      "relation_loss": 0.007190887816250324,
      "selected_pairs": 26302.5,
      "state_query_norm": 759.6861328125,
      "state_relation_loss": 0.0009067562932614237,
      "step": 4400,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.007190887816250324,
      "weighted_relation_loss": 0.007190887816250324
    },
    {
      "action_query_norm": 425.5254852294922,
      "action_relation_loss": 0.006317879958078265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.125350795135641,
      "flow_matching_loss": 0.06065329201519489,
      "grad_norm": 0.13777896761894226,
      "learning_rate": 9.745133395827993e-05,
      "loss": 0.0679,
      "relation_loss": 0.0072233402635902165,
      "selected_pairs": 27102.3,
      "state_query_norm": 765.0242553710938,
      "state_relation_loss": 0.000905460276408121,
      "step": 4410,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0072233402635902165,
      "weighted_relation_loss": 0.0072233402635902165
    },
    {
      "action_query_norm": 427.06199645996094,
      "action_relation_loss": 0.007000537961721421,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.134705332086062,
      "flow_matching_loss": 0.059606447070837024,
      "grad_norm": 0.1487598717212677,
      "learning_rate": 9.743393291077572e-05,
      "loss": 0.0675,
      "relation_loss": 0.007887809071689844,
      "selected_pairs": 26930.6,
      "state_query_norm": 759.1626525878906,
      "state_relation_loss": 0.0008872711681760847,
      "step": 4420,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.007887809071689844,
      "weighted_relation_loss": 0.007887809071689844
    },
    {
      "action_query_norm": 429.78684692382814,
      "action_relation_loss": 0.006222377810627222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.144059869036482,
      "flow_matching_loss": 0.06029266230762005,
      "grad_norm": 0.15874136984348297,
      "learning_rate": 9.741647422657028e-05,
      "loss": 0.0674,
      "relation_loss": 0.007138583343476057,
      "selected_pairs": 26530.8,
      "state_query_norm": 751.4119079589843,
      "state_relation_loss": 0.0009162054513581098,
      "step": 4430,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.007138583343476057,
      "weighted_relation_loss": 0.007138583343476057
    },
    {
      "action_query_norm": 426.2756591796875,
      "action_relation_loss": 0.006566833006218076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.153414405986903,
      "flow_matching_loss": 0.058605782315135004,
      "grad_norm": 0.19265146553516388,
      "learning_rate": 9.739895792687758e-05,
      "loss": 0.0661,
      "relation_loss": 0.007511289184913039,
      "selected_pairs": 26760.8,
      "state_query_norm": 755.2073486328125,
      "state_relation_loss": 0.0009444562660064548,
      "step": 4440,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.007511289184913039,
      "weighted_relation_loss": 0.007511289184913039
    },
    {
      "action_query_norm": 433.95623168945315,
      "action_relation_loss": 0.006237017503008246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.1627689429373245,
      "flow_matching_loss": 0.058323674649000165,
      "grad_norm": 0.1758352667093277,
      "learning_rate": 9.738138403298157e-05,
      "loss": 0.0655,
      "relation_loss": 0.007164891017600894,
      "selected_pairs": 26949.0,
      "state_query_norm": 760.7566589355469,
      "state_relation_loss": 0.0009278735611587763,
      "step": 4450,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007164891017600894,
      "weighted_relation_loss": 0.007164891017600894
    },
    {
      "action_query_norm": 427.6121551513672,
      "action_relation_loss": 0.005938132060691714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.172123479887746,
      "flow_matching_loss": 0.059673726186156274,
      "grad_norm": 0.1973683089017868,
      "learning_rate": 9.736375256623619e-05,
      "loss": 0.0665,
      "relation_loss": 0.006826874008402229,
      "selected_pairs": 27174.7,
      "state_query_norm": 759.7732543945312,
      "state_relation_loss": 0.0008887419593520462,
      "step": 4460,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006826874008402229,
      "weighted_relation_loss": 0.006826874008402229
    },
    {
      "action_query_norm": 422.86325988769534,
      "action_relation_loss": 0.00602517812512815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.181478016838167,
      "flow_matching_loss": 0.06031770296394825,
      "grad_norm": 0.14696769416332245,
      "learning_rate": 9.734606354806533e-05,
      "loss": 0.0672,
      "relation_loss": 0.00689725331030786,
      "selected_pairs": 26616.6,
      "state_query_norm": 757.6165405273438,
      "state_relation_loss": 0.0008720751910004764,
      "step": 4470,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00689725331030786,
      "weighted_relation_loss": 0.00689725331030786
    },
    {
      "action_query_norm": 421.1580474853516,
      "action_relation_loss": 0.006244421238079667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.190832553788588,
      "flow_matching_loss": 0.05932631827890873,
      "grad_norm": 0.19781723618507385,
      "learning_rate": 9.73283169999628e-05,
      "loss": 0.0664,
      "relation_loss": 0.007078635506331921,
      "selected_pairs": 26854.5,
      "state_query_norm": 750.9466369628906,
      "state_relation_loss": 0.0008342142158653587,
      "step": 4480,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.007078635506331921,
      "weighted_relation_loss": 0.007078635506331921
    },
    {
      "action_query_norm": 422.299609375,
      "action_relation_loss": 0.005985591700300575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.200187090739009,
      "flow_matching_loss": 0.058817650005221364,
      "grad_norm": 0.14454175531864166,
      "learning_rate": 9.731051294349238e-05,
      "loss": 0.0657,
      "relation_loss": 0.0068350061774253845,
      "selected_pairs": 26718.2,
      "state_query_norm": 759.1274169921875,
      "state_relation_loss": 0.0008494145004078746,
      "step": 4490,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0068350061774253845,
      "weighted_relation_loss": 0.0068350061774253845
    },
    {
      "action_query_norm": 427.99554138183595,
      "action_relation_loss": 0.005966425593942404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.209541627689429,
      "flow_matching_loss": 0.057987839356064795,
      "grad_norm": 0.13559556007385254,
      "learning_rate": 9.729265140028762e-05,
      "loss": 0.0648,
      "relation_loss": 0.006815571244806052,
      "selected_pairs": 26734.2,
      "state_query_norm": 750.8484924316406,
      "state_relation_loss": 0.0008491456916090101,
      "step": 4500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006815571244806052,
      "weighted_relation_loss": 0.006815571244806052
    },
    {
      "action_query_norm": 433.2817321777344,
      "action_relation_loss": 0.005898884916678071,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.21889616463985,
      "flow_matching_loss": 0.05776314623653889,
      "grad_norm": 0.1465674787759781,
      "learning_rate": 9.727473239205201e-05,
      "loss": 0.0646,
      "relation_loss": 0.006810298515483737,
      "selected_pairs": 27220.2,
      "state_query_norm": 759.3407836914063,
      "state_relation_loss": 0.0009114135813433677,
      "step": 4510,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.006810298515483737,
      "weighted_relation_loss": 0.006810298515483737
    },
    {
      "action_query_norm": 428.44451293945315,
      "action_relation_loss": 0.006039456790313124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.228250701590271,
      "flow_matching_loss": 0.060725574195384976,
      "grad_norm": 0.14254318177700043,
      "learning_rate": 9.725675594055883e-05,
      "loss": 0.0676,
      "relation_loss": 0.006916133686900139,
      "selected_pairs": 26849.1,
      "state_query_norm": 753.6101318359375,
      "state_relation_loss": 0.0008766768674831838,
      "step": 4520,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006916133686900139,
      "weighted_relation_loss": 0.006916133686900139
    },
    {
      "action_query_norm": 427.2253112792969,
      "action_relation_loss": 0.0063586637377738954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.237605238540692,
      "flow_matching_loss": 0.05911871194839478,
      "grad_norm": 0.15641500055789948,
      "learning_rate": 9.723872206765116e-05,
      "loss": 0.0663,
      "relation_loss": 0.007206439645960927,
      "selected_pairs": 26843.8,
      "state_query_norm": 765.104736328125,
      "state_relation_loss": 0.0008477759722154588,
      "step": 4530,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.007206439645960927,
      "weighted_relation_loss": 0.007206439645960927
    },
    {
      "action_query_norm": 423.5145660400391,
      "action_relation_loss": 0.006200152868404985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.2469597754911135,
      "flow_matching_loss": 0.059545783698558806,
      "grad_norm": 0.1373068392276764,
      "learning_rate": 9.722063079524185e-05,
      "loss": 0.0666,
      "relation_loss": 0.007066063769161701,
      "selected_pairs": 26571.7,
      "state_query_norm": 767.6974853515625,
      "state_relation_loss": 0.0008659108832944184,
      "step": 4540,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.007066063769161701,
      "weighted_relation_loss": 0.007066063769161701
    },
    {
      "action_query_norm": 423.22803955078126,
      "action_relation_loss": 0.006494453875347972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.256314312441534,
      "flow_matching_loss": 0.059685487672686575,
      "grad_norm": 0.14807020127773285,
      "learning_rate": 9.720248214531351e-05,
      "loss": 0.067,
      "relation_loss": 0.007341995928436517,
      "selected_pairs": 27244.9,
      "state_query_norm": 760.6480529785156,
      "state_relation_loss": 0.0008475420065224171,
      "step": 4550,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.007341995928436517,
      "weighted_relation_loss": 0.007341995928436517
    },
    {
      "action_query_norm": 431.71409301757814,
      "action_relation_loss": 0.006328437710180879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.265668849391955,
      "flow_matching_loss": 0.05878040827810764,
      "grad_norm": 0.20983383059501648,
      "learning_rate": 9.718427613991848e-05,
      "loss": 0.066,
      "relation_loss": 0.007173926196992397,
      "selected_pairs": 26466.5,
      "state_query_norm": 761.6416687011719,
      "state_relation_loss": 0.0008454884926322847,
      "step": 4560,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.007173926196992397,
      "weighted_relation_loss": 0.007173926196992397
    },
    {
      "action_query_norm": 430.511962890625,
      "action_relation_loss": 0.005824685888364911,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.275023386342376,
      "flow_matching_loss": 0.06114187389612198,
      "grad_norm": 0.17346909642219543,
      "learning_rate": 9.716601280117873e-05,
      "loss": 0.0679,
      "relation_loss": 0.006724075879901648,
      "selected_pairs": 26393.0,
      "state_query_norm": 767.6717956542968,
      "state_relation_loss": 0.0008993899857159704,
      "step": 4570,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.006724075879901648,
      "weighted_relation_loss": 0.006724075879901648
    },
    {
      "action_query_norm": 428.3722778320313,
      "action_relation_loss": 0.006003778427839279,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.284377923292797,
      "flow_matching_loss": 0.059116951748728754,
      "grad_norm": 0.17775170505046844,
      "learning_rate": 9.714769215128596e-05,
      "loss": 0.066,
      "relation_loss": 0.0068792798556387424,
      "selected_pairs": 27292.4,
      "state_query_norm": 767.3436828613281,
      "state_relation_loss": 0.0008755013986956328,
      "step": 4580,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0068792798556387424,
      "weighted_relation_loss": 0.0068792798556387424
    },
    {
      "action_query_norm": 430.1119415283203,
      "action_relation_loss": 0.006087076989933849,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.293732460243218,
      "flow_matching_loss": 0.061618533730506894,
      "grad_norm": 0.21867156028747559,
      "learning_rate": 9.712931421250152e-05,
      "loss": 0.0686,
      "relation_loss": 0.006973867723718286,
      "selected_pairs": 26826.6,
      "state_query_norm": 772.6036499023437,
      "state_relation_loss": 0.0008867907745297998,
      "step": 4590,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006973867723718286,
      "weighted_relation_loss": 0.006973867723718286
    },
    {
      "action_query_norm": 421.1217041015625,
      "action_relation_loss": 0.005985646462067962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.303086997193639,
      "flow_matching_loss": 0.06048437356948853,
      "grad_norm": 0.17740345001220703,
      "learning_rate": 9.711087900715627e-05,
      "loss": 0.0673,
      "relation_loss": 0.006858866848051548,
      "selected_pairs": 27035.3,
      "state_query_norm": 766.477587890625,
      "state_relation_loss": 0.0008732204267289489,
      "step": 4600,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.006858866848051548,
      "weighted_relation_loss": 0.006858866848051548
    },
    {
      "action_query_norm": 427.84214782714844,
      "action_relation_loss": 0.00613219067454338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.31244153414406,
      "flow_matching_loss": 0.061556149274110794,
      "grad_norm": 0.18319247663021088,
      "learning_rate": 9.709238655765078e-05,
      "loss": 0.0686,
      "relation_loss": 0.007010318525135517,
      "selected_pairs": 27133.0,
      "state_query_norm": 776.798974609375,
      "state_relation_loss": 0.0008781278622336686,
      "step": 4610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.007010318525135517,
      "weighted_relation_loss": 0.007010318525135517
    },
    {
      "action_query_norm": 428.7943145751953,
      "action_relation_loss": 0.006523905461654067,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.3217960710944805,
      "flow_matching_loss": 0.0594113290309906,
      "grad_norm": 0.2754274606704712,
      "learning_rate": 9.707383688645511e-05,
      "loss": 0.0668,
      "relation_loss": 0.0074144316837191585,
      "selected_pairs": 27479.4,
      "state_query_norm": 787.3234436035157,
      "state_relation_loss": 0.0008905262395273895,
      "step": 4620,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0074144316837191585,
      "weighted_relation_loss": 0.0074144316837191585
    },
    {
      "action_query_norm": 431.3693481445313,
      "action_relation_loss": 0.006363493576645851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.331150608044902,
      "flow_matching_loss": 0.059362102672457694,
      "grad_norm": 0.1945480853319168,
      "learning_rate": 9.705523001610883e-05,
      "loss": 0.0666,
      "relation_loss": 0.007276571821421385,
      "selected_pairs": 27172.1,
      "state_query_norm": 778.6607849121094,
      "state_relation_loss": 0.0009130782855208963,
      "step": 4630,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.007276571821421385,
      "weighted_relation_loss": 0.007276571821421385
    },
    {
      "action_query_norm": 419.57672424316405,
      "action_relation_loss": 0.00629053832963109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.340505144995323,
      "flow_matching_loss": 0.0607470728456974,
      "grad_norm": 0.24294808506965637,
      "learning_rate": 9.703656596922107e-05,
      "loss": 0.0679,
      "relation_loss": 0.0071699026506394144,
      "selected_pairs": 26385.8,
      "state_query_norm": 784.3706115722656,
      "state_relation_loss": 0.0008793642977252602,
      "step": 4640,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0071699026506394144,
      "weighted_relation_loss": 0.0071699026506394144
    },
    {
      "action_query_norm": 430.29563293457034,
      "action_relation_loss": 0.006190485041588545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.349859681945744,
      "flow_matching_loss": 0.061448191106319425,
      "grad_norm": 0.19656306505203247,
      "learning_rate": 9.70178447684704e-05,
      "loss": 0.0685,
      "relation_loss": 0.007040499895811081,
      "selected_pairs": 26946.1,
      "state_query_norm": 787.6009094238282,
      "state_relation_loss": 0.0008500147610902786,
      "step": 4650,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007040499895811081,
      "weighted_relation_loss": 0.007040499895811081
    },
    {
      "action_query_norm": 427.8468322753906,
      "action_relation_loss": 0.006403185147792101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.359214218896165,
      "flow_matching_loss": 0.05945614166557789,
      "grad_norm": 0.18063503503799438,
      "learning_rate": 9.699906643660483e-05,
      "loss": 0.0667,
      "relation_loss": 0.007236656174063683,
      "selected_pairs": 26876.2,
      "state_query_norm": 764.1964477539062,
      "state_relation_loss": 0.0008334710262715816,
      "step": 4660,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.007236656174063683,
      "weighted_relation_loss": 0.007236656174063683
    },
    {
      "action_query_norm": 422.49717102050784,
      "action_relation_loss": 0.006083998829126358,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.368568755846586,
      "flow_matching_loss": 0.05702238865196705,
      "grad_norm": 0.12416362017393112,
      "learning_rate": 9.698023099644185e-05,
      "loss": 0.064,
      "relation_loss": 0.006950163329020142,
      "selected_pairs": 27174.4,
      "state_query_norm": 761.7341857910156,
      "state_relation_loss": 0.0008661645755637437,
      "step": 4670,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006950163329020142,
      "weighted_relation_loss": 0.006950163329020142
    },
    {
      "action_query_norm": 427.6350128173828,
      "action_relation_loss": 0.006038391776382923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.377923292797006,
      "flow_matching_loss": 0.05956118367612362,
      "grad_norm": 0.15618248283863068,
      "learning_rate": 9.696133847086823e-05,
      "loss": 0.0665,
      "relation_loss": 0.006889499677345157,
      "selected_pairs": 26758.8,
      "state_query_norm": 764.196044921875,
      "state_relation_loss": 0.000851107930066064,
      "step": 4680,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.006889499677345157,
      "weighted_relation_loss": 0.006889499677345157
    },
    {
      "action_query_norm": 427.98528442382815,
      "action_relation_loss": 0.00587990372441709,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.387277829747427,
      "flow_matching_loss": 0.05945146195590496,
      "grad_norm": 0.15553216636180878,
      "learning_rate": 9.694238888284022e-05,
      "loss": 0.0662,
      "relation_loss": 0.006729829963296652,
      "selected_pairs": 27203.0,
      "state_query_norm": 772.3720825195312,
      "state_relation_loss": 0.0008499262388795614,
      "step": 4690,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.006729829963296652,
      "weighted_relation_loss": 0.006729829963296652
    },
    {
      "action_query_norm": 430.1715057373047,
      "action_relation_loss": 0.005988501384854317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.396632366697848,
      "flow_matching_loss": 0.057611022517085074,
      "grad_norm": 0.14037556946277618,
      "learning_rate": 9.692338225538333e-05,
      "loss": 0.0645,
      "relation_loss": 0.0068553362507373095,
      "selected_pairs": 26989.5,
      "state_query_norm": 763.5241333007813,
      "state_relation_loss": 0.0008668349008075892,
      "step": 4700,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0068553362507373095,
      "weighted_relation_loss": 0.0068553362507373095
    },
    {
      "action_query_norm": 425.8192626953125,
      "action_relation_loss": 0.006346972752362489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.4059869036482695,
      "flow_matching_loss": 0.0581196416169405,
      "grad_norm": 0.15934205055236816,
      "learning_rate": 9.690431861159241e-05,
      "loss": 0.0653,
      "relation_loss": 0.007219541491940618,
      "selected_pairs": 26391.4,
      "state_query_norm": 765.6478698730468,
      "state_relation_loss": 0.0008725687279365957,
      "step": 4710,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.007219541491940618,
      "weighted_relation_loss": 0.007219541491940618
    },
    {
      "action_query_norm": 434.39781494140624,
      "action_relation_loss": 0.006174721149727702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.415341440598691,
      "flow_matching_loss": 0.06086670532822609,
      "grad_norm": 0.15966713428497314,
      "learning_rate": 9.688519797463161e-05,
      "loss": 0.0679,
      "relation_loss": 0.007039722520858049,
      "selected_pairs": 26615.1,
      "state_query_norm": 773.839697265625,
      "state_relation_loss": 0.0008650013711303473,
      "step": 4720,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.007039722520858049,
      "weighted_relation_loss": 0.007039722520858049
    },
    {
      "action_query_norm": 431.5443908691406,
      "action_relation_loss": 0.006066871108487248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.424695977549112,
      "flow_matching_loss": 0.05441422238945961,
      "grad_norm": 0.15899202227592468,
      "learning_rate": 9.686602036773426e-05,
      "loss": 0.0613,
      "relation_loss": 0.00693029910326004,
      "selected_pairs": 26551.4,
      "state_query_norm": 778.4251953125,
      "state_relation_loss": 0.0008634280180558562,
      "step": 4730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.00693029910326004,
      "weighted_relation_loss": 0.00693029910326004
    },
    {
      "action_query_norm": 433.8604370117188,
      "action_relation_loss": 0.006257356982678175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.434050514499532,
      "flow_matching_loss": 0.05814659185707569,
      "grad_norm": 0.16099463403224945,
      "learning_rate": 9.684678581420302e-05,
      "loss": 0.0652,
      "relation_loss": 0.007096384465694427,
      "selected_pairs": 26685.8,
      "state_query_norm": 778.8965576171875,
      "state_relation_loss": 0.0008390274364501238,
      "step": 4740,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007096384465694427,
      "weighted_relation_loss": 0.007096384465694427
    },
    {
      "action_query_norm": 437.0643035888672,
      "action_relation_loss": 0.0060970294289290905,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.443405051449953,
      "flow_matching_loss": 0.06005348265171051,
      "grad_norm": 0.15838061273097992,
      "learning_rate": 9.682749433740962e-05,
      "loss": 0.067,
      "relation_loss": 0.006956075690686703,
      "selected_pairs": 26479.2,
      "state_query_norm": 773.4578918457031,
      "state_relation_loss": 0.0008590462617576123,
      "step": 4750,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.006956075690686703,
      "weighted_relation_loss": 0.006956075690686703
    },
    {
      "action_query_norm": 430.719287109375,
      "action_relation_loss": 0.006353129586204887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.452759588400374,
      "flow_matching_loss": 0.05863657109439373,
      "grad_norm": 0.14369837939739227,
      "learning_rate": 9.680814596079507e-05,
      "loss": 0.0658,
      "relation_loss": 0.007187190977856517,
      "selected_pairs": 26325.6,
      "state_query_norm": 781.1455383300781,
      "state_relation_loss": 0.0008340613741893322,
      "step": 4760,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.007187190977856517,
      "weighted_relation_loss": 0.007187190977856517
    },
    {
      "action_query_norm": 432.2760772705078,
      "action_relation_loss": 0.006387614691630006,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.462114125350795,
      "flow_matching_loss": 0.06167535297572613,
      "grad_norm": 0.1987888365983963,
      "learning_rate": 9.678874070786945e-05,
      "loss": 0.0689,
      "relation_loss": 0.007219954440370202,
      "selected_pairs": 26525.3,
      "state_query_norm": 778.3137634277343,
      "state_relation_loss": 0.0008323398011270911,
      "step": 4770,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.007219954440370202,
      "weighted_relation_loss": 0.007219954440370202
    },
    {
      "action_query_norm": 433.5343353271484,
      "action_relation_loss": 0.006419890699908138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.471468662301216,
      "flow_matching_loss": 0.06461027786135673,
      "grad_norm": 0.2561970353126526,
      "learning_rate": 9.676927860221199e-05,
      "loss": 0.0719,
      "relation_loss": 0.007269630068913102,
      "selected_pairs": 26920.2,
      "state_query_norm": 778.4724243164062,
      "state_relation_loss": 0.0008497393690049648,
      "step": 4780,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.007269630068913102,
      "weighted_relation_loss": 0.007269630068913102
    },
    {
      "action_query_norm": 428.81357727050784,
      "action_relation_loss": 0.006402520416304469,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.480823199251637,
      "flow_matching_loss": 0.06269493848085403,
      "grad_norm": 0.20385801792144775,
      "learning_rate": 9.674975966747097e-05,
      "loss": 0.07,
      "relation_loss": 0.007338332198560238,
      "selected_pairs": 26915.8,
      "state_query_norm": 781.3493041992188,
      "state_relation_loss": 0.0009358117124065757,
      "step": 4790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.007338332198560238,
      "weighted_relation_loss": 0.007338332198560238
    },
    {
      "action_query_norm": 432.59336242675784,
      "action_relation_loss": 0.006389084318652749,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.490177736202058,
      "flow_matching_loss": 0.062408578023314475,
      "grad_norm": 0.18722742795944214,
      "learning_rate": 9.673018392736374e-05,
      "loss": 0.0697,
      "relation_loss": 0.007291840109974146,
      "selected_pairs": 26436.9,
      "state_query_norm": 786.2225280761719,
      "state_relation_loss": 0.0009027557796798646,
      "step": 4800,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.007291840109974146,
      "weighted_relation_loss": 0.007291840109974146
    },
    {
      "action_query_norm": 427.1769165039062,
      "action_relation_loss": 0.0061473171692341564,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.499532273152479,
      "flow_matching_loss": 0.05990623235702515,
      "grad_norm": 0.1496957391500473,
      "learning_rate": 9.671055140567667e-05,
      "loss": 0.067,
      "relation_loss": 0.007054650178179145,
      "selected_pairs": 27087.2,
      "state_query_norm": 776.9870239257813,
      "state_relation_loss": 0.0009073330380488187,
      "step": 4810,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.007054650178179145,
      "weighted_relation_loss": 0.007054650178179145
    },
    {
      "action_query_norm": 431.34327392578126,
      "action_relation_loss": 0.0061450975481420755,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.5088868101029,
      "flow_matching_loss": 0.060694117471575736,
      "grad_norm": 0.14410927891731262,
      "learning_rate": 9.669086212626511e-05,
      "loss": 0.0677,
      "relation_loss": 0.007043318171054125,
      "selected_pairs": 27337.9,
      "state_query_norm": 770.2145324707031,
      "state_relation_loss": 0.0008982206578366459,
      "step": 4820,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.007043318171054125,
      "weighted_relation_loss": 0.007043318171054125
    },
    {
      "action_query_norm": 429.11513366699216,
      "action_relation_loss": 0.006098166201263666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.518241347053321,
      "flow_matching_loss": 0.062278001382946965,
      "grad_norm": 0.15861061215400696,
      "learning_rate": 9.667111611305341e-05,
      "loss": 0.0693,
      "relation_loss": 0.006980479834601283,
      "selected_pairs": 27014.6,
      "state_query_norm": 782.0607421875,
      "state_relation_loss": 0.0008823136682622135,
      "step": 4830,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.006980479834601283,
      "weighted_relation_loss": 0.006980479834601283
    },
    {
      "action_query_norm": 433.0005676269531,
      "action_relation_loss": 0.005703412601724267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.527595884003742,
      "flow_matching_loss": 0.059609904512763025,
      "grad_norm": 0.1645684689283371,
      "learning_rate": 9.665131339003486e-05,
      "loss": 0.0662,
      "relation_loss": 0.006564458552747965,
      "selected_pairs": 27315.2,
      "state_query_norm": 790.3461669921875,
      "state_relation_loss": 0.0008610459859482944,
      "step": 4840,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.006564458552747965,
      "weighted_relation_loss": 0.006564458552747965
    },
    {
      "action_query_norm": 428.1277374267578,
      "action_relation_loss": 0.006295628333464265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.536950420954163,
      "flow_matching_loss": 0.056507451459765434,
      "grad_norm": 0.1599641889333725,
      "learning_rate": 9.663145398127158e-05,
      "loss": 0.0637,
      "relation_loss": 0.00717430878430605,
      "selected_pairs": 26943.8,
      "state_query_norm": 767.1972961425781,
      "state_relation_loss": 0.0008786804857663811,
      "step": 4850,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00717430878430605,
      "weighted_relation_loss": 0.00717430878430605
    },
    {
      "action_query_norm": 431.1368438720703,
      "action_relation_loss": 0.005684844870120287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.546304957904583,
      "flow_matching_loss": 0.05855197757482529,
      "grad_norm": 0.12389852106571198,
      "learning_rate": 9.661153791089467e-05,
      "loss": 0.0651,
      "relation_loss": 0.0065521488431841135,
      "selected_pairs": 26363.4,
      "state_query_norm": 795.6692016601562,
      "state_relation_loss": 0.0008673039847053587,
      "step": 4860,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0065521488431841135,
      "weighted_relation_loss": 0.0065521488431841135
    },
    {
      "action_query_norm": 431.4805023193359,
      "action_relation_loss": 0.005836617341265083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.555659494855004,
      "flow_matching_loss": 0.05500219762325287,
      "grad_norm": 0.15328863263130188,
      "learning_rate": 9.659156520310402e-05,
      "loss": 0.0617,
      "relation_loss": 0.006710148276761174,
      "selected_pairs": 26178.7,
      "state_query_norm": 797.3598937988281,
      "state_relation_loss": 0.0008735309005714953,
      "step": 4870,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.006710148276761174,
      "weighted_relation_loss": 0.006710148276761174
    },
    {
      "action_query_norm": 435.4593536376953,
      "action_relation_loss": 0.006042108591645956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.5650140318054255,
      "flow_matching_loss": 0.05891784764826298,
      "grad_norm": 0.14320215582847595,
      "learning_rate": 9.657153588216835e-05,
      "loss": 0.0658,
      "relation_loss": 0.006888816272839904,
      "selected_pairs": 26435.9,
      "state_query_norm": 790.6340881347656,
      "state_relation_loss": 0.0008467077219393105,
      "step": 4880,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.006888816272839904,
      "weighted_relation_loss": 0.006888816272839904
    },
    {
      "action_query_norm": 433.2730377197266,
      "action_relation_loss": 0.005869532562792301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.574368568755847,
      "flow_matching_loss": 0.060882818698883054,
      "grad_norm": 0.20129895210266113,
      "learning_rate": 9.655144997242516e-05,
      "loss": 0.0676,
      "relation_loss": 0.006698622880503535,
      "selected_pairs": 27193.1,
      "state_query_norm": 782.4518249511718,
      "state_relation_loss": 0.0008290903468150645,
      "step": 4890,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.006698622880503535,
      "weighted_relation_loss": 0.006698622880503535
    },
    {
      "action_query_norm": 430.5964385986328,
      "action_relation_loss": 0.005937970150262118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.583723105706268,
      "flow_matching_loss": 0.05899882353842258,
      "grad_norm": 0.1704712212085724,
      "learning_rate": 9.653130749828075e-05,
      "loss": 0.0658,
      "relation_loss": 0.00679692616686225,
      "selected_pairs": 27151.5,
      "state_query_norm": 775.0173889160156,
      "state_relation_loss": 0.0008589559991378337,
      "step": 4900,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.00679692616686225,
      "weighted_relation_loss": 0.00679692616686225
    },
    {
      "action_query_norm": 434.19066162109374,
      "action_relation_loss": 0.006211857497692108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.593077642656689,
      "flow_matching_loss": 0.058649509027600286,
      "grad_norm": 0.13163235783576965,
      "learning_rate": 9.65111084842101e-05,
      "loss": 0.0658,
      "relation_loss": 0.007105628727003932,
      "selected_pairs": 26505.1,
      "state_query_norm": 777.1243469238282,
      "state_relation_loss": 0.0008937712467741221,
      "step": 4910,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.007105628727003932,
      "weighted_relation_loss": 0.007105628727003932
    },
    {
      "action_query_norm": 435.5994506835938,
      "action_relation_loss": 0.005972906062379479,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.60243217960711,
      "flow_matching_loss": 0.05941057838499546,
      "grad_norm": 0.20003364980220795,
      "learning_rate": 9.649085295475695e-05,
      "loss": 0.0662,
      "relation_loss": 0.006828499585390091,
      "selected_pairs": 26832.7,
      "state_query_norm": 793.9248840332032,
      "state_relation_loss": 0.0008555936045013368,
      "step": 4920,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006828499585390091,
      "weighted_relation_loss": 0.006828499585390091
    },
    {
      "action_query_norm": 439.06318359375,
      "action_relation_loss": 0.006053099595010281,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.61178671655753,
      "flow_matching_loss": 0.05436150133609772,
      "grad_norm": 0.12216457724571228,
      "learning_rate": 9.647054093453365e-05,
      "loss": 0.0613,
      "relation_loss": 0.006941931555047631,
      "selected_pairs": 27038.6,
      "state_query_norm": 799.5560668945312,
      "state_relation_loss": 0.0008888319600373507,
      "step": 4930,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.006941931555047631,
      "weighted_relation_loss": 0.006941931555047631
    },
    {
      "action_query_norm": 438.48914489746096,
      "action_relation_loss": 0.0064504531677812334,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.621141253507951,
      "flow_matching_loss": 0.05804116614162922,
      "grad_norm": 0.1326068490743637,
      "learning_rate": 9.645017244822123e-05,
      "loss": 0.0654,
      "relation_loss": 0.0073388033546507355,
      "selected_pairs": 27036.7,
      "state_query_norm": 798.9297058105469,
      "state_relation_loss": 0.0008883502741809934,
      "step": 4940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0073388033546507355,
      "weighted_relation_loss": 0.0073388033546507355
    },
    {
      "action_query_norm": 438.4155212402344,
      "action_relation_loss": 0.006033795978873968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.630495790458372,
      "flow_matching_loss": 0.057179196551442144,
      "grad_norm": 0.2017369121313095,
      "learning_rate": 9.642974752056931e-05,
      "loss": 0.064,
      "relation_loss": 0.0068504434078931805,
      "selected_pairs": 26715.9,
      "state_query_norm": 804.9440490722657,
      "state_relation_loss": 0.0008166473940946162,
      "step": 4950,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0068504434078931805,
      "weighted_relation_loss": 0.0068504434078931805
    },
    {
      "action_query_norm": 438.59669799804686,
      "action_relation_loss": 0.006282864324748516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.639850327408793,
      "flow_matching_loss": 0.05999842621386051,
      "grad_norm": 0.14520935714244843,
      "learning_rate": 9.640926617639613e-05,
      "loss": 0.0671,
      "relation_loss": 0.007105425326153636,
      "selected_pairs": 26277.7,
      "state_query_norm": 802.8848510742188,
      "state_relation_loss": 0.0008225609839428216,
      "step": 4960,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.007105425326153636,
      "weighted_relation_loss": 0.007105425326153636
    },
    {
      "action_query_norm": 436.5287231445312,
      "action_relation_loss": 0.006292568380013108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.6492048643592145,
      "flow_matching_loss": 0.05973059386014938,
      "grad_norm": 0.1533179134130478,
      "learning_rate": 9.638872844058843e-05,
      "loss": 0.0669,
      "relation_loss": 0.007150170253589749,
      "selected_pairs": 26803.0,
      "state_query_norm": 801.6776611328125,
      "state_relation_loss": 0.0008576019085012376,
      "step": 4970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007150170253589749,
      "weighted_relation_loss": 0.007150170253589749
    },
    {
      "action_query_norm": 439.8993316650391,
      "action_relation_loss": 0.005916609801352024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.658559401309635,
      "flow_matching_loss": 0.059461765736341474,
      "grad_norm": 0.17962174117565155,
      "learning_rate": 9.63681343381015e-05,
      "loss": 0.0662,
      "relation_loss": 0.006786846928298474,
      "selected_pairs": 26901.1,
      "state_query_norm": 788.5127807617188,
      "state_relation_loss": 0.0008702371211256832,
      "step": 4980,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006786846928298474,
      "weighted_relation_loss": 0.006786846928298474
    },
    {
      "action_query_norm": 440.35059509277346,
      "action_relation_loss": 0.005956314923241735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.667913938260056,
      "flow_matching_loss": 0.06064799018204212,
      "grad_norm": 0.15543922781944275,
      "learning_rate": 9.634748389395914e-05,
      "loss": 0.0674,
      "relation_loss": 0.0068007488269358875,
      "selected_pairs": 26502.4,
      "state_query_norm": 791.6955810546875,
      "state_relation_loss": 0.0008444340142887086,
      "step": 4990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0068007488269358875,
      "weighted_relation_loss": 0.0068007488269358875
    },
    {
      "action_query_norm": 443.9473937988281,
      "action_relation_loss": 0.0060817394871264694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.677268475210477,
      "flow_matching_loss": 0.0602375328540802,
      "grad_norm": 0.15648050606250763,
      "learning_rate": 9.632677713325353e-05,
      "loss": 0.0672,
      "relation_loss": 0.006940178386867046,
      "selected_pairs": 27039.4,
      "state_query_norm": 790.6825866699219,
      "state_relation_loss": 0.0008584389579482377,
      "step": 5000,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.006940178386867046,
      "weighted_relation_loss": 0.006940178386867046
    },
    {
      "action_query_norm": 437.757763671875,
      "action_relation_loss": 0.006063412129878998,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.686623012160898,
      "flow_matching_loss": 0.056786779686808585,
      "grad_norm": 0.1805265098810196,
      "learning_rate": 9.63060140811454e-05,
      "loss": 0.0637,
      "relation_loss": 0.006932232342660427,
      "selected_pairs": 27536.3,
      "state_query_norm": 790.1994934082031,
      "state_relation_loss": 0.000868820259347558,
      "step": 5010,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.006932232342660427,
      "weighted_relation_loss": 0.006932232342660427
    },
    {
      "action_query_norm": 436.0645812988281,
      "action_relation_loss": 0.00605693063698709,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.695977549111319,
      "flow_matching_loss": 0.0554230734705925,
      "grad_norm": 0.19536441564559937,
      "learning_rate": 9.628519476286379e-05,
      "loss": 0.0623,
      "relation_loss": 0.006904550269246102,
      "selected_pairs": 26875.0,
      "state_query_norm": 790.4802612304687,
      "state_relation_loss": 0.0008476197079289704,
      "step": 5020,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.006904550269246102,
      "weighted_relation_loss": 0.006904550269246102
    },
    {
      "action_query_norm": 441.3907989501953,
      "action_relation_loss": 0.0059419802855700254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.70533208606174,
      "flow_matching_loss": 0.05590357519686222,
      "grad_norm": 0.18319910764694214,
      "learning_rate": 9.626431920370612e-05,
      "loss": 0.0627,
      "relation_loss": 0.006763010565191507,
      "selected_pairs": 27154.3,
      "state_query_norm": 794.1040161132812,
      "state_relation_loss": 0.0008210302970837801,
      "step": 5030,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.006763010565191507,
      "weighted_relation_loss": 0.006763010565191507
    },
    {
      "action_query_norm": 442.2090118408203,
      "action_relation_loss": 0.005784286465495825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.714686623012161,
      "flow_matching_loss": 0.05810193195939064,
      "grad_norm": 0.2355799674987793,
      "learning_rate": 9.624338742903819e-05,
      "loss": 0.0647,
      "relation_loss": 0.006631068419665098,
      "selected_pairs": 27257.5,
      "state_query_norm": 798.3844909667969,
      "state_relation_loss": 0.0008467819716315717,
      "step": 5040,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.006631068419665098,
      "weighted_relation_loss": 0.006631068419665098
    },
    {
      "action_query_norm": 444.95458068847654,
      "action_relation_loss": 0.006178525416180492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.724041159962582,
      "flow_matching_loss": 0.058842433989048,
      "grad_norm": 0.25817015767097473,
      "learning_rate": 9.622239946429406e-05,
      "loss": 0.0658,
      "relation_loss": 0.006987325893715024,
      "selected_pairs": 26686.5,
      "state_query_norm": 800.1330810546875,
      "state_relation_loss": 0.0008088005124591291,
      "step": 5050,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006987325893715024,
      "weighted_relation_loss": 0.006987325893715024
    },
    {
      "action_query_norm": 446.5823150634766,
      "action_relation_loss": 0.006073114834725857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.733395696913003,
      "flow_matching_loss": 0.06035243608057499,
      "grad_norm": 0.24767418205738068,
      "learning_rate": 9.620135533497609e-05,
      "loss": 0.0673,
      "relation_loss": 0.006904649594798684,
      "selected_pairs": 26639.2,
      "state_query_norm": 791.7032470703125,
      "state_relation_loss": 0.0008315347542520613,
      "step": 5060,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.006904649594798684,
      "weighted_relation_loss": 0.006904649594798684
    },
    {
      "action_query_norm": 440.8127044677734,
      "action_relation_loss": 0.005908680288121105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.742750233863424,
      "flow_matching_loss": 0.0565594881772995,
      "grad_norm": 0.19832517206668854,
      "learning_rate": 9.61802550666549e-05,
      "loss": 0.0633,
      "relation_loss": 0.006737181497737765,
      "selected_pairs": 26802.0,
      "state_query_norm": 799.5123229980469,
      "state_relation_loss": 0.0008285012561827898,
      "step": 5070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.006737181497737765,
      "weighted_relation_loss": 0.006737181497737765
    },
    {
      "action_query_norm": 447.49482421875,
      "action_relation_loss": 0.0055651661939919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.752104770813845,
      "flow_matching_loss": 0.05691812187433243,
      "grad_norm": 0.19701829552650452,
      "learning_rate": 9.615909868496928e-05,
      "loss": 0.0633,
      "relation_loss": 0.006399046070873738,
      "selected_pairs": 27221.0,
      "state_query_norm": 813.4734313964843,
      "state_relation_loss": 0.0008338798594195396,
      "step": 5080,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.006399046070873738,
      "weighted_relation_loss": 0.006399046070873738
    },
    {
      "action_query_norm": 438.21754760742186,
      "action_relation_loss": 0.006077870726585388,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.761459307764266,
      "flow_matching_loss": 0.0585881806910038,
      "grad_norm": 0.18586444854736328,
      "learning_rate": 9.613788621562622e-05,
      "loss": 0.0655,
      "relation_loss": 0.00688983490690589,
      "selected_pairs": 26737.6,
      "state_query_norm": 805.3558898925781,
      "state_relation_loss": 0.0008119642094243318,
      "step": 5090,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00688983490690589,
      "weighted_relation_loss": 0.00688983490690589
    },
    {
      "action_query_norm": 437.2102478027344,
      "action_relation_loss": 0.00586237171664834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.770813844714686,
      "flow_matching_loss": 0.060083257406949996,
      "grad_norm": 0.25451424717903137,
      "learning_rate": 9.611661768440093e-05,
      "loss": 0.0668,
      "relation_loss": 0.00668056053109467,
      "selected_pairs": 26997.2,
      "state_query_norm": 803.5092712402344,
      "state_relation_loss": 0.0008181888028047979,
      "step": 5100,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00668056053109467,
      "weighted_relation_loss": 0.00668056053109467
    },
    {
      "action_query_norm": 442.4995544433594,
      "action_relation_loss": 0.005964576825499534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.780168381665107,
      "flow_matching_loss": 0.05680765509605408,
      "grad_norm": 0.1786988228559494,
      "learning_rate": 9.609529311713661e-05,
      "loss": 0.0636,
      "relation_loss": 0.00679857013747096,
      "selected_pairs": 27201.1,
      "state_query_norm": 791.8078247070313,
      "state_relation_loss": 0.0008339933236129581,
      "step": 5110,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00679857013747096,
      "weighted_relation_loss": 0.00679857013747096
    },
    {
      "action_query_norm": 449.738134765625,
      "action_relation_loss": 0.006070346618071198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.789522918615528,
      "flow_matching_loss": 0.05791683383285999,
      "grad_norm": 0.12644366919994354,
      "learning_rate": 9.607391253974466e-05,
      "loss": 0.0648,
      "relation_loss": 0.006912722671404481,
      "selected_pairs": 27084.9,
      "state_query_norm": 792.3953308105469,
      "state_relation_loss": 0.0008423760591540486,
      "step": 5120,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.006912722671404481,
      "weighted_relation_loss": 0.006912722671404481
    },
    {
      "action_query_norm": 441.2373901367188,
      "action_relation_loss": 0.005701656313613057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.7988774555659495,
      "flow_matching_loss": 0.05688490942120552,
      "grad_norm": 0.17900502681732178,
      "learning_rate": 9.605247597820448e-05,
      "loss": 0.0634,
      "relation_loss": 0.006548255961388349,
      "selected_pairs": 26688.3,
      "state_query_norm": 784.6526000976562,
      "state_relation_loss": 0.0008465995662845672,
      "step": 5130,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006548255961388349,
      "weighted_relation_loss": 0.006548255961388349
    },
    {
      "action_query_norm": 439.4871795654297,
      "action_relation_loss": 0.006069963239133358,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.808231992516371,
      "flow_matching_loss": 0.05761708356440067,
      "grad_norm": 0.18078450858592987,
      "learning_rate": 9.603098345856354e-05,
      "loss": 0.0645,
      "relation_loss": 0.006884023547172546,
      "selected_pairs": 26894.6,
      "state_query_norm": 788.4248352050781,
      "state_relation_loss": 0.0008140602789353579,
      "step": 5140,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006884023547172546,
      "weighted_relation_loss": 0.006884023547172546
    },
    {
      "action_query_norm": 447.44481201171874,
      "action_relation_loss": 0.005695032468065619,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.817586529466792,
      "flow_matching_loss": 0.05843028128147125,
      "grad_norm": 0.18635611236095428,
      "learning_rate": 9.600943500693725e-05,
      "loss": 0.065,
      "relation_loss": 0.006544667202979327,
      "selected_pairs": 26767.9,
      "state_query_norm": 797.0773742675781,
      "state_relation_loss": 0.0008496346767060458,
      "step": 5150,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.006544667202979327,
      "weighted_relation_loss": 0.006544667202979327
    },
    {
      "action_query_norm": 451.5291015625,
      "action_relation_loss": 0.005674049397930503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.826941066417213,
      "flow_matching_loss": 0.056899357587099075,
      "grad_norm": 0.1465732753276825,
      "learning_rate": 9.598783064950902e-05,
      "loss": 0.0634,
      "relation_loss": 0.00651156073436141,
      "selected_pairs": 26736.7,
      "state_query_norm": 806.5416687011718,
      "state_relation_loss": 0.0008375113131478429,
      "step": 5160,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00651156073436141,
      "weighted_relation_loss": 0.00651156073436141
    },
    {
      "action_query_norm": 445.52100524902346,
      "action_relation_loss": 0.005609003081917763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.836295603367633,
      "flow_matching_loss": 0.05507675632834434,
      "grad_norm": 0.20275644958019257,
      "learning_rate": 9.596617041253018e-05,
      "loss": 0.0615,
      "relation_loss": 0.006436400255188346,
      "selected_pairs": 26747.3,
      "state_query_norm": 799.619482421875,
      "state_relation_loss": 0.0008273971034213901,
      "step": 5170,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006436400255188346,
      "weighted_relation_loss": 0.006436400255188346
    },
    {
      "action_query_norm": 449.5955474853516,
      "action_relation_loss": 0.005980088980868459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.845650140318054,
      "flow_matching_loss": 0.056815820187330245,
      "grad_norm": 0.16824820637702942,
      "learning_rate": 9.594445432231996e-05,
      "loss": 0.0636,
      "relation_loss": 0.0068090977147221565,
      "selected_pairs": 26881.2,
      "state_query_norm": 799.17470703125,
      "state_relation_loss": 0.0008290087687782943,
      "step": 5180,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0068090977147221565,
      "weighted_relation_loss": 0.0068090977147221565
    },
    {
      "action_query_norm": 446.14618225097655,
      "action_relation_loss": 0.005685238633304834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.855004677268475,
      "flow_matching_loss": 0.05766451209783554,
      "grad_norm": 0.21647927165031433,
      "learning_rate": 9.592268240526547e-05,
      "loss": 0.0642,
      "relation_loss": 0.006530987052246928,
      "selected_pairs": 27504.2,
      "state_query_norm": 798.2338684082031,
      "state_relation_loss": 0.00084574842476286,
      "step": 5190,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.006530987052246928,
      "weighted_relation_loss": 0.006530987052246928
    },
    {
      "action_query_norm": 446.21257934570315,
      "action_relation_loss": 0.005841830000281334,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.864359214218896,
      "flow_matching_loss": 0.05694802850484848,
      "grad_norm": 0.1678769439458847,
      "learning_rate": 9.590085468782162e-05,
      "loss": 0.0636,
      "relation_loss": 0.0066629437729716304,
      "selected_pairs": 26300.9,
      "state_query_norm": 818.67607421875,
      "state_relation_loss": 0.0008211137843318284,
      "step": 5200,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0066629437729716304,
      "weighted_relation_loss": 0.0066629437729716304
    },
    {
      "action_query_norm": 445.9340484619141,
      "action_relation_loss": 0.005601152358576655,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.873713751169317,
      "flow_matching_loss": 0.05814447999000549,
      "grad_norm": 0.11718056350946426,
      "learning_rate": 9.587897119651116e-05,
      "loss": 0.0646,
      "relation_loss": 0.0064141553360968825,
      "selected_pairs": 27092.1,
      "state_query_norm": 803.2413635253906,
      "state_relation_loss": 0.0008130029949825257,
      "step": 5210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0064141553360968825,
      "weighted_relation_loss": 0.0064141553360968825
    },
    {
      "action_query_norm": 441.2641204833984,
      "action_relation_loss": 0.005570190725848079,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.883068288119738,
      "flow_matching_loss": 0.05779024139046669,
      "grad_norm": 0.14378748834133148,
      "learning_rate": 9.585703195792459e-05,
      "loss": 0.0642,
      "relation_loss": 0.006400300841778517,
      "selected_pairs": 26572.1,
      "state_query_norm": 806.8657165527344,
      "state_relation_loss": 0.0008301101217512042,
      "step": 5220,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.006400300841778517,
      "weighted_relation_loss": 0.006400300841778517
    },
    {
      "action_query_norm": 448.9681854248047,
      "action_relation_loss": 0.005966610182076693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.892422825070159,
      "flow_matching_loss": 0.05680064745247364,
      "grad_norm": 0.13879330456256866,
      "learning_rate": 9.583503699872016e-05,
      "loss": 0.0636,
      "relation_loss": 0.006802439503371715,
      "selected_pairs": 26461.9,
      "state_query_norm": 810.2661499023437,
      "state_relation_loss": 0.0008358292689081281,
      "step": 5230,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.006802439503371715,
      "weighted_relation_loss": 0.006802439503371715
    },
    {
      "action_query_norm": 448.3871795654297,
      "action_relation_loss": 0.005880145262926817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.90177736202058,
      "flow_matching_loss": 0.05911844670772552,
      "grad_norm": 0.16145285964012146,
      "learning_rate": 9.581298634562381e-05,
      "loss": 0.0658,
      "relation_loss": 0.006708968384191394,
      "selected_pairs": 26825.0,
      "state_query_norm": 808.1297973632812,
      "state_relation_loss": 0.0008288231620099396,
      "step": 5240,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006708968384191394,
      "weighted_relation_loss": 0.006708968384191394
    },
    {
      "action_query_norm": 450.1983215332031,
      "action_relation_loss": 0.006011597299948335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.911131898971001,
      "flow_matching_loss": 0.058589131012558936,
      "grad_norm": 0.11638545989990234,
      "learning_rate": 9.579088002542917e-05,
      "loss": 0.0654,
      "relation_loss": 0.006814314890652895,
      "selected_pairs": 27242.4,
      "state_query_norm": 803.0595336914063,
      "state_relation_loss": 0.000802717637270689,
      "step": 5250,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.006814314890652895,
      "weighted_relation_loss": 0.006814314890652895
    },
    {
      "action_query_norm": 452.6897003173828,
      "action_relation_loss": 0.005565186915919185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.920486435921422,
      "flow_matching_loss": 0.057758231088519095,
      "grad_norm": 0.1535252332687378,
      "learning_rate": 9.57687180649975e-05,
      "loss": 0.0641,
      "relation_loss": 0.006380970356985927,
      "selected_pairs": 26881.6,
      "state_query_norm": 820.3510925292969,
      "state_relation_loss": 0.0008157834643498063,
      "step": 5260,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.006380970356985927,
      "weighted_relation_loss": 0.006380970356985927
    },
    {
      "action_query_norm": 449.121435546875,
      "action_relation_loss": 0.005765683623030782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.929840972871843,
      "flow_matching_loss": 0.05582754798233509,
      "grad_norm": 0.17398014664649963,
      "learning_rate": 9.574650049125768e-05,
      "loss": 0.0624,
      "relation_loss": 0.006575603922829032,
      "selected_pairs": 26687.0,
      "state_query_norm": 826.9128662109375,
      "state_relation_loss": 0.0008099202299490571,
      "step": 5270,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006575603922829032,
      "weighted_relation_loss": 0.006575603922829032
    },
    {
      "action_query_norm": 444.847265625,
      "action_relation_loss": 0.005670687509700656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.939195509822264,
      "flow_matching_loss": 0.0562750231474638,
      "grad_norm": 0.2104674130678177,
      "learning_rate": 9.572422733120614e-05,
      "loss": 0.0628,
      "relation_loss": 0.006478593358770013,
      "selected_pairs": 27336.7,
      "state_query_norm": 795.5181091308593,
      "state_relation_loss": 0.000807905854890123,
      "step": 5280,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.006478593358770013,
      "weighted_relation_loss": 0.006478593358770013
    },
    {
      "action_query_norm": 444.01647033691404,
      "action_relation_loss": 0.0062157079111784695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.948550046772684,
      "flow_matching_loss": 0.05895146168768406,
      "grad_norm": 0.18596334755420685,
      "learning_rate": 9.570189861190689e-05,
      "loss": 0.066,
      "relation_loss": 0.007021623943001032,
      "selected_pairs": 26801.0,
      "state_query_norm": 815.1355224609375,
      "state_relation_loss": 0.0008059160725679248,
      "step": 5290,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007021623943001032,
      "weighted_relation_loss": 0.007021623943001032
    },
    {
      "action_query_norm": 446.3655303955078,
      "action_relation_loss": 0.00572627717629075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.9579045837231055,
      "flow_matching_loss": 0.05535387247800827,
      "grad_norm": 0.14011244475841522,
      "learning_rate": 9.56795143604914e-05,
      "loss": 0.0619,
      "relation_loss": 0.006565611204132437,
      "selected_pairs": 26387.4,
      "state_query_norm": 799.1859069824219,
      "state_relation_loss": 0.000839334080228582,
      "step": 5300,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.006565611204132437,
      "weighted_relation_loss": 0.006565611204132437
    },
    {
      "action_query_norm": 449.37217102050784,
      "action_relation_loss": 0.005974059039726853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.967259120673527,
      "flow_matching_loss": 0.05571749284863472,
      "grad_norm": 0.10532345622777939,
      "learning_rate": 9.565707460415869e-05,
      "loss": 0.0625,
      "relation_loss": 0.006779744196683169,
      "selected_pairs": 26897.7,
      "state_query_norm": 825.6490966796875,
      "state_relation_loss": 0.0008056851744186133,
      "step": 5310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006779744196683169,
      "weighted_relation_loss": 0.006779744196683169
    },
    {
      "action_query_norm": 452.0156188964844,
      "action_relation_loss": 0.005613654665648937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.976613657623948,
      "flow_matching_loss": 0.054593116790056226,
      "grad_norm": 0.13179215788841248,
      "learning_rate": 9.563457937017515e-05,
      "loss": 0.061,
      "relation_loss": 0.006438233563676477,
      "selected_pairs": 26850.6,
      "state_query_norm": 808.0746398925781,
      "state_relation_loss": 0.0008245789562352001,
      "step": 5320,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006438233563676477,
      "weighted_relation_loss": 0.006438233563676477
    },
    {
      "action_query_norm": 453.77460327148435,
      "action_relation_loss": 0.005592872947454452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.985968194574369,
      "flow_matching_loss": 0.05652292408049107,
      "grad_norm": 0.1376078575849533,
      "learning_rate": 9.56120286858746e-05,
      "loss": 0.063,
      "relation_loss": 0.006437959568575025,
      "selected_pairs": 27596.7,
      "state_query_norm": 812.4867858886719,
      "state_relation_loss": 0.0008450866735074669,
      "step": 5330,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.006437959568575025,
      "weighted_relation_loss": 0.006437959568575025
    },
    {
      "action_query_norm": 448.39019775390625,
      "action_relation_loss": 0.006110551999881864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.99532273152479,
      "flow_matching_loss": 0.058592090755701064,
      "grad_norm": 0.13265933096408844,
      "learning_rate": 9.558942257865829e-05,
      "loss": 0.0655,
      "relation_loss": 0.006897468026727438,
      "selected_pairs": 27059.2,
      "state_query_norm": 806.2742919921875,
      "state_relation_loss": 0.0007869160093832761,
      "step": 5340,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.006897468026727438,
      "weighted_relation_loss": 0.006897468026727438
    },
    {
      "action_query_norm": 444.9855438232422,
      "action_relation_loss": 0.005587719939649105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.00467726847521,
      "flow_matching_loss": 0.05275074802339077,
      "grad_norm": 0.13041895627975464,
      "learning_rate": 9.556676107599472e-05,
      "loss": 0.0592,
      "relation_loss": 0.006459022406488657,
      "selected_pairs": 23362.0,
      "state_query_norm": 813.3554138183594,
      "state_relation_loss": 0.0008713025134056806,
      "step": 5350,
      "valid_action_samples": 210.1,
      "weighted_auxiliary_loss": 0.006459022406488657,
      "weighted_relation_loss": 0.006459022406488657
    },
    {
      "action_query_norm": 451.6586944580078,
      "action_relation_loss": 0.005482711317017675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.014031805425631,
      "flow_matching_loss": 0.056419917941093446,
      "grad_norm": 0.16852009296417236,
      "learning_rate": 9.554404420541978e-05,
      "loss": 0.0628,
      "relation_loss": 0.006341870361939073,
      "selected_pairs": 26740.2,
      "state_query_norm": 814.6473937988281,
      "state_relation_loss": 0.0008591590158175677,
      "step": 5360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006341870361939073,
      "weighted_relation_loss": 0.006341870361939073
    },
    {
      "action_query_norm": 451.6543823242188,
      "action_relation_loss": 0.005802818527445197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.023386342376052,
      "flow_matching_loss": 0.05424342453479767,
      "grad_norm": 0.1563342958688736,
      "learning_rate": 9.55212719945366e-05,
      "loss": 0.0609,
      "relation_loss": 0.006637037079781294,
      "selected_pairs": 26661.4,
      "state_query_norm": 818.1913024902344,
      "state_relation_loss": 0.0008342185465153307,
      "step": 5370,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.006637037079781294,
      "weighted_relation_loss": 0.006637037079781294
    },
    {
      "action_query_norm": 455.49060974121096,
      "action_relation_loss": 0.005669622682034969,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.032740879326473,
      "flow_matching_loss": 0.055117100849747656,
      "grad_norm": 0.1824905276298523,
      "learning_rate": 9.549844447101559e-05,
      "loss": 0.0616,
      "relation_loss": 0.006491708755493164,
      "selected_pairs": 27112.7,
      "state_query_norm": 813.2259582519531,
      "state_relation_loss": 0.0008220860967412591,
      "step": 5380,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.006491708755493164,
      "weighted_relation_loss": 0.006491708755493164
    },
    {
      "action_query_norm": 456.06798400878904,
      "action_relation_loss": 0.005599220888689161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.0420954162768945,
      "flow_matching_loss": 0.055988019704818724,
      "grad_norm": 0.14249087870121002,
      "learning_rate": 9.547556166259433e-05,
      "loss": 0.0624,
      "relation_loss": 0.006420363625511527,
      "selected_pairs": 27224.7,
      "state_query_norm": 802.0097534179688,
      "state_relation_loss": 0.0008211426844354719,
      "step": 5390,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.006420363625511527,
      "weighted_relation_loss": 0.006420363625511527
    },
    {
      "action_query_norm": 458.4372528076172,
      "action_relation_loss": 0.005591599084436893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.051449953227316,
      "flow_matching_loss": 0.052772851288318635,
      "grad_norm": 0.17385096848011017,
      "learning_rate": 9.545262359707756e-05,
      "loss": 0.0592,
      "relation_loss": 0.006400020187720656,
      "selected_pairs": 26507.0,
      "state_query_norm": 818.2531555175781,
      "state_relation_loss": 0.0008084210567176342,
      "step": 5400,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.006400020187720656,
      "weighted_relation_loss": 0.006400020187720656
    },
    {
      "action_query_norm": 452.9245574951172,
      "action_relation_loss": 0.00530710001476109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.060804490177736,
      "flow_matching_loss": 0.054820914193987845,
      "grad_norm": 0.1548786461353302,
      "learning_rate": 9.542963030233724e-05,
      "loss": 0.0609,
      "relation_loss": 0.006099164765328169,
      "selected_pairs": 26779.8,
      "state_query_norm": 808.0880432128906,
      "state_relation_loss": 0.0007920648204162717,
      "step": 5410,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.006099164765328169,
      "weighted_relation_loss": 0.006099164765328169
    },
    {
      "action_query_norm": 451.3801666259766,
      "action_relation_loss": 0.005530362017452717,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.070159027128157,
      "flow_matching_loss": 0.05373186394572258,
      "grad_norm": 0.17652955651283264,
      "learning_rate": 9.540658180631237e-05,
      "loss": 0.06,
      "relation_loss": 0.00629412466660142,
      "selected_pairs": 26569.3,
      "state_query_norm": 803.6675964355469,
      "state_relation_loss": 0.0007637626025825739,
      "step": 5420,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.00629412466660142,
      "weighted_relation_loss": 0.00629412466660142
    },
    {
      "action_query_norm": 459.8363433837891,
      "action_relation_loss": 0.004974704142659903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.079513564078578,
      "flow_matching_loss": 0.05181461237370968,
      "grad_norm": 0.26396864652633667,
      "learning_rate": 9.538347813700904e-05,
      "loss": 0.0576,
      "relation_loss": 0.005740052089095116,
      "selected_pairs": 26971.9,
      "state_query_norm": 813.33955078125,
      "state_relation_loss": 0.0007653479464352131,
      "step": 5430,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.005740052089095116,
      "weighted_relation_loss": 0.005740052089095116
    },
    {
      "action_query_norm": 454.97292175292966,
      "action_relation_loss": 0.005259239207953214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.088868101028999,
      "flow_matching_loss": 0.05632600225508213,
      "grad_norm": 0.17540930211544037,
      "learning_rate": 9.536031932250036e-05,
      "loss": 0.0624,
      "relation_loss": 0.006024037348106503,
      "selected_pairs": 26619.8,
      "state_query_norm": 822.6370239257812,
      "state_relation_loss": 0.0007647981692571193,
      "step": 5440,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.006024037348106503,
      "weighted_relation_loss": 0.006024037348106503
    },
    {
      "action_query_norm": 458.4640808105469,
      "action_relation_loss": 0.0052427278831601145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.09822263797942,
      "flow_matching_loss": 0.05332571417093277,
      "grad_norm": 0.19135352969169617,
      "learning_rate": 9.533710539092653e-05,
      "loss": 0.0593,
      "relation_loss": 0.006018765037879348,
      "selected_pairs": 26205.6,
      "state_query_norm": 827.4262084960938,
      "state_relation_loss": 0.0007760371605399996,
      "step": 5450,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.006018765037879348,
      "weighted_relation_loss": 0.006018765037879348
    },
    {
      "action_query_norm": 456.0533508300781,
      "action_relation_loss": 0.0055557283572852615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.107577174929841,
      "flow_matching_loss": 0.05377525202929974,
      "grad_norm": 0.13046713173389435,
      "learning_rate": 9.531383637049464e-05,
      "loss": 0.0601,
      "relation_loss": 0.00632336107082665,
      "selected_pairs": 26509.4,
      "state_query_norm": 812.9847534179687,
      "state_relation_loss": 0.0007676326727960259,
      "step": 5460,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.00632336107082665,
      "weighted_relation_loss": 0.00632336107082665
    },
    {
      "action_query_norm": 455.5186401367188,
      "action_relation_loss": 0.005215960927307606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.1169317118802615,
      "flow_matching_loss": 0.055601205304265024,
      "grad_norm": 0.15333211421966553,
      "learning_rate": 9.529051228947875e-05,
      "loss": 0.0616,
      "relation_loss": 0.006019258359447122,
      "selected_pairs": 26646.4,
      "state_query_norm": 801.8196411132812,
      "state_relation_loss": 0.000803297437960282,
      "step": 5470,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.006019258359447122,
      "weighted_relation_loss": 0.006019258359447122
    },
    {
      "action_query_norm": 452.3578582763672,
      "action_relation_loss": 0.004972144030034542,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.126286248830683,
      "flow_matching_loss": 0.05320807509124279,
      "grad_norm": 0.16912753880023956,
      "learning_rate": 9.52671331762198e-05,
      "loss": 0.0589,
      "relation_loss": 0.005724605359137058,
      "selected_pairs": 27361.6,
      "state_query_norm": 820.6231872558594,
      "state_relation_loss": 0.0007524612883571535,
      "step": 5480,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.005724605359137058,
      "weighted_relation_loss": 0.005724605359137058
    },
    {
      "action_query_norm": 451.74342956542966,
      "action_relation_loss": 0.0054356221109628676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.135640785781104,
      "flow_matching_loss": 0.05434213913977146,
      "grad_norm": 0.15050943195819855,
      "learning_rate": 9.524369905912565e-05,
      "loss": 0.0606,
      "relation_loss": 0.00620938753709197,
      "selected_pairs": 26319.3,
      "state_query_norm": 823.6432800292969,
      "state_relation_loss": 0.0007737653213553131,
      "step": 5490,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.00620938753709197,
      "weighted_relation_loss": 0.00620938753709197
    },
    {
      "action_query_norm": 455.9176483154297,
      "action_relation_loss": 0.00523705598898232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.144995322731525,
      "flow_matching_loss": 0.053143317252397536,
      "grad_norm": 0.15188460052013397,
      "learning_rate": 9.522020996667092e-05,
      "loss": 0.0592,
      "relation_loss": 0.006023693596944213,
      "selected_pairs": 26922.7,
      "state_query_norm": 825.2766906738282,
      "state_relation_loss": 0.0007866375730372965,
      "step": 5500,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006023693596944213,
      "weighted_relation_loss": 0.006023693596944213
    },
    {
      "action_query_norm": 453.5549743652344,
      "action_relation_loss": 0.005248401127755642,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.154349859681946,
      "flow_matching_loss": 0.05443600043654442,
      "grad_norm": 0.12335394322872162,
      "learning_rate": 9.519666592739709e-05,
      "loss": 0.0605,
      "relation_loss": 0.006047882372513413,
      "selected_pairs": 27312.2,
      "state_query_norm": 814.4089660644531,
      "state_relation_loss": 0.0007994812796823681,
      "step": 5510,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.006047882372513413,
      "weighted_relation_loss": 0.006047882372513413
    },
    {
      "action_query_norm": 458.7919494628906,
      "action_relation_loss": 0.005270446371287108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.163704396632367,
      "flow_matching_loss": 0.054640442878007886,
      "grad_norm": 0.1444796770811081,
      "learning_rate": 9.517306696991241e-05,
      "loss": 0.0607,
      "relation_loss": 0.006077220337465406,
      "selected_pairs": 26408.9,
      "state_query_norm": 819.8843688964844,
      "state_relation_loss": 0.0008067739137914032,
      "step": 5520,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.006077220337465406,
      "weighted_relation_loss": 0.006077220337465406
    },
    {
      "action_query_norm": 455.3528594970703,
      "action_relation_loss": 0.005448880884796381,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.173058933582787,
      "flow_matching_loss": 0.052738790214061734,
      "grad_norm": 0.1695551872253418,
      "learning_rate": 9.51494131228918e-05,
      "loss": 0.059,
      "relation_loss": 0.006250377791002393,
      "selected_pairs": 26663.8,
      "state_query_norm": 825.0988891601562,
      "state_relation_loss": 0.0008014969236683101,
      "step": 5530,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.006250377791002393,
      "weighted_relation_loss": 0.006250377791002393
    },
    {
      "action_query_norm": 458.68585510253905,
      "action_relation_loss": 0.005248253140598536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.182413470533208,
      "flow_matching_loss": 0.05393458604812622,
      "grad_norm": 0.12991954386234283,
      "learning_rate": 9.512570441507695e-05,
      "loss": 0.0599,
      "relation_loss": 0.006003776425495743,
      "selected_pairs": 27129.1,
      "state_query_norm": 834.3459594726562,
      "state_relation_loss": 0.000755523203406483,
      "step": 5540,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.006003776425495743,
      "weighted_relation_loss": 0.006003776425495743
    },
    {
      "action_query_norm": 462.0420440673828,
      "action_relation_loss": 0.005171243147924542,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.191768007483629,
      "flow_matching_loss": 0.056142573803663255,
      "grad_norm": 0.15967132151126862,
      "learning_rate": 9.510194087527615e-05,
      "loss": 0.0621,
      "relation_loss": 0.005970064224675298,
      "selected_pairs": 26548.2,
      "state_query_norm": 825.8800354003906,
      "state_relation_loss": 0.0007988210942130536,
      "step": 5550,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005970064224675298,
      "weighted_relation_loss": 0.005970064224675298
    },
    {
      "action_query_norm": 458.8299255371094,
      "action_relation_loss": 0.005133936181664467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.2011225444340505,
      "flow_matching_loss": 0.05494342669844628,
      "grad_norm": 0.14856763184070587,
      "learning_rate": 9.507812253236435e-05,
      "loss": 0.0608,
      "relation_loss": 0.005905828019604087,
      "selected_pairs": 27571.6,
      "state_query_norm": 839.3164916992188,
      "state_relation_loss": 0.0007718918728642166,
      "step": 5560,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.005905828019604087,
      "weighted_relation_loss": 0.005905828019604087
    },
    {
      "action_query_norm": 458.9759155273438,
      "action_relation_loss": 0.005528114316985011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.210477081384472,
      "flow_matching_loss": 0.05426554456353187,
      "grad_norm": 0.17057138681411743,
      "learning_rate": 9.505424941528309e-05,
      "loss": 0.0606,
      "relation_loss": 0.006288785766810179,
      "selected_pairs": 26526.3,
      "state_query_norm": 823.3413269042969,
      "state_relation_loss": 0.0007606714090798051,
      "step": 5570,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.006288785766810179,
      "weighted_relation_loss": 0.006288785766810179
    },
    {
      "action_query_norm": 455.3581909179687,
      "action_relation_loss": 0.005546430684626102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.219831618334893,
      "flow_matching_loss": 0.056169285997748376,
      "grad_norm": 0.17131821811199188,
      "learning_rate": 9.503032155304046e-05,
      "loss": 0.0625,
      "relation_loss": 0.006304579693824053,
      "selected_pairs": 26946.0,
      "state_query_norm": 823.6471496582031,
      "state_relation_loss": 0.0007581490033771842,
      "step": 5580,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006304579693824053,
      "weighted_relation_loss": 0.006304579693824053
    },
    {
      "action_query_norm": 457.743701171875,
      "action_relation_loss": 0.005449847178533673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.229186155285314,
      "flow_matching_loss": 0.05474234968423843,
      "grad_norm": 0.17192527651786804,
      "learning_rate": 9.500633897471106e-05,
      "loss": 0.0609,
      "relation_loss": 0.006197783537209034,
      "selected_pairs": 27174.9,
      "state_query_norm": 826.6268981933594,
      "state_relation_loss": 0.000747936381958425,
      "step": 5590,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006197783537209034,
      "weighted_relation_loss": 0.006197783537209034
    },
    {
      "action_query_norm": 462.97455444335935,
      "action_relation_loss": 0.0054767349734902385,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.238540692235734,
      "flow_matching_loss": 0.055483927205204966,
      "grad_norm": 0.2077712118625641,
      "learning_rate": 9.498230170943596e-05,
      "loss": 0.0617,
      "relation_loss": 0.006249548494815826,
      "selected_pairs": 26855.0,
      "state_query_norm": 822.1870239257812,
      "state_relation_loss": 0.000772813573712483,
      "step": 5600,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006249548494815826,
      "weighted_relation_loss": 0.006249548494815826
    },
    {
      "action_query_norm": 457.7295227050781,
      "action_relation_loss": 0.005213399743661284,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.247895229186155,
      "flow_matching_loss": 0.056330135092139244,
      "grad_norm": 0.16514165699481964,
      "learning_rate": 9.495820978642275e-05,
      "loss": 0.0623,
      "relation_loss": 0.005994348553940654,
      "selected_pairs": 27106.6,
      "state_query_norm": 818.8182250976563,
      "state_relation_loss": 0.0007809487986378371,
      "step": 5610,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005994348553940654,
      "weighted_relation_loss": 0.005994348553940654
    },
    {
      "action_query_norm": 461.968994140625,
      "action_relation_loss": 0.005168612813577056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.257249766136576,
      "flow_matching_loss": 0.053940891101956365,
      "grad_norm": 0.18315798044204712,
      "learning_rate": 9.493406323494535e-05,
      "loss": 0.0599,
      "relation_loss": 0.005952248582616448,
      "selected_pairs": 26876.7,
      "state_query_norm": 820.3842712402344,
      "state_relation_loss": 0.0007836358214262873,
      "step": 5620,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005952248582616448,
      "weighted_relation_loss": 0.005952248582616448
    },
    {
      "action_query_norm": 461.49205627441404,
      "action_relation_loss": 0.005666444962844253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.266604303086997,
      "flow_matching_loss": 0.055866959318518636,
      "grad_norm": 0.14748235046863556,
      "learning_rate": 9.490986208434413e-05,
      "loss": 0.0623,
      "relation_loss": 0.006429423391819,
      "selected_pairs": 26261.9,
      "state_query_norm": 821.2223876953125,
      "state_relation_loss": 0.0007629784638993442,
      "step": 5630,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.006429423391819,
      "weighted_relation_loss": 0.006429423391819
    },
    {
      "action_query_norm": 469.05357971191404,
      "action_relation_loss": 0.005336779030039906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.275958840037418,
      "flow_matching_loss": 0.05509271584451199,
      "grad_norm": 0.17559313774108887,
      "learning_rate": 9.488560636402577e-05,
      "loss": 0.0612,
      "relation_loss": 0.006092344876378775,
      "selected_pairs": 27202.2,
      "state_query_norm": 829.9695373535156,
      "state_relation_loss": 0.0007555658055935055,
      "step": 5640,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.006092344876378775,
      "weighted_relation_loss": 0.006092344876378775
    },
    {
      "action_query_norm": 467.2764892578125,
      "action_relation_loss": 0.005385007429867983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.285313376987839,
      "flow_matching_loss": 0.052440321817994115,
      "grad_norm": 0.15407738089561462,
      "learning_rate": 9.486129610346321e-05,
      "loss": 0.0586,
      "relation_loss": 0.006150069879367948,
      "selected_pairs": 27172.4,
      "state_query_norm": 824.1905517578125,
      "state_relation_loss": 0.0007650623912923038,
      "step": 5650,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006150069879367948,
      "weighted_relation_loss": 0.006150069879367948
    },
    {
      "action_query_norm": 466.6945556640625,
      "action_relation_loss": 0.005351125681772828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.29466791393826,
      "flow_matching_loss": 0.0553972240537405,
      "grad_norm": 0.1892116814851761,
      "learning_rate": 9.483693133219576e-05,
      "loss": 0.0615,
      "relation_loss": 0.006097082607448101,
      "selected_pairs": 26824.9,
      "state_query_norm": 835.391357421875,
      "state_relation_loss": 0.0007459570188075304,
      "step": 5660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006097082607448101,
      "weighted_relation_loss": 0.006097082607448101
    },
    {
      "action_query_norm": 470.6549774169922,
      "action_relation_loss": 0.005418550129979849,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.304022450888681,
      "flow_matching_loss": 0.05625520870089531,
      "grad_norm": 0.23926733434200287,
      "learning_rate": 9.481251207982888e-05,
      "loss": 0.0624,
      "relation_loss": 0.006193499267101288,
      "selected_pairs": 26458.3,
      "state_query_norm": 833.9419067382812,
      "state_relation_loss": 0.0007749491080176086,
      "step": 5670,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.006193499267101288,
      "weighted_relation_loss": 0.006193499267101288
    },
    {
      "action_query_norm": 467.8846740722656,
      "action_relation_loss": 0.005465161707252264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.313376987839102,
      "flow_matching_loss": 0.056678658723831175,
      "grad_norm": 0.17563840746879578,
      "learning_rate": 9.47880383760343e-05,
      "loss": 0.0629,
      "relation_loss": 0.006248239940032363,
      "selected_pairs": 26971.7,
      "state_query_norm": 824.7332580566406,
      "state_relation_loss": 0.0007830782094970345,
      "step": 5680,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.006248239940032363,
      "weighted_relation_loss": 0.006248239940032363
    },
    {
      "action_query_norm": 467.2576141357422,
      "action_relation_loss": 0.005355560127645731,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.322731524789523,
      "flow_matching_loss": 0.05591895803809166,
      "grad_norm": 0.10741808265447617,
      "learning_rate": 9.476351025054983e-05,
      "loss": 0.0621,
      "relation_loss": 0.006131690321490168,
      "selected_pairs": 27018.2,
      "state_query_norm": 813.9004577636719,
      "state_relation_loss": 0.0007761302345898003,
      "step": 5690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.006131690321490168,
      "weighted_relation_loss": 0.006131690321490168
    },
    {
      "action_query_norm": 474.46221923828125,
      "action_relation_loss": 0.005311819445341826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.332086061739944,
      "flow_matching_loss": 0.05602300502359867,
      "grad_norm": 0.14519643783569336,
      "learning_rate": 9.473892773317952e-05,
      "loss": 0.0621,
      "relation_loss": 0.00608236500993371,
      "selected_pairs": 26873.0,
      "state_query_norm": 833.3880737304687,
      "state_relation_loss": 0.0007705455413088202,
      "step": 5700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00608236500993371,
      "weighted_relation_loss": 0.00608236500993371
    },
    {
      "action_query_norm": 458.54059448242185,
      "action_relation_loss": 0.0056037584785372015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.341440598690365,
      "flow_matching_loss": 0.05298620015382767,
      "grad_norm": 0.15533186495304108,
      "learning_rate": 9.471429085379338e-05,
      "loss": 0.0594,
      "relation_loss": 0.0063767136540263895,
      "selected_pairs": 26276.1,
      "state_query_norm": 837.5284912109375,
      "state_relation_loss": 0.0007729551871307194,
      "step": 5710,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0063767136540263895,
      "weighted_relation_loss": 0.0063767136540263895
    },
    {
      "action_query_norm": 468.70799865722654,
      "action_relation_loss": 0.005566051416099072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.3507951356407855,
      "flow_matching_loss": 0.052684539929032326,
      "grad_norm": 0.14024826884269714,
      "learning_rate": 9.468959964232757e-05,
      "loss": 0.059,
      "relation_loss": 0.0063159879297018055,
      "selected_pairs": 27154.9,
      "state_query_norm": 830.03525390625,
      "state_relation_loss": 0.0007499364844989031,
      "step": 5720,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0063159879297018055,
      "weighted_relation_loss": 0.0063159879297018055
    },
    {
      "action_query_norm": 465.05446166992186,
      "action_relation_loss": 0.0057342374231666325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.360149672591207,
      "flow_matching_loss": 0.05546387024223805,
      "grad_norm": 0.15440554916858673,
      "learning_rate": 9.466485412878425e-05,
      "loss": 0.062,
      "relation_loss": 0.006510100560262799,
      "selected_pairs": 26553.9,
      "state_query_norm": 829.2384826660157,
      "state_relation_loss": 0.0007758630381431431,
      "step": 5730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.006510100560262799,
      "weighted_relation_loss": 0.006510100560262799
    },
    {
      "action_query_norm": 466.4154357910156,
      "action_relation_loss": 0.005273229395970702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.369504209541628,
      "flow_matching_loss": 0.0531740490347147,
      "grad_norm": 0.1791875660419464,
      "learning_rate": 9.464005434323154e-05,
      "loss": 0.0592,
      "relation_loss": 0.006062890030443668,
      "selected_pairs": 26803.3,
      "state_query_norm": 826.4356811523437,
      "state_relation_loss": 0.000789660750888288,
      "step": 5740,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.006062890030443668,
      "weighted_relation_loss": 0.006062890030443668
    },
    {
      "action_query_norm": 465.12633361816404,
      "action_relation_loss": 0.005973938480019569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.378858746492049,
      "flow_matching_loss": 0.05520947463810444,
      "grad_norm": 0.20063480734825134,
      "learning_rate": 9.461520031580352e-05,
      "loss": 0.062,
      "relation_loss": 0.0067499698605388405,
      "selected_pairs": 27059.6,
      "state_query_norm": 832.0767639160156,
      "state_relation_loss": 0.0007760313397739083,
      "step": 5750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0067499698605388405,
      "weighted_relation_loss": 0.0067499698605388405
    },
    {
      "action_query_norm": 473.17548217773435,
      "action_relation_loss": 0.005344421369954944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.38821328344247,
      "flow_matching_loss": 0.05336463414132595,
      "grad_norm": 0.17944860458374023,
      "learning_rate": 9.459029207670019e-05,
      "loss": 0.0595,
      "relation_loss": 0.006113353976979851,
      "selected_pairs": 27181.1,
      "state_query_norm": 835.135009765625,
      "state_relation_loss": 0.0007689325779210776,
      "step": 5760,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006113353976979851,
      "weighted_relation_loss": 0.006113353976979851
    },
    {
      "action_query_norm": 468.6238067626953,
      "action_relation_loss": 0.005216214060783386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.397567820392891,
      "flow_matching_loss": 0.05564799793064594,
      "grad_norm": 0.17820943892002106,
      "learning_rate": 9.456532965618737e-05,
      "loss": 0.0617,
      "relation_loss": 0.006008894275873899,
      "selected_pairs": 26613.9,
      "state_query_norm": 844.5414184570312,
      "state_relation_loss": 0.0007926801801659167,
      "step": 5770,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.006008894275873899,
      "weighted_relation_loss": 0.006008894275873899
    },
    {
      "action_query_norm": 469.9561401367188,
      "action_relation_loss": 0.00537467272952199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.406922357343311,
      "flow_matching_loss": 0.05596620254218578,
      "grad_norm": 0.20911112427711487,
      "learning_rate": 9.454031308459681e-05,
      "loss": 0.0621,
      "relation_loss": 0.0061525693628937,
      "selected_pairs": 27269.6,
      "state_query_norm": 838.7640014648438,
      "state_relation_loss": 0.0007778965577017516,
      "step": 5780,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0061525693628937,
      "weighted_relation_loss": 0.0061525693628937
    },
    {
      "action_query_norm": 468.64532165527345,
      "action_relation_loss": 0.00531531018204987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.416276894293732,
      "flow_matching_loss": 0.053438080847263335,
      "grad_norm": 0.18384529650211334,
      "learning_rate": 9.451524239232595e-05,
      "loss": 0.0595,
      "relation_loss": 0.006047809729352593,
      "selected_pairs": 26415.6,
      "state_query_norm": 839.9993408203125,
      "state_relation_loss": 0.0007324995880480855,
      "step": 5790,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.006047809729352593,
      "weighted_relation_loss": 0.006047809729352593
    },
    {
      "action_query_norm": 467.1901397705078,
      "action_relation_loss": 0.005115330638363958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.425631431244153,
      "flow_matching_loss": 0.0563743457198143,
      "grad_norm": 0.1771247237920761,
      "learning_rate": 9.449011760983809e-05,
      "loss": 0.0623,
      "relation_loss": 0.005909057892858982,
      "selected_pairs": 26490.2,
      "state_query_norm": 829.8153503417968,
      "state_relation_loss": 0.0007937271671835333,
      "step": 5800,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005909057892858982,
      "weighted_relation_loss": 0.005909057892858982
    },
    {
      "action_query_norm": 469.8861480712891,
      "action_relation_loss": 0.0051685955841094255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.4349859681945745,
      "flow_matching_loss": 0.05300217941403389,
      "grad_norm": 0.174229696393013,
      "learning_rate": 9.446493876766218e-05,
      "loss": 0.059,
      "relation_loss": 0.0059545118361711506,
      "selected_pairs": 27287.8,
      "state_query_norm": 842.7120483398437,
      "state_relation_loss": 0.0007859162811655551,
      "step": 5810,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0059545118361711506,
      "weighted_relation_loss": 0.0059545118361711506
    },
    {
      "action_query_norm": 466.00458984375,
      "action_relation_loss": 0.0052925815340131525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.444340505144996,
      "flow_matching_loss": 0.055828011408448217,
      "grad_norm": 0.1406564712524414,
      "learning_rate": 9.44397058963929e-05,
      "loss": 0.0619,
      "relation_loss": 0.0060693829786032435,
      "selected_pairs": 27059.5,
      "state_query_norm": 843.8024353027344,
      "state_relation_loss": 0.000776801467873156,
      "step": 5820,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0060693829786032435,
      "weighted_relation_loss": 0.0060693829786032435
    },
    {
      "action_query_norm": 462.4892852783203,
      "action_relation_loss": 0.0055858772713690994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.453695042095417,
      "flow_matching_loss": 0.0561879463493824,
      "grad_norm": 0.1258949637413025,
      "learning_rate": 9.441441902669056e-05,
      "loss": 0.0625,
      "relation_loss": 0.006356418365612626,
      "selected_pairs": 26595.7,
      "state_query_norm": 830.7036926269532,
      "state_relation_loss": 0.0007705410884227603,
      "step": 5830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.006356418365612626,
      "weighted_relation_loss": 0.006356418365612626
    },
    {
      "action_query_norm": 470.3818664550781,
      "action_relation_loss": 0.005479542352259159,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.463049579045837,
      "flow_matching_loss": 0.05720206312835217,
      "grad_norm": 0.1395164132118225,
      "learning_rate": 9.43890781892811e-05,
      "loss": 0.0635,
      "relation_loss": 0.00626224591396749,
      "selected_pairs": 27014.4,
      "state_query_norm": 839.7881042480469,
      "state_relation_loss": 0.0007827036024536937,
      "step": 5840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.00626224591396749,
      "weighted_relation_loss": 0.00626224591396749
    },
    {
      "action_query_norm": 469.2292175292969,
      "action_relation_loss": 0.005313166696578264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.472404115996258,
      "flow_matching_loss": 0.05361331515014171,
      "grad_norm": 0.14176174998283386,
      "learning_rate": 9.436368341495603e-05,
      "loss": 0.0597,
      "relation_loss": 0.006088001653552055,
      "selected_pairs": 26949.8,
      "state_query_norm": 840.0862548828125,
      "state_relation_loss": 0.0007748349686153233,
      "step": 5850,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006088001653552055,
      "weighted_relation_loss": 0.006088001653552055
    },
    {
      "action_query_norm": 465.41236572265626,
      "action_relation_loss": 0.005568659910932183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.481758652946679,
      "flow_matching_loss": 0.05494086854159832,
      "grad_norm": 0.17052949965000153,
      "learning_rate": 9.43382347345724e-05,
      "loss": 0.0613,
      "relation_loss": 0.006377017544582486,
      "selected_pairs": 27335.9,
      "state_query_norm": 837.4747741699218,
      "state_relation_loss": 0.0008083575870841741,
      "step": 5860,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.006377017544582486,
      "weighted_relation_loss": 0.006377017544582486
    },
    {
      "action_query_norm": 464.873828125,
      "action_relation_loss": 0.005382371554151178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.4911131898971,
      "flow_matching_loss": 0.05339914858341217,
      "grad_norm": 0.11837467551231384,
      "learning_rate": 9.431273217905273e-05,
      "loss": 0.0596,
      "relation_loss": 0.006153704272583127,
      "selected_pairs": 26948.9,
      "state_query_norm": 846.9619323730469,
      "state_relation_loss": 0.000771332741715014,
      "step": 5870,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006153704272583127,
      "weighted_relation_loss": 0.006153704272583127
    },
    {
      "action_query_norm": 473.73614807128905,
      "action_relation_loss": 0.005064943712204695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.500467726847521,
      "flow_matching_loss": 0.05195935294032097,
      "grad_norm": 0.1318739801645279,
      "learning_rate": 9.428717577938504e-05,
      "loss": 0.0578,
      "relation_loss": 0.005819746758788824,
      "selected_pairs": 26688.4,
      "state_query_norm": 842.0244812011719,
      "state_relation_loss": 0.0007548030116595328,
      "step": 5880,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005819746758788824,
      "weighted_relation_loss": 0.005819746758788824
    },
    {
      "action_query_norm": 476.9430847167969,
      "action_relation_loss": 0.005042365100234747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.509822263797942,
      "flow_matching_loss": 0.054543886333703995,
      "grad_norm": 0.152011439204216,
      "learning_rate": 9.426156556662276e-05,
      "loss": 0.0603,
      "relation_loss": 0.0057931256480515,
      "selected_pairs": 26641.6,
      "state_query_norm": 858.592724609375,
      "state_relation_loss": 0.0007507605478167534,
      "step": 5890,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0057931256480515,
      "weighted_relation_loss": 0.0057931256480515
    },
    {
      "action_query_norm": 467.5791839599609,
      "action_relation_loss": 0.005314667942002416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.519176800748363,
      "flow_matching_loss": 0.054591409489512445,
      "grad_norm": 0.15728791058063507,
      "learning_rate": 9.423590157188474e-05,
      "loss": 0.0606,
      "relation_loss": 0.006041801394894719,
      "selected_pairs": 26851.1,
      "state_query_norm": 840.6946044921875,
      "state_relation_loss": 0.000727133423788473,
      "step": 5900,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006041801394894719,
      "weighted_relation_loss": 0.006041801394894719
    },
    {
      "action_query_norm": 472.2828308105469,
      "action_relation_loss": 0.005192613555118441,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.528531337698784,
      "flow_matching_loss": 0.054881739616394046,
      "grad_norm": 0.12282660603523254,
      "learning_rate": 9.421018382635513e-05,
      "loss": 0.0608,
      "relation_loss": 0.005960326083004475,
      "selected_pairs": 26630.4,
      "state_query_norm": 847.7914245605468,
      "state_relation_loss": 0.0007677124929614365,
      "step": 5910,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005960326083004475,
      "weighted_relation_loss": 0.005960326083004475
    },
    {
      "action_query_norm": 473.9700225830078,
      "action_relation_loss": 0.005180678842589259,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.537885874649205,
      "flow_matching_loss": 0.05381385572254658,
      "grad_norm": 0.12714020907878876,
      "learning_rate": 9.418441236128343e-05,
      "loss": 0.0597,
      "relation_loss": 0.005915902089327574,
      "selected_pairs": 26711.6,
      "state_query_norm": 855.0571350097656,
      "state_relation_loss": 0.0007352232176344841,
      "step": 5920,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005915902089327574,
      "weighted_relation_loss": 0.005915902089327574
    },
    {
      "action_query_norm": 469.3923065185547,
      "action_relation_loss": 0.005227384902536869,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.547240411599626,
      "flow_matching_loss": 0.05263854973018169,
      "grad_norm": 0.14610961079597473,
      "learning_rate": 9.41585872079844e-05,
      "loss": 0.0586,
      "relation_loss": 0.005964904511347413,
      "selected_pairs": 27744.1,
      "state_query_norm": 842.3016906738281,
      "state_relation_loss": 0.0007375195564236492,
      "step": 5930,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.005964904511347413,
      "weighted_relation_loss": 0.005964904511347413
    },
    {
      "action_query_norm": 465.8488037109375,
      "action_relation_loss": 0.005196061171591282,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.556594948550047,
      "flow_matching_loss": 0.052957385405898096,
      "grad_norm": 0.14380818605422974,
      "learning_rate": 9.413270839783802e-05,
      "loss": 0.0589,
      "relation_loss": 0.005948197096586227,
      "selected_pairs": 26741.4,
      "state_query_norm": 840.1978637695313,
      "state_relation_loss": 0.0007521359133534133,
      "step": 5940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005948197096586227,
      "weighted_relation_loss": 0.005948197096586227
    },
    {
      "action_query_norm": 478.2765777587891,
      "action_relation_loss": 0.005018647154793143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.565949485500468,
      "flow_matching_loss": 0.0537079207599163,
      "grad_norm": 0.15882998704910278,
      "learning_rate": 9.41067759622895e-05,
      "loss": 0.0594,
      "relation_loss": 0.005732351122424007,
      "selected_pairs": 26776.6,
      "state_query_norm": 855.2108337402344,
      "state_relation_loss": 0.0007137040665838867,
      "step": 5950,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005732351122424007,
      "weighted_relation_loss": 0.005732351122424007
    },
    {
      "action_query_norm": 481.2045013427734,
      "action_relation_loss": 0.0051527534611523155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.575304022450888,
      "flow_matching_loss": 0.0530270803719759,
      "grad_norm": 0.15960603952407837,
      "learning_rate": 9.408078993284917e-05,
      "loss": 0.0589,
      "relation_loss": 0.005878587532788515,
      "selected_pairs": 26529.1,
      "state_query_norm": 857.4685180664062,
      "state_relation_loss": 0.0007258340541739017,
      "step": 5960,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.005878587532788515,
      "weighted_relation_loss": 0.005878587532788515
    },
    {
      "action_query_norm": 475.44219360351565,
      "action_relation_loss": 0.005174315487965942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.584658559401309,
      "flow_matching_loss": 0.054932742938399316,
      "grad_norm": 0.15074174106121063,
      "learning_rate": 9.405475034109254e-05,
      "loss": 0.0609,
      "relation_loss": 0.005946570634841919,
      "selected_pairs": 26644.7,
      "state_query_norm": 867.2316467285157,
      "state_relation_loss": 0.0007722551818005741,
      "step": 5970,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.005946570634841919,
      "weighted_relation_loss": 0.005946570634841919
    },
    {
      "action_query_norm": 478.0249328613281,
      "action_relation_loss": 0.00504305143840611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.5940130963517305,
      "flow_matching_loss": 0.05173889435827732,
      "grad_norm": 0.16748292744159698,
      "learning_rate": 9.402865721866015e-05,
      "loss": 0.0575,
      "relation_loss": 0.005773041071370244,
      "selected_pairs": 26301.0,
      "state_query_norm": 844.0310791015625,
      "state_relation_loss": 0.0007299896504264324,
      "step": 5980,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.005773041071370244,
      "weighted_relation_loss": 0.005773041071370244
    },
    {
      "action_query_norm": 484.2657470703125,
      "action_relation_loss": 0.005009410437196493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.603367633302152,
      "flow_matching_loss": 0.0521736815571785,
      "grad_norm": 0.14467976987361908,
      "learning_rate": 9.400251059725762e-05,
      "loss": 0.0579,
      "relation_loss": 0.005745933344587684,
      "selected_pairs": 26803.1,
      "state_query_norm": 842.8569519042969,
      "state_relation_loss": 0.0007365229132119567,
      "step": 5990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005745933344587684,
      "weighted_relation_loss": 0.005745933344587684
    },
    {
      "action_query_norm": 470.38294982910156,
      "action_relation_loss": 0.005311468336731196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.612722170252573,
      "flow_matching_loss": 0.05459766313433647,
      "grad_norm": 0.15081936120986938,
      "learning_rate": 9.397631050865554e-05,
      "loss": 0.0606,
      "relation_loss": 0.0060282938182353975,
      "selected_pairs": 26915.3,
      "state_query_norm": 864.0981140136719,
      "state_relation_loss": 0.0007168254989665002,
      "step": 6000,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0060282938182353975,
      "weighted_relation_loss": 0.0060282938182353975
    },
    {
      "action_query_norm": 471.78180236816405,
      "action_relation_loss": 0.0053554278332740065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.622076707202994,
      "flow_matching_loss": 0.05330076962709427,
      "grad_norm": 0.13099080324172974,
      "learning_rate": 9.395005698468949e-05,
      "loss": 0.0594,
      "relation_loss": 0.0061116405762732025,
      "selected_pairs": 27026.6,
      "state_query_norm": 851.2313537597656,
      "state_relation_loss": 0.0007562127022538335,
      "step": 6010,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0061116405762732025,
      "weighted_relation_loss": 0.0061116405762732025
    },
    {
      "action_query_norm": 471.91624450683594,
      "action_relation_loss": 0.005327718611806631,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.631431244153415,
      "flow_matching_loss": 0.05315525829792023,
      "grad_norm": 0.2227051705121994,
      "learning_rate": 9.392375005726e-05,
      "loss": 0.0592,
      "relation_loss": 0.0060491005424410105,
      "selected_pairs": 27200.2,
      "state_query_norm": 877.3921569824219,
      "state_relation_loss": 0.0007213819189928472,
      "step": 6020,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0060491005424410105,
      "weighted_relation_loss": 0.0060491005424410105
    },
    {
      "action_query_norm": 473.7975189208984,
      "action_relation_loss": 0.00505557069554925,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.640785781103835,
      "flow_matching_loss": 0.051817454397678375,
      "grad_norm": 0.17184138298034668,
      "learning_rate": 9.389738975833242e-05,
      "loss": 0.0576,
      "relation_loss": 0.005804294673725963,
      "selected_pairs": 26696.7,
      "state_query_norm": 863.3275329589844,
      "state_relation_loss": 0.0007487239956390113,
      "step": 6030,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005804294673725963,
      "weighted_relation_loss": 0.005804294673725963
    },
    {
      "action_query_norm": 477.28301391601565,
      "action_relation_loss": 0.005287117976695299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.650140318054256,
      "flow_matching_loss": 0.05385294407606125,
      "grad_norm": 0.18891513347625732,
      "learning_rate": 9.387097611993707e-05,
      "loss": 0.0599,
      "relation_loss": 0.006010074866935611,
      "selected_pairs": 26919.0,
      "state_query_norm": 855.2084106445312,
      "state_relation_loss": 0.0007229568960610777,
      "step": 6040,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006010074866935611,
      "weighted_relation_loss": 0.006010074866935611
    },
    {
      "action_query_norm": 473.3245086669922,
      "action_relation_loss": 0.0054121390916407105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.659494855004677,
      "flow_matching_loss": 0.05447323732078076,
      "grad_norm": 0.152494415640831,
      "learning_rate": 9.384450917416894e-05,
      "loss": 0.0607,
      "relation_loss": 0.006179752014577389,
      "selected_pairs": 27009.5,
      "state_query_norm": 864.4318176269531,
      "state_relation_loss": 0.0007676128763705492,
      "step": 6050,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.006179752014577389,
      "weighted_relation_loss": 0.006179752014577389
    },
    {
      "action_query_norm": 475.64979248046876,
      "action_relation_loss": 0.005133641930297017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.668849391955098,
      "flow_matching_loss": 0.054361872002482416,
      "grad_norm": 0.13073262572288513,
      "learning_rate": 9.381798895318792e-05,
      "loss": 0.0603,
      "relation_loss": 0.005904457671567798,
      "selected_pairs": 26828.1,
      "state_query_norm": 856.4446960449219,
      "state_relation_loss": 0.0007708156947046519,
      "step": 6060,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005904457671567798,
      "weighted_relation_loss": 0.005904457671567798
    },
    {
      "action_query_norm": 473.2085327148437,
      "action_relation_loss": 0.005086251627653837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.6782039289055195,
      "flow_matching_loss": 0.05193582065403461,
      "grad_norm": 0.1842425912618637,
      "learning_rate": 9.379141548921853e-05,
      "loss": 0.0578,
      "relation_loss": 0.005821184068918228,
      "selected_pairs": 27220.6,
      "state_query_norm": 852.1326599121094,
      "state_relation_loss": 0.0007349324354436248,
      "step": 6070,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.005821184068918228,
      "weighted_relation_loss": 0.005821184068918228
    },
    {
      "action_query_norm": 464.3480499267578,
      "action_relation_loss": 0.005114008160308004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.68755846585594,
      "flow_matching_loss": 0.0564173024147749,
      "grad_norm": 0.11968554556369781,
      "learning_rate": 9.376478881455009e-05,
      "loss": 0.0622,
      "relation_loss": 0.005829934682697058,
      "selected_pairs": 27269.6,
      "state_query_norm": 853.5963989257813,
      "state_relation_loss": 0.0007159265631344169,
      "step": 6080,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005829934682697058,
      "weighted_relation_loss": 0.005829934682697058
    },
    {
      "action_query_norm": 468.0746124267578,
      "action_relation_loss": 0.005063014663755893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.696913002806361,
      "flow_matching_loss": 0.054031073674559595,
      "grad_norm": 0.1185663640499115,
      "learning_rate": 9.373810896153647e-05,
      "loss": 0.0598,
      "relation_loss": 0.005773181188851595,
      "selected_pairs": 27342.2,
      "state_query_norm": 850.3782348632812,
      "state_relation_loss": 0.0007101665425579995,
      "step": 6090,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.005773181188851595,
      "weighted_relation_loss": 0.005773181188851595
    },
    {
      "action_query_norm": 478.0910675048828,
      "action_relation_loss": 0.004827944235876203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.706267539756782,
      "flow_matching_loss": 0.05486654937267303,
      "grad_norm": 0.15778806805610657,
      "learning_rate": 9.371137596259623e-05,
      "loss": 0.0604,
      "relation_loss": 0.005541460867971182,
      "selected_pairs": 27346.2,
      "state_query_norm": 860.6823547363281,
      "state_relation_loss": 0.0007135166088119149,
      "step": 6100,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.005541460867971182,
      "weighted_relation_loss": 0.005541460867971182
    },
    {
      "action_query_norm": 480.369677734375,
      "action_relation_loss": 0.0052900174632668495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.715622076707203,
      "flow_matching_loss": 0.05564243830740452,
      "grad_norm": 0.1686747819185257,
      "learning_rate": 9.368458985021248e-05,
      "loss": 0.0617,
      "relation_loss": 0.006037489790469408,
      "selected_pairs": 26277.3,
      "state_query_norm": 871.5285949707031,
      "state_relation_loss": 0.0007474722864571959,
      "step": 6110,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.006037489790469408,
      "weighted_relation_loss": 0.006037489790469408
    },
    {
      "action_query_norm": 476.5117980957031,
      "action_relation_loss": 0.005299742054194212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.724976613657624,
      "flow_matching_loss": 0.056403474882245064,
      "grad_norm": 0.18061181902885437,
      "learning_rate": 9.365775065693287e-05,
      "loss": 0.0625,
      "relation_loss": 0.00605675783008337,
      "selected_pairs": 26673.4,
      "state_query_norm": 857.583544921875,
      "state_relation_loss": 0.0007570157817099243,
      "step": 6120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00605675783008337,
      "weighted_relation_loss": 0.00605675783008337
    },
    {
      "action_query_norm": 475.1204467773438,
      "action_relation_loss": 0.005254966905340552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.734331150608045,
      "flow_matching_loss": 0.05304862409830093,
      "grad_norm": 0.16158875823020935,
      "learning_rate": 9.363085841536957e-05,
      "loss": 0.0591,
      "relation_loss": 0.0060053889639675615,
      "selected_pairs": 27148.0,
      "state_query_norm": 847.0239135742188,
      "state_relation_loss": 0.0007504221401177346,
      "step": 6130,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0060053889639675615,
      "weighted_relation_loss": 0.0060053889639675615
    },
    {
      "action_query_norm": 476.7716400146484,
      "action_relation_loss": 0.005204573832452297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.743685687558466,
      "flow_matching_loss": 0.05340455956757069,
      "grad_norm": 0.11820542067289352,
      "learning_rate": 9.360391315819917e-05,
      "loss": 0.0594,
      "relation_loss": 0.005951399728655815,
      "selected_pairs": 27009.9,
      "state_query_norm": 842.7771850585938,
      "state_relation_loss": 0.0007468259136658162,
      "step": 6140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.005951399728655815,
      "weighted_relation_loss": 0.005951399728655815
    },
    {
      "action_query_norm": 478.64625854492186,
      "action_relation_loss": 0.0049792179372161625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.7530402245088865,
      "flow_matching_loss": 0.052140241116285325,
      "grad_norm": 0.14792285859584808,
      "learning_rate": 9.357691491816269e-05,
      "loss": 0.0578,
      "relation_loss": 0.005704059870913625,
      "selected_pairs": 27107.1,
      "state_query_norm": 854.6438537597656,
      "state_relation_loss": 0.0007248419802635908,
      "step": 6150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005704059870913625,
      "weighted_relation_loss": 0.005704059870913625
    },
    {
      "action_query_norm": 478.58037719726565,
      "action_relation_loss": 0.00498176459223032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.762394761459308,
      "flow_matching_loss": 0.05242051854729653,
      "grad_norm": 0.14184020459651947,
      "learning_rate": 9.354986372806557e-05,
      "loss": 0.0581,
      "relation_loss": 0.0057014207355678085,
      "selected_pairs": 26159.3,
      "state_query_norm": 862.0396362304688,
      "state_relation_loss": 0.0007196561200544238,
      "step": 6160,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0057014207355678085,
      "weighted_relation_loss": 0.0057014207355678085
    },
    {
      "action_query_norm": 477.4999664306641,
      "action_relation_loss": 0.004973091883584857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.771749298409729,
      "flow_matching_loss": 0.05009227842092514,
      "grad_norm": 0.16267287731170654,
      "learning_rate": 9.352275962077752e-05,
      "loss": 0.0558,
      "relation_loss": 0.005713463574647903,
      "selected_pairs": 26877.3,
      "state_query_norm": 846.4531677246093,
      "state_relation_loss": 0.0007403716736007482,
      "step": 6170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005713463574647903,
      "weighted_relation_loss": 0.005713463574647903
    },
    {
      "action_query_norm": 482.5142852783203,
      "action_relation_loss": 0.004980779089964926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.78110383536015,
      "flow_matching_loss": 0.051833585277199744,
      "grad_norm": 0.15221378207206726,
      "learning_rate": 9.349560262923262e-05,
      "loss": 0.0575,
      "relation_loss": 0.0057014379184693095,
      "selected_pairs": 26895.0,
      "state_query_norm": 851.28564453125,
      "state_relation_loss": 0.0007206587877590209,
      "step": 6180,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0057014379184693095,
      "weighted_relation_loss": 0.0057014379184693095
    },
    {
      "action_query_norm": 489.3518402099609,
      "action_relation_loss": 0.005156535701826215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.790458372310571,
      "flow_matching_loss": 0.0510355144739151,
      "grad_norm": 0.14909175038337708,
      "learning_rate": 9.346839278642913e-05,
      "loss": 0.0569,
      "relation_loss": 0.005899801524356008,
      "selected_pairs": 26651.4,
      "state_query_norm": 857.7672973632813,
      "state_relation_loss": 0.0007432659098412841,
      "step": 6190,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.005899801524356008,
      "weighted_relation_loss": 0.005899801524356008
    },
    {
      "action_query_norm": 490.8187194824219,
      "action_relation_loss": 0.0050509095191955565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.799812909260991,
      "flow_matching_loss": 0.05172931328415871,
      "grad_norm": 0.14860183000564575,
      "learning_rate": 9.344113012542963e-05,
      "loss": 0.0575,
      "relation_loss": 0.005775970220565796,
      "selected_pairs": 26765.2,
      "state_query_norm": 872.1420715332031,
      "state_relation_loss": 0.0007250607595779002,
      "step": 6200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.005775970220565796,
      "weighted_relation_loss": 0.005775970220565796
    },
    {
      "action_query_norm": 487.2393035888672,
      "action_relation_loss": 0.004988230764865875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.809167446211412,
      "flow_matching_loss": 0.0532674677670002,
      "grad_norm": 0.13901805877685547,
      "learning_rate": 9.341381467936079e-05,
      "loss": 0.059,
      "relation_loss": 0.00573272742331028,
      "selected_pairs": 27359.0,
      "state_query_norm": 856.9559387207031,
      "state_relation_loss": 0.0007444967166520655,
      "step": 6210,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.00573272742331028,
      "weighted_relation_loss": 0.00573272742331028
    },
    {
      "action_query_norm": 484.7963531494141,
      "action_relation_loss": 0.005162856774404645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.818521983161833,
      "flow_matching_loss": 0.05397437885403633,
      "grad_norm": 0.14940552413463593,
      "learning_rate": 9.338644648141346e-05,
      "loss": 0.0599,
      "relation_loss": 0.0059048178605735305,
      "selected_pairs": 26343.6,
      "state_query_norm": 880.0715148925781,
      "state_relation_loss": 0.0007419610628858209,
      "step": 6220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0059048178605735305,
      "weighted_relation_loss": 0.0059048178605735305
    },
    {
      "action_query_norm": 482.72760314941405,
      "action_relation_loss": 0.005082461796700954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.827876520112254,
      "flow_matching_loss": 0.052703120931983,
      "grad_norm": 0.14924202859401703,
      "learning_rate": 9.335902556484257e-05,
      "loss": 0.0585,
      "relation_loss": 0.005814466951414942,
      "selected_pairs": 27335.0,
      "state_query_norm": 863.8123352050782,
      "state_relation_loss": 0.000732005067402497,
      "step": 6230,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.005814466951414942,
      "weighted_relation_loss": 0.005814466951414942
    },
    {
      "action_query_norm": 479.4371307373047,
      "action_relation_loss": 0.004796852823346853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.8372310570626755,
      "flow_matching_loss": 0.05395405255258083,
      "grad_norm": 0.18206952512264252,
      "learning_rate": 9.333155196296712e-05,
      "loss": 0.0594,
      "relation_loss": 0.005491363303735852,
      "selected_pairs": 26590.5,
      "state_query_norm": 874.754443359375,
      "state_relation_loss": 0.0006945104978512973,
      "step": 6240,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.005491363303735852,
      "weighted_relation_loss": 0.005491363303735852
    },
    {
      "action_query_norm": 480.09963073730466,
      "action_relation_loss": 0.005026297643780708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.846585594013097,
      "flow_matching_loss": 0.05410178489983082,
      "grad_norm": 0.17936943471431732,
      "learning_rate": 9.330402570917016e-05,
      "loss": 0.0598,
      "relation_loss": 0.005743918241932988,
      "selected_pairs": 27206.4,
      "state_query_norm": 863.5077209472656,
      "state_relation_loss": 0.0007176205399446189,
      "step": 6250,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.005743918241932988,
      "weighted_relation_loss": 0.005743918241932988
    },
    {
      "action_query_norm": 479.8121856689453,
      "action_relation_loss": 0.0048673192504793406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.855940130963518,
      "flow_matching_loss": 0.053176290541887286,
      "grad_norm": 0.1406252682209015,
      "learning_rate": 9.327644683689865e-05,
      "loss": 0.0588,
      "relation_loss": 0.005595484841614961,
      "selected_pairs": 26826.8,
      "state_query_norm": 862.6741149902343,
      "state_relation_loss": 0.0007281655271071941,
      "step": 6260,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005595484841614961,
      "weighted_relation_loss": 0.005595484841614961
    },
    {
      "action_query_norm": 482.49480895996095,
      "action_relation_loss": 0.004835339263081551,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.865294667913938,
      "flow_matching_loss": 0.05256818495690822,
      "grad_norm": 0.13104765117168427,
      "learning_rate": 9.324881537966354e-05,
      "loss": 0.0581,
      "relation_loss": 0.005580549500882626,
      "selected_pairs": 26804.9,
      "state_query_norm": 859.7901306152344,
      "state_relation_loss": 0.0007452102028764785,
      "step": 6270,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005580549500882626,
      "weighted_relation_loss": 0.005580549500882626
    },
    {
      "action_query_norm": 480.1590118408203,
      "action_relation_loss": 0.005178187321871519,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.874649204864359,
      "flow_matching_loss": 0.0541033037006855,
      "grad_norm": 0.13681402802467346,
      "learning_rate": 9.322113137103964e-05,
      "loss": 0.06,
      "relation_loss": 0.005926024680957198,
      "selected_pairs": 26894.4,
      "state_query_norm": 875.1646240234375,
      "state_relation_loss": 0.0007478373183403165,
      "step": 6280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005926024680957198,
      "weighted_relation_loss": 0.005926024680957198
    },
    {
      "action_query_norm": 478.5542785644531,
      "action_relation_loss": 0.005027100397273898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.88400374181478,
      "flow_matching_loss": 0.05434209704399109,
      "grad_norm": 0.16151508688926697,
      "learning_rate": 9.319339484466564e-05,
      "loss": 0.0601,
      "relation_loss": 0.00573173575103283,
      "selected_pairs": 26984.7,
      "state_query_norm": 858.963232421875,
      "state_relation_loss": 0.0007046352839097381,
      "step": 6290,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00573173575103283,
      "weighted_relation_loss": 0.00573173575103283
    },
    {
      "action_query_norm": 470.6531036376953,
      "action_relation_loss": 0.005038583651185036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.893358278765201,
      "flow_matching_loss": 0.053406591340899465,
      "grad_norm": 0.1514822393655777,
      "learning_rate": 9.316560583424403e-05,
      "loss": 0.0592,
      "relation_loss": 0.005758047010749579,
      "selected_pairs": 26828.0,
      "state_query_norm": 848.2870422363281,
      "state_relation_loss": 0.0007194633421022445,
      "step": 6300,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005758047010749579,
      "weighted_relation_loss": 0.005758047010749579
    },
    {
      "action_query_norm": 469.5035095214844,
      "action_relation_loss": 0.004873245069757104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.902712815715622,
      "flow_matching_loss": 0.053991790115833285,
      "grad_norm": 0.15646934509277344,
      "learning_rate": 9.313776437354108e-05,
      "loss": 0.0596,
      "relation_loss": 0.005586153734475375,
      "selected_pairs": 27181.5,
      "state_query_norm": 865.0834533691407,
      "state_relation_loss": 0.0007129086414352059,
      "step": 6310,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.005586153734475375,
      "weighted_relation_loss": 0.005586153734475375
    },
    {
      "action_query_norm": 478.0000335693359,
      "action_relation_loss": 0.004879901465028524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.9120673526660426,
      "flow_matching_loss": 0.05301421135663986,
      "grad_norm": 0.16243061423301697,
      "learning_rate": 9.31098704963868e-05,
      "loss": 0.0586,
      "relation_loss": 0.005623655114322901,
      "selected_pairs": 26965.1,
      "state_query_norm": 865.1051940917969,
      "state_relation_loss": 0.0007437536085490137,
      "step": 6320,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.005623655114322901,
      "weighted_relation_loss": 0.005623655114322901
    },
    {
      "action_query_norm": 483.5193359375,
      "action_relation_loss": 0.0050544542726129295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.921421889616464,
      "flow_matching_loss": 0.050887460634112355,
      "grad_norm": 0.15816763043403625,
      "learning_rate": 9.308192423667485e-05,
      "loss": 0.0567,
      "relation_loss": 0.005777340987697244,
      "selected_pairs": 26893.4,
      "state_query_norm": 855.0672668457031,
      "state_relation_loss": 0.0007228867034427822,
      "step": 6330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005777340987697244,
      "weighted_relation_loss": 0.005777340987697244
    },
    {
      "action_query_norm": 484.56841735839845,
      "action_relation_loss": 0.005240146163851022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.930776426566885,
      "flow_matching_loss": 0.05334760099649429,
      "grad_norm": 0.13786724209785461,
      "learning_rate": 9.305392562836262e-05,
      "loss": 0.0593,
      "relation_loss": 0.005972923059016466,
      "selected_pairs": 27224.4,
      "state_query_norm": 867.2612487792969,
      "state_relation_loss": 0.0007327769475523382,
      "step": 6340,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.005972923059016466,
      "weighted_relation_loss": 0.005972923059016466
    },
    {
      "action_query_norm": 483.4077392578125,
      "action_relation_loss": 0.005163508560508489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.940130963517306,
      "flow_matching_loss": 0.052266566827893254,
      "grad_norm": 0.20075905323028564,
      "learning_rate": 9.3025874705471e-05,
      "loss": 0.0582,
      "relation_loss": 0.00590708558447659,
      "selected_pairs": 26664.6,
      "state_query_norm": 856.0344360351562,
      "state_relation_loss": 0.0007435770065058022,
      "step": 6350,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00590708558447659,
      "weighted_relation_loss": 0.00590708558447659
    },
    {
      "action_query_norm": 474.8812622070312,
      "action_relation_loss": 0.0047948378138244156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.949485500467727,
      "flow_matching_loss": 0.05324158146977424,
      "grad_norm": 0.21285966038703918,
      "learning_rate": 9.299777150208457e-05,
      "loss": 0.0588,
      "relation_loss": 0.005530682438984513,
      "selected_pairs": 26341.9,
      "state_query_norm": 863.1805419921875,
      "state_relation_loss": 0.0007358446600846946,
      "step": 6360,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.005530682438984513,
      "weighted_relation_loss": 0.005530682438984513
    },
    {
      "action_query_norm": 482.64214782714845,
      "action_relation_loss": 0.005109423352405429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.958840037418148,
      "flow_matching_loss": 0.05273670069873333,
      "grad_norm": 0.19934779405593872,
      "learning_rate": 9.296961605235133e-05,
      "loss": 0.0586,
      "relation_loss": 0.005848778085783124,
      "selected_pairs": 27452.1,
      "state_query_norm": 872.6587097167969,
      "state_relation_loss": 0.0007393547624815255,
      "step": 6370,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.005848778085783124,
      "weighted_relation_loss": 0.005848778085783124
    },
    {
      "action_query_norm": 479.19990234375,
      "action_relation_loss": 0.0051165463402867315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.968194574368569,
      "flow_matching_loss": 0.05352006815373898,
      "grad_norm": 0.1776047796010971,
      "learning_rate": 9.29414083904828e-05,
      "loss": 0.0594,
      "relation_loss": 0.0058400042355060576,
      "selected_pairs": 27312.7,
      "state_query_norm": 883.0159606933594,
      "state_relation_loss": 0.0007234578835777939,
      "step": 6380,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0058400042355060576,
      "weighted_relation_loss": 0.0058400042355060576
    },
    {
      "action_query_norm": 484.17112426757814,
      "action_relation_loss": 0.004988884180784225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.977549111318989,
      "flow_matching_loss": 0.05502255819737911,
      "grad_norm": 0.15671215951442719,
      "learning_rate": 9.291314855075397e-05,
      "loss": 0.0608,
      "relation_loss": 0.005741688841953874,
      "selected_pairs": 26806.3,
      "state_query_norm": 869.6231994628906,
      "state_relation_loss": 0.0007528046669904142,
      "step": 6390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005741688841953874,
      "weighted_relation_loss": 0.005741688841953874
    },
    {
      "action_query_norm": 492.3347473144531,
      "action_relation_loss": 0.004884851071983576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.9869036482694105,
      "flow_matching_loss": 0.05394231639802456,
      "grad_norm": 0.17040711641311646,
      "learning_rate": 9.288483656750322e-05,
      "loss": 0.0595,
      "relation_loss": 0.005590220773592591,
      "selected_pairs": 26798.4,
      "state_query_norm": 875.0891723632812,
      "state_relation_loss": 0.0007053697074297816,
      "step": 6400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005590220773592591,
      "weighted_relation_loss": 0.005590220773592591
    },
    {
      "action_query_norm": 484.9919952392578,
      "action_relation_loss": 0.005089999921619892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.996258185219832,
      "flow_matching_loss": 0.05257516615092754,
      "grad_norm": 0.17929556965827942,
      "learning_rate": 9.285647247513225e-05,
      "loss": 0.0584,
      "relation_loss": 0.005830747587606311,
      "selected_pairs": 26619.8,
      "state_query_norm": 874.0923217773437,
      "state_relation_loss": 0.0007407477241940796,
      "step": 6410,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005830747587606311,
      "weighted_relation_loss": 0.005830747587606311
    },
    {
      "action_query_norm": 486.87493896484375,
      "action_relation_loss": 0.004997379425913095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.005612722170253,
      "flow_matching_loss": 0.05114921182394028,
      "grad_norm": 0.14712552726268768,
      "learning_rate": 9.282805630810614e-05,
      "loss": 0.0569,
      "relation_loss": 0.005718148592859506,
      "selected_pairs": 23926.7,
      "state_query_norm": 879.7644592285156,
      "state_relation_loss": 0.0007207691262010485,
      "step": 6420,
      "valid_action_samples": 212.7,
      "weighted_auxiliary_loss": 0.005718148592859506,
      "weighted_relation_loss": 0.005718148592859506
    },
    {
      "action_query_norm": 489.16159973144534,
      "action_relation_loss": 0.0049700484611094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.014967259120674,
      "flow_matching_loss": 0.05158201120793819,
      "grad_norm": 0.15651161968708038,
      "learning_rate": 9.279958810095317e-05,
      "loss": 0.0573,
      "relation_loss": 0.005707523738965392,
      "selected_pairs": 26690.3,
      "state_query_norm": 881.3081481933593,
      "state_relation_loss": 0.0007374752371106297,
      "step": 6430,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005707523738965392,
      "weighted_relation_loss": 0.005707523738965392
    },
    {
      "action_query_norm": 481.5690093994141,
      "action_relation_loss": 0.004823976522311568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.024321796071095,
      "flow_matching_loss": 0.052566719427704814,
      "grad_norm": 0.15035800635814667,
      "learning_rate": 9.277106788826494e-05,
      "loss": 0.0581,
      "relation_loss": 0.005540875485166907,
      "selected_pairs": 26732.6,
      "state_query_norm": 864.1220458984375,
      "state_relation_loss": 0.0007168989744968712,
      "step": 6440,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005540875485166907,
      "weighted_relation_loss": 0.005540875485166907
    },
    {
      "action_query_norm": 479.2961791992187,
      "action_relation_loss": 0.004742534132674336,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.033676333021515,
      "flow_matching_loss": 0.05050456337630749,
      "grad_norm": 0.21205313503742218,
      "learning_rate": 9.274249570469618e-05,
      "loss": 0.0559,
      "relation_loss": 0.005439406586810946,
      "selected_pairs": 27291.4,
      "state_query_norm": 889.3051635742188,
      "state_relation_loss": 0.0006968724250327796,
      "step": 6450,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005439406586810946,
      "weighted_relation_loss": 0.005439406586810946
    },
    {
      "action_query_norm": 483.1458465576172,
      "action_relation_loss": 0.0048355442006140946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.043030869971936,
      "flow_matching_loss": 0.05326439067721367,
      "grad_norm": 0.1583632528781891,
      "learning_rate": 9.271387158496476e-05,
      "loss": 0.0588,
      "relation_loss": 0.005569527111947537,
      "selected_pairs": 27152.3,
      "state_query_norm": 883.9408996582031,
      "state_relation_loss": 0.0007339829404372722,
      "step": 6460,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.005569527111947537,
      "weighted_relation_loss": 0.005569527111947537
    },
    {
      "action_query_norm": 490.1340576171875,
      "action_relation_loss": 0.004939849535003304,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.052385406922357,
      "flow_matching_loss": 0.0518708061426878,
      "grad_norm": 0.13470147550106049,
      "learning_rate": 9.268519556385173e-05,
      "loss": 0.0575,
      "relation_loss": 0.0056472871452569965,
      "selected_pairs": 27086.6,
      "state_query_norm": 888.0078002929688,
      "state_relation_loss": 0.0007074375986121595,
      "step": 6470,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0056472871452569965,
      "weighted_relation_loss": 0.0056472871452569965
    },
    {
      "action_query_norm": 492.0332061767578,
      "action_relation_loss": 0.004782508825883269,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.061739943872778,
      "flow_matching_loss": 0.05266715474426746,
      "grad_norm": 0.11611851304769516,
      "learning_rate": 9.265646767620113e-05,
      "loss": 0.0582,
      "relation_loss": 0.005502604926005006,
      "selected_pairs": 27080.1,
      "state_query_norm": 870.4493103027344,
      "state_relation_loss": 0.0007200960244517773,
      "step": 6480,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005502604926005006,
      "weighted_relation_loss": 0.005502604926005006
    },
    {
      "action_query_norm": 492.5807373046875,
      "action_relation_loss": 0.004659044649451971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.0710944808231995,
      "flow_matching_loss": 0.05134519562125206,
      "grad_norm": 0.16066418588161469,
      "learning_rate": 9.262768795692006e-05,
      "loss": 0.0567,
      "relation_loss": 0.005397759703919291,
      "selected_pairs": 26996.7,
      "state_query_norm": 887.8652893066406,
      "state_relation_loss": 0.0007387150835711509,
      "step": 6490,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.005397759703919291,
      "weighted_relation_loss": 0.005397759703919291
    },
    {
      "action_query_norm": 490.54090270996096,
      "action_relation_loss": 0.004884669557213783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.080449017773621,
      "flow_matching_loss": 0.05238926038146019,
      "grad_norm": 0.1740349978208542,
      "learning_rate": 9.259885644097861e-05,
      "loss": 0.058,
      "relation_loss": 0.005633750231936574,
      "selected_pairs": 27008.8,
      "state_query_norm": 877.7345825195313,
      "state_relation_loss": 0.0007490806514397263,
      "step": 6500,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.005633750231936574,
      "weighted_relation_loss": 0.005633750231936574
    },
    {
      "action_query_norm": 484.59900817871096,
      "action_relation_loss": 0.004895105725154281,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.089803554724041,
      "flow_matching_loss": 0.05226634331047535,
      "grad_norm": 0.1474115252494812,
      "learning_rate": 9.256997316340976e-05,
      "loss": 0.0579,
      "relation_loss": 0.005587272625416517,
      "selected_pairs": 26528.5,
      "state_query_norm": 884.4525085449219,
      "state_relation_loss": 0.0006921669293660671,
      "step": 6510,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.005587272625416517,
      "weighted_relation_loss": 0.005587272625416517
    },
    {
      "action_query_norm": 489.30095825195315,
      "action_relation_loss": 0.004794761678203941,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.099158091674462,
      "flow_matching_loss": 0.05027394108474255,
      "grad_norm": 0.11129391193389893,
      "learning_rate": 9.254103815930943e-05,
      "loss": 0.0558,
      "relation_loss": 0.005519619258120656,
      "selected_pairs": 26506.6,
      "state_query_norm": 872.3284301757812,
      "state_relation_loss": 0.0007248575799167156,
      "step": 6520,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.005519619258120656,
      "weighted_relation_loss": 0.005519619258120656
    },
    {
      "action_query_norm": 493.8390625,
      "action_relation_loss": 0.004657819587737322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.108512628624883,
      "flow_matching_loss": 0.04980265311896801,
      "grad_norm": 0.12910011410713196,
      "learning_rate": 9.251205146383636e-05,
      "loss": 0.0552,
      "relation_loss": 0.005355869187042117,
      "selected_pairs": 26625.1,
      "state_query_norm": 899.4694702148438,
      "state_relation_loss": 0.0006980496284086258,
      "step": 6530,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005355869187042117,
      "weighted_relation_loss": 0.005355869187042117
    },
    {
      "action_query_norm": 486.9023803710937,
      "action_relation_loss": 0.004555209330283106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.117867165575304,
      "flow_matching_loss": 0.05237943232059479,
      "grad_norm": 0.13120563328266144,
      "learning_rate": 9.248301311221216e-05,
      "loss": 0.0576,
      "relation_loss": 0.005230382364243269,
      "selected_pairs": 27504.8,
      "state_query_norm": 884.6662353515625,
      "state_relation_loss": 0.0006751729582902044,
      "step": 6540,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.005230382364243269,
      "weighted_relation_loss": 0.005230382364243269
    },
    {
      "action_query_norm": 492.42335815429686,
      "action_relation_loss": 0.004491617437452078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.127221702525725,
      "flow_matching_loss": 0.049281572178006175,
      "grad_norm": 0.10903660207986832,
      "learning_rate": 9.245392313972115e-05,
      "loss": 0.0545,
      "relation_loss": 0.005181792983785271,
      "selected_pairs": 26917.4,
      "state_query_norm": 876.4909362792969,
      "state_relation_loss": 0.0006901754415594042,
      "step": 6550,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005181792983785271,
      "weighted_relation_loss": 0.005181792983785271
    },
    {
      "action_query_norm": 481.88865661621094,
      "action_relation_loss": 0.00466402848251164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.136576239476146,
      "flow_matching_loss": 0.0519346185028553,
      "grad_norm": 0.13406610488891602,
      "learning_rate": 9.242478158171037e-05,
      "loss": 0.0573,
      "relation_loss": 0.005361558916047215,
      "selected_pairs": 26735.5,
      "state_query_norm": 872.9086181640625,
      "state_relation_loss": 0.0006975305092055351,
      "step": 6560,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005361558916047215,
      "weighted_relation_loss": 0.005361558916047215
    },
    {
      "action_query_norm": 489.548828125,
      "action_relation_loss": 0.004661851329728961,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.1459307764265665,
      "flow_matching_loss": 0.051151742413640025,
      "grad_norm": 0.15014389157295227,
      "learning_rate": 9.239558847358957e-05,
      "loss": 0.0565,
      "relation_loss": 0.005374691635370255,
      "selected_pairs": 26903.7,
      "state_query_norm": 873.1643798828125,
      "state_relation_loss": 0.0007128402998205274,
      "step": 6570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005374691635370255,
      "weighted_relation_loss": 0.005374691635370255
    },
    {
      "action_query_norm": 487.0616180419922,
      "action_relation_loss": 0.004710411466658115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.155285313376988,
      "flow_matching_loss": 0.050497499853372575,
      "grad_norm": 0.18722863495349884,
      "learning_rate": 9.236634385083114e-05,
      "loss": 0.0559,
      "relation_loss": 0.005412560794502497,
      "selected_pairs": 26816.6,
      "state_query_norm": 886.8289794921875,
      "state_relation_loss": 0.0007021493162028492,
      "step": 6580,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005412560794502497,
      "weighted_relation_loss": 0.005412560794502497
    },
    {
      "action_query_norm": 494.80052185058594,
      "action_relation_loss": 0.004620319604873658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.164639850327409,
      "flow_matching_loss": 0.05258212499320507,
      "grad_norm": 0.1450129747390747,
      "learning_rate": 9.233704774897005e-05,
      "loss": 0.0579,
      "relation_loss": 0.0053045756183564665,
      "selected_pairs": 27292.0,
      "state_query_norm": 886.6431640625,
      "state_relation_loss": 0.0006842559610959143,
      "step": 6590,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0053045756183564665,
      "weighted_relation_loss": 0.0053045756183564665
    },
    {
      "action_query_norm": 490.04942932128904,
      "action_relation_loss": 0.004663331434130669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.17399438727783,
      "flow_matching_loss": 0.05020551159977913,
      "grad_norm": 0.13345447182655334,
      "learning_rate": 9.230770020360382e-05,
      "loss": 0.0556,
      "relation_loss": 0.005353419575840235,
      "selected_pairs": 26785.9,
      "state_query_norm": 891.9218139648438,
      "state_relation_loss": 0.0006900881824549287,
      "step": 6600,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005353419575840235,
      "weighted_relation_loss": 0.005353419575840235
    },
    {
      "action_query_norm": 487.2314788818359,
      "action_relation_loss": 0.004686871776357293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.183348924228251,
      "flow_matching_loss": 0.050823478400707243,
      "grad_norm": 0.11896675825119019,
      "learning_rate": 9.227830125039247e-05,
      "loss": 0.0562,
      "relation_loss": 0.005373452790081501,
      "selected_pairs": 27116.3,
      "state_query_norm": 891.167529296875,
      "state_relation_loss": 0.0006865810602903367,
      "step": 6610,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005373452790081501,
      "weighted_relation_loss": 0.005373452790081501
    },
    {
      "action_query_norm": 489.57465209960935,
      "action_relation_loss": 0.0047434406820684675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.192703461178672,
      "flow_matching_loss": 0.049368150532245636,
      "grad_norm": 0.11376562714576721,
      "learning_rate": 9.224885092505853e-05,
      "loss": 0.0548,
      "relation_loss": 0.005421458464115858,
      "selected_pairs": 26898.4,
      "state_query_norm": 901.6181457519531,
      "state_relation_loss": 0.0006780177762266248,
      "step": 6620,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005421458464115858,
      "weighted_relation_loss": 0.005421458464115858
    },
    {
      "action_query_norm": 502.3519256591797,
      "action_relation_loss": 0.004541963897645473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.202057998129092,
      "flow_matching_loss": 0.05172547735273838,
      "grad_norm": 0.10645370930433273,
      "learning_rate": 9.221934926338691e-05,
      "loss": 0.0569,
      "relation_loss": 0.005212092632427811,
      "selected_pairs": 26701.0,
      "state_query_norm": 887.385107421875,
      "state_relation_loss": 0.0006701287464238703,
      "step": 6630,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005212092632427811,
      "weighted_relation_loss": 0.005212092632427811
    },
    {
      "action_query_norm": 489.6275238037109,
      "action_relation_loss": 0.00454197502695024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.211412535079513,
      "flow_matching_loss": 0.04999616891145706,
      "grad_norm": 0.1590372771024704,
      "learning_rate": 9.21897963012249e-05,
      "loss": 0.0552,
      "relation_loss": 0.005199174117296934,
      "selected_pairs": 26737.8,
      "state_query_norm": 893.9970825195312,
      "state_relation_loss": 0.0006571990961674601,
      "step": 6640,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005199174117296934,
      "weighted_relation_loss": 0.005199174117296934
    },
    {
      "action_query_norm": 484.49334106445315,
      "action_relation_loss": 0.004812978114932776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.220767072029934,
      "flow_matching_loss": 0.05284057967364788,
      "grad_norm": 0.17616470158100128,
      "learning_rate": 9.216019207448217e-05,
      "loss": 0.0583,
      "relation_loss": 0.005495040025562048,
      "selected_pairs": 26277.3,
      "state_query_norm": 886.1223510742187,
      "state_relation_loss": 0.0006820618873462081,
      "step": 6650,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.005495040025562048,
      "weighted_relation_loss": 0.005495040025562048
    },
    {
      "action_query_norm": 490.6021453857422,
      "action_relation_loss": 0.004519834090024233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.2301216089803555,
      "flow_matching_loss": 0.049835069477558135,
      "grad_norm": 0.16292567551136017,
      "learning_rate": 9.21305366191306e-05,
      "loss": 0.0551,
      "relation_loss": 0.005220562033355236,
      "selected_pairs": 27033.8,
      "state_query_norm": 889.7712890625,
      "state_relation_loss": 0.0007007280481047928,
      "step": 6660,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.005220562033355236,
      "weighted_relation_loss": 0.005220562033355236
    },
    {
      "action_query_norm": 484.66407165527346,
      "action_relation_loss": 0.004490616638213396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.239476145930777,
      "flow_matching_loss": 0.051854036003351214,
      "grad_norm": 0.14722111821174622,
      "learning_rate": 9.21008299712044e-05,
      "loss": 0.057,
      "relation_loss": 0.005173224583268165,
      "selected_pairs": 26571.6,
      "state_query_norm": 882.8757446289062,
      "state_relation_loss": 0.000682607910130173,
      "step": 6670,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.005173224583268165,
      "weighted_relation_loss": 0.005173224583268165
    },
    {
      "action_query_norm": 488.5275451660156,
      "action_relation_loss": 0.004615340055897832,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.248830682881198,
      "flow_matching_loss": 0.050236478447914124,
      "grad_norm": 0.21869970858097076,
      "learning_rate": 9.207107216679993e-05,
      "loss": 0.0555,
      "relation_loss": 0.0053024882916361095,
      "selected_pairs": 26779.7,
      "state_query_norm": 878.7551940917969,
      "state_relation_loss": 0.0006871482823044061,
      "step": 6680,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0053024882916361095,
      "weighted_relation_loss": 0.0053024882916361095
    },
    {
      "action_query_norm": 491.3839965820313,
      "action_relation_loss": 0.004616005113348365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.258185219831619,
      "flow_matching_loss": 0.05013022944331169,
      "grad_norm": 0.19932223856449127,
      "learning_rate": 9.204126324207574e-05,
      "loss": 0.0554,
      "relation_loss": 0.00528726247139275,
      "selected_pairs": 26410.5,
      "state_query_norm": 887.5479187011719,
      "state_relation_loss": 0.0006712572532705963,
      "step": 6690,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.00528726247139275,
      "weighted_relation_loss": 0.00528726247139275
    },
    {
      "action_query_norm": 498.83391723632815,
      "action_relation_loss": 0.004545217100530863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.267539756782039,
      "flow_matching_loss": 0.0518291849642992,
      "grad_norm": 0.19017882645130157,
      "learning_rate": 9.201140323325248e-05,
      "loss": 0.0571,
      "relation_loss": 0.005254886718466878,
      "selected_pairs": 26851.0,
      "state_query_norm": 883.7269714355468,
      "state_relation_loss": 0.0007096695946529508,
      "step": 6700,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.005254886718466878,
      "weighted_relation_loss": 0.005254886718466878
    },
    {
      "action_query_norm": 501.269970703125,
      "action_relation_loss": 0.004547096602618694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.27689429373246,
      "flow_matching_loss": 0.0538205124437809,
      "grad_norm": 0.17176808416843414,
      "learning_rate": 9.198149217661287e-05,
      "loss": 0.0591,
      "relation_loss": 0.0052572950720787045,
      "selected_pairs": 26691.3,
      "state_query_norm": 890.1633850097656,
      "state_relation_loss": 0.0007101984752807766,
      "step": 6710,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0052572950720787045,
      "weighted_relation_loss": 0.0052572950720787045
    },
    {
      "action_query_norm": 501.00705871582034,
      "action_relation_loss": 0.004400992952287197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.286248830682881,
      "flow_matching_loss": 0.049191610887646675,
      "grad_norm": 0.19181221723556519,
      "learning_rate": 9.195153010850167e-05,
      "loss": 0.0543,
      "relation_loss": 0.005092982249334454,
      "selected_pairs": 27102.4,
      "state_query_norm": 890.2480773925781,
      "state_relation_loss": 0.0006919893552549183,
      "step": 6720,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005092982249334454,
      "weighted_relation_loss": 0.005092982249334454
    },
    {
      "action_query_norm": 499.2842681884766,
      "action_relation_loss": 0.004617689875885844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.295603367633302,
      "flow_matching_loss": 0.0509688425809145,
      "grad_norm": 0.21335290372371674,
      "learning_rate": 9.192151706532562e-05,
      "loss": 0.0563,
      "relation_loss": 0.005361208878457546,
      "selected_pairs": 27313.7,
      "state_query_norm": 884.1291625976562,
      "state_relation_loss": 0.0007435190491378307,
      "step": 6730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.005361208878457546,
      "weighted_relation_loss": 0.005361208878457546
    },
    {
      "action_query_norm": 489.7004791259766,
      "action_relation_loss": 0.004596414184197784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.304957904583723,
      "flow_matching_loss": 0.049367272108793256,
      "grad_norm": 0.15173573791980743,
      "learning_rate": 9.189145308355337e-05,
      "loss": 0.0547,
      "relation_loss": 0.005318502243608236,
      "selected_pairs": 27173.3,
      "state_query_norm": 892.5024291992188,
      "state_relation_loss": 0.0007220879953820259,
      "step": 6740,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.005318502243608236,
      "weighted_relation_loss": 0.005318502243608236
    },
    {
      "action_query_norm": 492.1643463134766,
      "action_relation_loss": 0.004401222057640553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.314312441534144,
      "flow_matching_loss": 0.048738385736942294,
      "grad_norm": 0.1754804253578186,
      "learning_rate": 9.186133819971555e-05,
      "loss": 0.0538,
      "relation_loss": 0.005108090816065669,
      "selected_pairs": 26939.5,
      "state_query_norm": 892.3381103515625,
      "state_relation_loss": 0.0007068687584251166,
      "step": 6750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.005108090816065669,
      "weighted_relation_loss": 0.005108090816065669
    },
    {
      "action_query_norm": 488.1174255371094,
      "action_relation_loss": 0.004652719246223569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.323666978484565,
      "flow_matching_loss": 0.050094280391931534,
      "grad_norm": 0.12123844772577286,
      "learning_rate": 9.183117245040455e-05,
      "loss": 0.0555,
      "relation_loss": 0.005356103600934148,
      "selected_pairs": 26715.1,
      "state_query_norm": 897.2271789550781,
      "state_relation_loss": 0.0007033844129182398,
      "step": 6760,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005356103600934148,
      "weighted_relation_loss": 0.005356103600934148
    },
    {
      "action_query_norm": 496.49413146972654,
      "action_relation_loss": 0.004578161239624024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.333021515434986,
      "flow_matching_loss": 0.050341499596834184,
      "grad_norm": 0.15709322690963745,
      "learning_rate": 9.180095587227458e-05,
      "loss": 0.0556,
      "relation_loss": 0.005273362156003714,
      "selected_pairs": 27057.8,
      "state_query_norm": 875.4489624023438,
      "state_relation_loss": 0.000695200910558924,
      "step": 6770,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.005273362156003714,
      "weighted_relation_loss": 0.005273362156003714
    },
    {
      "action_query_norm": 490.1014831542969,
      "action_relation_loss": 0.004511296236887574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.342376052385407,
      "flow_matching_loss": 0.05120731331408024,
      "grad_norm": 0.22218406200408936,
      "learning_rate": 9.177068850204167e-05,
      "loss": 0.0564,
      "relation_loss": 0.005195552995428443,
      "selected_pairs": 26408.9,
      "state_query_norm": 885.9524658203125,
      "state_relation_loss": 0.000684256840031594,
      "step": 6780,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.005195552995428443,
      "weighted_relation_loss": 0.005195552995428443
    },
    {
      "action_query_norm": 500.2740447998047,
      "action_relation_loss": 0.004524660296738148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.351730589335828,
      "flow_matching_loss": 0.049877088889479636,
      "grad_norm": 0.17695492506027222,
      "learning_rate": 9.174037037648351e-05,
      "loss": 0.0551,
      "relation_loss": 0.005248410161584616,
      "selected_pairs": 26945.2,
      "state_query_norm": 894.2887878417969,
      "state_relation_loss": 0.000723749870667234,
      "step": 6790,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.005248410161584616,
      "weighted_relation_loss": 0.005248410161584616
    },
    {
      "action_query_norm": 497.5658203125,
      "action_relation_loss": 0.004505920736119151,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.361085126286249,
      "flow_matching_loss": 0.05015245862305164,
      "grad_norm": 0.202272430062294,
      "learning_rate": 9.171000153243947e-05,
      "loss": 0.0553,
      "relation_loss": 0.005196748254820705,
      "selected_pairs": 26706.7,
      "state_query_norm": 895.5948852539062,
      "state_relation_loss": 0.0006908274895977229,
      "step": 6800,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005196748254820705,
      "weighted_relation_loss": 0.005196748254820705
    },
    {
      "action_query_norm": 498.6901397705078,
      "action_relation_loss": 0.004594680527225137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.37043966323667,
      "flow_matching_loss": 0.051705758273601535,
      "grad_norm": 0.1634700745344162,
      "learning_rate": 9.167958200681057e-05,
      "loss": 0.057,
      "relation_loss": 0.005289579974487424,
      "selected_pairs": 26779.7,
      "state_query_norm": 895.0541625976563,
      "state_relation_loss": 0.0006948993948753923,
      "step": 6810,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005289579974487424,
      "weighted_relation_loss": 0.005289579974487424
    },
    {
      "action_query_norm": 501.65098876953124,
      "action_relation_loss": 0.004398907674476505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.37979420018709,
      "flow_matching_loss": 0.050210225582122806,
      "grad_norm": 0.11490604281425476,
      "learning_rate": 9.164911183655943e-05,
      "loss": 0.0553,
      "relation_loss": 0.005080853262916208,
      "selected_pairs": 26875.8,
      "state_query_norm": 890.6318176269531,
      "state_relation_loss": 0.0006819456350058318,
      "step": 6820,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005080853262916208,
      "weighted_relation_loss": 0.005080853262916208
    },
    {
      "action_query_norm": 501.3747497558594,
      "action_relation_loss": 0.004448704980313778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.3891487371375115,
      "flow_matching_loss": 0.04960234463214874,
      "grad_norm": 0.17479094862937927,
      "learning_rate": 9.161859105871013e-05,
      "loss": 0.0547,
      "relation_loss": 0.0051377896219491955,
      "selected_pairs": 27458.2,
      "state_query_norm": 882.4052490234375,
      "state_relation_loss": 0.0006890846358146519,
      "step": 6830,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0051377896219491955,
      "weighted_relation_loss": 0.0051377896219491955
    },
    {
      "action_query_norm": 492.81382751464844,
      "action_relation_loss": 0.004216966056264937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.398503274087933,
      "flow_matching_loss": 0.050017811357975006,
      "grad_norm": 0.16065643727779388,
      "learning_rate": 9.158801971034832e-05,
      "loss": 0.0549,
      "relation_loss": 0.004890246270224452,
      "selected_pairs": 27055.6,
      "state_query_norm": 884.3946472167969,
      "state_relation_loss": 0.0006732802139595151,
      "step": 6840,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004890246270224452,
      "weighted_relation_loss": 0.004890246270224452
    },
    {
      "action_query_norm": 490.2684692382812,
      "action_relation_loss": 0.004661925928667187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.407857811038354,
      "flow_matching_loss": 0.048853788524866104,
      "grad_norm": 0.17519915103912354,
      "learning_rate": 9.155739782862107e-05,
      "loss": 0.0542,
      "relation_loss": 0.00532471714541316,
      "selected_pairs": 26551.6,
      "state_query_norm": 893.2152404785156,
      "state_relation_loss": 0.0006627911527175457,
      "step": 6850,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.00532471714541316,
      "weighted_relation_loss": 0.00532471714541316
    },
    {
      "action_query_norm": 499.7798706054688,
      "action_relation_loss": 0.004702491778880357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.417212347988775,
      "flow_matching_loss": 0.04958325996994972,
      "grad_norm": 0.1684318333864212,
      "learning_rate": 9.152672545073687e-05,
      "loss": 0.055,
      "relation_loss": 0.005377156706526875,
      "selected_pairs": 26784.5,
      "state_query_norm": 894.2111938476562,
      "state_relation_loss": 0.0006746649218257516,
      "step": 6860,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005377156706526875,
      "weighted_relation_loss": 0.005377156706526875
    },
    {
      "action_query_norm": 496.0365264892578,
      "action_relation_loss": 0.0045959562296047805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.426566884939196,
      "flow_matching_loss": 0.05282326899468899,
      "grad_norm": 0.1708952635526657,
      "learning_rate": 9.149600261396551e-05,
      "loss": 0.0581,
      "relation_loss": 0.0052708994131535295,
      "selected_pairs": 26546.5,
      "state_query_norm": 908.7303649902344,
      "state_relation_loss": 0.0006749431951902807,
      "step": 6870,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0052708994131535295,
      "weighted_relation_loss": 0.0052708994131535295
    },
    {
      "action_query_norm": 490.6993835449219,
      "action_relation_loss": 0.0043828011956065895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.435921421889616,
      "flow_matching_loss": 0.04750762730836868,
      "grad_norm": 0.14753054082393646,
      "learning_rate": 9.146522935563817e-05,
      "loss": 0.0526,
      "relation_loss": 0.0050694150384515526,
      "selected_pairs": 27434.6,
      "state_query_norm": 903.3293823242187,
      "state_relation_loss": 0.0006866137788165361,
      "step": 6880,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0050694150384515526,
      "weighted_relation_loss": 0.0050694150384515526
    },
    {
      "action_query_norm": 493.8893524169922,
      "action_relation_loss": 0.0044105200096964834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.445275958840037,
      "flow_matching_loss": 0.04983228407800198,
      "grad_norm": 0.1306886225938797,
      "learning_rate": 9.143440571314723e-05,
      "loss": 0.0549,
      "relation_loss": 0.005085789831355214,
      "selected_pairs": 26661.8,
      "state_query_norm": 915.1377258300781,
      "state_relation_loss": 0.0006752698973286897,
      "step": 6890,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.005085789831355214,
      "weighted_relation_loss": 0.005085789831355214
    },
    {
      "action_query_norm": 492.9794189453125,
      "action_relation_loss": 0.004638906894251704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.454630495790458,
      "flow_matching_loss": 0.05081675015389919,
      "grad_norm": 0.14219459891319275,
      "learning_rate": 9.140353172394636e-05,
      "loss": 0.0562,
      "relation_loss": 0.005348239606246352,
      "selected_pairs": 26711.1,
      "state_query_norm": 904.7443786621094,
      "state_relation_loss": 0.0007093327527400106,
      "step": 6900,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005348239606246352,
      "weighted_relation_loss": 0.005348239606246352
    },
    {
      "action_query_norm": 492.4263641357422,
      "action_relation_loss": 0.004911355115473271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.463985032740879,
      "flow_matching_loss": 0.05122901052236557,
      "grad_norm": 0.1336318701505661,
      "learning_rate": 9.137260742555032e-05,
      "loss": 0.0568,
      "relation_loss": 0.005603106785565614,
      "selected_pairs": 27269.6,
      "state_query_norm": 910.1798767089844,
      "state_relation_loss": 0.0006917516700923442,
      "step": 6910,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005603106785565614,
      "weighted_relation_loss": 0.005603106785565614
    },
    {
      "action_query_norm": 502.477197265625,
      "action_relation_loss": 0.004624171275645494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.4733395696913005,
      "flow_matching_loss": 0.05159914121031761,
      "grad_norm": 0.1495935022830963,
      "learning_rate": 9.13416328555351e-05,
      "loss": 0.0569,
      "relation_loss": 0.005307314172387123,
      "selected_pairs": 26964.6,
      "state_query_norm": 900.272412109375,
      "state_relation_loss": 0.0006831429025623947,
      "step": 6920,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.005307314172387123,
      "weighted_relation_loss": 0.005307314172387123
    },
    {
      "action_query_norm": 494.518896484375,
      "action_relation_loss": 0.004537284281104803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.482694106641722,
      "flow_matching_loss": 0.05059564672410488,
      "grad_norm": 0.14415623247623444,
      "learning_rate": 9.131060805153769e-05,
      "loss": 0.0558,
      "relation_loss": 0.005212011933326721,
      "selected_pairs": 26418.6,
      "state_query_norm": 890.4906372070312,
      "state_relation_loss": 0.0006747276172973215,
      "step": 6930,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.005212011933326721,
      "weighted_relation_loss": 0.005212011933326721
    },
    {
      "action_query_norm": 494.43408203125,
      "action_relation_loss": 0.004507814301177859,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.492048643592142,
      "flow_matching_loss": 0.05103384330868721,
      "grad_norm": 0.11865690350532532,
      "learning_rate": 9.127953305125618e-05,
      "loss": 0.0562,
      "relation_loss": 0.005177323240786791,
      "selected_pairs": 26249.2,
      "state_query_norm": 896.6086547851562,
      "state_relation_loss": 0.0006695088988635689,
      "step": 6940,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.005177323240786791,
      "weighted_relation_loss": 0.005177323240786791
    },
    {
      "action_query_norm": 489.4101318359375,
      "action_relation_loss": 0.004446367989294231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.501403180542563,
      "flow_matching_loss": 0.04949055910110474,
      "grad_norm": 0.15840639173984528,
      "learning_rate": 9.124840789244958e-05,
      "loss": 0.0547,
      "relation_loss": 0.005187411839142442,
      "selected_pairs": 26640.2,
      "state_query_norm": 899.7942199707031,
      "state_relation_loss": 0.000741043797461316,
      "step": 6950,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.005187411839142442,
      "weighted_relation_loss": 0.005187411839142442
    },
    {
      "action_query_norm": 494.4933349609375,
      "action_relation_loss": 0.004548757709562779,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.510757717492984,
      "flow_matching_loss": 0.05043845884501934,
      "grad_norm": 0.14009948074817657,
      "learning_rate": 9.121723261293794e-05,
      "loss": 0.0557,
      "relation_loss": 0.005234326515346765,
      "selected_pairs": 26802.6,
      "state_query_norm": 903.8337646484375,
      "state_relation_loss": 0.0006855688465293497,
      "step": 6960,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005234326515346765,
      "weighted_relation_loss": 0.005234326515346765
    },
    {
      "action_query_norm": 505.87904052734376,
      "action_relation_loss": 0.004447106900624931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.520112254443405,
      "flow_matching_loss": 0.04693428128957748,
      "grad_norm": 0.12029014527797699,
      "learning_rate": 9.118600725060214e-05,
      "loss": 0.052,
      "relation_loss": 0.005114505998790264,
      "selected_pairs": 27102.4,
      "state_query_norm": 905.0012756347656,
      "state_relation_loss": 0.0006673990574199706,
      "step": 6970,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005114505998790264,
      "weighted_relation_loss": 0.005114505998790264
    },
    {
      "action_query_norm": 503.2234771728516,
      "action_relation_loss": 0.004226786782965064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.529466791393826,
      "flow_matching_loss": 0.04999247118830681,
      "grad_norm": 0.15667396783828735,
      "learning_rate": 9.115473184338392e-05,
      "loss": 0.0549,
      "relation_loss": 0.004904076829552651,
      "selected_pairs": 27131.1,
      "state_query_norm": 900.3777709960938,
      "state_relation_loss": 0.000677290017483756,
      "step": 6980,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004904076829552651,
      "weighted_relation_loss": 0.004904076829552651
    },
    {
      "action_query_norm": 506.4732269287109,
      "action_relation_loss": 0.004326013172976673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.538821328344247,
      "flow_matching_loss": 0.05089491493999958,
      "grad_norm": 0.16398711502552032,
      "learning_rate": 9.11234064292859e-05,
      "loss": 0.0559,
      "relation_loss": 0.004979829071089625,
      "selected_pairs": 26278.2,
      "state_query_norm": 914.0571716308593,
      "state_relation_loss": 0.0006538158864714205,
      "step": 6990,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.004979829071089625,
      "weighted_relation_loss": 0.004979829071089625
    },
    {
      "action_query_norm": 512.6618927001953,
      "action_relation_loss": 0.004598494479432702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.5481758652946676,
      "flow_matching_loss": 0.050392452999949454,
      "grad_norm": 0.17199863493442535,
      "learning_rate": 9.109203104637137e-05,
      "loss": 0.0557,
      "relation_loss": 0.00527463792823255,
      "selected_pairs": 26967.1,
      "state_query_norm": 911.6544860839844,
      "state_relation_loss": 0.0006761435128282755,
      "step": 7000,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.00527463792823255,
      "weighted_relation_loss": 0.00527463792823255
    },
    {
      "action_query_norm": 506.1993774414062,
      "action_relation_loss": 0.004911844618618488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.557530402245089,
      "flow_matching_loss": 0.053175130113959315,
      "grad_norm": 0.16419795155525208,
      "learning_rate": 9.10606057327644e-05,
      "loss": 0.0588,
      "relation_loss": 0.005578299053013325,
      "selected_pairs": 27085.2,
      "state_query_norm": 903.3116455078125,
      "state_relation_loss": 0.0006664544576779008,
      "step": 7010,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005578299053013325,
      "weighted_relation_loss": 0.005578299053013325
    },
    {
      "action_query_norm": 497.3437835693359,
      "action_relation_loss": 0.004424958513118327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.56688493919551,
      "flow_matching_loss": 0.04987495504319668,
      "grad_norm": 0.1182982474565506,
      "learning_rate": 9.10291305266497e-05,
      "loss": 0.055,
      "relation_loss": 0.005093555012717843,
      "selected_pairs": 26570.0,
      "state_query_norm": 902.5046813964843,
      "state_relation_loss": 0.0006685965461656451,
      "step": 7020,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.005093555012717843,
      "weighted_relation_loss": 0.005093555012717843
    },
    {
      "action_query_norm": 506.16826782226565,
      "action_relation_loss": 0.004427641443908214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.576239476145931,
      "flow_matching_loss": 0.05146990455687046,
      "grad_norm": 0.1566743552684784,
      "learning_rate": 9.099760546627261e-05,
      "loss": 0.0566,
      "relation_loss": 0.005125963920727372,
      "selected_pairs": 26781.5,
      "state_query_norm": 899.9922241210937,
      "state_relation_loss": 0.0006983224302530289,
      "step": 7030,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005125963920727372,
      "weighted_relation_loss": 0.005125963920727372
    },
    {
      "action_query_norm": 505.33457946777344,
      "action_relation_loss": 0.0043306232430040835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.585594013096352,
      "flow_matching_loss": 0.050825875252485275,
      "grad_norm": 0.16383826732635498,
      "learning_rate": 9.096603058993907e-05,
      "loss": 0.0558,
      "relation_loss": 0.0050055789295583965,
      "selected_pairs": 26627.0,
      "state_query_norm": 914.1032287597657,
      "state_relation_loss": 0.0006749557272996753,
      "step": 7040,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0050055789295583965,
      "weighted_relation_loss": 0.0050055789295583965
    },
    {
      "action_query_norm": 498.05051879882814,
      "action_relation_loss": 0.004515804955735803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.594948550046773,
      "flow_matching_loss": 0.05056200549006462,
      "grad_norm": 0.18183760344982147,
      "learning_rate": 9.093440593601552e-05,
      "loss": 0.0557,
      "relation_loss": 0.005181068973615766,
      "selected_pairs": 26485.3,
      "state_query_norm": 913.12919921875,
      "state_relation_loss": 0.0006652640644460916,
      "step": 7050,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005181068973615766,
      "weighted_relation_loss": 0.005181068973615766
    },
    {
      "action_query_norm": 497.44557495117186,
      "action_relation_loss": 0.0045109635684639215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.604303086997193,
      "flow_matching_loss": 0.047136425599455835,
      "grad_norm": 0.17206406593322754,
      "learning_rate": 9.090273154292889e-05,
      "loss": 0.0523,
      "relation_loss": 0.005200137663632631,
      "selected_pairs": 26897.0,
      "state_query_norm": 898.2735046386719,
      "state_relation_loss": 0.000689174042781815,
      "step": 7060,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005200137663632631,
      "weighted_relation_loss": 0.005200137663632631
    },
    {
      "action_query_norm": 496.38023071289064,
      "action_relation_loss": 0.004453487554565072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.613657623947614,
      "flow_matching_loss": 0.051065389066934586,
      "grad_norm": 0.11332204192876816,
      "learning_rate": 9.087100744916656e-05,
      "loss": 0.0562,
      "relation_loss": 0.005091306334361434,
      "selected_pairs": 26925.3,
      "state_query_norm": 902.2001770019531,
      "state_relation_loss": 0.0006378187681548297,
      "step": 7070,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005091306334361434,
      "weighted_relation_loss": 0.005091306334361434
    },
    {
      "action_query_norm": 489.42772521972654,
      "action_relation_loss": 0.004419788136146963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.6230121608980355,
      "flow_matching_loss": 0.05021502450108528,
      "grad_norm": 0.14641812443733215,
      "learning_rate": 9.08392336932763e-05,
      "loss": 0.0553,
      "relation_loss": 0.0050754176918417215,
      "selected_pairs": 27554.9,
      "state_query_norm": 905.7035766601563,
      "state_relation_loss": 0.0006556295556947589,
      "step": 7080,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0050754176918417215,
      "weighted_relation_loss": 0.0050754176918417215
    },
    {
      "action_query_norm": 495.6088409423828,
      "action_relation_loss": 0.004254299611784518,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.632366697848457,
      "flow_matching_loss": 0.04894715510308743,
      "grad_norm": 0.1598253846168518,
      "learning_rate": 9.080741031386619e-05,
      "loss": 0.0538,
      "relation_loss": 0.004884825833141804,
      "selected_pairs": 26898.2,
      "state_query_norm": 920.3513549804687,
      "state_relation_loss": 0.0006305262155365199,
      "step": 7090,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.004884825833141804,
      "weighted_relation_loss": 0.004884825833141804
    },
    {
      "action_query_norm": 505.39385986328125,
      "action_relation_loss": 0.00422819871455431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.641721234798878,
      "flow_matching_loss": 0.04896603189408779,
      "grad_norm": 0.17160533368587494,
      "learning_rate": 9.077553734960468e-05,
      "loss": 0.0539,
      "relation_loss": 0.0048958037979900835,
      "selected_pairs": 26696.2,
      "state_query_norm": 912.34638671875,
      "state_relation_loss": 0.0006676051416434348,
      "step": 7100,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0048958037979900835,
      "weighted_relation_loss": 0.0048958037979900835
    },
    {
      "action_query_norm": 504.1013549804687,
      "action_relation_loss": 0.004462240426801145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.651075771749299,
      "flow_matching_loss": 0.04599820300936699,
      "grad_norm": 0.15452952682971954,
      "learning_rate": 9.074361483922041e-05,
      "loss": 0.0511,
      "relation_loss": 0.0051200530957430605,
      "selected_pairs": 27149.6,
      "state_query_norm": 908.4715209960938,
      "state_relation_loss": 0.0006578126398380846,
      "step": 7110,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0051200530957430605,
      "weighted_relation_loss": 0.0051200530957430605
    },
    {
      "action_query_norm": 500.9526794433594,
      "action_relation_loss": 0.004554146504960954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.66043030869972,
      "flow_matching_loss": 0.04851620718836784,
      "grad_norm": 0.11983925104141235,
      "learning_rate": 9.071164282150224e-05,
      "loss": 0.0537,
      "relation_loss": 0.0052309607621282336,
      "selected_pairs": 26895.7,
      "state_query_norm": 907.1453491210938,
      "state_relation_loss": 0.0006768142862711102,
      "step": 7120,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0052309607621282336,
      "weighted_relation_loss": 0.0052309607621282336
    },
    {
      "action_query_norm": 499.7546051025391,
      "action_relation_loss": 0.004315801267512143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.66978484565014,
      "flow_matching_loss": 0.04896114394068718,
      "grad_norm": 0.12210920453071594,
      "learning_rate": 9.067962133529919e-05,
      "loss": 0.054,
      "relation_loss": 0.004995212377980352,
      "selected_pairs": 26642.1,
      "state_query_norm": 907.811328125,
      "state_relation_loss": 0.0006794112152419985,
      "step": 7130,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.004995212377980352,
      "weighted_relation_loss": 0.004995212377980352
    },
    {
      "action_query_norm": 495.9558990478516,
      "action_relation_loss": 0.004614911181852222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.679139382600561,
      "flow_matching_loss": 0.0479378305375576,
      "grad_norm": 0.1424679309129715,
      "learning_rate": 9.064755041952036e-05,
      "loss": 0.0532,
      "relation_loss": 0.005311349593102932,
      "selected_pairs": 26922.8,
      "state_query_norm": 912.4452575683594,
      "state_relation_loss": 0.0006964383937884122,
      "step": 7140,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005311349593102932,
      "weighted_relation_loss": 0.005311349593102932
    },
    {
      "action_query_norm": 492.44074096679685,
      "action_relation_loss": 0.0044516148744151,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.688493919550982,
      "flow_matching_loss": 0.05101001150906086,
      "grad_norm": 0.15911267697811127,
      "learning_rate": 9.061543011313497e-05,
      "loss": 0.0562,
      "relation_loss": 0.005145550426095724,
      "selected_pairs": 26389.7,
      "state_query_norm": 899.5484924316406,
      "state_relation_loss": 0.000693935516756028,
      "step": 7150,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.005145550426095724,
      "weighted_relation_loss": 0.005145550426095724
    },
    {
      "action_query_norm": 500.93300170898436,
      "action_relation_loss": 0.0044515690067783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.697848456501403,
      "flow_matching_loss": 0.05188452303409576,
      "grad_norm": 0.12820442020893097,
      "learning_rate": 9.05832604551722e-05,
      "loss": 0.057,
      "relation_loss": 0.005128572694957256,
      "selected_pairs": 27289.1,
      "state_query_norm": 918.4021484375,
      "state_relation_loss": 0.0006770036765374243,
      "step": 7160,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005128572694957256,
      "weighted_relation_loss": 0.005128572694957256
    },
    {
      "action_query_norm": 509.45627746582034,
      "action_relation_loss": 0.004445683071389794,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.7072029934518245,
      "flow_matching_loss": 0.04672493375837803,
      "grad_norm": 0.11026696860790253,
      "learning_rate": 9.055104148472124e-05,
      "loss": 0.0519,
      "relation_loss": 0.005137619376182556,
      "selected_pairs": 26709.2,
      "state_query_norm": 916.8740417480469,
      "state_relation_loss": 0.000691936316434294,
      "step": 7170,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005137619376182556,
      "weighted_relation_loss": 0.005137619376182556
    },
    {
      "action_query_norm": 510.0453674316406,
      "action_relation_loss": 0.004399213474243879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.716557530402245,
      "flow_matching_loss": 0.0500279750674963,
      "grad_norm": 0.14080972969532013,
      "learning_rate": 9.051877324093114e-05,
      "loss": 0.0551,
      "relation_loss": 0.00508202644996345,
      "selected_pairs": 26963.6,
      "state_query_norm": 934.5365112304687,
      "state_relation_loss": 0.0006828129407949746,
      "step": 7180,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.00508202644996345,
      "weighted_relation_loss": 0.00508202644996345
    },
    {
      "action_query_norm": 507.67865600585935,
      "action_relation_loss": 0.004145735967904329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.725912067352666,
      "flow_matching_loss": 0.04762193523347378,
      "grad_norm": 0.15805192291736603,
      "learning_rate": 9.04864557630109e-05,
      "loss": 0.0524,
      "relation_loss": 0.004817933076992631,
      "selected_pairs": 26618.7,
      "state_query_norm": 921.2562377929687,
      "state_relation_loss": 0.0006721971207298338,
      "step": 7190,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.004817933076992631,
      "weighted_relation_loss": 0.004817933076992631
    },
    {
      "action_query_norm": 503.83405151367185,
      "action_relation_loss": 0.004175020405091346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.735266604303087,
      "flow_matching_loss": 0.05227475203573704,
      "grad_norm": 0.1847992241382599,
      "learning_rate": 9.045408909022928e-05,
      "loss": 0.0571,
      "relation_loss": 0.004821416269987821,
      "selected_pairs": 27004.5,
      "state_query_norm": 914.513818359375,
      "state_relation_loss": 0.0006463958183303475,
      "step": 7200,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004821416269987821,
      "weighted_relation_loss": 0.004821416269987821
    },
    {
      "action_query_norm": 510.2468780517578,
      "action_relation_loss": 0.00435223444364965,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.744621141253508,
      "flow_matching_loss": 0.05008913092315197,
      "grad_norm": 0.21597015857696533,
      "learning_rate": 9.042167326191484e-05,
      "loss": 0.0551,
      "relation_loss": 0.005002657044678926,
      "selected_pairs": 26989.9,
      "state_query_norm": 928.4822998046875,
      "state_relation_loss": 0.0006504225486423821,
      "step": 7210,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.005002657044678926,
      "weighted_relation_loss": 0.005002657044678926
    },
    {
      "action_query_norm": 508.5022888183594,
      "action_relation_loss": 0.004347295197658241,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.753975678203929,
      "flow_matching_loss": 0.05117356218397617,
      "grad_norm": 0.13026906549930573,
      "learning_rate": 9.038920831745587e-05,
      "loss": 0.0562,
      "relation_loss": 0.005023653013631701,
      "selected_pairs": 27083.9,
      "state_query_norm": 911.0066284179687,
      "state_relation_loss": 0.0006763578276149929,
      "step": 7220,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005023653013631701,
      "weighted_relation_loss": 0.005023653013631701
    },
    {
      "action_query_norm": 511.10093688964844,
      "action_relation_loss": 0.004297797288745642,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.76333021515435,
      "flow_matching_loss": 0.05061821825802326,
      "grad_norm": 0.13398109376430511,
      "learning_rate": 9.035669429630036e-05,
      "loss": 0.0556,
      "relation_loss": 0.004967557219788432,
      "selected_pairs": 26840.5,
      "state_query_norm": 919.3254089355469,
      "state_relation_loss": 0.0006697599194012583,
      "step": 7230,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.004967557219788432,
      "weighted_relation_loss": 0.004967557219788432
    },
    {
      "action_query_norm": 506.5530090332031,
      "action_relation_loss": 0.004340606881305576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.772684752104771,
      "flow_matching_loss": 0.04873910583555698,
      "grad_norm": 0.1304948329925537,
      "learning_rate": 9.032413123795588e-05,
      "loss": 0.0537,
      "relation_loss": 0.004973451746627689,
      "selected_pairs": 26490.1,
      "state_query_norm": 911.7620910644531,
      "state_relation_loss": 0.000632844801293686,
      "step": 7240,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.004973451746627689,
      "weighted_relation_loss": 0.004973451746627689
    },
    {
      "action_query_norm": 508.8490936279297,
      "action_relation_loss": 0.004206532123498618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.7820392890551915,
      "flow_matching_loss": 0.04826834239065647,
      "grad_norm": 0.17360082268714905,
      "learning_rate": 9.029151918198962e-05,
      "loss": 0.0531,
      "relation_loss": 0.0048795138951390985,
      "selected_pairs": 26466.0,
      "state_query_norm": 905.6014587402344,
      "state_relation_loss": 0.0006729817891027778,
      "step": 7250,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0048795138951390985,
      "weighted_relation_loss": 0.0048795138951390985
    },
    {
      "action_query_norm": 510.3006988525391,
      "action_relation_loss": 0.004208910465240479,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.791393826005613,
      "flow_matching_loss": 0.049794838204979895,
      "grad_norm": 0.16206690669059753,
      "learning_rate": 9.025885816802833e-05,
      "loss": 0.0546,
      "relation_loss": 0.004853428341448307,
      "selected_pairs": 26639.9,
      "state_query_norm": 914.1415588378907,
      "state_relation_loss": 0.0006445179111324251,
      "step": 7260,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.004853428341448307,
      "weighted_relation_loss": 0.004853428341448307
    },
    {
      "action_query_norm": 503.2195526123047,
      "action_relation_loss": 0.004421752481721342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.800748362956034,
      "flow_matching_loss": 0.04806279167532921,
      "grad_norm": 0.15712380409240723,
      "learning_rate": 9.02261482357582e-05,
      "loss": 0.0531,
      "relation_loss": 0.005070327781140804,
      "selected_pairs": 26826.3,
      "state_query_norm": 926.5881591796875,
      "state_relation_loss": 0.0006485752470325679,
      "step": 7270,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005070327781140804,
      "weighted_relation_loss": 0.005070327781140804
    },
    {
      "action_query_norm": 499.5351959228516,
      "action_relation_loss": 0.0043936468660831455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.810102899906455,
      "flow_matching_loss": 0.048252389580011365,
      "grad_norm": 0.1240411251783371,
      "learning_rate": 9.019338942492485e-05,
      "loss": 0.0533,
      "relation_loss": 0.005059087183326483,
      "selected_pairs": 26438.1,
      "state_query_norm": 910.9183532714844,
      "state_relation_loss": 0.0006654402997810394,
      "step": 7280,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.005059087183326483,
      "weighted_relation_loss": 0.005059087183326483
    },
    {
      "action_query_norm": 504.2654541015625,
      "action_relation_loss": 0.004353971150703728,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.819457436856876,
      "flow_matching_loss": 0.05127390064299107,
      "grad_norm": 0.1217980682849884,
      "learning_rate": 9.01605817753334e-05,
      "loss": 0.0563,
      "relation_loss": 0.005002522747963667,
      "selected_pairs": 27134.6,
      "state_query_norm": 930.8236755371094,
      "state_relation_loss": 0.0006485516845714301,
      "step": 7290,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.005002522747963667,
      "weighted_relation_loss": 0.005002522747963667
    },
    {
      "action_query_norm": 508.1477447509766,
      "action_relation_loss": 0.0041782698594033715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.828811973807296,
      "flow_matching_loss": 0.05000911764800549,
      "grad_norm": 0.11811535060405731,
      "learning_rate": 9.012772532684818e-05,
      "loss": 0.0548,
      "relation_loss": 0.004840586148202419,
      "selected_pairs": 26850.7,
      "state_query_norm": 932.7414428710938,
      "state_relation_loss": 0.0006623162771575153,
      "step": 7300,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004840586148202419,
      "weighted_relation_loss": 0.004840586148202419
    },
    {
      "action_query_norm": 499.8562286376953,
      "action_relation_loss": 0.004479476111009717,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.838166510757717,
      "flow_matching_loss": 0.04859404042363167,
      "grad_norm": 0.09843884408473969,
      "learning_rate": 9.00948201193929e-05,
      "loss": 0.0537,
      "relation_loss": 0.0051402742974460125,
      "selected_pairs": 26733.4,
      "state_query_norm": 920.2340942382813,
      "state_relation_loss": 0.0006607982097193599,
      "step": 7310,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0051402742974460125,
      "weighted_relation_loss": 0.0051402742974460125
    },
    {
      "action_query_norm": 501.92881164550784,
      "action_relation_loss": 0.0044651766773313286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.847521047708138,
      "flow_matching_loss": 0.05109967552125454,
      "grad_norm": 0.1483508199453354,
      "learning_rate": 9.006186619295047e-05,
      "loss": 0.0562,
      "relation_loss": 0.005128100933507085,
      "selected_pairs": 26828.8,
      "state_query_norm": 915.2103332519531,
      "state_relation_loss": 0.000662924291100353,
      "step": 7320,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005128100933507085,
      "weighted_relation_loss": 0.005128100933507085
    },
    {
      "action_query_norm": 501.3848083496094,
      "action_relation_loss": 0.004443084169179201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.856875584658559,
      "flow_matching_loss": 0.051204510033130646,
      "grad_norm": 0.1356075555086136,
      "learning_rate": 9.002886358756305e-05,
      "loss": 0.0563,
      "relation_loss": 0.005119977332651615,
      "selected_pairs": 27571.3,
      "state_query_norm": 923.8186584472656,
      "state_relation_loss": 0.0006768931809347123,
      "step": 7330,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.005119977332651615,
      "weighted_relation_loss": 0.005119977332651615
    },
    {
      "action_query_norm": 504.0192443847656,
      "action_relation_loss": 0.004145434452220797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.8662301216089805,
      "flow_matching_loss": 0.04885188527405262,
      "grad_norm": 0.14869798719882965,
      "learning_rate": 8.999581234333188e-05,
      "loss": 0.0537,
      "relation_loss": 0.004822615161538124,
      "selected_pairs": 27056.8,
      "state_query_norm": 924.5421325683594,
      "state_relation_loss": 0.0006771806918550282,
      "step": 7340,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004822615161538124,
      "weighted_relation_loss": 0.004822615161538124
    },
    {
      "action_query_norm": 514.4196533203125,
      "action_relation_loss": 0.004331066971644759,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.875584658559402,
      "flow_matching_loss": 0.04910500608384609,
      "grad_norm": 0.1376146674156189,
      "learning_rate": 8.996271250041734e-05,
      "loss": 0.0542,
      "relation_loss": 0.005053903069347143,
      "selected_pairs": 26802.9,
      "state_query_norm": 904.0765258789063,
      "state_relation_loss": 0.0007228361209854484,
      "step": 7350,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005053903069347143,
      "weighted_relation_loss": 0.005053903069347143
    },
    {
      "action_query_norm": 515.0720092773438,
      "action_relation_loss": 0.004355379403568805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.884939195509823,
      "flow_matching_loss": 0.05043304972350597,
      "grad_norm": 0.14824633300304413,
      "learning_rate": 8.99295640990389e-05,
      "loss": 0.0555,
      "relation_loss": 0.005044558877125382,
      "selected_pairs": 27599.3,
      "state_query_norm": 928.5692138671875,
      "state_relation_loss": 0.0006891795026604086,
      "step": 7360,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.005044558877125382,
      "weighted_relation_loss": 0.005044558877125382
    },
    {
      "action_query_norm": 524.7114440917969,
      "action_relation_loss": 0.004275306919589638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.894293732460243,
      "flow_matching_loss": 0.047676043957471846,
      "grad_norm": 0.12718549370765686,
      "learning_rate": 8.989636717947494e-05,
      "loss": 0.0526,
      "relation_loss": 0.0049343433696776625,
      "selected_pairs": 27337.1,
      "state_query_norm": 921.4943725585938,
      "state_relation_loss": 0.0006590364209841937,
      "step": 7370,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0049343433696776625,
      "weighted_relation_loss": 0.0049343433696776625
    },
    {
      "action_query_norm": 506.43641052246096,
      "action_relation_loss": 0.004500704584643245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.903648269410664,
      "flow_matching_loss": 0.04912163093686104,
      "grad_norm": 0.15517404675483704,
      "learning_rate": 8.986312178206291e-05,
      "loss": 0.0543,
      "relation_loss": 0.005192489363253117,
      "selected_pairs": 26689.2,
      "state_query_norm": 918.8499694824219,
      "state_relation_loss": 0.0006917847611475735,
      "step": 7380,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005192489363253117,
      "weighted_relation_loss": 0.005192489363253117
    },
    {
      "action_query_norm": 504.58335876464844,
      "action_relation_loss": 0.004231941630132496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.913002806361085,
      "flow_matching_loss": 0.049244602024555204,
      "grad_norm": 0.1530858427286148,
      "learning_rate": 8.982982794719904e-05,
      "loss": 0.0541,
      "relation_loss": 0.004892588267102838,
      "selected_pairs": 26234.2,
      "state_query_norm": 926.776025390625,
      "state_relation_loss": 0.0006606466020457446,
      "step": 7390,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.004892588267102838,
      "weighted_relation_loss": 0.004892588267102838
    },
    {
      "action_query_norm": 521.2083801269531,
      "action_relation_loss": 0.004038054659031331,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.922357343311506,
      "flow_matching_loss": 0.050132586434483525,
      "grad_norm": 0.16375181078910828,
      "learning_rate": 8.979648571533852e-05,
      "loss": 0.0548,
      "relation_loss": 0.004699181858450174,
      "selected_pairs": 27414.5,
      "state_query_norm": 934.8110717773437,
      "state_relation_loss": 0.0006611272459849715,
      "step": 7400,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.004699181858450174,
      "weighted_relation_loss": 0.004699181858450174
    },
    {
      "action_query_norm": 512.4388549804687,
      "action_relation_loss": 0.004236335773020983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.931711880261927,
      "flow_matching_loss": 0.046394043043255805,
      "grad_norm": 0.14580628275871277,
      "learning_rate": 8.976309512699528e-05,
      "loss": 0.0513,
      "relation_loss": 0.004888086160644889,
      "selected_pairs": 26938.9,
      "state_query_norm": 923.2300598144532,
      "state_relation_loss": 0.0006517504050862044,
      "step": 7410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004888086160644889,
      "weighted_relation_loss": 0.004888086160644889
    },
    {
      "action_query_norm": 513.0909088134765,
      "action_relation_loss": 0.004314055619761348,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.9410664172123475,
      "flow_matching_loss": 0.048546279594302175,
      "grad_norm": 0.13248424232006073,
      "learning_rate": 8.972965622274205e-05,
      "loss": 0.0535,
      "relation_loss": 0.0049630958586931225,
      "selected_pairs": 26712.5,
      "state_query_norm": 917.2020385742187,
      "state_relation_loss": 0.0006490402505733073,
      "step": 7420,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0049630958586931225,
      "weighted_relation_loss": 0.0049630958586931225
    },
    {
      "action_query_norm": 512.42421875,
      "action_relation_loss": 0.0041332303313538436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.950420954162769,
      "flow_matching_loss": 0.04887900911271572,
      "grad_norm": 0.13856686651706696,
      "learning_rate": 8.969616904321026e-05,
      "loss": 0.0536,
      "relation_loss": 0.0047595596872270106,
      "selected_pairs": 26823.7,
      "state_query_norm": 928.1165466308594,
      "state_relation_loss": 0.000626329320948571,
      "step": 7430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0047595596872270106,
      "weighted_relation_loss": 0.0047595596872270106
    },
    {
      "action_query_norm": 512.3883880615234,
      "action_relation_loss": 0.004468551627360285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.95977549111319,
      "flow_matching_loss": 0.04702824503183365,
      "grad_norm": 0.15175467729568481,
      "learning_rate": 8.966263362908998e-05,
      "loss": 0.0522,
      "relation_loss": 0.005131481587886811,
      "selected_pairs": 26712.4,
      "state_query_norm": 928.4260803222656,
      "state_relation_loss": 0.0006629299081396311,
      "step": 7440,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005131481587886811,
      "weighted_relation_loss": 0.005131481587886811
    },
    {
      "action_query_norm": 511.7614288330078,
      "action_relation_loss": 0.0042983239982277155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.969130028063611,
      "flow_matching_loss": 0.0511312410235405,
      "grad_norm": 0.16169188916683197,
      "learning_rate": 8.962905002112989e-05,
      "loss": 0.0561,
      "relation_loss": 0.0049610810820013285,
      "selected_pairs": 26691.8,
      "state_query_norm": 928.5058410644531,
      "state_relation_loss": 0.0006627571070566774,
      "step": 7450,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0049610810820013285,
      "weighted_relation_loss": 0.0049610810820013285
    },
    {
      "action_query_norm": 509.2166320800781,
      "action_relation_loss": 0.0043889486463740465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.978484565014032,
      "flow_matching_loss": 0.048549574241042136,
      "grad_norm": 0.16636337339878082,
      "learning_rate": 8.959541826013724e-05,
      "loss": 0.0536,
      "relation_loss": 0.0050385492388159035,
      "selected_pairs": 26746.0,
      "state_query_norm": 914.8980590820313,
      "state_relation_loss": 0.0006496005808003247,
      "step": 7460,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0050385492388159035,
      "weighted_relation_loss": 0.0050385492388159035
    },
    {
      "action_query_norm": 516.1145935058594,
      "action_relation_loss": 0.004121672851033509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.987839101964453,
      "flow_matching_loss": 0.04813133254647255,
      "grad_norm": 0.18662109971046448,
      "learning_rate": 8.956173838697781e-05,
      "loss": 0.0529,
      "relation_loss": 0.0047849081456661224,
      "selected_pairs": 26899.5,
      "state_query_norm": 922.1029174804687,
      "state_relation_loss": 0.0006632353237364441,
      "step": 7470,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0047849081456661224,
      "weighted_relation_loss": 0.0047849081456661224
    },
    {
      "action_query_norm": 520.3564758300781,
      "action_relation_loss": 0.004182634362950921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.997193638914874,
      "flow_matching_loss": 0.048438633978366855,
      "grad_norm": 0.14026769995689392,
      "learning_rate": 8.952801044257581e-05,
      "loss": 0.0533,
      "relation_loss": 0.004813986783847213,
      "selected_pairs": 26529.9,
      "state_query_norm": 930.1165161132812,
      "state_relation_loss": 0.0006313523626886308,
      "step": 7480,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.004813986783847213,
      "weighted_relation_loss": 0.004813986783847213
    },
    {
      "action_query_norm": 515.8049713134766,
      "action_relation_loss": 0.00419939283747226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.006548175865294,
      "flow_matching_loss": 0.05000877045094967,
      "grad_norm": 0.1821301132440567,
      "learning_rate": 8.949423446791387e-05,
      "loss": 0.0549,
      "relation_loss": 0.0048601004295051094,
      "selected_pairs": 24532.8,
      "state_query_norm": 920.6722290039063,
      "state_relation_loss": 0.0006607075745705515,
      "step": 7490,
      "valid_action_samples": 215.2,
      "weighted_auxiliary_loss": 0.0048601004295051094,
      "weighted_relation_loss": 0.0048601004295051094
    },
    {
      "action_query_norm": 515.3446838378907,
      "action_relation_loss": 0.003828165493905544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.015902712815715,
      "flow_matching_loss": 0.04595098346471786,
      "grad_norm": 0.13093148171901703,
      "learning_rate": 8.9460410504033e-05,
      "loss": 0.0504,
      "relation_loss": 0.00446727816015482,
      "selected_pairs": 26719.0,
      "state_query_norm": 931.5148132324218,
      "state_relation_loss": 0.0006391126487869769,
      "step": 7500,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00446727816015482,
      "weighted_relation_loss": 0.00446727816015482
    },
    {
      "action_query_norm": 515.4019104003906,
      "action_relation_loss": 0.004091408848762512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.0252572497661365,
      "flow_matching_loss": 0.04705873392522335,
      "grad_norm": 0.11923756450414658,
      "learning_rate": 8.942653859203248e-05,
      "loss": 0.0518,
      "relation_loss": 0.004756104853004217,
      "selected_pairs": 27063.8,
      "state_query_norm": 911.3247985839844,
      "state_relation_loss": 0.0006646958645433187,
      "step": 7510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004756104853004217,
      "weighted_relation_loss": 0.004756104853004217
    },
    {
      "action_query_norm": 512.157583618164,
      "action_relation_loss": 0.004077299847267568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.034611786716558,
      "flow_matching_loss": 0.047374877706170083,
      "grad_norm": 0.13340242207050323,
      "learning_rate": 8.93926187730699e-05,
      "loss": 0.0521,
      "relation_loss": 0.0047212076839059595,
      "selected_pairs": 27530.2,
      "state_query_norm": 908.2484252929687,
      "state_relation_loss": 0.0006439079239498824,
      "step": 7520,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0047212076839059595,
      "weighted_relation_loss": 0.0047212076839059595
    },
    {
      "action_query_norm": 517.6708251953125,
      "action_relation_loss": 0.004076504032127559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.043966323666979,
      "flow_matching_loss": 0.04857328310608864,
      "grad_norm": 0.1068955808877945,
      "learning_rate": 8.935865108836102e-05,
      "loss": 0.0533,
      "relation_loss": 0.004742853064090014,
      "selected_pairs": 26868.9,
      "state_query_norm": 911.4627075195312,
      "state_relation_loss": 0.000666349078528583,
      "step": 7530,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004742853064090014,
      "weighted_relation_loss": 0.004742853064090014
    },
    {
      "action_query_norm": 512.5985595703125,
      "action_relation_loss": 0.004100571270100772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.0533208606174,
      "flow_matching_loss": 0.04895846657454968,
      "grad_norm": 0.14323687553405762,
      "learning_rate": 8.932463557917981e-05,
      "loss": 0.0537,
      "relation_loss": 0.004735350329428911,
      "selected_pairs": 27058.6,
      "state_query_norm": 926.0817199707031,
      "state_relation_loss": 0.0006347790942527354,
      "step": 7540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004735350329428911,
      "weighted_relation_loss": 0.004735350329428911
    },
    {
      "action_query_norm": 507.9543853759766,
      "action_relation_loss": 0.0039671319769695405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.06267539756782,
      "flow_matching_loss": 0.04762296788394451,
      "grad_norm": 0.16680261492729187,
      "learning_rate": 8.929057228685829e-05,
      "loss": 0.0522,
      "relation_loss": 0.004607529798522592,
      "selected_pairs": 26162.2,
      "state_query_norm": 923.0666625976562,
      "state_relation_loss": 0.0006403978040907532,
      "step": 7550,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.004607529798522592,
      "weighted_relation_loss": 0.004607529798522592
    },
    {
      "action_query_norm": 509.15618591308595,
      "action_relation_loss": 0.003977893642149866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.072029934518241,
      "flow_matching_loss": 0.04457953944802284,
      "grad_norm": 0.1520862579345703,
      "learning_rate": 8.925646125278658e-05,
      "loss": 0.0492,
      "relation_loss": 0.004604031378403306,
      "selected_pairs": 26683.5,
      "state_query_norm": 919.259375,
      "state_relation_loss": 0.000626137648941949,
      "step": 7560,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004604031378403306,
      "weighted_relation_loss": 0.004604031378403306
    },
    {
      "action_query_norm": 516.11845703125,
      "action_relation_loss": 0.0039889566134661434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.081384471468662,
      "flow_matching_loss": 0.04874416403472424,
      "grad_norm": 0.13655230402946472,
      "learning_rate": 8.92223025184128e-05,
      "loss": 0.0534,
      "relation_loss": 0.004620982566848397,
      "selected_pairs": 26848.8,
      "state_query_norm": 920.7138977050781,
      "state_relation_loss": 0.0006320260756183416,
      "step": 7570,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004620982566848397,
      "weighted_relation_loss": 0.004620982566848397
    },
    {
      "action_query_norm": 514.8310668945312,
      "action_relation_loss": 0.004165097372606397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.090739008419083,
      "flow_matching_loss": 0.0450119610875845,
      "grad_norm": 0.12172723561525345,
      "learning_rate": 8.918809612524303e-05,
      "loss": 0.0498,
      "relation_loss": 0.004835597332566976,
      "selected_pairs": 26575.9,
      "state_query_norm": 924.4200134277344,
      "state_relation_loss": 0.0006704999366775155,
      "step": 7580,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004835597332566976,
      "weighted_relation_loss": 0.004835597332566976
    },
    {
      "action_query_norm": 520.4588073730469,
      "action_relation_loss": 0.004206071281805635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.100093545369504,
      "flow_matching_loss": 0.04617603309452534,
      "grad_norm": 0.12897713482379913,
      "learning_rate": 8.91538421148413e-05,
      "loss": 0.051,
      "relation_loss": 0.004858222976326943,
      "selected_pairs": 26421.2,
      "state_query_norm": 919.8436279296875,
      "state_relation_loss": 0.0006521517178043723,
      "step": 7590,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004858222976326943,
      "weighted_relation_loss": 0.004858222976326943
    },
    {
      "action_query_norm": 519.7188995361328,
      "action_relation_loss": 0.0040599244181066755,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1094480823199255,
      "flow_matching_loss": 0.047138508409261703,
      "grad_norm": 0.11881684511899948,
      "learning_rate": 8.91195405288294e-05,
      "loss": 0.0518,
      "relation_loss": 0.004695325111970306,
      "selected_pairs": 26797.9,
      "state_query_norm": 936.0305297851562,
      "state_relation_loss": 0.0006354007462505251,
      "step": 7600,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.004695325111970306,
      "weighted_relation_loss": 0.004695325111970306
    },
    {
      "action_query_norm": 510.4640625,
      "action_relation_loss": 0.004035857575945556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.118802619270346,
      "flow_matching_loss": 0.049322858452796936,
      "grad_norm": 0.16653645038604736,
      "learning_rate": 8.908519140888705e-05,
      "loss": 0.054,
      "relation_loss": 0.0046799791045486925,
      "selected_pairs": 26929.9,
      "state_query_norm": 931.3586791992187,
      "state_relation_loss": 0.0006441215751692653,
      "step": 7610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0046799791045486925,
      "weighted_relation_loss": 0.0046799791045486925
    },
    {
      "action_query_norm": 514.8649719238281,
      "action_relation_loss": 0.004126556077972054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.128157156220767,
      "flow_matching_loss": 0.045211400091648105,
      "grad_norm": 0.13159061968326569,
      "learning_rate": 8.905079479675164e-05,
      "loss": 0.05,
      "relation_loss": 0.0047484983690083025,
      "selected_pairs": 26339.0,
      "state_query_norm": 931.4051330566406,
      "state_relation_loss": 0.0006219423201400786,
      "step": 7620,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0047484983690083025,
      "weighted_relation_loss": 0.0047484983690083025
    },
    {
      "action_query_norm": 519.7519592285156,
      "action_relation_loss": 0.0040880969492718576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.137511693171188,
      "flow_matching_loss": 0.0460042167454958,
      "grad_norm": 0.16998425126075745,
      "learning_rate": 8.901635073421831e-05,
      "loss": 0.0508,
      "relation_loss": 0.004745960747823119,
      "selected_pairs": 27088.2,
      "state_query_norm": 930.9759399414063,
      "state_relation_loss": 0.0006578637403436005,
      "step": 7630,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.004745960747823119,
      "weighted_relation_loss": 0.004745960747823119
    },
    {
      "action_query_norm": 514.0924011230469,
      "action_relation_loss": 0.003923501004464924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.146866230121609,
      "flow_matching_loss": 0.04692248478531837,
      "grad_norm": 0.16330835223197937,
      "learning_rate": 8.898185926313983e-05,
      "loss": 0.0515,
      "relation_loss": 0.004574545519426465,
      "selected_pairs": 26827.2,
      "state_query_norm": 922.6629211425782,
      "state_relation_loss": 0.0006510445207823067,
      "step": 7640,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.004574545519426465,
      "weighted_relation_loss": 0.004574545519426465
    },
    {
      "action_query_norm": 514.5109710693359,
      "action_relation_loss": 0.0039076462620869275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.15622076707203,
      "flow_matching_loss": 0.046257587522268294,
      "grad_norm": 0.12398504465818405,
      "learning_rate": 8.894732042542659e-05,
      "loss": 0.0508,
      "relation_loss": 0.004554359009489417,
      "selected_pairs": 27271.9,
      "state_query_norm": 922.7795959472656,
      "state_relation_loss": 0.0006467127648647874,
      "step": 7650,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004554359009489417,
      "weighted_relation_loss": 0.004554359009489417
    },
    {
      "action_query_norm": 510.9230041503906,
      "action_relation_loss": 0.003849264374002814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.165575304022451,
      "flow_matching_loss": 0.04735092520713806,
      "grad_norm": 0.14947015047073364,
      "learning_rate": 8.891273426304654e-05,
      "loss": 0.0519,
      "relation_loss": 0.004502235539257526,
      "selected_pairs": 26946.4,
      "state_query_norm": 929.4930603027344,
      "state_relation_loss": 0.0006529711070470511,
      "step": 7660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004502235539257526,
      "weighted_relation_loss": 0.004502235539257526
    },
    {
      "action_query_norm": 511.3505157470703,
      "action_relation_loss": 0.0038975358242169023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1749298409728715,
      "flow_matching_loss": 0.04361540153622627,
      "grad_norm": 0.1287873089313507,
      "learning_rate": 8.887810081802515e-05,
      "loss": 0.0482,
      "relation_loss": 0.00454428531229496,
      "selected_pairs": 26694.4,
      "state_query_norm": 919.5005432128906,
      "state_relation_loss": 0.0006467494822572917,
      "step": 7670,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00454428531229496,
      "weighted_relation_loss": 0.00454428531229496
    },
    {
      "action_query_norm": 512.7161499023438,
      "action_relation_loss": 0.0040524456650018695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1842843779232926,
      "flow_matching_loss": 0.04659770056605339,
      "grad_norm": 0.11022771894931793,
      "learning_rate": 8.884342013244529e-05,
      "loss": 0.0513,
      "relation_loss": 0.004689437430351972,
      "selected_pairs": 27314.0,
      "state_query_norm": 923.1420532226563,
      "state_relation_loss": 0.0006369917828124017,
      "step": 7680,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.004689437430351972,
      "weighted_relation_loss": 0.004689437430351972
    },
    {
      "action_query_norm": 511.3255279541016,
      "action_relation_loss": 0.004031734052114188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.193638914873714,
      "flow_matching_loss": 0.047326554730534556,
      "grad_norm": 0.13338667154312134,
      "learning_rate": 8.880869224844725e-05,
      "loss": 0.052,
      "relation_loss": 0.0046612135600298645,
      "selected_pairs": 26902.5,
      "state_query_norm": 928.1290100097656,
      "state_relation_loss": 0.000629479467170313,
      "step": 7690,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0046612135600298645,
      "weighted_relation_loss": 0.0046612135600298645
    },
    {
      "action_query_norm": 514.1151763916016,
      "action_relation_loss": 0.004100810969248414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.202993451824135,
      "flow_matching_loss": 0.04758962951600552,
      "grad_norm": 0.14628303050994873,
      "learning_rate": 8.877391720822875e-05,
      "loss": 0.0523,
      "relation_loss": 0.004749232763424515,
      "selected_pairs": 26876.9,
      "state_query_norm": 931.9595275878906,
      "state_relation_loss": 0.000648421747609973,
      "step": 7700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004749232763424515,
      "weighted_relation_loss": 0.004749232763424515
    },
    {
      "action_query_norm": 515.1523956298828,
      "action_relation_loss": 0.003982981690205634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.212347988774556,
      "flow_matching_loss": 0.04620596021413803,
      "grad_norm": 0.11630793660879135,
      "learning_rate": 8.873909505404468e-05,
      "loss": 0.0508,
      "relation_loss": 0.004602666338905692,
      "selected_pairs": 26846.2,
      "state_query_norm": 925.1207763671875,
      "state_relation_loss": 0.0006196847010869533,
      "step": 7710,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004602666338905692,
      "weighted_relation_loss": 0.004602666338905692
    },
    {
      "action_query_norm": 512.416943359375,
      "action_relation_loss": 0.004029577458277344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.221702525724977,
      "flow_matching_loss": 0.045327603816986084,
      "grad_norm": 0.12044466286897659,
      "learning_rate": 8.870422582820726e-05,
      "loss": 0.05,
      "relation_loss": 0.004661560384556651,
      "selected_pairs": 26801.9,
      "state_query_norm": 930.3442138671875,
      "state_relation_loss": 0.0006319829379208386,
      "step": 7720,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.004661560384556651,
      "weighted_relation_loss": 0.004661560384556651
    },
    {
      "action_query_norm": 521.7852325439453,
      "action_relation_loss": 0.00406543689314276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.231057062675397,
      "flow_matching_loss": 0.047096189111471176,
      "grad_norm": 0.14469225704669952,
      "learning_rate": 8.866930957308588e-05,
      "loss": 0.0518,
      "relation_loss": 0.004723256407305598,
      "selected_pairs": 26943.4,
      "state_query_norm": 924.4437622070312,
      "state_relation_loss": 0.0006578195374459028,
      "step": 7730,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004723256407305598,
      "weighted_relation_loss": 0.004723256407305598
    },
    {
      "action_query_norm": 528.5474914550781,
      "action_relation_loss": 0.004068176262080669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.240411599625818,
      "flow_matching_loss": 0.046379143744707106,
      "grad_norm": 0.14111188054084778,
      "learning_rate": 8.86343463311071e-05,
      "loss": 0.0511,
      "relation_loss": 0.004719288367778063,
      "selected_pairs": 26987.8,
      "state_query_norm": 913.33212890625,
      "state_relation_loss": 0.0006511120125651359,
      "step": 7740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004719288367778063,
      "weighted_relation_loss": 0.004719288367778063
    },
    {
      "action_query_norm": 527.9998840332031,
      "action_relation_loss": 0.004105764511041343,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.249766136576239,
      "flow_matching_loss": 0.0463164247572422,
      "grad_norm": 0.15669897198677063,
      "learning_rate": 8.859933614475452e-05,
      "loss": 0.0511,
      "relation_loss": 0.004768395703285933,
      "selected_pairs": 27200.3,
      "state_query_norm": 935.9756408691406,
      "state_relation_loss": 0.0006626311398576946,
      "step": 7750,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.004768395703285933,
      "weighted_relation_loss": 0.004768395703285933
    },
    {
      "action_query_norm": 524.8763488769531,
      "action_relation_loss": 0.003929354413412511,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.2591206735266605,
      "flow_matching_loss": 0.046283993497490884,
      "grad_norm": 0.19470064342021942,
      "learning_rate": 8.85642790565689e-05,
      "loss": 0.0508,
      "relation_loss": 0.004552929196506739,
      "selected_pairs": 27312.9,
      "state_query_norm": 925.9018310546875,
      "state_relation_loss": 0.0006235747132450342,
      "step": 7760,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.004552929196506739,
      "weighted_relation_loss": 0.004552929196506739
    },
    {
      "action_query_norm": 526.6052642822266,
      "action_relation_loss": 0.0040900143794715405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.268475210477082,
      "flow_matching_loss": 0.046697523817420004,
      "grad_norm": 0.13995759189128876,
      "learning_rate": 8.852917510914782e-05,
      "loss": 0.0514,
      "relation_loss": 0.004729216452687979,
      "selected_pairs": 27549.8,
      "state_query_norm": 927.8949462890625,
      "state_relation_loss": 0.0006392021430656313,
      "step": 7770,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.004729216452687979,
      "weighted_relation_loss": 0.004729216452687979
    },
    {
      "action_query_norm": 523.257211303711,
      "action_relation_loss": 0.003956609941087663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.277829747427503,
      "flow_matching_loss": 0.047533102333545685,
      "grad_norm": 0.14056380093097687,
      "learning_rate": 8.84940243451459e-05,
      "loss": 0.0521,
      "relation_loss": 0.004555981745943427,
      "selected_pairs": 27103.3,
      "state_query_norm": 935.4669006347656,
      "state_relation_loss": 0.0005993718048557639,
      "step": 7780,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.004555981745943427,
      "weighted_relation_loss": 0.004555981745943427
    },
    {
      "action_query_norm": 529.2887084960937,
      "action_relation_loss": 0.0038929481524974108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.287184284377924,
      "flow_matching_loss": 0.049003676325082776,
      "grad_norm": 0.11078283190727234,
      "learning_rate": 8.845882680727469e-05,
      "loss": 0.0535,
      "relation_loss": 0.004542436124756933,
      "selected_pairs": 27064.9,
      "state_query_norm": 939.1418090820313,
      "state_relation_loss": 0.0006494879839010536,
      "step": 7790,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004542436124756933,
      "weighted_relation_loss": 0.004542436124756933
    },
    {
      "action_query_norm": 523.7294372558594,
      "action_relation_loss": 0.0040657027857378125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.296538821328344,
      "flow_matching_loss": 0.04702260866761208,
      "grad_norm": 0.11325513571500778,
      "learning_rate": 8.842358253830246e-05,
      "loss": 0.0517,
      "relation_loss": 0.00468513541854918,
      "selected_pairs": 26664.8,
      "state_query_norm": 941.437451171875,
      "state_relation_loss": 0.0006194326793774963,
      "step": 7800,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00468513541854918,
      "weighted_relation_loss": 0.00468513541854918
    },
    {
      "action_query_norm": 517.7862426757813,
      "action_relation_loss": 0.003972657606936991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.305893358278765,
      "flow_matching_loss": 0.04834022931754589,
      "grad_norm": 0.1473883092403412,
      "learning_rate": 8.838829158105434e-05,
      "loss": 0.0529,
      "relation_loss": 0.00459204912185669,
      "selected_pairs": 26546.3,
      "state_query_norm": 925.3837890625,
      "state_relation_loss": 0.0006193915382027626,
      "step": 7810,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.00459204912185669,
      "weighted_relation_loss": 0.00459204912185669
    },
    {
      "action_query_norm": 519.4896789550781,
      "action_relation_loss": 0.004027042118832469,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.315247895229186,
      "flow_matching_loss": 0.04608314633369446,
      "grad_norm": 0.13389287889003754,
      "learning_rate": 8.835295397841217e-05,
      "loss": 0.0507,
      "relation_loss": 0.004655892634764314,
      "selected_pairs": 26577.2,
      "state_query_norm": 940.2551818847656,
      "state_relation_loss": 0.0006288504868280142,
      "step": 7820,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004655892634764314,
      "weighted_relation_loss": 0.004655892634764314
    },
    {
      "action_query_norm": 513.9511413574219,
      "action_relation_loss": 0.003967474703676999,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.324602432179607,
      "flow_matching_loss": 0.04430330209434032,
      "grad_norm": 0.1262970268726349,
      "learning_rate": 8.831756977331447e-05,
      "loss": 0.0489,
      "relation_loss": 0.004609988536685705,
      "selected_pairs": 26386.2,
      "state_query_norm": 949.1316528320312,
      "state_relation_loss": 0.0006425138330087065,
      "step": 7830,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.004609988536685705,
      "weighted_relation_loss": 0.004609988536685705
    },
    {
      "action_query_norm": 515.589697265625,
      "action_relation_loss": 0.0040871792705729606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.333956969130028,
      "flow_matching_loss": 0.04849584959447384,
      "grad_norm": 0.11979864537715912,
      "learning_rate": 8.828213900875638e-05,
      "loss": 0.0532,
      "relation_loss": 0.004718262096866965,
      "selected_pairs": 26689.0,
      "state_query_norm": 929.613134765625,
      "state_relation_loss": 0.0006310828612186015,
      "step": 7840,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004718262096866965,
      "weighted_relation_loss": 0.004718262096866965
    },
    {
      "action_query_norm": 518.4766784667969,
      "action_relation_loss": 0.003950229985639453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.343311506080449,
      "flow_matching_loss": 0.04592828564345837,
      "grad_norm": 0.1467534452676773,
      "learning_rate": 8.824666172778964e-05,
      "loss": 0.0505,
      "relation_loss": 0.004588539712131023,
      "selected_pairs": 26765.9,
      "state_query_norm": 923.3991271972657,
      "state_relation_loss": 0.0006383096799254418,
      "step": 7850,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004588539712131023,
      "weighted_relation_loss": 0.004588539712131023
    },
    {
      "action_query_norm": 518.5912536621094,
      "action_relation_loss": 0.0041004466125741604,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.35266604303087,
      "flow_matching_loss": 0.04766035750508309,
      "grad_norm": 0.1475106179714203,
      "learning_rate": 8.821113797352245e-05,
      "loss": 0.0524,
      "relation_loss": 0.004750349558889866,
      "selected_pairs": 26852.2,
      "state_query_norm": 934.4881103515625,
      "state_relation_loss": 0.0006499029637780041,
      "step": 7860,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004750349558889866,
      "weighted_relation_loss": 0.004750349558889866
    },
    {
      "action_query_norm": 524.9134735107422,
      "action_relation_loss": 0.003949271398596466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.362020579981291,
      "flow_matching_loss": 0.045225202292203906,
      "grad_norm": 0.15234726667404175,
      "learning_rate": 8.817556778911956e-05,
      "loss": 0.0498,
      "relation_loss": 0.004588138684630394,
      "selected_pairs": 26671.0,
      "state_query_norm": 946.6626831054688,
      "state_relation_loss": 0.0006388673034962266,
      "step": 7870,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004588138684630394,
      "weighted_relation_loss": 0.004588138684630394
    },
    {
      "action_query_norm": 522.8946838378906,
      "action_relation_loss": 0.003822307987138629,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.371375116931712,
      "flow_matching_loss": 0.04385448656976223,
      "grad_norm": 0.14578551054000854,
      "learning_rate": 8.81399512178021e-05,
      "loss": 0.0483,
      "relation_loss": 0.004450303083285689,
      "selected_pairs": 26755.8,
      "state_query_norm": 935.297412109375,
      "state_relation_loss": 0.0006279951485339552,
      "step": 7880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004450303083285689,
      "weighted_relation_loss": 0.004450303083285689
    },
    {
      "action_query_norm": 523.7557312011719,
      "action_relation_loss": 0.003926800168119371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.380729653882133,
      "flow_matching_loss": 0.044671355932950976,
      "grad_norm": 0.1377589851617813,
      "learning_rate": 8.810428830284752e-05,
      "loss": 0.0492,
      "relation_loss": 0.004522110102698207,
      "selected_pairs": 26500.4,
      "state_query_norm": 938.9575012207031,
      "state_relation_loss": 0.0005953098705504089,
      "step": 7890,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.004522110102698207,
      "weighted_relation_loss": 0.004522110102698207
    },
    {
      "action_query_norm": 524.0446228027344,
      "action_relation_loss": 0.0036584598943591117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.390084190832554,
      "flow_matching_loss": 0.04637015797197819,
      "grad_norm": 0.13771624863147736,
      "learning_rate": 8.806857908758967e-05,
      "loss": 0.0507,
      "relation_loss": 0.004287360049784183,
      "selected_pairs": 26536.6,
      "state_query_norm": 932.7638488769531,
      "state_relation_loss": 0.0006289001554250717,
      "step": 7900,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.004287360049784183,
      "weighted_relation_loss": 0.004287360049784183
    },
    {
      "action_query_norm": 523.6795043945312,
      "action_relation_loss": 0.0038544081849977373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.399438727782975,
      "flow_matching_loss": 0.04441789500415325,
      "grad_norm": 0.10741928219795227,
      "learning_rate": 8.80328236154186e-05,
      "loss": 0.0489,
      "relation_loss": 0.004455517185851932,
      "selected_pairs": 26556.1,
      "state_query_norm": 936.2706665039062,
      "state_relation_loss": 0.0006011090765241534,
      "step": 7910,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.004455517185851932,
      "weighted_relation_loss": 0.004455517185851932
    },
    {
      "action_query_norm": 527.511083984375,
      "action_relation_loss": 0.00390121981035918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.408793264733395,
      "flow_matching_loss": 0.04560843445360661,
      "grad_norm": 0.10620895773172379,
      "learning_rate": 8.799702192978056e-05,
      "loss": 0.0501,
      "relation_loss": 0.004519156273454427,
      "selected_pairs": 25996.9,
      "state_query_norm": 941.6592895507813,
      "state_relation_loss": 0.0006179364921990782,
      "step": 7920,
      "valid_action_samples": 228.4,
      "weighted_auxiliary_loss": 0.004519156273454427,
      "weighted_relation_loss": 0.004519156273454427
    },
    {
      "action_query_norm": 523.0976440429688,
      "action_relation_loss": 0.0039995158789679405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.4181478016838165,
      "flow_matching_loss": 0.04520392045378685,
      "grad_norm": 0.10763239860534668,
      "learning_rate": 8.7961174074178e-05,
      "loss": 0.0498,
      "relation_loss": 0.004597704997286201,
      "selected_pairs": 26708.8,
      "state_query_norm": 937.4336669921875,
      "state_relation_loss": 0.0005981890892144293,
      "step": 7930,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.004597704997286201,
      "weighted_relation_loss": 0.004597704997286201
    },
    {
      "action_query_norm": 518.9557708740234,
      "action_relation_loss": 0.0038583210902288557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.427502338634238,
      "flow_matching_loss": 0.04454657882452011,
      "grad_norm": 0.13084383308887482,
      "learning_rate": 8.792528009216942e-05,
      "loss": 0.049,
      "relation_loss": 0.004480999102815986,
      "selected_pairs": 26739.5,
      "state_query_norm": 956.0146728515625,
      "state_relation_loss": 0.0006226780125871301,
      "step": 7940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004480999102815986,
      "weighted_relation_loss": 0.004480999102815986
    },
    {
      "action_query_norm": 517.4895080566406,
      "action_relation_loss": 0.0037773519987240435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.436856875584659,
      "flow_matching_loss": 0.046036606281995775,
      "grad_norm": 0.11724791675806046,
      "learning_rate": 8.788934002736943e-05,
      "loss": 0.0504,
      "relation_loss": 0.004397428897209466,
      "selected_pairs": 26670.3,
      "state_query_norm": 929.8171325683594,
      "state_relation_loss": 0.0006200769043061882,
      "step": 7950,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004397428897209466,
      "weighted_relation_loss": 0.004397428897209466
    },
    {
      "action_query_norm": 525.6163543701172,
      "action_relation_loss": 0.003929098160006106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.44621141253508,
      "flow_matching_loss": 0.048157722875475886,
      "grad_norm": 0.11779699474573135,
      "learning_rate": 8.785335392344857e-05,
      "loss": 0.0527,
      "relation_loss": 0.0045437539462000135,
      "selected_pairs": 26644.4,
      "state_query_norm": 958.6058715820312,
      "state_relation_loss": 0.0006146558269392699,
      "step": 7960,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0045437539462000135,
      "weighted_relation_loss": 0.0045437539462000135
    },
    {
      "action_query_norm": 523.9324401855469,
      "action_relation_loss": 0.0038888967130333183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.4555659494855,
      "flow_matching_loss": 0.04744897186756134,
      "grad_norm": 0.1182965561747551,
      "learning_rate": 8.781732182413336e-05,
      "loss": 0.052,
      "relation_loss": 0.00452163340523839,
      "selected_pairs": 26787.1,
      "state_query_norm": 942.5304504394531,
      "state_relation_loss": 0.0006327367329504341,
      "step": 7970,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.00452163340523839,
      "weighted_relation_loss": 0.00452163340523839
    },
    {
      "action_query_norm": 527.8507446289062,
      "action_relation_loss": 0.0038052676944062115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.464920486435921,
      "flow_matching_loss": 0.046778169274330136,
      "grad_norm": 0.14884454011917114,
      "learning_rate": 8.778124377320618e-05,
      "loss": 0.0512,
      "relation_loss": 0.0044457293581217526,
      "selected_pairs": 26275.4,
      "state_query_norm": 950.0121520996094,
      "state_relation_loss": 0.0006404617335647345,
      "step": 7980,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0044457293581217526,
      "weighted_relation_loss": 0.0044457293581217526
    },
    {
      "action_query_norm": 529.8316711425781,
      "action_relation_loss": 0.0038073206786066295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.474275023386342,
      "flow_matching_loss": 0.04567047990858555,
      "grad_norm": 0.1512722223997116,
      "learning_rate": 8.774511981450529e-05,
      "loss": 0.0501,
      "relation_loss": 0.004420643392950297,
      "selected_pairs": 26989.2,
      "state_query_norm": 953.4545288085938,
      "state_relation_loss": 0.0006133227434474975,
      "step": 7990,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004420643392950297,
      "weighted_relation_loss": 0.004420643392950297
    },
    {
      "action_query_norm": 527.6935363769531,
      "action_relation_loss": 0.003777579008601606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.483629560336763,
      "flow_matching_loss": 0.048735448718070985,
      "grad_norm": 0.14668193459510803,
      "learning_rate": 8.770894999192467e-05,
      "loss": 0.0531,
      "relation_loss": 0.004406402865424752,
      "selected_pairs": 26742.7,
      "state_query_norm": 942.1049865722656,
      "state_relation_loss": 0.0006288237927947193,
      "step": 8000,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004406402865424752,
      "weighted_relation_loss": 0.004406402865424752
    },
    {
      "action_query_norm": 527.9415283203125,
      "action_relation_loss": 0.003722730511799455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.492984097287184,
      "flow_matching_loss": 0.04565029442310333,
      "grad_norm": 0.13073040544986725,
      "learning_rate": 8.767273434941412e-05,
      "loss": 0.05,
      "relation_loss": 0.004331588372588157,
      "selected_pairs": 26966.4,
      "state_query_norm": 949.1980041503906,
      "state_relation_loss": 0.0006088578491471707,
      "step": 8010,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.004331588372588157,
      "weighted_relation_loss": 0.004331588372588157
    },
    {
      "action_query_norm": 517.8497619628906,
      "action_relation_loss": 0.00375060155056417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.5023386342376055,
      "flow_matching_loss": 0.049091894179582596,
      "grad_norm": 0.09614181518554688,
      "learning_rate": 8.7636472930979e-05,
      "loss": 0.0535,
      "relation_loss": 0.004373769788071513,
      "selected_pairs": 26736.2,
      "state_query_norm": 951.0515930175782,
      "state_relation_loss": 0.0006231682666111738,
      "step": 8020,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004373769788071513,
      "weighted_relation_loss": 0.004373769788071513
    },
    {
      "action_query_norm": 520.2145751953125,
      "action_relation_loss": 0.00372809236869216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.511693171188027,
      "flow_matching_loss": 0.045771818980574606,
      "grad_norm": 0.16740542650222778,
      "learning_rate": 8.76001657806804e-05,
      "loss": 0.0501,
      "relation_loss": 0.004362330911681056,
      "selected_pairs": 27082.3,
      "state_query_norm": 928.4939697265625,
      "state_relation_loss": 0.0006342385662719607,
      "step": 8030,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.004362330911681056,
      "weighted_relation_loss": 0.004362330911681056
    },
    {
      "action_query_norm": 524.1717346191406,
      "action_relation_loss": 0.003880334459245205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.521047708138447,
      "flow_matching_loss": 0.0463707584887743,
      "grad_norm": 0.17343761026859283,
      "learning_rate": 8.756381294263491e-05,
      "loss": 0.0509,
      "relation_loss": 0.004512317618355155,
      "selected_pairs": 26593.9,
      "state_query_norm": 953.6648681640625,
      "state_relation_loss": 0.000631983153289184,
      "step": 8040,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004512317618355155,
      "weighted_relation_loss": 0.004512317618355155
    },
    {
      "action_query_norm": 526.3343933105468,
      "action_relation_loss": 0.003886175691150129,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.530402245088868,
      "flow_matching_loss": 0.04485687427222729,
      "grad_norm": 0.11429870128631592,
      "learning_rate": 8.752741446101464e-05,
      "loss": 0.0494,
      "relation_loss": 0.004500855505466461,
      "selected_pairs": 26991.3,
      "state_query_norm": 951.0267211914063,
      "state_relation_loss": 0.0006146798375993967,
      "step": 8050,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004500855505466461,
      "weighted_relation_loss": 0.004500855505466461
    },
    {
      "action_query_norm": 520.1774383544922,
      "action_relation_loss": 0.003774738730862737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.539756782039289,
      "flow_matching_loss": 0.045749160647392276,
      "grad_norm": 0.2023480236530304,
      "learning_rate": 8.749097038004722e-05,
      "loss": 0.0501,
      "relation_loss": 0.004361889604479074,
      "selected_pairs": 27456.2,
      "state_query_norm": 944.2058715820312,
      "state_relation_loss": 0.0005871509201824665,
      "step": 8060,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.004361889604479074,
      "weighted_relation_loss": 0.004361889604479074
    },
    {
      "action_query_norm": 528.1732849121094,
      "action_relation_loss": 0.0038461589720100164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.54911131898971,
      "flow_matching_loss": 0.045958439260721205,
      "grad_norm": 0.18390096724033356,
      "learning_rate": 8.745448074401561e-05,
      "loss": 0.0504,
      "relation_loss": 0.004443357372656465,
      "selected_pairs": 26803.1,
      "state_query_norm": 959.2000732421875,
      "state_relation_loss": 0.0005971984704956412,
      "step": 8070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.004443357372656465,
      "weighted_relation_loss": 0.004443357372656465
    },
    {
      "action_query_norm": 530.2654174804687,
      "action_relation_loss": 0.003859643149189651,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.558465855940131,
      "flow_matching_loss": 0.047632810100913046,
      "grad_norm": 0.21205921471118927,
      "learning_rate": 8.741794559725818e-05,
      "loss": 0.0521,
      "relation_loss": 0.00444524441845715,
      "selected_pairs": 27639.3,
      "state_query_norm": 954.0643859863281,
      "state_relation_loss": 0.0005856011935975403,
      "step": 8080,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.00444524441845715,
      "weighted_relation_loss": 0.00444524441845715
    },
    {
      "action_query_norm": 527.6815887451172,
      "action_relation_loss": 0.0038160757860168813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.567820392890552,
      "flow_matching_loss": 0.045935345068573954,
      "grad_norm": 0.15865270793437958,
      "learning_rate": 8.738136498416856e-05,
      "loss": 0.0504,
      "relation_loss": 0.004450913425534964,
      "selected_pairs": 26410.0,
      "state_query_norm": 943.0098083496093,
      "state_relation_loss": 0.000634837691904977,
      "step": 8090,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004450913425534964,
      "weighted_relation_loss": 0.004450913425534964
    },
    {
      "action_query_norm": 522.2746154785157,
      "action_relation_loss": 0.003759830794297159,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.5771749298409725,
      "flow_matching_loss": 0.045838237181305884,
      "grad_norm": 0.15723463892936707,
      "learning_rate": 8.734473894919564e-05,
      "loss": 0.0502,
      "relation_loss": 0.004367028828710318,
      "selected_pairs": 27361.5,
      "state_query_norm": 956.2963684082031,
      "state_relation_loss": 0.0006071979238186032,
      "step": 8100,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.004367028828710318,
      "weighted_relation_loss": 0.004367028828710318
    },
    {
      "action_query_norm": 509.65460815429685,
      "action_relation_loss": 0.003849510964937508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.586529466791394,
      "flow_matching_loss": 0.0458855751901865,
      "grad_norm": 0.18802976608276367,
      "learning_rate": 8.730806753684353e-05,
      "loss": 0.0503,
      "relation_loss": 0.004456914565525949,
      "selected_pairs": 26233.7,
      "state_query_norm": 952.3001770019531,
      "state_relation_loss": 0.000607403606409207,
      "step": 8110,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.004456914565525949,
      "weighted_relation_loss": 0.004456914565525949
    },
    {
      "action_query_norm": 513.826123046875,
      "action_relation_loss": 0.003956642327830195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.595884003741815,
      "flow_matching_loss": 0.047934343293309215,
      "grad_norm": 0.19753620028495789,
      "learning_rate": 8.727135079167144e-05,
      "loss": 0.0525,
      "relation_loss": 0.004589314758777619,
      "selected_pairs": 26593.6,
      "state_query_norm": 959.2075378417969,
      "state_relation_loss": 0.0006326724833343178,
      "step": 8120,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004589314758777619,
      "weighted_relation_loss": 0.004589314758777619
    },
    {
      "action_query_norm": 521.5445922851562,
      "action_relation_loss": 0.00402541586663574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.605238540692236,
      "flow_matching_loss": 0.04583433009684086,
      "grad_norm": 0.12366708368062973,
      "learning_rate": 8.723458875829368e-05,
      "loss": 0.0505,
      "relation_loss": 0.004680114472284913,
      "selected_pairs": 27317.2,
      "state_query_norm": 937.0751953125,
      "state_relation_loss": 0.0006546986638568342,
      "step": 8130,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.004680114472284913,
      "weighted_relation_loss": 0.004680114472284913
    },
    {
      "action_query_norm": 525.4060180664062,
      "action_relation_loss": 0.0040133180096745494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.614593077642657,
      "flow_matching_loss": 0.04604966789484024,
      "grad_norm": 0.16718286275863647,
      "learning_rate": 8.719778148137958e-05,
      "loss": 0.0507,
      "relation_loss": 0.004643405228853226,
      "selected_pairs": 27218.6,
      "state_query_norm": 955.8975524902344,
      "state_relation_loss": 0.0006300872075371445,
      "step": 8140,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.004643405228853226,
      "weighted_relation_loss": 0.004643405228853226
    },
    {
      "action_query_norm": 526.844580078125,
      "action_relation_loss": 0.0038124519400298596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.623947614593078,
      "flow_matching_loss": 0.04819395653903484,
      "grad_norm": 0.12034691125154495,
      "learning_rate": 8.716092900565346e-05,
      "loss": 0.0526,
      "relation_loss": 0.004445585096254945,
      "selected_pairs": 27339.6,
      "state_query_norm": 957.8034606933594,
      "state_relation_loss": 0.0006331331562250853,
      "step": 8150,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.004445585096254945,
      "weighted_relation_loss": 0.004445585096254945
    },
    {
      "action_query_norm": 527.2497314453125,
      "action_relation_loss": 0.003910005977377296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.633302151543498,
      "flow_matching_loss": 0.04517250172793865,
      "grad_norm": 0.10483146458864212,
      "learning_rate": 8.712403137589456e-05,
      "loss": 0.0497,
      "relation_loss": 0.004575505992397666,
      "selected_pairs": 26443.3,
      "state_query_norm": 947.7626647949219,
      "state_relation_loss": 0.0006655000732280314,
      "step": 8160,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004575505992397666,
      "weighted_relation_loss": 0.004575505992397666
    },
    {
      "action_query_norm": 528.6126403808594,
      "action_relation_loss": 0.003812119481153786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.642656688493919,
      "flow_matching_loss": 0.04336623065173626,
      "grad_norm": 0.14041218161582947,
      "learning_rate": 8.708708863693697e-05,
      "loss": 0.0478,
      "relation_loss": 0.00446985773742199,
      "selected_pairs": 26592.6,
      "state_query_norm": 950.5677917480468,
      "state_relation_loss": 0.0006577382504474371,
      "step": 8170,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.00446985773742199,
      "weighted_relation_loss": 0.00446985773742199
    },
    {
      "action_query_norm": 538.0149047851562,
      "action_relation_loss": 0.003695158986374736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.65201122544434,
      "flow_matching_loss": 0.0461704146116972,
      "grad_norm": 0.17924228310585022,
      "learning_rate": 8.70501008336696e-05,
      "loss": 0.0505,
      "relation_loss": 0.004331409931182861,
      "selected_pairs": 26625.7,
      "state_query_norm": 963.1275390625,
      "state_relation_loss": 0.000636250973911956,
      "step": 8180,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.004331409931182861,
      "weighted_relation_loss": 0.004331409931182861
    },
    {
      "action_query_norm": 529.8655822753906,
      "action_relation_loss": 0.003997420473024249,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.6613657623947615,
      "flow_matching_loss": 0.04565142542123794,
      "grad_norm": 0.12469110637903214,
      "learning_rate": 8.701306801103611e-05,
      "loss": 0.0503,
      "relation_loss": 0.00461810699198395,
      "selected_pairs": 26687.9,
      "state_query_norm": 975.1375610351563,
      "state_relation_loss": 0.000620686454931274,
      "step": 8190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00461810699198395,
      "weighted_relation_loss": 0.00461810699198395
    },
    {
      "action_query_norm": 537.3650512695312,
      "action_relation_loss": 0.004009532951749861,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.670720299345183,
      "flow_matching_loss": 0.04796731285750866,
      "grad_norm": 0.13072435557842255,
      "learning_rate": 8.697599021403489e-05,
      "loss": 0.0526,
      "relation_loss": 0.004657566547393799,
      "selected_pairs": 26757.4,
      "state_query_norm": 958.7047668457031,
      "state_relation_loss": 0.0006480335898231715,
      "step": 8200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004657566547393799,
      "weighted_relation_loss": 0.004657566547393799
    },
    {
      "action_query_norm": 527.3225555419922,
      "action_relation_loss": 0.0038408600492402913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.680074836295604,
      "flow_matching_loss": 0.046033330261707306,
      "grad_norm": 0.1432667076587677,
      "learning_rate": 8.693886748771896e-05,
      "loss": 0.0505,
      "relation_loss": 0.004472321039065719,
      "selected_pairs": 26844.6,
      "state_query_norm": 950.7819519042969,
      "state_relation_loss": 0.0006314609665423632,
      "step": 8210,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004472321039065719,
      "weighted_relation_loss": 0.004472321039065719
    },
    {
      "action_query_norm": 522.8514953613281,
      "action_relation_loss": 0.0037761164829134943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.689429373246024,
      "flow_matching_loss": 0.04357806332409382,
      "grad_norm": 0.12097101658582687,
      "learning_rate": 8.690169987719593e-05,
      "loss": 0.048,
      "relation_loss": 0.004399337992072106,
      "selected_pairs": 26694.2,
      "state_query_norm": 951.4782775878906,
      "state_relation_loss": 0.0006232214975170792,
      "step": 8220,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004399337992072106,
      "weighted_relation_loss": 0.004399337992072106
    },
    {
      "action_query_norm": 525.2640747070312,
      "action_relation_loss": 0.003891175193712115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.698783910196445,
      "flow_matching_loss": 0.048077859356999396,
      "grad_norm": 0.10522059351205826,
      "learning_rate": 8.686448742762791e-05,
      "loss": 0.0526,
      "relation_loss": 0.0044934943318367,
      "selected_pairs": 27404.6,
      "state_query_norm": 956.9010437011718,
      "state_relation_loss": 0.0006023191148415208,
      "step": 8230,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0044934943318367,
      "weighted_relation_loss": 0.0044934943318367
    },
    {
      "action_query_norm": 532.0315246582031,
      "action_relation_loss": 0.0038012113654986024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.708138447146866,
      "flow_matching_loss": 0.04502856209874153,
      "grad_norm": 0.16248907148838043,
      "learning_rate": 8.682723018423161e-05,
      "loss": 0.0495,
      "relation_loss": 0.004458381561562419,
      "selected_pairs": 26593.1,
      "state_query_norm": 958.7580749511719,
      "state_relation_loss": 0.0006571700796484947,
      "step": 8240,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004458381561562419,
      "weighted_relation_loss": 0.004458381561562419
    },
    {
      "action_query_norm": 539.1028747558594,
      "action_relation_loss": 0.0036745429504662754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.717492984097287,
      "flow_matching_loss": 0.044858097285032275,
      "grad_norm": 0.1516617089509964,
      "learning_rate": 8.678992819227804e-05,
      "loss": 0.0492,
      "relation_loss": 0.0043023241683840755,
      "selected_pairs": 26852.4,
      "state_query_norm": 964.4057556152344,
      "state_relation_loss": 0.0006277812062762678,
      "step": 8250,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0043023241683840755,
      "weighted_relation_loss": 0.0043023241683840755
    },
    {
      "action_query_norm": 530.4803466796875,
      "action_relation_loss": 0.003695951006375253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.726847521047708,
      "flow_matching_loss": 0.044401464983820915,
      "grad_norm": 0.1526845097541809,
      "learning_rate": 8.675258149709265e-05,
      "loss": 0.0487,
      "relation_loss": 0.0042880745604634285,
      "selected_pairs": 26673.4,
      "state_query_norm": 944.9651794433594,
      "state_relation_loss": 0.0005921235599089414,
      "step": 8260,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0042880745604634285,
      "weighted_relation_loss": 0.0042880745604634285
    },
    {
      "action_query_norm": 529.2898315429687,
      "action_relation_loss": 0.003777149738743901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.736202057998129,
      "flow_matching_loss": 0.04446127712726593,
      "grad_norm": 0.1478595733642578,
      "learning_rate": 8.67151901440552e-05,
      "loss": 0.0489,
      "relation_loss": 0.00439485078677535,
      "selected_pairs": 26626.7,
      "state_query_norm": 947.2582153320312,
      "state_relation_loss": 0.000617701024748385,
      "step": 8270,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00439485078677535,
      "weighted_relation_loss": 0.00439485078677535
    },
    {
      "action_query_norm": 526.7798706054688,
      "action_relation_loss": 0.003830328886397183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.74555659494855,
      "flow_matching_loss": 0.04600383788347244,
      "grad_norm": 0.1466798633337021,
      "learning_rate": 8.667775417859971e-05,
      "loss": 0.0504,
      "relation_loss": 0.004416847275570035,
      "selected_pairs": 27274.3,
      "state_query_norm": 954.8776611328125,
      "state_relation_loss": 0.0005865183949936182,
      "step": 8280,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004416847275570035,
      "weighted_relation_loss": 0.004416847275570035
    },
    {
      "action_query_norm": 527.7296813964844,
      "action_relation_loss": 0.0038165083155035973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.754911131898971,
      "flow_matching_loss": 0.04466906376183033,
      "grad_norm": 0.14577335119247437,
      "learning_rate": 8.664027364621441e-05,
      "loss": 0.0491,
      "relation_loss": 0.00441261432133615,
      "selected_pairs": 27181.9,
      "state_query_norm": 949.3770202636719,
      "state_relation_loss": 0.0005961059767287224,
      "step": 8290,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.00441261432133615,
      "weighted_relation_loss": 0.00441261432133615
    },
    {
      "action_query_norm": 531.8586608886719,
      "action_relation_loss": 0.0036842938512563705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.764265668849392,
      "flow_matching_loss": 0.04375079944729805,
      "grad_norm": 0.09674638509750366,
      "learning_rate": 8.660274859244168e-05,
      "loss": 0.048,
      "relation_loss": 0.0042829851154237986,
      "selected_pairs": 27406.4,
      "state_query_norm": 950.5733947753906,
      "state_relation_loss": 0.0005986912525258959,
      "step": 8300,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0042829851154237986,
      "weighted_relation_loss": 0.0042829851154237986
    },
    {
      "action_query_norm": 535.525048828125,
      "action_relation_loss": 0.003648552601225674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.773620205799813,
      "flow_matching_loss": 0.04541836157441139,
      "grad_norm": 0.15159979462623596,
      "learning_rate": 8.656517906287799e-05,
      "loss": 0.0497,
      "relation_loss": 0.0042447969550266865,
      "selected_pairs": 27082.1,
      "state_query_norm": 959.1208557128906,
      "state_relation_loss": 0.0005962443072348833,
      "step": 8310,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0042447969550266865,
      "weighted_relation_loss": 0.0042447969550266865
    },
    {
      "action_query_norm": 533.5879760742188,
      "action_relation_loss": 0.0037983704125508664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.782974742750234,
      "flow_matching_loss": 0.048194879293441774,
      "grad_norm": 0.14756983518600464,
      "learning_rate": 8.652756510317387e-05,
      "loss": 0.0526,
      "relation_loss": 0.0043855234514921905,
      "selected_pairs": 27179.3,
      "state_query_norm": 971.3963439941406,
      "state_relation_loss": 0.0005871530913282186,
      "step": 8320,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0043855234514921905,
      "weighted_relation_loss": 0.0043855234514921905
    },
    {
      "action_query_norm": 526.1124450683594,
      "action_relation_loss": 0.003733927686698735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.792329279700655,
      "flow_matching_loss": 0.04453525990247727,
      "grad_norm": 0.17185191810131073,
      "learning_rate": 8.648990675903382e-05,
      "loss": 0.0489,
      "relation_loss": 0.004327564779669046,
      "selected_pairs": 27013.8,
      "state_query_norm": 964.9368835449219,
      "state_relation_loss": 0.0005936370696872472,
      "step": 8330,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004327564779669046,
      "weighted_relation_loss": 0.004327564779669046
    },
    {
      "action_query_norm": 529.0268920898437,
      "action_relation_loss": 0.003611071896739304,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.801683816651076,
      "flow_matching_loss": 0.04603711254894734,
      "grad_norm": 0.12825067341327667,
      "learning_rate": 8.64522040762163e-05,
      "loss": 0.0502,
      "relation_loss": 0.004207918886095285,
      "selected_pairs": 27175.6,
      "state_query_norm": 962.7179748535157,
      "state_relation_loss": 0.0005968469427898526,
      "step": 8340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.004207918886095285,
      "weighted_relation_loss": 0.004207918886095285
    },
    {
      "action_query_norm": 534.3068115234375,
      "action_relation_loss": 0.003771249554120004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8110383536014965,
      "flow_matching_loss": 0.044796520844101906,
      "grad_norm": 0.14542225003242493,
      "learning_rate": 8.64144571005336e-05,
      "loss": 0.0492,
      "relation_loss": 0.004381239833310247,
      "selected_pairs": 26490.7,
      "state_query_norm": 968.454833984375,
      "state_relation_loss": 0.0006099903432186693,
      "step": 8350,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.004381239833310247,
      "weighted_relation_loss": 0.004381239833310247
    },
    {
      "action_query_norm": 530.0711608886719,
      "action_relation_loss": 0.00393716364633292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8203928905519176,
      "flow_matching_loss": 0.04494825676083565,
      "grad_norm": 0.11859716475009918,
      "learning_rate": 8.637666587785184e-05,
      "loss": 0.0495,
      "relation_loss": 0.0045592047739773985,
      "selected_pairs": 26507.8,
      "state_query_norm": 959.0075805664062,
      "state_relation_loss": 0.0006220411683898419,
      "step": 8360,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0045592047739773985,
      "weighted_relation_loss": 0.0045592047739773985
    },
    {
      "action_query_norm": 531.0871643066406,
      "action_relation_loss": 0.0036294997436925767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.829747427502339,
      "flow_matching_loss": 0.04556017816066742,
      "grad_norm": 0.12812325358390808,
      "learning_rate": 8.633883045409096e-05,
      "loss": 0.0498,
      "relation_loss": 0.004239895334467292,
      "selected_pairs": 26873.5,
      "state_query_norm": 953.7480895996093,
      "state_relation_loss": 0.0006103956024162471,
      "step": 8370,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004239895334467292,
      "weighted_relation_loss": 0.004239895334467292
    },
    {
      "action_query_norm": 528.6736083984375,
      "action_relation_loss": 0.003926711878739298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.83910196445276,
      "flow_matching_loss": 0.04628663510084152,
      "grad_norm": 0.15955455601215363,
      "learning_rate": 8.630095087522457e-05,
      "loss": 0.0508,
      "relation_loss": 0.004540586657822132,
      "selected_pairs": 26778.5,
      "state_query_norm": 964.4176696777344,
      "state_relation_loss": 0.0006138747616205364,
      "step": 8380,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004540586657822132,
      "weighted_relation_loss": 0.004540586657822132
    },
    {
      "action_query_norm": 527.0779296875,
      "action_relation_loss": 0.0038152704713866115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.848456501403181,
      "flow_matching_loss": 0.04821031503379345,
      "grad_norm": 0.12975198030471802,
      "learning_rate": 8.62630271872799e-05,
      "loss": 0.0526,
      "relation_loss": 0.00440807556733489,
      "selected_pairs": 26247.1,
      "state_query_norm": 965.4575378417969,
      "state_relation_loss": 0.0005928051890805363,
      "step": 8390,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.00440807556733489,
      "weighted_relation_loss": 0.00440807556733489
    },
    {
      "action_query_norm": 531.8990234375,
      "action_relation_loss": 0.0037172949174419045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.857811038353601,
      "flow_matching_loss": 0.04374498538672924,
      "grad_norm": 0.12321890890598297,
      "learning_rate": 8.62250594363378e-05,
      "loss": 0.0481,
      "relation_loss": 0.0043169274926185604,
      "selected_pairs": 26954.1,
      "state_query_norm": 963.9604309082031,
      "state_relation_loss": 0.0005996326042804867,
      "step": 8400,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0043169274926185604,
      "weighted_relation_loss": 0.0043169274926185604
    },
    {
      "action_query_norm": 532.1080505371094,
      "action_relation_loss": 0.0036997976480051876,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.867165575304022,
      "flow_matching_loss": 0.04439230561256409,
      "grad_norm": 0.16060538589954376,
      "learning_rate": 8.61870476685327e-05,
      "loss": 0.0487,
      "relation_loss": 0.004296068521216512,
      "selected_pairs": 26431.7,
      "state_query_norm": 969.4242309570312,
      "state_relation_loss": 0.0005962709314189851,
      "step": 8410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004296068521216512,
      "weighted_relation_loss": 0.004296068521216512
    },
    {
      "action_query_norm": 535.8018493652344,
      "action_relation_loss": 0.003642965597100556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.876520112254443,
      "flow_matching_loss": 0.045651298761367795,
      "grad_norm": 0.17124029994010925,
      "learning_rate": 8.614899193005248e-05,
      "loss": 0.0499,
      "relation_loss": 0.004230802739039063,
      "selected_pairs": 27131.2,
      "state_query_norm": 965.3698364257813,
      "state_relation_loss": 0.0005878370313439519,
      "step": 8420,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004230802739039063,
      "weighted_relation_loss": 0.004230802739039063
    },
    {
      "action_query_norm": 533.3310852050781,
      "action_relation_loss": 0.003642136603593826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.885874649204864,
      "flow_matching_loss": 0.044610587134957316,
      "grad_norm": 0.18994228541851044,
      "learning_rate": 8.611089226713842e-05,
      "loss": 0.0489,
      "relation_loss": 0.004244261141866446,
      "selected_pairs": 27084.8,
      "state_query_norm": 972.8459106445313,
      "state_relation_loss": 0.0006021244742441922,
      "step": 8430,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.004244261141866446,
      "weighted_relation_loss": 0.004244261141866446
    },
    {
      "action_query_norm": 532.4483764648437,
      "action_relation_loss": 0.0036897154757753016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8952291861552855,
      "flow_matching_loss": 0.043427969142794606,
      "grad_norm": 0.16291646659374237,
      "learning_rate": 8.607274872608521e-05,
      "loss": 0.0477,
      "relation_loss": 0.004295548633672297,
      "selected_pairs": 27010.7,
      "state_query_norm": 947.5291259765625,
      "state_relation_loss": 0.0006058331404346972,
      "step": 8440,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004295548633672297,
      "weighted_relation_loss": 0.004295548633672297
    },
    {
      "action_query_norm": 528.3003601074219,
      "action_relation_loss": 0.003696316247805953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.904583723105707,
      "flow_matching_loss": 0.043209591135382655,
      "grad_norm": 0.18459519743919373,
      "learning_rate": 8.603456135324088e-05,
      "loss": 0.0475,
      "relation_loss": 0.004320243932306766,
      "selected_pairs": 26807.3,
      "state_query_norm": 961.6986022949219,
      "state_relation_loss": 0.000623927655396983,
      "step": 8450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.004320243932306766,
      "weighted_relation_loss": 0.004320243932306766
    },
    {
      "action_query_norm": 532.8232971191406,
      "action_relation_loss": 0.0038003177382051946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.913938260056128,
      "flow_matching_loss": 0.0442930094897747,
      "grad_norm": 0.10889822989702225,
      "learning_rate": 8.599633019500665e-05,
      "loss": 0.0487,
      "relation_loss": 0.004399413987994194,
      "selected_pairs": 26547.9,
      "state_query_norm": 975.9125244140625,
      "state_relation_loss": 0.0005990962730720639,
      "step": 8460,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.004399413987994194,
      "weighted_relation_loss": 0.004399413987994194
    },
    {
      "action_query_norm": 534.9040771484375,
      "action_relation_loss": 0.003630471182987094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.923292797006548,
      "flow_matching_loss": 0.04573645293712616,
      "grad_norm": 0.15162397921085358,
      "learning_rate": 8.595805529783702e-05,
      "loss": 0.0499,
      "relation_loss": 0.004195081721991301,
      "selected_pairs": 27523.1,
      "state_query_norm": 976.0023254394531,
      "state_relation_loss": 0.0005646105099003762,
      "step": 8470,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.004195081721991301,
      "weighted_relation_loss": 0.004195081721991301
    },
    {
      "action_query_norm": 538.3071228027344,
      "action_relation_loss": 0.0036702460609376432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.932647333956969,
      "flow_matching_loss": 0.047136899828910825,
      "grad_norm": 0.14490555226802826,
      "learning_rate": 8.591973670823958e-05,
      "loss": 0.0514,
      "relation_loss": 0.004265060578472912,
      "selected_pairs": 26875.1,
      "state_query_norm": 962.8697143554688,
      "state_relation_loss": 0.000594814564101398,
      "step": 8480,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004265060578472912,
      "weighted_relation_loss": 0.004265060578472912
    },
    {
      "action_query_norm": 539.9623779296875,
      "action_relation_loss": 0.003798790741711855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.94200187090739,
      "flow_matching_loss": 0.04507466331124306,
      "grad_norm": 0.15236863493919373,
      "learning_rate": 8.588137447277503e-05,
      "loss": 0.0495,
      "relation_loss": 0.004407424479722977,
      "selected_pairs": 27265.4,
      "state_query_norm": 953.7912902832031,
      "state_relation_loss": 0.0006086337612941861,
      "step": 8490,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004407424479722977,
      "weighted_relation_loss": 0.004407424479722977
    },
    {
      "action_query_norm": 541.593017578125,
      "action_relation_loss": 0.0037547912215813994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.951356407857811,
      "flow_matching_loss": 0.04479712396860123,
      "grad_norm": 0.16289517283439636,
      "learning_rate": 8.584296863805709e-05,
      "loss": 0.0492,
      "relation_loss": 0.004358819406479597,
      "selected_pairs": 26437.1,
      "state_query_norm": 961.3785827636718,
      "state_relation_loss": 0.0006040282314643264,
      "step": 8500,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004358819406479597,
      "weighted_relation_loss": 0.004358819406479597
    },
    {
      "action_query_norm": 549.1244384765625,
      "action_relation_loss": 0.0038298537721857427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.960710944808232,
      "flow_matching_loss": 0.04442523494362831,
      "grad_norm": 0.18233013153076172,
      "learning_rate": 8.580451925075248e-05,
      "loss": 0.0489,
      "relation_loss": 0.004435187391936779,
      "selected_pairs": 27197.1,
      "state_query_norm": 966.9839416503906,
      "state_relation_loss": 0.0006053336488548666,
      "step": 8510,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.004435187391936779,
      "weighted_relation_loss": 0.004435187391936779
    },
    {
      "action_query_norm": 539.6074340820312,
      "action_relation_loss": 0.0037015322828665375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.9700654817586525,
      "flow_matching_loss": 0.046307656541466714,
      "grad_norm": 0.16265913844108582,
      "learning_rate": 8.576602635758085e-05,
      "loss": 0.0506,
      "relation_loss": 0.004305143933743238,
      "selected_pairs": 26778.9,
      "state_query_norm": 980.1029602050781,
      "state_relation_loss": 0.0006036117032635957,
      "step": 8520,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004305143933743238,
      "weighted_relation_loss": 0.004305143933743238
    },
    {
      "action_query_norm": 539.3402954101563,
      "action_relation_loss": 0.0037768000038340688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.979420018709074,
      "flow_matching_loss": 0.04468836262822151,
      "grad_norm": 0.1637653410434723,
      "learning_rate": 8.572749000531469e-05,
      "loss": 0.0491,
      "relation_loss": 0.0043682293500751255,
      "selected_pairs": 26967.7,
      "state_query_norm": 974.5851196289062,
      "state_relation_loss": 0.0005914294044487179,
      "step": 8530,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0043682293500751255,
      "weighted_relation_loss": 0.0043682293500751255
    },
    {
      "action_query_norm": 542.9591369628906,
      "action_relation_loss": 0.0038176597328856586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.988774555659495,
      "flow_matching_loss": 0.04393411614000797,
      "grad_norm": 0.19471432268619537,
      "learning_rate": 8.568891024077926e-05,
      "loss": 0.0483,
      "relation_loss": 0.0044093753676861525,
      "selected_pairs": 26921.0,
      "state_query_norm": 982.0920837402343,
      "state_relation_loss": 0.0005917156988289208,
      "step": 8540,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0044093753676861525,
      "weighted_relation_loss": 0.0044093753676861525
    },
    {
      "action_query_norm": 539.4157653808594,
      "action_relation_loss": 0.0036370187764987348,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.998129092609916,
      "flow_matching_loss": 0.04560359790921211,
      "grad_norm": 0.1700173318386078,
      "learning_rate": 8.565028711085265e-05,
      "loss": 0.0498,
      "relation_loss": 0.00422620412427932,
      "selected_pairs": 27362.5,
      "state_query_norm": 962.9297546386719,
      "state_relation_loss": 0.0005891853768844157,
      "step": 8550,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.00422620412427932,
      "weighted_relation_loss": 0.00422620412427932
    },
    {
      "action_query_norm": 536.9965393066407,
      "action_relation_loss": 0.0034806053154170513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.007483629560337,
      "flow_matching_loss": 0.044509631767869,
      "grad_norm": 0.19301845133304596,
      "learning_rate": 8.561162066246561e-05,
      "loss": 0.0486,
      "relation_loss": 0.004103639582172036,
      "selected_pairs": 24460.0,
      "state_query_norm": 954.226513671875,
      "state_relation_loss": 0.0006230342900380492,
      "step": 8560,
      "valid_action_samples": 214.9,
      "weighted_auxiliary_loss": 0.004103639582172036,
      "weighted_relation_loss": 0.004103639582172036
    },
    {
      "action_query_norm": 542.380615234375,
      "action_relation_loss": 0.0038236964959651233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.016838166510757,
      "flow_matching_loss": 0.043947724252939226,
      "grad_norm": 0.18280519545078278,
      "learning_rate": 8.557291094260152e-05,
      "loss": 0.0484,
      "relation_loss": 0.004402282368391752,
      "selected_pairs": 27131.2,
      "state_query_norm": 983.1006530761719,
      "state_relation_loss": 0.0005785859131719917,
      "step": 8570,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004402282368391752,
      "weighted_relation_loss": 0.004402282368391752
    },
    {
      "action_query_norm": 540.9336853027344,
      "action_relation_loss": 0.0037005732767283915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.026192703461179,
      "flow_matching_loss": 0.04312824234366417,
      "grad_norm": 0.1622137427330017,
      "learning_rate": 8.553415799829629e-05,
      "loss": 0.0474,
      "relation_loss": 0.004283533338457346,
      "selected_pairs": 26905.3,
      "state_query_norm": 974.1979248046875,
      "state_relation_loss": 0.0005829600093420594,
      "step": 8580,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.004283533338457346,
      "weighted_relation_loss": 0.004283533338457346
    },
    {
      "action_query_norm": 534.5365295410156,
      "action_relation_loss": 0.0034972863039001824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.0355472404116,
      "flow_matching_loss": 0.04760482423007488,
      "grad_norm": 0.14468695223331451,
      "learning_rate": 8.549536187663847e-05,
      "loss": 0.0517,
      "relation_loss": 0.0041159338550642134,
      "selected_pairs": 26179.7,
      "state_query_norm": 971.8159301757812,
      "state_relation_loss": 0.0006186475045979023,
      "step": 8590,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0041159338550642134,
      "weighted_relation_loss": 0.0041159338550642134
    },
    {
      "action_query_norm": 531.2682312011718,
      "action_relation_loss": 0.003483295044861734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.044901777362021,
      "flow_matching_loss": 0.04466441608965397,
      "grad_norm": 0.11855269223451614,
      "learning_rate": 8.545652262476897e-05,
      "loss": 0.0488,
      "relation_loss": 0.004105669585987926,
      "selected_pairs": 27154.8,
      "state_query_norm": 973.3913879394531,
      "state_relation_loss": 0.0006223745294846594,
      "step": 8600,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.004105669585987926,
      "weighted_relation_loss": 0.004105669585987926
    },
    {
      "action_query_norm": 533.5787109375,
      "action_relation_loss": 0.0034807235933840275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.054256314312441,
      "flow_matching_loss": 0.04460561238229275,
      "grad_norm": 0.11504379659891129,
      "learning_rate": 8.541764028988115e-05,
      "loss": 0.0487,
      "relation_loss": 0.004063369520008564,
      "selected_pairs": 27336.8,
      "state_query_norm": 973.7261291503906,
      "state_relation_loss": 0.0005826459149830043,
      "step": 8610,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.004063369520008564,
      "weighted_relation_loss": 0.004063369520008564
    },
    {
      "action_query_norm": 529.0368408203125,
      "action_relation_loss": 0.003761655976995826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.063610851262862,
      "flow_matching_loss": 0.042204834148287775,
      "grad_norm": 0.14466692507266998,
      "learning_rate": 8.537871491922071e-05,
      "loss": 0.0466,
      "relation_loss": 0.004346324154175818,
      "selected_pairs": 27225.2,
      "state_query_norm": 968.3954528808594,
      "state_relation_loss": 0.0005846681597176939,
      "step": 8620,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.004346324154175818,
      "weighted_relation_loss": 0.004346324154175818
    },
    {
      "action_query_norm": 534.2260681152344,
      "action_relation_loss": 0.0035139696206897496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.072965388213284,
      "flow_matching_loss": 0.04521840251982212,
      "grad_norm": 0.15069249272346497,
      "learning_rate": 8.533974656008566e-05,
      "loss": 0.0493,
      "relation_loss": 0.004096649098210036,
      "selected_pairs": 27011.4,
      "state_query_norm": 952.151318359375,
      "state_relation_loss": 0.0005826795182656497,
      "step": 8630,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004096649098210036,
      "weighted_relation_loss": 0.004096649098210036
    },
    {
      "action_query_norm": 531.9232604980468,
      "action_relation_loss": 0.0034408337669447064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.082319925163704,
      "flow_matching_loss": 0.04609651044011116,
      "grad_norm": 0.15417221188545227,
      "learning_rate": 8.530073525982621e-05,
      "loss": 0.0501,
      "relation_loss": 0.004022900178097188,
      "selected_pairs": 26873.3,
      "state_query_norm": 974.6927917480468,
      "state_relation_loss": 0.0005820663704071194,
      "step": 8640,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004022900178097188,
      "weighted_relation_loss": 0.004022900178097188
    },
    {
      "action_query_norm": 531.3268005371094,
      "action_relation_loss": 0.0033398889470845463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.091674462114126,
      "flow_matching_loss": 0.043350831046700476,
      "grad_norm": 0.12598194181919098,
      "learning_rate": 8.526168106584476e-05,
      "loss": 0.0473,
      "relation_loss": 0.003932884894311428,
      "selected_pairs": 26410.9,
      "state_query_norm": 974.1847106933594,
      "state_relation_loss": 0.0005929959646891802,
      "step": 8650,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.003932884894311428,
      "weighted_relation_loss": 0.003932884894311428
    },
    {
      "action_query_norm": 532.8646850585938,
      "action_relation_loss": 0.0035848803585395218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.101028999064546,
      "flow_matching_loss": 0.04482566229999065,
      "grad_norm": 0.17875170707702637,
      "learning_rate": 8.522258402559586e-05,
      "loss": 0.049,
      "relation_loss": 0.004195763659663498,
      "selected_pairs": 26229.1,
      "state_query_norm": 981.6729431152344,
      "state_relation_loss": 0.0006108833535108715,
      "step": 8660,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.004195763659663498,
      "weighted_relation_loss": 0.004195763659663498
    },
    {
      "action_query_norm": 535.9295959472656,
      "action_relation_loss": 0.0033550110179930925,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.110383536014968,
      "flow_matching_loss": 0.043320132046937944,
      "grad_norm": 0.12954629957675934,
      "learning_rate": 8.518344418658613e-05,
      "loss": 0.0473,
      "relation_loss": 0.003981084935367108,
      "selected_pairs": 26738.3,
      "state_query_norm": 976.4923645019531,
      "state_relation_loss": 0.0006260739057324827,
      "step": 8670,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003981084935367108,
      "weighted_relation_loss": 0.003981084935367108
    },
    {
      "action_query_norm": 532.7612365722656,
      "action_relation_loss": 0.003429344017058611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.119738072965388,
      "flow_matching_loss": 0.04212569072842598,
      "grad_norm": 0.1313677430152893,
      "learning_rate": 8.51442615963741e-05,
      "loss": 0.0461,
      "relation_loss": 0.004024054203182459,
      "selected_pairs": 27067.9,
      "state_query_norm": 974.1256774902344,
      "state_relation_loss": 0.0005947101861238479,
      "step": 8680,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004024054203182459,
      "weighted_relation_loss": 0.004024054203182459
    },
    {
      "action_query_norm": 533.1077026367187,
      "action_relation_loss": 0.003331877081654966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.129092609915809,
      "flow_matching_loss": 0.04206236749887467,
      "grad_norm": 0.12035423517227173,
      "learning_rate": 8.510503630257034e-05,
      "loss": 0.046,
      "relation_loss": 0.003947244165465236,
      "selected_pairs": 27263.8,
      "state_query_norm": 980.670166015625,
      "state_relation_loss": 0.0006153670779895038,
      "step": 8690,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003947244165465236,
      "weighted_relation_loss": 0.003947244165465236
    },
    {
      "action_query_norm": 539.5626708984375,
      "action_relation_loss": 0.003577204095199704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.13844714686623,
      "flow_matching_loss": 0.0436857283115387,
      "grad_norm": 0.10184080898761749,
      "learning_rate": 8.50657683528373e-05,
      "loss": 0.0478,
      "relation_loss": 0.004155518161132932,
      "selected_pairs": 27174.1,
      "state_query_norm": 971.0243835449219,
      "state_relation_loss": 0.0005783140484709292,
      "step": 8700,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.004155518161132932,
      "weighted_relation_loss": 0.004155518161132932
    },
    {
      "action_query_norm": 546.9137023925781,
      "action_relation_loss": 0.0034637321485206487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.14780168381665,
      "flow_matching_loss": 0.04513604156672955,
      "grad_norm": 0.11421088129281998,
      "learning_rate": 8.502645779488922e-05,
      "loss": 0.0492,
      "relation_loss": 0.004041435499675572,
      "selected_pairs": 27053.9,
      "state_query_norm": 977.7712219238281,
      "state_relation_loss": 0.0005777033162303269,
      "step": 8710,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004041435499675572,
      "weighted_relation_loss": 0.004041435499675572
    },
    {
      "action_query_norm": 535.1461181640625,
      "action_relation_loss": 0.0036725018871948123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.157156220767073,
      "flow_matching_loss": 0.04373280815780163,
      "grad_norm": 0.12219451367855072,
      "learning_rate": 8.498710467649213e-05,
      "loss": 0.048,
      "relation_loss": 0.004267009161412716,
      "selected_pairs": 26684.5,
      "state_query_norm": 977.7541870117187,
      "state_relation_loss": 0.0005945072567556054,
      "step": 8720,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004267009161412716,
      "weighted_relation_loss": 0.004267009161412716
    },
    {
      "action_query_norm": 539.7436401367188,
      "action_relation_loss": 0.0035074243089184163,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.166510757717493,
      "flow_matching_loss": 0.04296662323176861,
      "grad_norm": 0.1615505963563919,
      "learning_rate": 8.494770904546381e-05,
      "loss": 0.0471,
      "relation_loss": 0.0040969971101731065,
      "selected_pairs": 27179.5,
      "state_query_norm": 980.5438415527344,
      "state_relation_loss": 0.0005895727779716253,
      "step": 8730,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0040969971101731065,
      "weighted_relation_loss": 0.0040969971101731065
    },
    {
      "action_query_norm": 542.8858032226562,
      "action_relation_loss": 0.0034573789685964583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.175865294667913,
      "flow_matching_loss": 0.04419735968112946,
      "grad_norm": 0.1273488849401474,
      "learning_rate": 8.490827094967363e-05,
      "loss": 0.0482,
      "relation_loss": 0.004032950336113572,
      "selected_pairs": 26640.7,
      "state_query_norm": 966.8453979492188,
      "state_relation_loss": 0.0005755713500548154,
      "step": 8740,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.004032950336113572,
      "weighted_relation_loss": 0.004032950336113572
    },
    {
      "action_query_norm": 555.8037048339844,
      "action_relation_loss": 0.0035361475544050335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.185219831618335,
      "flow_matching_loss": 0.043201635405421254,
      "grad_norm": 0.15912599861621857,
      "learning_rate": 8.486879043704263e-05,
      "loss": 0.0473,
      "relation_loss": 0.004118938976898789,
      "selected_pairs": 26914.1,
      "state_query_norm": 983.3524536132812,
      "state_relation_loss": 0.0005827913817483931,
      "step": 8750,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004118938976898789,
      "weighted_relation_loss": 0.004118938976898789
    },
    {
      "action_query_norm": 539.7726745605469,
      "action_relation_loss": 0.003521754057146609,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.194574368568755,
      "flow_matching_loss": 0.043412952497601506,
      "grad_norm": 0.14124032855033875,
      "learning_rate": 8.482926755554332e-05,
      "loss": 0.0475,
      "relation_loss": 0.0040986193111166355,
      "selected_pairs": 26615.2,
      "state_query_norm": 978.2705139160156,
      "state_relation_loss": 0.0005768652423284948,
      "step": 8760,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0040986193111166355,
      "weighted_relation_loss": 0.0040986193111166355
    },
    {
      "action_query_norm": 526.9055908203125,
      "action_relation_loss": 0.0035393604077398775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.203928905519177,
      "flow_matching_loss": 0.04507113173604012,
      "grad_norm": 0.1443656086921692,
      "learning_rate": 8.478970235319975e-05,
      "loss": 0.0492,
      "relation_loss": 0.004091032547876239,
      "selected_pairs": 26764.3,
      "state_query_norm": 971.6598754882813,
      "state_relation_loss": 0.0005516721750609577,
      "step": 8770,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004091032547876239,
      "weighted_relation_loss": 0.004091032547876239
    },
    {
      "action_query_norm": 541.2082885742187,
      "action_relation_loss": 0.0036921762395650147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.213283442469598,
      "flow_matching_loss": 0.043031053617596626,
      "grad_norm": 0.12329087406396866,
      "learning_rate": 8.475009487808737e-05,
      "loss": 0.0473,
      "relation_loss": 0.004275073576718569,
      "selected_pairs": 26619.6,
      "state_query_norm": 991.4114501953125,
      "state_relation_loss": 0.00058289734297432,
      "step": 8780,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.004275073576718569,
      "weighted_relation_loss": 0.004275073576718569
    },
    {
      "action_query_norm": 543.347900390625,
      "action_relation_loss": 0.0035590824438259004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.22263797942002,
      "flow_matching_loss": 0.04377204142510891,
      "grad_norm": 0.11785334348678589,
      "learning_rate": 8.471044517833298e-05,
      "loss": 0.0479,
      "relation_loss": 0.004169820575043559,
      "selected_pairs": 27043.0,
      "state_query_norm": 990.4769104003906,
      "state_relation_loss": 0.0006107381545007229,
      "step": 8790,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004169820575043559,
      "weighted_relation_loss": 0.004169820575043559
    },
    {
      "action_query_norm": 541.713525390625,
      "action_relation_loss": 0.0035356319276615977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.23199251637044,
      "flow_matching_loss": 0.04071085005998611,
      "grad_norm": 0.12471303343772888,
      "learning_rate": 8.467075330211474e-05,
      "loss": 0.0449,
      "relation_loss": 0.004151571076363325,
      "selected_pairs": 26878.7,
      "state_query_norm": 965.4760620117188,
      "state_relation_loss": 0.0006159391195978969,
      "step": 8800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004151571076363325,
      "weighted_relation_loss": 0.004151571076363325
    },
    {
      "action_query_norm": 540.8876342773438,
      "action_relation_loss": 0.003477901080623269,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.24134705332086,
      "flow_matching_loss": 0.043120278418064116,
      "grad_norm": 0.09984950721263885,
      "learning_rate": 8.463101929766198e-05,
      "loss": 0.0472,
      "relation_loss": 0.004065473820082844,
      "selected_pairs": 27102.2,
      "state_query_norm": 989.280615234375,
      "state_relation_loss": 0.0005875727394595743,
      "step": 8810,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.004065473820082844,
      "weighted_relation_loss": 0.004065473820082844
    },
    {
      "action_query_norm": 537.4061340332031,
      "action_relation_loss": 0.0034473537001758815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.250701590271282,
      "flow_matching_loss": 0.04276677593588829,
      "grad_norm": 0.12686200439929962,
      "learning_rate": 8.459124321325528e-05,
      "loss": 0.0468,
      "relation_loss": 0.004046070436015725,
      "selected_pairs": 26433.8,
      "state_query_norm": 977.5634948730469,
      "state_relation_loss": 0.000598716747481376,
      "step": 8820,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004046070436015725,
      "weighted_relation_loss": 0.004046070436015725
    },
    {
      "action_query_norm": 542.8905212402344,
      "action_relation_loss": 0.0032562628155574204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.260056127221702,
      "flow_matching_loss": 0.042055669054389,
      "grad_norm": 0.1241050735116005,
      "learning_rate": 8.455142509722634e-05,
      "loss": 0.0459,
      "relation_loss": 0.003837440605275333,
      "selected_pairs": 26970.5,
      "state_query_norm": 975.8557434082031,
      "state_relation_loss": 0.0005811777606140822,
      "step": 8830,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003837440605275333,
      "weighted_relation_loss": 0.003837440605275333
    },
    {
      "action_query_norm": 547.812890625,
      "action_relation_loss": 0.0032996461726725102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.269410664172124,
      "flow_matching_loss": 0.04313613101840019,
      "grad_norm": 0.13858556747436523,
      "learning_rate": 8.451156499795792e-05,
      "loss": 0.047,
      "relation_loss": 0.00388825333211571,
      "selected_pairs": 26829.3,
      "state_query_norm": 988.4313049316406,
      "state_relation_loss": 0.0005886071652639657,
      "step": 8840,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.00388825333211571,
      "weighted_relation_loss": 0.00388825333211571
    },
    {
      "action_query_norm": 545.07724609375,
      "action_relation_loss": 0.0032714282628148795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.278765201122544,
      "flow_matching_loss": 0.04246821478009224,
      "grad_norm": 0.1526123583316803,
      "learning_rate": 8.447166296388381e-05,
      "loss": 0.0463,
      "relation_loss": 0.0038250223733484745,
      "selected_pairs": 26342.9,
      "state_query_norm": 992.9361511230469,
      "state_relation_loss": 0.000553594104712829,
      "step": 8850,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0038250223733484745,
      "weighted_relation_loss": 0.0038250223733484745
    },
    {
      "action_query_norm": 543.6814086914062,
      "action_relation_loss": 0.0035588672384619714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.288119738072965,
      "flow_matching_loss": 0.04537310190498829,
      "grad_norm": 0.12339430302381516,
      "learning_rate": 8.443171904348872e-05,
      "loss": 0.0495,
      "relation_loss": 0.004144959058612585,
      "selected_pairs": 27356.5,
      "state_query_norm": 977.2559936523437,
      "state_relation_loss": 0.0005860918085090816,
      "step": 8860,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.004144959058612585,
      "weighted_relation_loss": 0.004144959058612585
    },
    {
      "action_query_norm": 547.97802734375,
      "action_relation_loss": 0.0034005559980869295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.297474275023387,
      "flow_matching_loss": 0.04379488490521908,
      "grad_norm": 0.1508467197418213,
      "learning_rate": 8.43917332853083e-05,
      "loss": 0.0478,
      "relation_loss": 0.003994097211398184,
      "selected_pairs": 27200.2,
      "state_query_norm": 973.3297241210937,
      "state_relation_loss": 0.0005935412191320211,
      "step": 8870,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.003994097211398184,
      "weighted_relation_loss": 0.003994097211398184
    },
    {
      "action_query_norm": 553.5084533691406,
      "action_relation_loss": 0.0033213001443073153,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.306828811973807,
      "flow_matching_loss": 0.04104297086596489,
      "grad_norm": 0.13772685825824738,
      "learning_rate": 8.435170573792901e-05,
      "loss": 0.0449,
      "relation_loss": 0.0039020106894895436,
      "selected_pairs": 26805.0,
      "state_query_norm": 980.0825073242188,
      "state_relation_loss": 0.0005807105335406959,
      "step": 8880,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0039020106894895436,
      "weighted_relation_loss": 0.0039020106894895436
    },
    {
      "action_query_norm": 551.1997009277344,
      "action_relation_loss": 0.003454527840949595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.316183348924229,
      "flow_matching_loss": 0.04413051977753639,
      "grad_norm": 0.14920999109745026,
      "learning_rate": 8.431163644998808e-05,
      "loss": 0.0482,
      "relation_loss": 0.004019706160761416,
      "selected_pairs": 27124.4,
      "state_query_norm": 986.53720703125,
      "state_relation_loss": 0.0005651782790664584,
      "step": 8890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004019706160761416,
      "weighted_relation_loss": 0.004019706160761416
    },
    {
      "action_query_norm": 550.3625427246094,
      "action_relation_loss": 0.0034495755564421414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.325537885874649,
      "flow_matching_loss": 0.043306955695152284,
      "grad_norm": 0.1121906265616417,
      "learning_rate": 8.42715254701735e-05,
      "loss": 0.0473,
      "relation_loss": 0.004039718699641526,
      "selected_pairs": 26869.0,
      "state_query_norm": 973.3388549804688,
      "state_relation_loss": 0.0005901431548409164,
      "step": 8900,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004039718699641526,
      "weighted_relation_loss": 0.004039718699641526
    },
    {
      "action_query_norm": 550.0984680175782,
      "action_relation_loss": 0.003239515866152942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.334892422825071,
      "flow_matching_loss": 0.04359620288014412,
      "grad_norm": 0.14222244918346405,
      "learning_rate": 8.42313728472239e-05,
      "loss": 0.0474,
      "relation_loss": 0.0038197403540834785,
      "selected_pairs": 27316.6,
      "state_query_norm": 969.5438293457031,
      "state_relation_loss": 0.0005802244937513024,
      "step": 8910,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0038197403540834785,
      "weighted_relation_loss": 0.0038197403540834785
    },
    {
      "action_query_norm": 542.8059387207031,
      "action_relation_loss": 0.00349065528716892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.344246959775491,
      "flow_matching_loss": 0.04289878681302071,
      "grad_norm": 0.13873931765556335,
      "learning_rate": 8.419117862992844e-05,
      "loss": 0.047,
      "relation_loss": 0.004062302038073539,
      "selected_pairs": 26898.3,
      "state_query_norm": 975.3366943359375,
      "state_relation_loss": 0.0005716467159800232,
      "step": 8920,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.004062302038073539,
      "weighted_relation_loss": 0.004062302038073539
    },
    {
      "action_query_norm": 552.3102722167969,
      "action_relation_loss": 0.0037581825628876684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.353601496725911,
      "flow_matching_loss": 0.04272651895880699,
      "grad_norm": 0.12712454795837402,
      "learning_rate": 8.415094286712694e-05,
      "loss": 0.047,
      "relation_loss": 0.004319615848362446,
      "selected_pairs": 26485.4,
      "state_query_norm": 989.3156616210938,
      "state_relation_loss": 0.0005614333844278008,
      "step": 8930,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.004319615848362446,
      "weighted_relation_loss": 0.004319615848362446
    },
    {
      "action_query_norm": 546.3638854980469,
      "action_relation_loss": 0.003467898303642869,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.362956033676333,
      "flow_matching_loss": 0.04225433543324471,
      "grad_norm": 0.1593431830406189,
      "learning_rate": 8.411066560770965e-05,
      "loss": 0.0463,
      "relation_loss": 0.004014161927625537,
      "selected_pairs": 26767.9,
      "state_query_norm": 983.0210266113281,
      "state_relation_loss": 0.0005462636472657323,
      "step": 8940,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004014161927625537,
      "weighted_relation_loss": 0.004014161927625537
    },
    {
      "action_query_norm": 541.7685546875,
      "action_relation_loss": 0.0033526191487908362,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.372310570626754,
      "flow_matching_loss": 0.04212495759129524,
      "grad_norm": 0.1409839689731598,
      "learning_rate": 8.407034690061721e-05,
      "loss": 0.0461,
      "relation_loss": 0.0039493443444371225,
      "selected_pairs": 26492.7,
      "state_query_norm": 980.96201171875,
      "state_relation_loss": 0.0005967252305708826,
      "step": 8950,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0039493443444371225,
      "weighted_relation_loss": 0.0039493443444371225
    },
    {
      "action_query_norm": 543.5930847167969,
      "action_relation_loss": 0.003475526091642678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.381665107577176,
      "flow_matching_loss": 0.042547698691487314,
      "grad_norm": 0.11868655681610107,
      "learning_rate": 8.402998679484067e-05,
      "loss": 0.0466,
      "relation_loss": 0.004046584758907556,
      "selected_pairs": 27612.4,
      "state_query_norm": 981.1063842773438,
      "state_relation_loss": 0.0005710586090572178,
      "step": 8960,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.004046584758907556,
      "weighted_relation_loss": 0.004046584758907556
    },
    {
      "action_query_norm": 545.9357055664062,
      "action_relation_loss": 0.003479123837314546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.391019644527596,
      "flow_matching_loss": 0.04238871335983276,
      "grad_norm": 0.12446427345275879,
      "learning_rate": 8.398958533942135e-05,
      "loss": 0.0464,
      "relation_loss": 0.0040568772470578555,
      "selected_pairs": 26857.1,
      "state_query_norm": 998.9783874511719,
      "state_relation_loss": 0.0005777534388471395,
      "step": 8970,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0040568772470578555,
      "weighted_relation_loss": 0.0040568772470578555
    },
    {
      "action_query_norm": 551.2203002929688,
      "action_relation_loss": 0.0033472083043307064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.400374181478018,
      "flow_matching_loss": 0.04042565077543259,
      "grad_norm": 0.10477223992347717,
      "learning_rate": 8.394914258345083e-05,
      "loss": 0.0443,
      "relation_loss": 0.0039047983940690757,
      "selected_pairs": 26212.4,
      "state_query_norm": 1011.211865234375,
      "state_relation_loss": 0.0005575900839176029,
      "step": 8980,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0039047983940690757,
      "weighted_relation_loss": 0.0039047983940690757
    },
    {
      "action_query_norm": 553.8463317871094,
      "action_relation_loss": 0.003572805295698345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.409728718428438,
      "flow_matching_loss": 0.041887576878070834,
      "grad_norm": 0.11442601680755615,
      "learning_rate": 8.390865857607089e-05,
      "loss": 0.046,
      "relation_loss": 0.004121716273948551,
      "selected_pairs": 26691.9,
      "state_query_norm": 1000.0075988769531,
      "state_relation_loss": 0.0005489109840709716,
      "step": 8990,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004121716273948551,
      "weighted_relation_loss": 0.004121716273948551
    },
    {
      "action_query_norm": 545.6077026367187,
      "action_relation_loss": 0.0034978134324774148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.419083255378858,
      "flow_matching_loss": 0.04393541775643826,
      "grad_norm": 0.11992964148521423,
      "learning_rate": 8.386813336647339e-05,
      "loss": 0.048,
      "relation_loss": 0.0041119955014437435,
      "selected_pairs": 26603.5,
      "state_query_norm": 987.8634887695313,
      "state_relation_loss": 0.0006141820922493935,
      "step": 9000,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0041119955014437435,
      "weighted_relation_loss": 0.0041119955014437435
    },
    {
      "action_query_norm": 549.0711669921875,
      "action_relation_loss": 0.0033879751805216073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.42843779232928,
      "flow_matching_loss": 0.042435580492019655,
      "grad_norm": 0.12338505685329437,
      "learning_rate": 8.38275670039003e-05,
      "loss": 0.0464,
      "relation_loss": 0.003973628208041191,
      "selected_pairs": 26645.4,
      "state_query_norm": 990.243505859375,
      "state_relation_loss": 0.0005856530275195837,
      "step": 9010,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.003973628208041191,
      "weighted_relation_loss": 0.003973628208041191
    },
    {
      "action_query_norm": 551.2844116210938,
      "action_relation_loss": 0.003393764444626868,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.4377923292797,
      "flow_matching_loss": 0.04240964688360691,
      "grad_norm": 0.11610587686300278,
      "learning_rate": 8.378695953764356e-05,
      "loss": 0.0464,
      "relation_loss": 0.003971210541203618,
      "selected_pairs": 26778.5,
      "state_query_norm": 982.9735595703125,
      "state_relation_loss": 0.0005774461023975164,
      "step": 9020,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003971210541203618,
      "weighted_relation_loss": 0.003971210541203618
    },
    {
      "action_query_norm": 553.6136840820312,
      "action_relation_loss": 0.003480224939994514,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.447146866230122,
      "flow_matching_loss": 0.04039696492254734,
      "grad_norm": 0.1377713978290558,
      "learning_rate": 8.374631101704509e-05,
      "loss": 0.0444,
      "relation_loss": 0.004051840538159013,
      "selected_pairs": 26071.9,
      "state_query_norm": 983.7896362304688,
      "state_relation_loss": 0.0005716156680136919,
      "step": 9030,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.004051840538159013,
      "weighted_relation_loss": 0.004051840538159013
    },
    {
      "action_query_norm": 548.9992553710938,
      "action_relation_loss": 0.003367119934409857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.456501403180543,
      "flow_matching_loss": 0.0415692899376154,
      "grad_norm": 0.16583213210105896,
      "learning_rate": 8.370562149149665e-05,
      "loss": 0.0455,
      "relation_loss": 0.003917260398156941,
      "selected_pairs": 26664.1,
      "state_query_norm": 985.7527954101563,
      "state_relation_loss": 0.000550140521954745,
      "step": 9040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003917260398156941,
      "weighted_relation_loss": 0.003917260398156941
    },
    {
      "action_query_norm": 548.7662414550781,
      "action_relation_loss": 0.003424159367568791,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.465855940130963,
      "flow_matching_loss": 0.04013263359665871,
      "grad_norm": 0.11410997062921524,
      "learning_rate": 8.366489101043989e-05,
      "loss": 0.0441,
      "relation_loss": 0.003970858105458319,
      "selected_pairs": 26919.4,
      "state_query_norm": 992.7849609375,
      "state_relation_loss": 0.0005466987553518265,
      "step": 9050,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.003970858105458319,
      "weighted_relation_loss": 0.003970858105458319
    },
    {
      "action_query_norm": 554.8182189941406,
      "action_relation_loss": 0.003445682115852833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.475210477081385,
      "flow_matching_loss": 0.04062117114663124,
      "grad_norm": 0.14752569794654846,
      "learning_rate": 8.362411962336614e-05,
      "loss": 0.0446,
      "relation_loss": 0.004023149865679443,
      "selected_pairs": 26577.9,
      "state_query_norm": 979.0880004882813,
      "state_relation_loss": 0.000577467744005844,
      "step": 9060,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004023149865679443,
      "weighted_relation_loss": 0.004023149865679443
    },
    {
      "action_query_norm": 534.1363159179688,
      "action_relation_loss": 0.003579130582511425,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.484565014031805,
      "flow_matching_loss": 0.045608242601156236,
      "grad_norm": 0.13639400899410248,
      "learning_rate": 8.358330737981651e-05,
      "loss": 0.0498,
      "relation_loss": 0.004144525807350874,
      "selected_pairs": 26577.8,
      "state_query_norm": 984.6842529296875,
      "state_relation_loss": 0.0005653952714055777,
      "step": 9070,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004144525807350874,
      "weighted_relation_loss": 0.004144525807350874
    },
    {
      "action_query_norm": 533.5847106933594,
      "action_relation_loss": 0.003586481744423509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.493919550982227,
      "flow_matching_loss": 0.04431584551930427,
      "grad_norm": 0.15665660798549652,
      "learning_rate": 8.35424543293817e-05,
      "loss": 0.0484,
      "relation_loss": 0.004134096344932914,
      "selected_pairs": 27126.9,
      "state_query_norm": 981.95634765625,
      "state_relation_loss": 0.0005476146296132355,
      "step": 9080,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004134096344932914,
      "weighted_relation_loss": 0.004134096344932914
    },
    {
      "action_query_norm": 546.3000854492187,
      "action_relation_loss": 0.0033927357755601404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.503274087932647,
      "flow_matching_loss": 0.042825596407055855,
      "grad_norm": 0.1564301699399948,
      "learning_rate": 8.350156052170206e-05,
      "loss": 0.0468,
      "relation_loss": 0.0039680462796241045,
      "selected_pairs": 26829.3,
      "state_query_norm": 991.5662658691406,
      "state_relation_loss": 0.0005753104691393673,
      "step": 9090,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0039680462796241045,
      "weighted_relation_loss": 0.0039680462796241045
    },
    {
      "action_query_norm": 548.0999938964844,
      "action_relation_loss": 0.0035550010157749056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.512628624883067,
      "flow_matching_loss": 0.04340166039764881,
      "grad_norm": 0.13688813149929047,
      "learning_rate": 8.346062600646738e-05,
      "loss": 0.0476,
      "relation_loss": 0.004159191320650279,
      "selected_pairs": 27296.8,
      "state_query_norm": 990.9430358886718,
      "state_relation_loss": 0.0006041903689038008,
      "step": 9100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.004159191320650279,
      "weighted_relation_loss": 0.004159191320650279
    },
    {
      "action_query_norm": 546.4554565429687,
      "action_relation_loss": 0.0032975059002637863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.52198316183349,
      "flow_matching_loss": 0.04638502448797226,
      "grad_norm": 0.13071417808532715,
      "learning_rate": 8.341965083341697e-05,
      "loss": 0.0503,
      "relation_loss": 0.0038910461589694023,
      "selected_pairs": 26893.0,
      "state_query_norm": 999.1722045898438,
      "state_relation_loss": 0.000593540258705616,
      "step": 9110,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0038910461589694023,
      "weighted_relation_loss": 0.0038910461589694023
    },
    {
      "action_query_norm": 553.163427734375,
      "action_relation_loss": 0.0033917462453246115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.53133769878391,
      "flow_matching_loss": 0.0414035726338625,
      "grad_norm": 0.14600512385368347,
      "learning_rate": 8.337863505233953e-05,
      "loss": 0.0454,
      "relation_loss": 0.003956585749983787,
      "selected_pairs": 26481.3,
      "state_query_norm": 996.8170837402344,
      "state_relation_loss": 0.000564839510479942,
      "step": 9120,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.003956585749983787,
      "weighted_relation_loss": 0.003956585749983787
    },
    {
      "action_query_norm": 553.1727600097656,
      "action_relation_loss": 0.003432031092233956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.540692235734332,
      "flow_matching_loss": 0.042651527747511866,
      "grad_norm": 0.12396778166294098,
      "learning_rate": 8.33375787130731e-05,
      "loss": 0.0466,
      "relation_loss": 0.003986458340659737,
      "selected_pairs": 26686.3,
      "state_query_norm": 993.9146423339844,
      "state_relation_loss": 0.0005544272833503782,
      "step": 9130,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.003986458340659737,
      "weighted_relation_loss": 0.003986458340659737
    },
    {
      "action_query_norm": 554.9112060546875,
      "action_relation_loss": 0.003435248089954257,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.550046772684752,
      "flow_matching_loss": 0.04318644478917122,
      "grad_norm": 0.16017447412014008,
      "learning_rate": 8.3296481865505e-05,
      "loss": 0.0472,
      "relation_loss": 0.003994910465553403,
      "selected_pairs": 26624.9,
      "state_query_norm": 985.0505859375,
      "state_relation_loss": 0.0005596623668679968,
      "step": 9140,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.003994910465553403,
      "weighted_relation_loss": 0.003994910465553403
    },
    {
      "action_query_norm": 549.9399169921875,
      "action_relation_loss": 0.0032231941586360335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.559401309635174,
      "flow_matching_loss": 0.041334738954901695,
      "grad_norm": 0.13055108487606049,
      "learning_rate": 8.325534455957181e-05,
      "loss": 0.0451,
      "relation_loss": 0.003792272228747606,
      "selected_pairs": 27106.1,
      "state_query_norm": 998.0852783203125,
      "state_relation_loss": 0.0005690780759323389,
      "step": 9150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.003792272228747606,
      "weighted_relation_loss": 0.003792272228747606
    },
    {
      "action_query_norm": 552.2555847167969,
      "action_relation_loss": 0.003356810240074992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.568755846585594,
      "flow_matching_loss": 0.04229737669229507,
      "grad_norm": 0.11896827816963196,
      "learning_rate": 8.321416684525917e-05,
      "loss": 0.0462,
      "relation_loss": 0.003912857081741094,
      "selected_pairs": 27087.5,
      "state_query_norm": 998.772509765625,
      "state_relation_loss": 0.0005560468649491668,
      "step": 9160,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.003912857081741094,
      "weighted_relation_loss": 0.003912857081741094
    },
    {
      "action_query_norm": 554.2039489746094,
      "action_relation_loss": 0.0033211656846106052,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.578110383536014,
      "flow_matching_loss": 0.043978850543498996,
      "grad_norm": 0.13550321757793427,
      "learning_rate": 8.317294877260193e-05,
      "loss": 0.0479,
      "relation_loss": 0.00390221003908664,
      "selected_pairs": 27199.0,
      "state_query_norm": 997.3952514648438,
      "state_relation_loss": 0.000581044371938333,
      "step": 9170,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00390221003908664,
      "weighted_relation_loss": 0.00390221003908664
    },
    {
      "action_query_norm": 556.0887634277344,
      "action_relation_loss": 0.0033908644458279015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.587464920486436,
      "flow_matching_loss": 0.03952331468462944,
      "grad_norm": 0.14029563963413239,
      "learning_rate": 8.313169039168394e-05,
      "loss": 0.0435,
      "relation_loss": 0.003949770447798073,
      "selected_pairs": 27482.4,
      "state_query_norm": 995.575244140625,
      "state_relation_loss": 0.0005589060310740024,
      "step": 9180,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.003949770447798073,
      "weighted_relation_loss": 0.003949770447798073
    },
    {
      "action_query_norm": 556.3617980957031,
      "action_relation_loss": 0.003370857797563076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.596819457436856,
      "flow_matching_loss": 0.043965934216976164,
      "grad_norm": 0.17758870124816895,
      "learning_rate": 8.3090391752638e-05,
      "loss": 0.0479,
      "relation_loss": 0.003941537020727992,
      "selected_pairs": 26741.5,
      "state_query_norm": 990.2359741210937,
      "state_relation_loss": 0.0005706792289856821,
      "step": 9190,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003941537020727992,
      "weighted_relation_loss": 0.003941537020727992
    },
    {
      "action_query_norm": 549.0210571289062,
      "action_relation_loss": 0.0032498379703611137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.606173994387278,
      "flow_matching_loss": 0.041313720121979716,
      "grad_norm": 0.13905392587184906,
      "learning_rate": 8.304905290564586e-05,
      "loss": 0.0451,
      "relation_loss": 0.0038342037703841926,
      "selected_pairs": 27266.2,
      "state_query_norm": 994.6820922851563,
      "state_relation_loss": 0.000584365805843845,
      "step": 9200,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0038342037703841926,
      "weighted_relation_loss": 0.0038342037703841926
    },
    {
      "action_query_norm": 549.5510925292969,
      "action_relation_loss": 0.0033473466290161015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.615528531337699,
      "flow_matching_loss": 0.04107506945729256,
      "grad_norm": 0.19396881759166718,
      "learning_rate": 8.300767390093813e-05,
      "loss": 0.045,
      "relation_loss": 0.003932237834669649,
      "selected_pairs": 27952.5,
      "state_query_norm": 997.5002502441406,
      "state_relation_loss": 0.0005848911765497178,
      "step": 9210,
      "valid_action_samples": 236.9,
      "weighted_auxiliary_loss": 0.003932237834669649,
      "weighted_relation_loss": 0.003932237834669649
    },
    {
      "action_query_norm": 550.4530456542968,
      "action_relation_loss": 0.0033421184634789824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.62488306828812,
      "flow_matching_loss": 0.04187066555023193,
      "grad_norm": 0.13805784285068512,
      "learning_rate": 8.296625478879417e-05,
      "loss": 0.0458,
      "relation_loss": 0.003949295450001955,
      "selected_pairs": 26711.9,
      "state_query_norm": 987.1495300292969,
      "state_relation_loss": 0.0006071769748814404,
      "step": 9220,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.003949295450001955,
      "weighted_relation_loss": 0.003949295450001955
    },
    {
      "action_query_norm": 556.0804321289063,
      "action_relation_loss": 0.003374145017005503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.63423760523854,
      "flow_matching_loss": 0.04122893027961254,
      "grad_norm": 0.12987491488456726,
      "learning_rate": 8.292479561954214e-05,
      "loss": 0.0452,
      "relation_loss": 0.003965492150746286,
      "selected_pairs": 27202.2,
      "state_query_norm": 1002.7022216796875,
      "state_relation_loss": 0.0005913471395615488,
      "step": 9230,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.003965492150746286,
      "weighted_relation_loss": 0.003965492150746286
    },
    {
      "action_query_norm": 555.3959777832031,
      "action_relation_loss": 0.003455666499212384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.643592142188961,
      "flow_matching_loss": 0.041277439892292024,
      "grad_norm": 0.10337214171886444,
      "learning_rate": 8.288329644355883e-05,
      "loss": 0.0453,
      "relation_loss": 0.004038891033269465,
      "selected_pairs": 26848.0,
      "state_query_norm": 998.1813720703125,
      "state_relation_loss": 0.0005832245107740164,
      "step": 9240,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004038891033269465,
      "weighted_relation_loss": 0.004038891033269465
    },
    {
      "action_query_norm": 559.6348266601562,
      "action_relation_loss": 0.0032584976172074674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.652946679139383,
      "flow_matching_loss": 0.04223010428249836,
      "grad_norm": 0.149591326713562,
      "learning_rate": 8.284175731126965e-05,
      "loss": 0.0461,
      "relation_loss": 0.0038695858558639882,
      "selected_pairs": 27155.9,
      "state_query_norm": 989.5505432128906,
      "state_relation_loss": 0.0006110882386565208,
      "step": 9250,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0038695858558639882,
      "weighted_relation_loss": 0.0038695858558639882
    },
    {
      "action_query_norm": 556.40478515625,
      "action_relation_loss": 0.003179538808763027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.662301216089803,
      "flow_matching_loss": 0.04009227752685547,
      "grad_norm": 0.13698168098926544,
      "learning_rate": 8.280017827314854e-05,
      "loss": 0.0438,
      "relation_loss": 0.0037378902547061444,
      "selected_pairs": 26880.4,
      "state_query_norm": 1006.1998657226562,
      "state_relation_loss": 0.0005583513848250732,
      "step": 9260,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0037378902547061444,
      "weighted_relation_loss": 0.0037378902547061444
    },
    {
      "action_query_norm": 563.4462707519531,
      "action_relation_loss": 0.0034245771821588278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.671655753040225,
      "flow_matching_loss": 0.045444510877132416,
      "grad_norm": 0.12009339034557343,
      "learning_rate": 8.2758559379718e-05,
      "loss": 0.0494,
      "relation_loss": 0.003990847384557128,
      "selected_pairs": 26986.5,
      "state_query_norm": 1005.420458984375,
      "state_relation_loss": 0.0005662702082190662,
      "step": 9270,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003990847384557128,
      "weighted_relation_loss": 0.003990847384557128
    },
    {
      "action_query_norm": 560.6691467285157,
      "action_relation_loss": 0.003333461144939065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.681010289990645,
      "flow_matching_loss": 0.04226514175534248,
      "grad_norm": 0.1549891233444214,
      "learning_rate": 8.271690068154887e-05,
      "loss": 0.0462,
      "relation_loss": 0.003907688078470528,
      "selected_pairs": 26872.2,
      "state_query_norm": 1006.3592163085938,
      "state_relation_loss": 0.0005742268986068666,
      "step": 9280,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.003907688078470528,
      "weighted_relation_loss": 0.003907688078470528
    },
    {
      "action_query_norm": 550.9599426269531,
      "action_relation_loss": 0.0033836888847872614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.690364826941066,
      "flow_matching_loss": 0.041888803243637085,
      "grad_norm": 0.12072302401065826,
      "learning_rate": 8.26752022292604e-05,
      "loss": 0.0458,
      "relation_loss": 0.003948901500552893,
      "selected_pairs": 27291.2,
      "state_query_norm": 1000.7525756835937,
      "state_relation_loss": 0.0005652126157656312,
      "step": 9290,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.003948901500552893,
      "weighted_relation_loss": 0.003948901500552893
    },
    {
      "action_query_norm": 558.9229736328125,
      "action_relation_loss": 0.003263807506300509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.699719363891488,
      "flow_matching_loss": 0.03957379162311554,
      "grad_norm": 0.09115283191204071,
      "learning_rate": 8.263346407352018e-05,
      "loss": 0.0434,
      "relation_loss": 0.003820504155009985,
      "selected_pairs": 26551.2,
      "state_query_norm": 992.2521362304688,
      "state_relation_loss": 0.0005566966516198591,
      "step": 9300,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003820504155009985,
      "weighted_relation_loss": 0.003820504155009985
    },
    {
      "action_query_norm": 559.9573059082031,
      "action_relation_loss": 0.0034002843080088497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.709073900841908,
      "flow_matching_loss": 0.04192681573331356,
      "grad_norm": 0.122779481112957,
      "learning_rate": 8.259168626504395e-05,
      "loss": 0.0459,
      "relation_loss": 0.003965255687944591,
      "selected_pairs": 27293.3,
      "state_query_norm": 1004.3297668457031,
      "state_relation_loss": 0.000564971350831911,
      "step": 9310,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.003965255687944591,
      "weighted_relation_loss": 0.003965255687944591
    },
    {
      "action_query_norm": 555.3008972167969,
      "action_relation_loss": 0.0032041049329563976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.71842843779233,
      "flow_matching_loss": 0.042171960696578026,
      "grad_norm": 0.10646789520978928,
      "learning_rate": 8.254986885459572e-05,
      "loss": 0.0459,
      "relation_loss": 0.0037542881444096567,
      "selected_pairs": 26667.1,
      "state_query_norm": 1005.5854614257812,
      "state_relation_loss": 0.0005501832376467064,
      "step": 9320,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0037542881444096567,
      "weighted_relation_loss": 0.0037542881444096567
    },
    {
      "action_query_norm": 553.7983825683593,
      "action_relation_loss": 0.0032198563450947402,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.72778297474275,
      "flow_matching_loss": 0.04294319152832031,
      "grad_norm": 0.11261028796434402,
      "learning_rate": 8.250801189298757e-05,
      "loss": 0.0468,
      "relation_loss": 0.003811966348439455,
      "selected_pairs": 26792.7,
      "state_query_norm": 981.4952819824218,
      "state_relation_loss": 0.0005921100149862468,
      "step": 9330,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003811966348439455,
      "weighted_relation_loss": 0.003811966348439455
    },
    {
      "action_query_norm": 551.3647155761719,
      "action_relation_loss": 0.003331701294519007,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.737137511693172,
      "flow_matching_loss": 0.041633487492799756,
      "grad_norm": 0.14673645794391632,
      "learning_rate": 8.246611543107967e-05,
      "loss": 0.0455,
      "relation_loss": 0.0039038769667968155,
      "selected_pairs": 26315.6,
      "state_query_norm": 997.8766723632813,
      "state_relation_loss": 0.0005721756198909134,
      "step": 9340,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0039038769667968155,
      "weighted_relation_loss": 0.0039038769667968155
    },
    {
      "action_query_norm": 552.2107299804687,
      "action_relation_loss": 0.00335298937279731,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.746492048643592,
      "flow_matching_loss": 0.04358184449374676,
      "grad_norm": 0.15298618376255035,
      "learning_rate": 8.242417951978018e-05,
      "loss": 0.0475,
      "relation_loss": 0.003889773227274418,
      "selected_pairs": 26851.8,
      "state_query_norm": 1003.7683410644531,
      "state_relation_loss": 0.0005367838602978736,
      "step": 9350,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003889773227274418,
      "weighted_relation_loss": 0.003889773227274418
    },
    {
      "action_query_norm": 561.5155883789063,
      "action_relation_loss": 0.0032679249532520773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.755846585594012,
      "flow_matching_loss": 0.04218414202332497,
      "grad_norm": 0.1223965436220169,
      "learning_rate": 8.238220421004517e-05,
      "loss": 0.046,
      "relation_loss": 0.00382993184030056,
      "selected_pairs": 26414.5,
      "state_query_norm": 996.4355773925781,
      "state_relation_loss": 0.0005620069045107811,
      "step": 9360,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.00382993184030056,
      "weighted_relation_loss": 0.00382993184030056
    },
    {
      "action_query_norm": 555.1390869140625,
      "action_relation_loss": 0.0034148884005844593,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.765201122544434,
      "flow_matching_loss": 0.042939276248216626,
      "grad_norm": 0.1571241170167923,
      "learning_rate": 8.234018955287859e-05,
      "loss": 0.0469,
      "relation_loss": 0.003981832484714687,
      "selected_pairs": 27082.6,
      "state_query_norm": 1003.9740295410156,
      "state_relation_loss": 0.0005669440783094615,
      "step": 9370,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.003981832484714687,
      "weighted_relation_loss": 0.003981832484714687
    },
    {
      "action_query_norm": 553.6563659667969,
      "action_relation_loss": 0.003396576223894954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.774555659494855,
      "flow_matching_loss": 0.04090852960944176,
      "grad_norm": 0.15773308277130127,
      "learning_rate": 8.229813559933224e-05,
      "loss": 0.0449,
      "relation_loss": 0.00396086007822305,
      "selected_pairs": 27225.4,
      "state_query_norm": 988.1735534667969,
      "state_relation_loss": 0.0005642837844789028,
      "step": 9380,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00396086007822305,
      "weighted_relation_loss": 0.00396086007822305
    },
    {
      "action_query_norm": 556.1097351074219,
      "action_relation_loss": 0.0032496386440470813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.783910196445277,
      "flow_matching_loss": 0.045295316725969315,
      "grad_norm": 0.15939468145370483,
      "learning_rate": 8.22560424005056e-05,
      "loss": 0.0491,
      "relation_loss": 0.0038284519221633674,
      "selected_pairs": 26422.0,
      "state_query_norm": 1008.2328063964844,
      "state_relation_loss": 0.0005788132722955198,
      "step": 9390,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0038284519221633674,
      "weighted_relation_loss": 0.0038284519221633674
    },
    {
      "action_query_norm": 559.2563354492188,
      "action_relation_loss": 0.0034504033625125884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.793264733395697,
      "flow_matching_loss": 0.0423170443624258,
      "grad_norm": 0.17682932317256927,
      "learning_rate": 8.22139100075459e-05,
      "loss": 0.0463,
      "relation_loss": 0.004019969166256488,
      "selected_pairs": 27243.0,
      "state_query_norm": 1004.4385314941406,
      "state_relation_loss": 0.0005695657862816006,
      "step": 9400,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.004019969166256488,
      "weighted_relation_loss": 0.004019969166256488
    },
    {
      "action_query_norm": 557.4080444335938,
      "action_relation_loss": 0.0032994797453284263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.802619270346117,
      "flow_matching_loss": 0.04416643641889095,
      "grad_norm": 0.12588536739349365,
      "learning_rate": 8.217173847164799e-05,
      "loss": 0.048,
      "relation_loss": 0.003824420226737857,
      "selected_pairs": 26548.0,
      "state_query_norm": 1016.5113708496094,
      "state_relation_loss": 0.0005249404523056001,
      "step": 9410,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003824420226737857,
      "weighted_relation_loss": 0.003824420226737857
    },
    {
      "action_query_norm": 559.597119140625,
      "action_relation_loss": 0.0032629608642309902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.811973807296539,
      "flow_matching_loss": 0.04206833466887474,
      "grad_norm": 0.1570904701948166,
      "learning_rate": 8.212952784405422e-05,
      "loss": 0.0459,
      "relation_loss": 0.0038230855483561755,
      "selected_pairs": 26180.6,
      "state_query_norm": 1006.2282104492188,
      "state_relation_loss": 0.0005601246550213546,
      "step": 9420,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0038230855483561755,
      "weighted_relation_loss": 0.0038230855483561755
    },
    {
      "action_query_norm": 558.1781799316407,
      "action_relation_loss": 0.0032768035074695946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.82132834424696,
      "flow_matching_loss": 0.04225524738430977,
      "grad_norm": 0.13533973693847656,
      "learning_rate": 8.208727817605451e-05,
      "loss": 0.0461,
      "relation_loss": 0.0038436790462583305,
      "selected_pairs": 26478.8,
      "state_query_norm": 1025.414208984375,
      "state_relation_loss": 0.0005668755882652476,
      "step": 9430,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0038436790462583305,
      "weighted_relation_loss": 0.0038436790462583305
    },
    {
      "action_query_norm": 560.5394226074219,
      "action_relation_loss": 0.003266953956335783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.830682881197381,
      "flow_matching_loss": 0.043409931287169454,
      "grad_norm": 0.1284053474664688,
      "learning_rate": 8.204498951898618e-05,
      "loss": 0.0472,
      "relation_loss": 0.003819417464546859,
      "selected_pairs": 26415.1,
      "state_query_norm": 1000.0353393554688,
      "state_relation_loss": 0.0005524635140318424,
      "step": 9440,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.003819417464546859,
      "weighted_relation_loss": 0.003819417464546859
    },
    {
      "action_query_norm": 562.1952514648438,
      "action_relation_loss": 0.003223808738403022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.840037418147801,
      "flow_matching_loss": 0.0420730609446764,
      "grad_norm": 0.16417309641838074,
      "learning_rate": 8.200266192423395e-05,
      "loss": 0.0459,
      "relation_loss": 0.0037770394701510668,
      "selected_pairs": 26411.2,
      "state_query_norm": 1007.428564453125,
      "state_relation_loss": 0.0005532307259272784,
      "step": 9450,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0037770394701510668,
      "weighted_relation_loss": 0.0037770394701510668
    },
    {
      "action_query_norm": 563.9052856445312,
      "action_relation_loss": 0.0032908694352954625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.849391955098223,
      "flow_matching_loss": 0.04154379293322563,
      "grad_norm": 0.14659863710403442,
      "learning_rate": 8.196029544322981e-05,
      "loss": 0.0454,
      "relation_loss": 0.0038149986183270814,
      "selected_pairs": 27088.9,
      "state_query_norm": 1006.9716674804688,
      "state_relation_loss": 0.0005241291422862559,
      "step": 9460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0038149986183270814,
      "weighted_relation_loss": 0.0038149986183270814
    },
    {
      "action_query_norm": 556.7891296386719,
      "action_relation_loss": 0.003328652190975845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.858746492048644,
      "flow_matching_loss": 0.04239146932959557,
      "grad_norm": 0.09701742231845856,
      "learning_rate": 8.191789012745305e-05,
      "loss": 0.0463,
      "relation_loss": 0.003895257553085685,
      "selected_pairs": 27009.4,
      "state_query_norm": 1007.4248596191406,
      "state_relation_loss": 0.0005666053737513721,
      "step": 9470,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.003895257553085685,
      "weighted_relation_loss": 0.003895257553085685
    },
    {
      "action_query_norm": 552.0458129882812,
      "action_relation_loss": 0.003255417617037892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.868101028999064,
      "flow_matching_loss": 0.0427290789783001,
      "grad_norm": 0.10491536557674408,
      "learning_rate": 8.187544602843014e-05,
      "loss": 0.0465,
      "relation_loss": 0.003816994116641581,
      "selected_pairs": 27410.5,
      "state_query_norm": 1004.6232482910157,
      "state_relation_loss": 0.0005615765403490514,
      "step": 9480,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.003816994116641581,
      "weighted_relation_loss": 0.003816994116641581
    },
    {
      "action_query_norm": 556.0442687988282,
      "action_relation_loss": 0.003322582063265145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.877455565949486,
      "flow_matching_loss": 0.04292430244386196,
      "grad_norm": 0.1411946713924408,
      "learning_rate": 8.183296319773466e-05,
      "loss": 0.0468,
      "relation_loss": 0.003871519863605499,
      "selected_pairs": 26856.2,
      "state_query_norm": 993.1449890136719,
      "state_relation_loss": 0.0005489376955665648,
      "step": 9490,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003871519863605499,
      "weighted_relation_loss": 0.003871519863605499
    },
    {
      "action_query_norm": 554.7248962402343,
      "action_relation_loss": 0.0034360333578661085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.886810102899906,
      "flow_matching_loss": 0.041689448058605194,
      "grad_norm": 0.1257874220609665,
      "learning_rate": 8.179044168698721e-05,
      "loss": 0.0457,
      "relation_loss": 0.003990411409176886,
      "selected_pairs": 27456.7,
      "state_query_norm": 988.6370788574219,
      "state_relation_loss": 0.0005543780425796285,
      "step": 9500,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.003990411409176886,
      "weighted_relation_loss": 0.003990411409176886
    },
    {
      "action_query_norm": 552.9196105957031,
      "action_relation_loss": 0.0032867805100977423,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.896164639850328,
      "flow_matching_loss": 0.03973102942109108,
      "grad_norm": 0.10397106409072876,
      "learning_rate": 8.174788154785548e-05,
      "loss": 0.0436,
      "relation_loss": 0.0038355195894837378,
      "selected_pairs": 26688.8,
      "state_query_norm": 1016.8511169433593,
      "state_relation_loss": 0.0005487390939379111,
      "step": 9510,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0038355195894837378,
      "weighted_relation_loss": 0.0038355195894837378
    },
    {
      "action_query_norm": 559.4782531738281,
      "action_relation_loss": 0.003311318322084844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.905519176800748,
      "flow_matching_loss": 0.04074000120162964,
      "grad_norm": 0.12949863076210022,
      "learning_rate": 8.170528283205403e-05,
      "loss": 0.0446,
      "relation_loss": 0.003870849753729999,
      "selected_pairs": 26759.9,
      "state_query_norm": 1024.859619140625,
      "state_relation_loss": 0.000559531501494348,
      "step": 9520,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.003870849753729999,
      "weighted_relation_loss": 0.003870849753729999
    },
    {
      "action_query_norm": 564.6876281738281,
      "action_relation_loss": 0.00321872616186738,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.91487371375117,
      "flow_matching_loss": 0.04139854274690151,
      "grad_norm": 0.11702084541320801,
      "learning_rate": 8.166264559134435e-05,
      "loss": 0.0452,
      "relation_loss": 0.0037728709634393454,
      "selected_pairs": 26708.5,
      "state_query_norm": 1007.7614685058594,
      "state_relation_loss": 0.0005541448073927314,
      "step": 9530,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0037728709634393454,
      "weighted_relation_loss": 0.0037728709634393454
    },
    {
      "action_query_norm": 560.6519836425781,
      "action_relation_loss": 0.0032292411662638186,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.92422825070159,
      "flow_matching_loss": 0.04030180312693119,
      "grad_norm": 0.10571271181106567,
      "learning_rate": 8.161996987753467e-05,
      "loss": 0.0441,
      "relation_loss": 0.0037823944119736554,
      "selected_pairs": 26968.2,
      "state_query_norm": 1010.1472534179687,
      "state_relation_loss": 0.0005531532631721347,
      "step": 9540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0037823944119736554,
      "weighted_relation_loss": 0.0037823944119736554
    },
    {
      "action_query_norm": 556.1283569335938,
      "action_relation_loss": 0.00310270837508142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.93358278765201,
      "flow_matching_loss": 0.042969264462590216,
      "grad_norm": 0.11005048453807831,
      "learning_rate": 8.157725574248e-05,
      "loss": 0.0466,
      "relation_loss": 0.0036320701241493226,
      "selected_pairs": 26371.3,
      "state_query_norm": 1001.1010864257812,
      "state_relation_loss": 0.0005293617228744552,
      "step": 9550,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0036320701241493226,
      "weighted_relation_loss": 0.0036320701241493226
    },
    {
      "action_query_norm": 557.3899536132812,
      "action_relation_loss": 0.003240291611291468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.942937324602433,
      "flow_matching_loss": 0.04191734232008457,
      "grad_norm": 0.09906919300556183,
      "learning_rate": 8.153450323808209e-05,
      "loss": 0.0457,
      "relation_loss": 0.003772311401553452,
      "selected_pairs": 26551.3,
      "state_query_norm": 1004.5155334472656,
      "state_relation_loss": 0.0005320197698893026,
      "step": 9560,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003772311401553452,
      "weighted_relation_loss": 0.003772311401553452
    },
    {
      "action_query_norm": 553.0848449707031,
      "action_relation_loss": 0.003228071518242359,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.952291861552853,
      "flow_matching_loss": 0.038464033231139186,
      "grad_norm": 0.15748560428619385,
      "learning_rate": 8.149171241628923e-05,
      "loss": 0.0422,
      "relation_loss": 0.0037664315197616814,
      "selected_pairs": 26832.9,
      "state_query_norm": 1006.7323364257812,
      "state_relation_loss": 0.0005383600248023868,
      "step": 9570,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0037664315197616814,
      "weighted_relation_loss": 0.0037664315197616814
    },
    {
      "action_query_norm": 553.2714416503907,
      "action_relation_loss": 0.0031720722327008843,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.961646398503275,
      "flow_matching_loss": 0.0424450945109129,
      "grad_norm": 0.16389495134353638,
      "learning_rate": 8.144888332909631e-05,
      "loss": 0.0462,
      "relation_loss": 0.003718210384249687,
      "selected_pairs": 26706.4,
      "state_query_norm": 1014.5457580566406,
      "state_relation_loss": 0.0005461381457280368,
      "step": 9580,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.003718210384249687,
      "weighted_relation_loss": 0.003718210384249687
    },
    {
      "action_query_norm": 550.5174621582031,
      "action_relation_loss": 0.0033366245217621325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.971000935453695,
      "flow_matching_loss": 0.043343938514590265,
      "grad_norm": 0.14495499432086945,
      "learning_rate": 8.14060160285447e-05,
      "loss": 0.0472,
      "relation_loss": 0.00390142067335546,
      "selected_pairs": 27293.5,
      "state_query_norm": 1011.5506713867187,
      "state_relation_loss": 0.0005647961690556258,
      "step": 9590,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.00390142067335546,
      "weighted_relation_loss": 0.00390142067335546
    },
    {
      "action_query_norm": 548.9722778320313,
      "action_relation_loss": 0.0033768446417525412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.980355472404115,
      "flow_matching_loss": 0.041609359160065654,
      "grad_norm": 0.11460935324430466,
      "learning_rate": 8.136311056672223e-05,
      "loss": 0.0455,
      "relation_loss": 0.0039225353859364985,
      "selected_pairs": 27040.8,
      "state_query_norm": 1020.2197326660156,
      "state_relation_loss": 0.000545690770377405,
      "step": 9600,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0039225353859364985,
      "weighted_relation_loss": 0.0039225353859364985
    },
    {
      "action_query_norm": 551.7501708984375,
      "action_relation_loss": 0.0033284461591392756,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.989710009354537,
      "flow_matching_loss": 0.040153414756059644,
      "grad_norm": 0.11006030440330505,
      "learning_rate": 8.132016699576308e-05,
      "loss": 0.044,
      "relation_loss": 0.00388201370369643,
      "selected_pairs": 26944.2,
      "state_query_norm": 1024.9250183105469,
      "state_relation_loss": 0.0005535675038117916,
      "step": 9610,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00388201370369643,
      "weighted_relation_loss": 0.00388201370369643
    },
    {
      "action_query_norm": 559.7614868164062,
      "action_relation_loss": 0.003177989274263382,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.999064546304957,
      "flow_matching_loss": 0.041895849630236626,
      "grad_norm": 0.09911860525608063,
      "learning_rate": 8.127718536784771e-05,
      "loss": 0.0456,
      "relation_loss": 0.0037249659886583685,
      "selected_pairs": 26407.7,
      "state_query_norm": 1009.315478515625,
      "state_relation_loss": 0.0005469767434988171,
      "step": 9620,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0037249659886583685,
      "weighted_relation_loss": 0.0037249659886583685
    },
    {
      "action_query_norm": 560.4214172363281,
      "action_relation_loss": 0.0029908480122685433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.00841908325538,
      "flow_matching_loss": 0.04016271159052849,
      "grad_norm": 0.15109315514564514,
      "learning_rate": 8.12341657352029e-05,
      "loss": 0.0437,
      "relation_loss": 0.0035113892517983913,
      "selected_pairs": 24759.8,
      "state_query_norm": 991.7206359863281,
      "state_relation_loss": 0.0005205412569921463,
      "step": 9630,
      "valid_action_samples": 216.1,
      "weighted_auxiliary_loss": 0.0035113892517983913,
      "weighted_relation_loss": 0.0035113892517983913
    },
    {
      "action_query_norm": 559.9179443359375,
      "action_relation_loss": 0.003005124907940626,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.0177736202058,
      "flow_matching_loss": 0.03997042030096054,
      "grad_norm": 0.1211642250418663,
      "learning_rate": 8.119110815010151e-05,
      "loss": 0.0435,
      "relation_loss": 0.00356105079408735,
      "selected_pairs": 26692.2,
      "state_query_norm": 1015.6693908691407,
      "state_relation_loss": 0.0005559258454013616,
      "step": 9640,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00356105079408735,
      "weighted_relation_loss": 0.00356105079408735
    },
    {
      "action_query_norm": 567.4121887207032,
      "action_relation_loss": 0.003019947768189013,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.027128157156222,
      "flow_matching_loss": 0.0409382950514555,
      "grad_norm": 0.1256401687860489,
      "learning_rate": 8.114801266486261e-05,
      "loss": 0.0445,
      "relation_loss": 0.0035565633792430163,
      "selected_pairs": 26873.4,
      "state_query_norm": 1002.2412780761719,
      "state_relation_loss": 0.0005366155819501728,
      "step": 9650,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0035565633792430163,
      "weighted_relation_loss": 0.0035565633792430163
    },
    {
      "action_query_norm": 565.6514282226562,
      "action_relation_loss": 0.0029787000268697737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.036482694106642,
      "flow_matching_loss": 0.04038695022463799,
      "grad_norm": 0.1378125101327896,
      "learning_rate": 8.110487933185122e-05,
      "loss": 0.0439,
      "relation_loss": 0.0035190138733014464,
      "selected_pairs": 26883.5,
      "state_query_norm": 1021.0841552734375,
      "state_relation_loss": 0.0005403138289693743,
      "step": 9660,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0035190138733014464,
      "weighted_relation_loss": 0.0035190138733014464
    },
    {
      "action_query_norm": 558.4425048828125,
      "action_relation_loss": 0.0030955821974202992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.045837231057062,
      "flow_matching_loss": 0.04053607508540154,
      "grad_norm": 0.14550113677978516,
      "learning_rate": 8.106170820347847e-05,
      "loss": 0.0442,
      "relation_loss": 0.003626950876787305,
      "selected_pairs": 26850.6,
      "state_query_norm": 1009.5698547363281,
      "state_relation_loss": 0.000531368688098155,
      "step": 9670,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003626950876787305,
      "weighted_relation_loss": 0.003626950876787305
    },
    {
      "action_query_norm": 564.5244384765625,
      "action_relation_loss": 0.0031275537330657244,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.055191768007484,
      "flow_matching_loss": 0.03834307752549648,
      "grad_norm": 0.12588070333003998,
      "learning_rate": 8.101849933220132e-05,
      "loss": 0.042,
      "relation_loss": 0.003660257742740214,
      "selected_pairs": 27031.8,
      "state_query_norm": 1013.6270629882813,
      "state_relation_loss": 0.0005327040300471709,
      "step": 9680,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003660257742740214,
      "weighted_relation_loss": 0.003660257742740214
    },
    {
      "action_query_norm": 565.0360412597656,
      "action_relation_loss": 0.0029817720875144004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.064546304957904,
      "flow_matching_loss": 0.04035240449011326,
      "grad_norm": 0.1611928641796112,
      "learning_rate": 8.097525277052264e-05,
      "loss": 0.0439,
      "relation_loss": 0.0035275893518701197,
      "selected_pairs": 27428.1,
      "state_query_norm": 1019.8797485351563,
      "state_relation_loss": 0.0005458172730868682,
      "step": 9690,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0035275893518701197,
      "weighted_relation_loss": 0.0035275893518701197
    },
    {
      "action_query_norm": 562.8030395507812,
      "action_relation_loss": 0.002990280161611736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.073900841908326,
      "flow_matching_loss": 0.04081925489008427,
      "grad_norm": 0.1241425946354866,
      "learning_rate": 8.093196857099106e-05,
      "loss": 0.0443,
      "relation_loss": 0.0035212487448006867,
      "selected_pairs": 27039.1,
      "state_query_norm": 1003.4727294921875,
      "state_relation_loss": 0.0005309685977408663,
      "step": 9700,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0035212487448006867,
      "weighted_relation_loss": 0.0035212487448006867
    },
    {
      "action_query_norm": 559.5898315429688,
      "action_relation_loss": 0.003017132147215307,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.083255378858746,
      "flow_matching_loss": 0.0391880739480257,
      "grad_norm": 0.13523641228675842,
      "learning_rate": 8.088864678620096e-05,
      "loss": 0.0428,
      "relation_loss": 0.00356930720154196,
      "selected_pairs": 27078.4,
      "state_query_norm": 1004.4044372558594,
      "state_relation_loss": 0.0005521750601474196,
      "step": 9710,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00356930720154196,
      "weighted_relation_loss": 0.00356930720154196
    },
    {
      "action_query_norm": 558.4697631835937,
      "action_relation_loss": 0.0030621053418144585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.092609915809167,
      "flow_matching_loss": 0.04045494273304939,
      "grad_norm": 0.10889661312103271,
      "learning_rate": 8.084528746879243e-05,
      "loss": 0.0441,
      "relation_loss": 0.0036016703117638825,
      "selected_pairs": 26664.5,
      "state_query_norm": 1014.4378295898438,
      "state_relation_loss": 0.0005395649583078921,
      "step": 9720,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0036016703117638825,
      "weighted_relation_loss": 0.0036016703117638825
    },
    {
      "action_query_norm": 568.2928466796875,
      "action_relation_loss": 0.0028668354032561185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.101964452759589,
      "flow_matching_loss": 0.03855947889387608,
      "grad_norm": 0.1095304861664772,
      "learning_rate": 8.080189067145108e-05,
      "loss": 0.042,
      "relation_loss": 0.0033908657962456344,
      "selected_pairs": 27248.0,
      "state_query_norm": 1004.3367309570312,
      "state_relation_loss": 0.000524030378437601,
      "step": 9730,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0033908657962456344,
      "weighted_relation_loss": 0.0033908657962456344
    },
    {
      "action_query_norm": 560.0877746582031,
      "action_relation_loss": 0.003134568058885634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.111318989710009,
      "flow_matching_loss": 0.039190280437469485,
      "grad_norm": 0.12785837054252625,
      "learning_rate": 8.075845644690813e-05,
      "loss": 0.0428,
      "relation_loss": 0.0036553909070789815,
      "selected_pairs": 26988.2,
      "state_query_norm": 999.97705078125,
      "state_relation_loss": 0.0005208228016272187,
      "step": 9740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0036553909070789815,
      "weighted_relation_loss": 0.0036553909070789815
    },
    {
      "action_query_norm": 560.21953125,
      "action_relation_loss": 0.0029519689036533237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.12067352666043,
      "flow_matching_loss": 0.03928518816828728,
      "grad_norm": 0.13549399375915527,
      "learning_rate": 8.07149848479403e-05,
      "loss": 0.0428,
      "relation_loss": 0.0034726294921711086,
      "selected_pairs": 26454.8,
      "state_query_norm": 1010.6527282714844,
      "state_relation_loss": 0.000520660585607402,
      "step": 9750,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0034726294921711086,
      "weighted_relation_loss": 0.0034726294921711086
    },
    {
      "action_query_norm": 566.6704040527344,
      "action_relation_loss": 0.0029471392976120115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.130028063610851,
      "flow_matching_loss": 0.04042433425784111,
      "grad_norm": 0.1315506547689438,
      "learning_rate": 8.067147592736962e-05,
      "loss": 0.0439,
      "relation_loss": 0.003469239198602736,
      "selected_pairs": 27063.0,
      "state_query_norm": 1007.6941162109375,
      "state_relation_loss": 0.0005220999417360872,
      "step": 9760,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003469239198602736,
      "weighted_relation_loss": 0.003469239198602736
    },
    {
      "action_query_norm": 568.9349060058594,
      "action_relation_loss": 0.0031316981185227633,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.139382600561273,
      "flow_matching_loss": 0.040152204036712644,
      "grad_norm": 0.12024048715829849,
      "learning_rate": 8.062792973806357e-05,
      "loss": 0.0438,
      "relation_loss": 0.0036521330941468476,
      "selected_pairs": 26458.3,
      "state_query_norm": 1024.3634521484375,
      "state_relation_loss": 0.0005204349465202541,
      "step": 9770,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0036521330941468476,
      "weighted_relation_loss": 0.0036521330941468476
    },
    {
      "action_query_norm": 568.8125,
      "action_relation_loss": 0.003052138234488666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.148737137511693,
      "flow_matching_loss": 0.03924481198191643,
      "grad_norm": 0.12406923621892929,
      "learning_rate": 8.058434633293485e-05,
      "loss": 0.0428,
      "relation_loss": 0.0035770424408838155,
      "selected_pairs": 26735.5,
      "state_query_norm": 999.0346069335938,
      "state_relation_loss": 0.0005249041947536171,
      "step": 9780,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0035770424408838155,
      "weighted_relation_loss": 0.0035770424408838155
    },
    {
      "action_query_norm": 563.4033996582032,
      "action_relation_loss": 0.0031162020051851867,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.158091674462113,
      "flow_matching_loss": 0.038242855295538905,
      "grad_norm": 0.1258315145969391,
      "learning_rate": 8.054072576494143e-05,
      "loss": 0.0419,
      "relation_loss": 0.0036445332458242773,
      "selected_pairs": 26499.6,
      "state_query_norm": 1008.2021057128907,
      "state_relation_loss": 0.0005283312639221549,
      "step": 9790,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0036445332458242773,
      "weighted_relation_loss": 0.0036445332458242773
    },
    {
      "action_query_norm": 565.0098754882813,
      "action_relation_loss": 0.0032487396616488694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.167446211412535,
      "flow_matching_loss": 0.04125081449747085,
      "grad_norm": 0.158834308385849,
      "learning_rate": 8.04970680870864e-05,
      "loss": 0.045,
      "relation_loss": 0.0037903203163295986,
      "selected_pairs": 27030.0,
      "state_query_norm": 1034.3328796386718,
      "state_relation_loss": 0.000541580596473068,
      "step": 9800,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0037903203163295986,
      "weighted_relation_loss": 0.0037903203163295986
    },
    {
      "action_query_norm": 561.49560546875,
      "action_relation_loss": 0.0030990973114967345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.176800748362956,
      "flow_matching_loss": 0.03936919569969177,
      "grad_norm": 0.14261086285114288,
      "learning_rate": 8.045337335241793e-05,
      "loss": 0.043,
      "relation_loss": 0.003644022555090487,
      "selected_pairs": 26800.4,
      "state_query_norm": 1013.6726684570312,
      "state_relation_loss": 0.0005449252552352845,
      "step": 9810,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.003644022555090487,
      "weighted_relation_loss": 0.003644022555090487
    },
    {
      "action_query_norm": 568.1686279296875,
      "action_relation_loss": 0.003068780549801886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.186155285313378,
      "flow_matching_loss": 0.04106721766293049,
      "grad_norm": 0.15039575099945068,
      "learning_rate": 8.040964161402932e-05,
      "loss": 0.0447,
      "relation_loss": 0.003596493904478848,
      "selected_pairs": 26989.0,
      "state_query_norm": 1015.1923217773438,
      "state_relation_loss": 0.0005277133692288771,
      "step": 9820,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003596493904478848,
      "weighted_relation_loss": 0.003596493904478848
    },
    {
      "action_query_norm": 563.8356323242188,
      "action_relation_loss": 0.0030837878352031113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.195509822263798,
      "flow_matching_loss": 0.04045993871986866,
      "grad_norm": 0.14517462253570557,
      "learning_rate": 8.036587292505869e-05,
      "loss": 0.0441,
      "relation_loss": 0.0036383698228746653,
      "selected_pairs": 27173.4,
      "state_query_norm": 1023.6918090820312,
      "state_relation_loss": 0.0005545819789404049,
      "step": 9830,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0036383698228746653,
      "weighted_relation_loss": 0.0036383698228746653
    },
    {
      "action_query_norm": 563.0954040527344,
      "action_relation_loss": 0.0030493496218696237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.204864359214218,
      "flow_matching_loss": 0.04034711644053459,
      "grad_norm": 0.1377430409193039,
      "learning_rate": 8.032206733868912e-05,
      "loss": 0.0439,
      "relation_loss": 0.0035764180356636643,
      "selected_pairs": 27010.4,
      "state_query_norm": 1036.0736145019532,
      "state_relation_loss": 0.0005270684283459559,
      "step": 9840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0035764180356636643,
      "weighted_relation_loss": 0.0035764180356636643
    },
    {
      "action_query_norm": 563.6783447265625,
      "action_relation_loss": 0.0031305431853979825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.21421889616464,
      "flow_matching_loss": 0.041195162758231166,
      "grad_norm": 0.15375657379627228,
      "learning_rate": 8.027822490814858e-05,
      "loss": 0.0448,
      "relation_loss": 0.003638674388639629,
      "selected_pairs": 26600.0,
      "state_query_norm": 1030.9631408691407,
      "state_relation_loss": 0.0005081311712274328,
      "step": 9850,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.003638674388639629,
      "weighted_relation_loss": 0.003638674388639629
    },
    {
      "action_query_norm": 569.2980102539062,
      "action_relation_loss": 0.00310614132322371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.22357343311506,
      "flow_matching_loss": 0.0411929115653038,
      "grad_norm": 0.10846293717622757,
      "learning_rate": 8.023434568670972e-05,
      "loss": 0.0448,
      "relation_loss": 0.0036316574783995747,
      "selected_pairs": 27033.0,
      "state_query_norm": 1020.2543518066407,
      "state_relation_loss": 0.0005255161580862477,
      "step": 9860,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0036316574783995747,
      "weighted_relation_loss": 0.0036316574783995747
    },
    {
      "action_query_norm": 573.8858764648437,
      "action_relation_loss": 0.0030593791510909797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.232927970065482,
      "flow_matching_loss": 0.04140405617654323,
      "grad_norm": 0.14652246236801147,
      "learning_rate": 8.019042972768992e-05,
      "loss": 0.045,
      "relation_loss": 0.003602531854994595,
      "selected_pairs": 27059.4,
      "state_query_norm": 1009.2606384277344,
      "state_relation_loss": 0.0005431526864413172,
      "step": 9870,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003602531854994595,
      "weighted_relation_loss": 0.003602531854994595
    },
    {
      "action_query_norm": 563.6106079101562,
      "action_relation_loss": 0.003108482644893229,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.242282507015902,
      "flow_matching_loss": 0.0372284771874547,
      "grad_norm": 0.16017505526542664,
      "learning_rate": 8.014647708445124e-05,
      "loss": 0.0409,
      "relation_loss": 0.00365060530602932,
      "selected_pairs": 27130.3,
      "state_query_norm": 1007.6447021484375,
      "state_relation_loss": 0.0005421226553153247,
      "step": 9880,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.00365060530602932,
      "weighted_relation_loss": 0.00365060530602932
    },
    {
      "action_query_norm": 569.708837890625,
      "action_relation_loss": 0.0031185034662485124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.251637043966324,
      "flow_matching_loss": 0.03988484553992748,
      "grad_norm": 0.1408877670764923,
      "learning_rate": 8.010248781040027e-05,
      "loss": 0.0435,
      "relation_loss": 0.00364089896902442,
      "selected_pairs": 26738.0,
      "state_query_norm": 1034.0307189941407,
      "state_relation_loss": 0.0005223954387474806,
      "step": 9890,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00364089896902442,
      "weighted_relation_loss": 0.00364089896902442
    },
    {
      "action_query_norm": 563.093310546875,
      "action_relation_loss": 0.003021990740671754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.260991580916745,
      "flow_matching_loss": 0.04202127158641815,
      "grad_norm": 0.13248883187770844,
      "learning_rate": 8.005846195898814e-05,
      "loss": 0.0456,
      "relation_loss": 0.0035599260358139873,
      "selected_pairs": 26387.5,
      "state_query_norm": 1004.732373046875,
      "state_relation_loss": 0.0005379353009629994,
      "step": 9900,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0035599260358139873,
      "weighted_relation_loss": 0.0035599260358139873
    },
    {
      "action_query_norm": 565.6777099609375,
      "action_relation_loss": 0.0030950745334848762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.270346117867165,
      "flow_matching_loss": 0.03878009505569935,
      "grad_norm": 0.14737559854984283,
      "learning_rate": 8.00143995837104e-05,
      "loss": 0.0424,
      "relation_loss": 0.0036437410162761807,
      "selected_pairs": 27361.5,
      "state_query_norm": 1020.2931335449218,
      "state_relation_loss": 0.0005486664886120707,
      "step": 9910,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0036437410162761807,
      "weighted_relation_loss": 0.0036437410162761807
    },
    {
      "action_query_norm": 563.5494812011718,
      "action_relation_loss": 0.003205077489838004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.279700654817587,
      "flow_matching_loss": 0.039921990036964415,
      "grad_norm": 0.15207381546497345,
      "learning_rate": 7.997030073810698e-05,
      "loss": 0.0437,
      "relation_loss": 0.003760047163814306,
      "selected_pairs": 27387.0,
      "state_query_norm": 1000.047216796875,
      "state_relation_loss": 0.0005549697030801326,
      "step": 9920,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.003760047163814306,
      "weighted_relation_loss": 0.003760047163814306
    },
    {
      "action_query_norm": 558.7274719238281,
      "action_relation_loss": 0.0030296727316454055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.289055191768007,
      "flow_matching_loss": 0.041547392308712,
      "grad_norm": 0.16700875759124756,
      "learning_rate": 7.992616547576217e-05,
      "loss": 0.0451,
      "relation_loss": 0.0035674469079822304,
      "selected_pairs": 26413.0,
      "state_query_norm": 1029.5193420410155,
      "state_relation_loss": 0.0005377741588745266,
      "step": 9930,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0035674469079822304,
      "weighted_relation_loss": 0.0035674469079822304
    },
    {
      "action_query_norm": 574.478466796875,
      "action_relation_loss": 0.002980783930979669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.298409728718429,
      "flow_matching_loss": 0.039762815833091734,
      "grad_norm": 0.16215664148330688,
      "learning_rate": 7.988199385030445e-05,
      "loss": 0.0433,
      "relation_loss": 0.003534931852482259,
      "selected_pairs": 26960.3,
      "state_query_norm": 1020.7365417480469,
      "state_relation_loss": 0.0005541479447856545,
      "step": 9940,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003534931852482259,
      "weighted_relation_loss": 0.003534931852482259
    },
    {
      "action_query_norm": 569.2502868652343,
      "action_relation_loss": 0.0031009413534775375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.30776426566885,
      "flow_matching_loss": 0.03952622227370739,
      "grad_norm": 0.13466964662075043,
      "learning_rate": 7.98377859154065e-05,
      "loss": 0.0432,
      "relation_loss": 0.0036410844884812834,
      "selected_pairs": 26821.6,
      "state_query_norm": 998.4271545410156,
      "state_relation_loss": 0.0005401430884376168,
      "step": 9950,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0036410844884812834,
      "weighted_relation_loss": 0.0036410844884812834
    },
    {
      "action_query_norm": 571.8309997558594,
      "action_relation_loss": 0.003068202827125788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.31711880261927,
      "flow_matching_loss": 0.04169608987867832,
      "grad_norm": 0.11939090490341187,
      "learning_rate": 7.979354172478516e-05,
      "loss": 0.0453,
      "relation_loss": 0.0036182948388159275,
      "selected_pairs": 26922.8,
      "state_query_norm": 1011.7553588867188,
      "state_relation_loss": 0.0005500919884070754,
      "step": 9960,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0036182948388159275,
      "weighted_relation_loss": 0.0036182948388159275
    },
    {
      "action_query_norm": 571.3349609375,
      "action_relation_loss": 0.003016367577947676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.326473339569691,
      "flow_matching_loss": 0.04255517981946468,
      "grad_norm": 0.09588981419801712,
      "learning_rate": 7.974926133220127e-05,
      "loss": 0.0461,
      "relation_loss": 0.0035838716197758914,
      "selected_pairs": 26801.1,
      "state_query_norm": 1014.7902465820313,
      "state_relation_loss": 0.0005675040243659169,
      "step": 9970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0035838716197758914,
      "weighted_relation_loss": 0.0035838716197758914
    },
    {
      "action_query_norm": 574.7902465820313,
      "action_relation_loss": 0.003120136843062937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.335827876520112,
      "flow_matching_loss": 0.04246480129659176,
      "grad_norm": 0.11581689119338989,
      "learning_rate": 7.970494479145967e-05,
      "loss": 0.0461,
      "relation_loss": 0.0036732615437358616,
      "selected_pairs": 26986.7,
      "state_query_norm": 1016.48291015625,
      "state_relation_loss": 0.000553124729776755,
      "step": 9980,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0036732615437358616,
      "weighted_relation_loss": 0.0036732615437358616
    },
    {
      "action_query_norm": 574.8994018554688,
      "action_relation_loss": 0.003055025194771588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.345182413470534,
      "flow_matching_loss": 0.0400150690227747,
      "grad_norm": 0.13821642100811005,
      "learning_rate": 7.966059215640919e-05,
      "loss": 0.0436,
      "relation_loss": 0.0036166618345305325,
      "selected_pairs": 27013.6,
      "state_query_norm": 1021.418701171875,
      "state_relation_loss": 0.0005616366164758801,
      "step": 9990,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0036166618345305325,
      "weighted_relation_loss": 0.0036166618345305325
    },
    {
      "action_query_norm": 572.7493957519531,
      "action_relation_loss": 0.003096339199692011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.354536950420954,
      "flow_matching_loss": 0.038231611251831055,
      "grad_norm": 0.10769647359848022,
      "learning_rate": 7.961620348094241e-05,
      "loss": 0.0419,
      "relation_loss": 0.003637776174582541,
      "selected_pairs": 26443.5,
      "state_query_norm": 1016.7017822265625,
      "state_relation_loss": 0.0005414369632489979,
      "step": 10000,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.003637776174582541,
      "weighted_relation_loss": 0.003637776174582541
    },
    {
      "action_query_norm": 567.2746032714844,
      "action_relation_loss": 0.0031807862222194673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.363891487371376,
      "flow_matching_loss": 0.04254019223153591,
      "grad_norm": 0.15685239434242249,
      "learning_rate": 7.957177881899577e-05,
      "loss": 0.0463,
      "relation_loss": 0.0037170667434111237,
      "selected_pairs": 25962.7,
      "state_query_norm": 1016.2889099121094,
      "state_relation_loss": 0.0005362805182812736,
      "step": 10010,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.0037170667434111237,
      "weighted_relation_loss": 0.0037170667434111237
    },
    {
      "action_query_norm": 565.3806823730469,
      "action_relation_loss": 0.003091058344580233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.373246024321796,
      "flow_matching_loss": 0.04097287841141224,
      "grad_norm": 0.11937982589006424,
      "learning_rate": 7.952731822454944e-05,
      "loss": 0.0446,
      "relation_loss": 0.003648437582887709,
      "selected_pairs": 27070.1,
      "state_query_norm": 1002.8889282226562,
      "state_relation_loss": 0.0005573792383074761,
      "step": 10020,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003648437582887709,
      "weighted_relation_loss": 0.003648437582887709
    },
    {
      "action_query_norm": 562.6949096679688,
      "action_relation_loss": 0.003119559190236032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.382600561272216,
      "flow_matching_loss": 0.0406427301466465,
      "grad_norm": 0.12170178443193436,
      "learning_rate": 7.948282175162722e-05,
      "loss": 0.0443,
      "relation_loss": 0.0036587070440873503,
      "selected_pairs": 26922.1,
      "state_query_norm": 1023.6469909667969,
      "state_relation_loss": 0.0005391478451201692,
      "step": 10030,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0036587070440873503,
      "weighted_relation_loss": 0.0036587070440873503
    },
    {
      "action_query_norm": 573.8922241210937,
      "action_relation_loss": 0.003206764278002083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.391955098222638,
      "flow_matching_loss": 0.03960405215620995,
      "grad_norm": 0.11800117790699005,
      "learning_rate": 7.943828945429652e-05,
      "loss": 0.0433,
      "relation_loss": 0.0037298842100426553,
      "selected_pairs": 26778.7,
      "state_query_norm": 1015.3623901367188,
      "state_relation_loss": 0.0005231199524132535,
      "step": 10040,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0037298842100426553,
      "weighted_relation_loss": 0.0037298842100426553
    },
    {
      "action_query_norm": 573.1947326660156,
      "action_relation_loss": 0.0029529579216614364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.401309635173059,
      "flow_matching_loss": 0.04156980700790882,
      "grad_norm": 0.09548933804035187,
      "learning_rate": 7.939372138666827e-05,
      "loss": 0.045,
      "relation_loss": 0.0034742013085633516,
      "selected_pairs": 26604.3,
      "state_query_norm": 1018.5814392089844,
      "state_relation_loss": 0.0005212433694396168,
      "step": 10050,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0034742013085633516,
      "weighted_relation_loss": 0.0034742013085633516
    },
    {
      "action_query_norm": 559.6249755859375,
      "action_relation_loss": 0.0029727087123319505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.41066417212348,
      "flow_matching_loss": 0.0376501752063632,
      "grad_norm": 0.11172764748334885,
      "learning_rate": 7.934911760289692e-05,
      "loss": 0.0412,
      "relation_loss": 0.003508494677953422,
      "selected_pairs": 26763.5,
      "state_query_norm": 1026.465838623047,
      "state_relation_loss": 0.0005357859801733867,
      "step": 10060,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.003508494677953422,
      "weighted_relation_loss": 0.003508494677953422
    },
    {
      "action_query_norm": 562.1607177734375,
      "action_relation_loss": 0.0032926470739766955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.4200187090739,
      "flow_matching_loss": 0.0419734176248312,
      "grad_norm": 0.1564335972070694,
      "learning_rate": 7.930447815718022e-05,
      "loss": 0.0458,
      "relation_loss": 0.0038246183423325418,
      "selected_pairs": 26992.9,
      "state_query_norm": 1033.200177001953,
      "state_relation_loss": 0.0005319712712662294,
      "step": 10070,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0038246183423325418,
      "weighted_relation_loss": 0.0038246183423325418
    },
    {
      "action_query_norm": 563.5837829589843,
      "action_relation_loss": 0.003094081860035658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.429373246024323,
      "flow_matching_loss": 0.038944365829229353,
      "grad_norm": 0.13859741389751434,
      "learning_rate": 7.925980310375934e-05,
      "loss": 0.0426,
      "relation_loss": 0.003610816854052246,
      "selected_pairs": 26988.3,
      "state_query_norm": 1038.4040466308593,
      "state_relation_loss": 0.0005167349852854386,
      "step": 10080,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003610816854052246,
      "weighted_relation_loss": 0.003610816854052246
    },
    {
      "action_query_norm": 574.5458862304688,
      "action_relation_loss": 0.0030355513561517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.438727782974743,
      "flow_matching_loss": 0.04114270992577076,
      "grad_norm": 0.12483018636703491,
      "learning_rate": 7.921509249691865e-05,
      "loss": 0.0447,
      "relation_loss": 0.0035649254685267808,
      "selected_pairs": 26924.7,
      "state_query_norm": 1040.522235107422,
      "state_relation_loss": 0.0005293740890920162,
      "step": 10090,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0035649254685267808,
      "weighted_relation_loss": 0.0035649254685267808
    },
    {
      "action_query_norm": 576.9600708007813,
      "action_relation_loss": 0.0029327939730137585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.448082319925163,
      "flow_matching_loss": 0.039910972490906714,
      "grad_norm": 0.10207889974117279,
      "learning_rate": 7.917034639098579e-05,
      "loss": 0.0434,
      "relation_loss": 0.0034579794155433774,
      "selected_pairs": 26757.3,
      "state_query_norm": 1023.9108581542969,
      "state_relation_loss": 0.0005251854512607679,
      "step": 10100,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0034579794155433774,
      "weighted_relation_loss": 0.0034579794155433774
    },
    {
      "action_query_norm": 574.4587463378906,
      "action_relation_loss": 0.0030461834743618966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.457436856875585,
      "flow_matching_loss": 0.039702728018164636,
      "grad_norm": 0.10545670241117477,
      "learning_rate": 7.912556484033146e-05,
      "loss": 0.0433,
      "relation_loss": 0.003564439108595252,
      "selected_pairs": 26947.2,
      "state_query_norm": 1022.7711669921875,
      "state_relation_loss": 0.0005182556575164199,
      "step": 10110,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.003564439108595252,
      "weighted_relation_loss": 0.003564439108595252
    },
    {
      "action_query_norm": 574.8014526367188,
      "action_relation_loss": 0.0030289598973467944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.466791393826005,
      "flow_matching_loss": 0.040310709923505786,
      "grad_norm": 0.11357544362545013,
      "learning_rate": 7.908074789936952e-05,
      "loss": 0.0439,
      "relation_loss": 0.003562743286602199,
      "selected_pairs": 26439.1,
      "state_query_norm": 1005.9628540039063,
      "state_relation_loss": 0.0005337833863450214,
      "step": 10120,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.003562743286602199,
      "weighted_relation_loss": 0.003562743286602199
    },
    {
      "action_query_norm": 576.7146240234375,
      "action_relation_loss": 0.0029479881282895803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.476145930776427,
      "flow_matching_loss": 0.04083816818892956,
      "grad_norm": 0.1427766978740692,
      "learning_rate": 7.903589562255673e-05,
      "loss": 0.0443,
      "relation_loss": 0.0034603183157742023,
      "selected_pairs": 27100.8,
      "state_query_norm": 1026.2715087890624,
      "state_relation_loss": 0.0005123301962157711,
      "step": 10130,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0034603183157742023,
      "weighted_relation_loss": 0.0034603183157742023
    },
    {
      "action_query_norm": 567.3924560546875,
      "action_relation_loss": 0.0031627719523385166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.485500467726848,
      "flow_matching_loss": 0.03956067785620689,
      "grad_norm": 0.1543733924627304,
      "learning_rate": 7.899100806439285e-05,
      "loss": 0.0432,
      "relation_loss": 0.003672938258387148,
      "selected_pairs": 26595.2,
      "state_query_norm": 1034.3492126464844,
      "state_relation_loss": 0.0005101663409732282,
      "step": 10140,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.003672938258387148,
      "weighted_relation_loss": 0.003672938258387148
    },
    {
      "action_query_norm": 571.6459899902344,
      "action_relation_loss": 0.00319326538592577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.494855004677268,
      "flow_matching_loss": 0.03907845411449671,
      "grad_norm": 0.18796664476394653,
      "learning_rate": 7.894608527942049e-05,
      "loss": 0.0428,
      "relation_loss": 0.0037106496281921865,
      "selected_pairs": 26553.8,
      "state_query_norm": 1011.8886291503907,
      "state_relation_loss": 0.0005173842335352674,
      "step": 10150,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0037106496281921865,
      "weighted_relation_loss": 0.0037106496281921865
    },
    {
      "action_query_norm": 577.1554382324218,
      "action_relation_loss": 0.002955740271136165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.50420954162769,
      "flow_matching_loss": 0.039829079806804654,
      "grad_norm": 0.14893127977848053,
      "learning_rate": 7.89011273222251e-05,
      "loss": 0.0433,
      "relation_loss": 0.0034846676746383308,
      "selected_pairs": 26991.3,
      "state_query_norm": 1022.4756530761719,
      "state_relation_loss": 0.0005289274005917833,
      "step": 10160,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0034846676746383308,
      "weighted_relation_loss": 0.0034846676746383308
    },
    {
      "action_query_norm": 571.609375,
      "action_relation_loss": 0.002940386580303311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.51356407857811,
      "flow_matching_loss": 0.03872067369520664,
      "grad_norm": 0.17626264691352844,
      "learning_rate": 7.885613424743482e-05,
      "loss": 0.0422,
      "relation_loss": 0.0034626900684088468,
      "selected_pairs": 26551.2,
      "state_query_norm": 1025.3439086914063,
      "state_relation_loss": 0.0005223034968366846,
      "step": 10170,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0034626900684088468,
      "weighted_relation_loss": 0.0034626900684088468
    },
    {
      "action_query_norm": 572.0151062011719,
      "action_relation_loss": 0.0030440923059359194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.522918615528532,
      "flow_matching_loss": 0.04122558832168579,
      "grad_norm": 0.12790977954864502,
      "learning_rate": 7.881110610972044e-05,
      "loss": 0.0448,
      "relation_loss": 0.003580496064387262,
      "selected_pairs": 27037.7,
      "state_query_norm": 1032.4681091308594,
      "state_relation_loss": 0.0005364037700928747,
      "step": 10180,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003580496064387262,
      "weighted_relation_loss": 0.003580496064387262
    },
    {
      "action_query_norm": 568.8158264160156,
      "action_relation_loss": 0.003098817286081612,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.532273152478952,
      "flow_matching_loss": 0.039553165435791016,
      "grad_norm": 0.15186405181884766,
      "learning_rate": 7.876604296379544e-05,
      "loss": 0.0432,
      "relation_loss": 0.003623780910857022,
      "selected_pairs": 26849.8,
      "state_query_norm": 1022.3883850097657,
      "state_relation_loss": 0.0005249636509688571,
      "step": 10190,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003623780910857022,
      "weighted_relation_loss": 0.003623780910857022
    },
    {
      "action_query_norm": 562.8159545898437,
      "action_relation_loss": 0.002903343574143946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.541627689429372,
      "flow_matching_loss": 0.03951599672436714,
      "grad_norm": 0.11599507927894592,
      "learning_rate": 7.872094486441579e-05,
      "loss": 0.043,
      "relation_loss": 0.0034405286191031338,
      "selected_pairs": 26728.3,
      "state_query_norm": 1023.5391662597656,
      "state_relation_loss": 0.0005371850624214858,
      "step": 10200,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0034405286191031338,
      "weighted_relation_loss": 0.0034405286191031338
    },
    {
      "action_query_norm": 570.0403381347656,
      "action_relation_loss": 0.0029725088039413094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.550982226379794,
      "flow_matching_loss": 0.04032497145235538,
      "grad_norm": 0.1493271291255951,
      "learning_rate": 7.867581186637991e-05,
      "loss": 0.0438,
      "relation_loss": 0.003503696876578033,
      "selected_pairs": 27325.5,
      "state_query_norm": 1026.0405578613281,
      "state_relation_loss": 0.0005311880639055743,
      "step": 10210,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.003503696876578033,
      "weighted_relation_loss": 0.003503696876578033
    },
    {
      "action_query_norm": 573.1463256835938,
      "action_relation_loss": 0.002892940677702427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.560336763330215,
      "flow_matching_loss": 0.03866332396864891,
      "grad_norm": 0.15073317289352417,
      "learning_rate": 7.863064402452866e-05,
      "loss": 0.0421,
      "relation_loss": 0.0034256954211741688,
      "selected_pairs": 26875.3,
      "state_query_norm": 1007.8794067382812,
      "state_relation_loss": 0.0005327547842171043,
      "step": 10220,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0034256954211741688,
      "weighted_relation_loss": 0.0034256954211741688
    },
    {
      "action_query_norm": 570.5025939941406,
      "action_relation_loss": 0.0029058302054181693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.569691300280637,
      "flow_matching_loss": 0.03935191221535206,
      "grad_norm": 0.12882313132286072,
      "learning_rate": 7.858544139374524e-05,
      "loss": 0.0428,
      "relation_loss": 0.0034239138010889293,
      "selected_pairs": 26479.4,
      "state_query_norm": 1020.8742736816406,
      "state_relation_loss": 0.0005180835723876953,
      "step": 10230,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0034239138010889293,
      "weighted_relation_loss": 0.0034239138010889293
    },
    {
      "action_query_norm": 576.5389526367187,
      "action_relation_loss": 0.0029705135617405175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.579045837231057,
      "flow_matching_loss": 0.03852740190923214,
      "grad_norm": 0.12647970020771027,
      "learning_rate": 7.854020402895508e-05,
      "loss": 0.042,
      "relation_loss": 0.00349657298065722,
      "selected_pairs": 27347.9,
      "state_query_norm": 1022.6256042480469,
      "state_relation_loss": 0.0005260593956336379,
      "step": 10240,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.00349657298065722,
      "weighted_relation_loss": 0.00349657298065722
    },
    {
      "action_query_norm": 570.9993286132812,
      "action_relation_loss": 0.0027900519082322715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.588400374181479,
      "flow_matching_loss": 0.03883907236158848,
      "grad_norm": 0.12229114025831223,
      "learning_rate": 7.849493198512587e-05,
      "loss": 0.0421,
      "relation_loss": 0.0032997441710904242,
      "selected_pairs": 26225.9,
      "state_query_norm": 1045.8112487792969,
      "state_relation_loss": 0.0005096922628581524,
      "step": 10250,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0032997441710904242,
      "weighted_relation_loss": 0.0032997441710904242
    },
    {
      "action_query_norm": 576.25419921875,
      "action_relation_loss": 0.002897094818763435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.597754911131899,
      "flow_matching_loss": 0.038874077051877974,
      "grad_norm": 0.13734158873558044,
      "learning_rate": 7.84496253172674e-05,
      "loss": 0.0423,
      "relation_loss": 0.003448928636498749,
      "selected_pairs": 26416.1,
      "state_query_norm": 1006.7274719238281,
      "state_relation_loss": 0.0005518338235560805,
      "step": 10260,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.003448928636498749,
      "weighted_relation_loss": 0.003448928636498749
    },
    {
      "action_query_norm": 572.5341003417968,
      "action_relation_loss": 0.002919224230572581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.60710944808232,
      "flow_matching_loss": 0.03781481832265854,
      "grad_norm": 0.12113689631223679,
      "learning_rate": 7.840428408043155e-05,
      "loss": 0.0413,
      "relation_loss": 0.0034680607495829463,
      "selected_pairs": 26923.1,
      "state_query_norm": 1014.6351318359375,
      "state_relation_loss": 0.0005488365655764938,
      "step": 10270,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0034680607495829463,
      "weighted_relation_loss": 0.0034680607495829463
    },
    {
      "action_query_norm": 569.191943359375,
      "action_relation_loss": 0.003088672808371484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.616463985032741,
      "flow_matching_loss": 0.03909649364650249,
      "grad_norm": 0.1549576371908188,
      "learning_rate": 7.835890832971218e-05,
      "loss": 0.0427,
      "relation_loss": 0.0036101733101531865,
      "selected_pairs": 26710.8,
      "state_query_norm": 1026.2001220703125,
      "state_relation_loss": 0.0005215004697674885,
      "step": 10280,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0036101733101531865,
      "weighted_relation_loss": 0.0036101733101531865
    },
    {
      "action_query_norm": 564.9037719726563,
      "action_relation_loss": 0.002907234593294561,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.625818521983161,
      "flow_matching_loss": 0.039565838128328326,
      "grad_norm": 0.1252710372209549,
      "learning_rate": 7.831349812024512e-05,
      "loss": 0.043,
      "relation_loss": 0.0034441775875166058,
      "selected_pairs": 27224.6,
      "state_query_norm": 1014.9341857910156,
      "state_relation_loss": 0.0005369429709389806,
      "step": 10290,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0034441775875166058,
      "weighted_relation_loss": 0.0034441775875166058
    },
    {
      "action_query_norm": 569.6801513671875,
      "action_relation_loss": 0.0030044791288673877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.635173058933583,
      "flow_matching_loss": 0.039204877242445944,
      "grad_norm": 0.15662819147109985,
      "learning_rate": 7.826805350720807e-05,
      "loss": 0.0427,
      "relation_loss": 0.0035041509196162223,
      "selected_pairs": 27132.2,
      "state_query_norm": 1048.477099609375,
      "state_relation_loss": 0.0004996717732865363,
      "step": 10300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0035041509196162223,
      "weighted_relation_loss": 0.0035041509196162223
    },
    {
      "action_query_norm": 575.3147827148438,
      "action_relation_loss": 0.002963569853454828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.644527595884004,
      "flow_matching_loss": 0.04194641634821892,
      "grad_norm": 0.12596780061721802,
      "learning_rate": 7.822257454582049e-05,
      "loss": 0.0454,
      "relation_loss": 0.003485400346107781,
      "selected_pairs": 27266.0,
      "state_query_norm": 1040.4733032226563,
      "state_relation_loss": 0.0005218305013841018,
      "step": 10310,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003485400346107781,
      "weighted_relation_loss": 0.003485400346107781
    },
    {
      "action_query_norm": 580.9404418945312,
      "action_relation_loss": 0.0029974442441016436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.653882132834426,
      "flow_matching_loss": 0.03882065266370773,
      "grad_norm": 0.14790353178977966,
      "learning_rate": 7.817706129134363e-05,
      "loss": 0.0423,
      "relation_loss": 0.0035151334712281825,
      "selected_pairs": 26299.1,
      "state_query_norm": 1039.2302307128907,
      "state_relation_loss": 0.0005176891922019422,
      "step": 10320,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0035151334712281825,
      "weighted_relation_loss": 0.0035151334712281825
    },
    {
      "action_query_norm": 573.0989685058594,
      "action_relation_loss": 0.0031227259896695614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.663236669784846,
      "flow_matching_loss": 0.040539173409342766,
      "grad_norm": 0.17598499357700348,
      "learning_rate": 7.813151379908036e-05,
      "loss": 0.0442,
      "relation_loss": 0.0036383661907166244,
      "selected_pairs": 27251.1,
      "state_query_norm": 1020.5908386230469,
      "state_relation_loss": 0.0005156401981366798,
      "step": 10330,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0036383661907166244,
      "weighted_relation_loss": 0.0036383661907166244
    },
    {
      "action_query_norm": 579.2519653320312,
      "action_relation_loss": 0.002848997595719993,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.672591206735266,
      "flow_matching_loss": 0.039389504492282866,
      "grad_norm": 0.12478603422641754,
      "learning_rate": 7.808593212437523e-05,
      "loss": 0.0428,
      "relation_loss": 0.0033699523890390994,
      "selected_pairs": 27365.9,
      "state_query_norm": 1054.060791015625,
      "state_relation_loss": 0.0005209548195125535,
      "step": 10340,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0033699523890390994,
      "weighted_relation_loss": 0.0033699523890390994
    },
    {
      "action_query_norm": 575.0722045898438,
      "action_relation_loss": 0.0030539716128259895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.681945743685688,
      "flow_matching_loss": 0.037942204251885416,
      "grad_norm": 0.13819986581802368,
      "learning_rate": 7.804031632261421e-05,
      "loss": 0.0415,
      "relation_loss": 0.0035673101898282765,
      "selected_pairs": 25797.4,
      "state_query_norm": 1029.143731689453,
      "state_relation_loss": 0.0005133385537192225,
      "step": 10350,
      "valid_action_samples": 227.6,
      "weighted_auxiliary_loss": 0.0035673101898282765,
      "weighted_relation_loss": 0.0035673101898282765
    },
    {
      "action_query_norm": 579.2841491699219,
      "action_relation_loss": 0.0028997734654694794,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.691300280636108,
      "flow_matching_loss": 0.04146587885916233,
      "grad_norm": 0.14658896625041962,
      "learning_rate": 7.799466644922484e-05,
      "loss": 0.0449,
      "relation_loss": 0.003400260512717068,
      "selected_pairs": 26994.8,
      "state_query_norm": 1034.7438232421875,
      "state_relation_loss": 0.0005004870326956734,
      "step": 10360,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003400260512717068,
      "weighted_relation_loss": 0.003400260512717068
    },
    {
      "action_query_norm": 571.8209533691406,
      "action_relation_loss": 0.0030097147217020394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.70065481758653,
      "flow_matching_loss": 0.03963577300310135,
      "grad_norm": 0.12387625873088837,
      "learning_rate": 7.794898255967602e-05,
      "loss": 0.0432,
      "relation_loss": 0.0035218587843701245,
      "selected_pairs": 26622.7,
      "state_query_norm": 1035.0505249023438,
      "state_relation_loss": 0.0005121440917719156,
      "step": 10370,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0035218587843701245,
      "weighted_relation_loss": 0.0035218587843701245
    },
    {
      "action_query_norm": 569.8168579101563,
      "action_relation_loss": 0.0031355522107332943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.71000935453695,
      "flow_matching_loss": 0.039151962473988534,
      "grad_norm": 0.13303181529045105,
      "learning_rate": 7.790326470947795e-05,
      "loss": 0.0428,
      "relation_loss": 0.0036528289783746004,
      "selected_pairs": 27182.3,
      "state_query_norm": 1022.8426452636719,
      "state_relation_loss": 0.000517276837490499,
      "step": 10380,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0036528289783746004,
      "weighted_relation_loss": 0.0036528289783746004
    },
    {
      "action_query_norm": 567.9633239746094,
      "action_relation_loss": 0.00286658292170614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.71936389148737,
      "flow_matching_loss": 0.03972375988960266,
      "grad_norm": 0.14285244047641754,
      "learning_rate": 7.785751295418217e-05,
      "loss": 0.0431,
      "relation_loss": 0.003369830781593919,
      "selected_pairs": 27687.7,
      "state_query_norm": 1011.2845947265625,
      "state_relation_loss": 0.0005032478336943313,
      "step": 10390,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.003369830781593919,
      "weighted_relation_loss": 0.003369830781593919
    },
    {
      "action_query_norm": 567.9064880371094,
      "action_relation_loss": 0.003096670308150351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.728718428437793,
      "flow_matching_loss": 0.03911119773983955,
      "grad_norm": 0.12458627671003342,
      "learning_rate": 7.781172734938136e-05,
      "loss": 0.0427,
      "relation_loss": 0.003610054380260408,
      "selected_pairs": 26716.3,
      "state_query_norm": 1035.2038940429688,
      "state_relation_loss": 0.0005133840197231621,
      "step": 10400,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.003610054380260408,
      "weighted_relation_loss": 0.003610054380260408
    },
    {
      "action_query_norm": 574.8490844726563,
      "action_relation_loss": 0.0028978212038055064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.738072965388213,
      "flow_matching_loss": 0.038355521485209466,
      "grad_norm": 0.15118414163589478,
      "learning_rate": 7.776590795070933e-05,
      "loss": 0.0418,
      "relation_loss": 0.0034301702864468096,
      "selected_pairs": 26946.6,
      "state_query_norm": 1027.1231750488282,
      "state_relation_loss": 0.000532349111745134,
      "step": 10410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0034301702864468096,
      "weighted_relation_loss": 0.0034301702864468096
    },
    {
      "action_query_norm": 578.398095703125,
      "action_relation_loss": 0.0028384374221786858,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.747427502338635,
      "flow_matching_loss": 0.037809199281036855,
      "grad_norm": 0.09941448271274567,
      "learning_rate": 7.772005481384099e-05,
      "loss": 0.0412,
      "relation_loss": 0.0033570374362170696,
      "selected_pairs": 26877.7,
      "state_query_norm": 1031.9016967773437,
      "state_relation_loss": 0.0005186000053072348,
      "step": 10420,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0033570374362170696,
      "weighted_relation_loss": 0.0033570374362170696
    },
    {
      "action_query_norm": 573.9805236816406,
      "action_relation_loss": 0.002998303901404142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.756782039289055,
      "flow_matching_loss": 0.03847942650318146,
      "grad_norm": 0.11581467092037201,
      "learning_rate": 7.767416799449223e-05,
      "loss": 0.042,
      "relation_loss": 0.0035280103562399744,
      "selected_pairs": 26782.8,
      "state_query_norm": 1017.850146484375,
      "state_relation_loss": 0.0005297064839396626,
      "step": 10430,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0035280103562399744,
      "weighted_relation_loss": 0.0035280103562399744
    },
    {
      "action_query_norm": 572.8916015625,
      "action_relation_loss": 0.0029456124641001223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.766136576239477,
      "flow_matching_loss": 0.038844496197998526,
      "grad_norm": 0.11846497654914856,
      "learning_rate": 7.762824754841984e-05,
      "loss": 0.0423,
      "relation_loss": 0.00346741343382746,
      "selected_pairs": 27200.9,
      "state_query_norm": 1030.9674438476563,
      "state_relation_loss": 0.000521800946444273,
      "step": 10440,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00346741343382746,
      "weighted_relation_loss": 0.00346741343382746
    },
    {
      "action_query_norm": 574.5427673339843,
      "action_relation_loss": 0.0029665195383131505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.775491113189897,
      "flow_matching_loss": 0.03876362144947052,
      "grad_norm": 0.15739421546459198,
      "learning_rate": 7.758229353142152e-05,
      "loss": 0.0422,
      "relation_loss": 0.0034826235380023716,
      "selected_pairs": 27150.5,
      "state_query_norm": 1029.023858642578,
      "state_relation_loss": 0.00051610397931654,
      "step": 10450,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0034826235380023716,
      "weighted_relation_loss": 0.0034826235380023716
    },
    {
      "action_query_norm": 572.18876953125,
      "action_relation_loss": 0.0030184530187398195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.784845650140317,
      "flow_matching_loss": 0.03752944096922874,
      "grad_norm": 0.15097635984420776,
      "learning_rate": 7.753630599933572e-05,
      "loss": 0.041,
      "relation_loss": 0.0035069282865151765,
      "selected_pairs": 26671.7,
      "state_query_norm": 1019.6273315429687,
      "state_relation_loss": 0.0004884752765065059,
      "step": 10460,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0035069282865151765,
      "weighted_relation_loss": 0.0035069282865151765
    },
    {
      "action_query_norm": 570.9335083007812,
      "action_relation_loss": 0.002899697283282876,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.79420018709074,
      "flow_matching_loss": 0.04046626016497612,
      "grad_norm": 0.11892124265432358,
      "learning_rate": 7.749028500804161e-05,
      "loss": 0.0438,
      "relation_loss": 0.0033815880538895725,
      "selected_pairs": 27389.4,
      "state_query_norm": 1046.0404357910156,
      "state_relation_loss": 0.00048189072695095094,
      "step": 10470,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0033815880538895725,
      "weighted_relation_loss": 0.0033815880538895725
    },
    {
      "action_query_norm": 577.3369140625,
      "action_relation_loss": 0.0029709691414609553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.80355472404116,
      "flow_matching_loss": 0.03908422030508518,
      "grad_norm": 0.16965310275554657,
      "learning_rate": 7.744423061345906e-05,
      "loss": 0.0426,
      "relation_loss": 0.0034721751464530824,
      "selected_pairs": 27174.5,
      "state_query_norm": 1033.3534851074219,
      "state_relation_loss": 0.0005012059846194461,
      "step": 10480,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0034721751464530824,
      "weighted_relation_loss": 0.0034721751464530824
    },
    {
      "action_query_norm": 577.5151000976563,
      "action_relation_loss": 0.0028582417173311113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.812909260991582,
      "flow_matching_loss": 0.03952766265720129,
      "grad_norm": 0.17885872721672058,
      "learning_rate": 7.73981428715485e-05,
      "loss": 0.0429,
      "relation_loss": 0.0033693339210003614,
      "selected_pairs": 26873.1,
      "state_query_norm": 1021.995166015625,
      "state_relation_loss": 0.0005110922444146127,
      "step": 10490,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0033693339210003614,
      "weighted_relation_loss": 0.0033693339210003614
    },
    {
      "action_query_norm": 576.7328857421875,
      "action_relation_loss": 0.002993247820995748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.822263797942002,
      "flow_matching_loss": 0.04005085490643978,
      "grad_norm": 0.20027944445610046,
      "learning_rate": 7.735202183831085e-05,
      "loss": 0.0436,
      "relation_loss": 0.00351250518579036,
      "selected_pairs": 27359.2,
      "state_query_norm": 1042.66904296875,
      "state_relation_loss": 0.0005192573211388662,
      "step": 10500,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.00351250518579036,
      "weighted_relation_loss": 0.00351250518579036
    },
    {
      "action_query_norm": 572.2126159667969,
      "action_relation_loss": 0.0030095010763034223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.831618334892422,
      "flow_matching_loss": 0.03864400889724493,
      "grad_norm": 0.18439587950706482,
      "learning_rate": 7.730586756978757e-05,
      "loss": 0.0422,
      "relation_loss": 0.0035097306594252587,
      "selected_pairs": 26695.9,
      "state_query_norm": 1043.0636901855469,
      "state_relation_loss": 0.0005002295976737515,
      "step": 10510,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0035097306594252587,
      "weighted_relation_loss": 0.0035097306594252587
    },
    {
      "action_query_norm": 573.2513427734375,
      "action_relation_loss": 0.0029195779468864204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.840972871842844,
      "flow_matching_loss": 0.03702600784599781,
      "grad_norm": 0.1582891047000885,
      "learning_rate": 7.72596801220604e-05,
      "loss": 0.0405,
      "relation_loss": 0.003453007712960243,
      "selected_pairs": 27227.3,
      "state_query_norm": 1034.4985595703124,
      "state_relation_loss": 0.0005334297515219077,
      "step": 10520,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.003453007712960243,
      "weighted_relation_loss": 0.003453007712960243
    },
    {
      "action_query_norm": 578.094189453125,
      "action_relation_loss": 0.0028008098481222985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.850327408793264,
      "flow_matching_loss": 0.03886319622397423,
      "grad_norm": 0.16664007306098938,
      "learning_rate": 7.721345955125147e-05,
      "loss": 0.0422,
      "relation_loss": 0.0033113542478531597,
      "selected_pairs": 27309.7,
      "state_query_norm": 1032.9440795898438,
      "state_relation_loss": 0.0005105443677166476,
      "step": 10530,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0033113542478531597,
      "weighted_relation_loss": 0.0033113542478531597
    },
    {
      "action_query_norm": 578.7240661621094,
      "action_relation_loss": 0.002922901138663292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.859681945743686,
      "flow_matching_loss": 0.036818161606788635,
      "grad_norm": 0.1526833027601242,
      "learning_rate": 7.71672059135231e-05,
      "loss": 0.0403,
      "relation_loss": 0.0034482138464227317,
      "selected_pairs": 26618.8,
      "state_query_norm": 1029.9028686523438,
      "state_relation_loss": 0.0005253127368632704,
      "step": 10540,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0034482138464227317,
      "weighted_relation_loss": 0.0034482138464227317
    },
    {
      "action_query_norm": 580.5031005859375,
      "action_relation_loss": 0.0028576386626809835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.869036482694106,
      "flow_matching_loss": 0.04093148037791252,
      "grad_norm": 0.11050985753536224,
      "learning_rate": 7.712091926507787e-05,
      "loss": 0.0443,
      "relation_loss": 0.003363448707386851,
      "selected_pairs": 27201.6,
      "state_query_norm": 1039.0513122558593,
      "state_relation_loss": 0.0005058100417954847,
      "step": 10550,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.003363448707386851,
      "weighted_relation_loss": 0.003363448707386851
    },
    {
      "action_query_norm": 585.5761352539063,
      "action_relation_loss": 0.002719356142915785,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.878391019644528,
      "flow_matching_loss": 0.037877120077610016,
      "grad_norm": 0.1328313648700714,
      "learning_rate": 7.70745996621584e-05,
      "loss": 0.0411,
      "relation_loss": 0.003245121333748102,
      "selected_pairs": 26745.2,
      "state_query_norm": 1026.5874328613281,
      "state_relation_loss": 0.0005257651704596355,
      "step": 10560,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003245121333748102,
      "weighted_relation_loss": 0.003245121333748102
    },
    {
      "action_query_norm": 573.6903137207031,
      "action_relation_loss": 0.0029166545951738953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.887745556594949,
      "flow_matching_loss": 0.03882316574454307,
      "grad_norm": 0.14476077258586884,
      "learning_rate": 7.702824716104735e-05,
      "loss": 0.0423,
      "relation_loss": 0.0034308719681575895,
      "selected_pairs": 26614.4,
      "state_query_norm": 1047.8802856445313,
      "state_relation_loss": 0.0005142173555213958,
      "step": 10570,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0034308719681575895,
      "weighted_relation_loss": 0.0034308719681575895
    },
    {
      "action_query_norm": 573.3966796875,
      "action_relation_loss": 0.0029083315283060076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.897100093545369,
      "flow_matching_loss": 0.0397309709340334,
      "grad_norm": 0.17183610796928406,
      "learning_rate": 7.698186181806743e-05,
      "loss": 0.0432,
      "relation_loss": 0.003424932691268623,
      "selected_pairs": 26298.9,
      "state_query_norm": 1040.1437377929688,
      "state_relation_loss": 0.0005166011629626155,
      "step": 10580,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.003424932691268623,
      "weighted_relation_loss": 0.003424932691268623
    },
    {
      "action_query_norm": 570.7764526367188,
      "action_relation_loss": 0.0031607785262167455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.90645463049579,
      "flow_matching_loss": 0.040141834318637847,
      "grad_norm": 0.12871664762496948,
      "learning_rate": 7.693544368958116e-05,
      "loss": 0.0438,
      "relation_loss": 0.003687571594491601,
      "selected_pairs": 26553.1,
      "state_query_norm": 1041.9046142578125,
      "state_relation_loss": 0.0005267930624540895,
      "step": 10590,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003687571594491601,
      "weighted_relation_loss": 0.003687571594491601
    },
    {
      "action_query_norm": 571.4447631835938,
      "action_relation_loss": 0.002950307144783437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.915809167446211,
      "flow_matching_loss": 0.03976629078388214,
      "grad_norm": 0.11760960519313812,
      "learning_rate": 7.688899283199096e-05,
      "loss": 0.0432,
      "relation_loss": 0.0034764314303174617,
      "selected_pairs": 26923.4,
      "state_query_norm": 1034.2349365234375,
      "state_relation_loss": 0.0005261242768028751,
      "step": 10600,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0034764314303174617,
      "weighted_relation_loss": 0.0034764314303174617
    },
    {
      "action_query_norm": 577.0715515136719,
      "action_relation_loss": 0.0028633818961679935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.925163704396633,
      "flow_matching_loss": 0.03934662938117981,
      "grad_norm": 0.15818284451961517,
      "learning_rate": 7.684250930173901e-05,
      "loss": 0.0427,
      "relation_loss": 0.003398223058320582,
      "selected_pairs": 26340.3,
      "state_query_norm": 1041.76865234375,
      "state_relation_loss": 0.0005348411446902901,
      "step": 10610,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.003398223058320582,
      "weighted_relation_loss": 0.003398223058320582
    },
    {
      "action_query_norm": 577.3201599121094,
      "action_relation_loss": 0.0028928406070917847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.934518241347053,
      "flow_matching_loss": 0.040355789661407473,
      "grad_norm": 0.120877206325531,
      "learning_rate": 7.679599315530718e-05,
      "loss": 0.0438,
      "relation_loss": 0.0034176698653027417,
      "selected_pairs": 26771.3,
      "state_query_norm": 1058.9779174804687,
      "state_relation_loss": 0.0005248292465694249,
      "step": 10620,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0034176698653027417,
      "weighted_relation_loss": 0.0034176698653027417
    },
    {
      "action_query_norm": 568.7537109375,
      "action_relation_loss": 0.0028550802962854505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.943872778297475,
      "flow_matching_loss": 0.04080682583153248,
      "grad_norm": 0.12081024795770645,
      "learning_rate": 7.674944444921695e-05,
      "loss": 0.0442,
      "relation_loss": 0.0033784123370423913,
      "selected_pairs": 26344.7,
      "state_query_norm": 1035.2439331054688,
      "state_relation_loss": 0.000523332052398473,
      "step": 10630,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0033784123370423913,
      "weighted_relation_loss": 0.0033784123370423913
    },
    {
      "action_query_norm": 580.0529357910157,
      "action_relation_loss": 0.0027627511881291866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.953227315247895,
      "flow_matching_loss": 0.037575660087168214,
      "grad_norm": 0.1451549381017685,
      "learning_rate": 7.670286324002944e-05,
      "loss": 0.0408,
      "relation_loss": 0.0032697264570742844,
      "selected_pairs": 27124.6,
      "state_query_norm": 1030.6995056152343,
      "state_relation_loss": 0.0005069753126008437,
      "step": 10640,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0032697264570742844,
      "weighted_relation_loss": 0.0032697264570742844
    },
    {
      "action_query_norm": 574.9779479980468,
      "action_relation_loss": 0.002915973449125886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.962581852198316,
      "flow_matching_loss": 0.03955503441393375,
      "grad_norm": 0.1250634342432022,
      "learning_rate": 7.665624958434514e-05,
      "loss": 0.043,
      "relation_loss": 0.003434453345835209,
      "selected_pairs": 26850.8,
      "state_query_norm": 1058.73701171875,
      "state_relation_loss": 0.0005184799229027703,
      "step": 10650,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003434453345835209,
      "weighted_relation_loss": 0.003434453345835209
    },
    {
      "action_query_norm": 584.4838623046875,
      "action_relation_loss": 0.0029210567474365233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.971936389148738,
      "flow_matching_loss": 0.040140556171536446,
      "grad_norm": 0.1509746015071869,
      "learning_rate": 7.66096035388041e-05,
      "loss": 0.0436,
      "relation_loss": 0.003432206157594919,
      "selected_pairs": 27058.9,
      "state_query_norm": 1046.3776000976563,
      "state_relation_loss": 0.0005111494363518431,
      "step": 10660,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003432206157594919,
      "weighted_relation_loss": 0.003432206157594919
    },
    {
      "action_query_norm": 573.5697326660156,
      "action_relation_loss": 0.003021849971264601,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.981290926099158,
      "flow_matching_loss": 0.03764980435371399,
      "grad_norm": 0.12715943157672882,
      "learning_rate": 7.656292516008563e-05,
      "loss": 0.0412,
      "relation_loss": 0.0035462477942928674,
      "selected_pairs": 25962.6,
      "state_query_norm": 1049.74208984375,
      "state_relation_loss": 0.0005243978288490325,
      "step": 10670,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.0035462477942928674,
      "weighted_relation_loss": 0.0035462477942928674
    },
    {
      "action_query_norm": 584.258056640625,
      "action_relation_loss": 0.0028857015538960697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.99064546304958,
      "flow_matching_loss": 0.040479031205177304,
      "grad_norm": 0.13331079483032227,
      "learning_rate": 7.651621450490837e-05,
      "loss": 0.0439,
      "relation_loss": 0.0033824668033048512,
      "selected_pairs": 27103.3,
      "state_query_norm": 1038.6276794433593,
      "state_relation_loss": 0.0004967652552295476,
      "step": 10680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0033824668033048512,
      "weighted_relation_loss": 0.0033824668033048512
    },
    {
      "action_query_norm": 587.2194458007813,
      "action_relation_loss": 0.002964497054927051,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.0,
      "flow_matching_loss": 0.03923434279859066,
      "grad_norm": 0.20457729697227478,
      "learning_rate": 7.646947163003017e-05,
      "loss": 0.0427,
      "relation_loss": 0.003469493403099477,
      "selected_pairs": 23529.8,
      "state_query_norm": 1039.961199951172,
      "state_relation_loss": 0.0005049963685451076,
      "step": 10690,
      "valid_action_samples": 210.5,
      "weighted_auxiliary_loss": 0.003469493403099477,
      "weighted_relation_loss": 0.003469493403099477
    },
    {
      "action_query_norm": 586.7861145019531,
      "action_relation_loss": 0.0026766726514324546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.00935453695042,
      "flow_matching_loss": 0.03564930558204651,
      "grad_norm": 0.11019870638847351,
      "learning_rate": 7.642269659224803e-05,
      "loss": 0.0388,
      "relation_loss": 0.0031853951513767242,
      "selected_pairs": 27135.3,
      "state_query_norm": 1046.695068359375,
      "state_relation_loss": 0.000508722517406568,
      "step": 10700,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0031853951513767242,
      "weighted_relation_loss": 0.0031853951513767242
    },
    {
      "action_query_norm": 585.5565673828125,
      "action_relation_loss": 0.0029123608488589524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.018709073900842,
      "flow_matching_loss": 0.03730071019381285,
      "grad_norm": 0.15744152665138245,
      "learning_rate": 7.637588944839803e-05,
      "loss": 0.0407,
      "relation_loss": 0.0034276835387572644,
      "selected_pairs": 26921.8,
      "state_query_norm": 1041.8314575195313,
      "state_relation_loss": 0.0005153227044502273,
      "step": 10710,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0034276835387572644,
      "weighted_relation_loss": 0.0034276835387572644
    },
    {
      "action_query_norm": 583.4116394042969,
      "action_relation_loss": 0.00281402007676661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.028063610851262,
      "flow_matching_loss": 0.03713334053754806,
      "grad_norm": 0.15258337557315826,
      "learning_rate": 7.632905025535528e-05,
      "loss": 0.0405,
      "relation_loss": 0.0033205282175913452,
      "selected_pairs": 26046.4,
      "state_query_norm": 1035.4375122070312,
      "state_relation_loss": 0.0005065081408247351,
      "step": 10720,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0033205282175913452,
      "weighted_relation_loss": 0.0033205282175913452
    },
    {
      "action_query_norm": 588.7040710449219,
      "action_relation_loss": 0.002841329015791416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.037418147801684,
      "flow_matching_loss": 0.03681127279996872,
      "grad_norm": 0.1133296936750412,
      "learning_rate": 7.628217907003378e-05,
      "loss": 0.0401,
      "relation_loss": 0.003336361772380769,
      "selected_pairs": 26847.7,
      "state_query_norm": 1034.4014404296875,
      "state_relation_loss": 0.0004950327711412683,
      "step": 10730,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003336361772380769,
      "weighted_relation_loss": 0.003336361772380769
    },
    {
      "action_query_norm": 588.2967529296875,
      "action_relation_loss": 0.0027347343508154153,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.046772684752105,
      "flow_matching_loss": 0.038078172504901885,
      "grad_norm": 0.11054793000221252,
      "learning_rate": 7.623527594938649e-05,
      "loss": 0.0413,
      "relation_loss": 0.0032014291267842055,
      "selected_pairs": 26786.3,
      "state_query_norm": 1042.5593872070312,
      "state_relation_loss": 0.0004666947439545766,
      "step": 10740,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0032014291267842055,
      "weighted_relation_loss": 0.0032014291267842055
    },
    {
      "action_query_norm": 582.9651733398438,
      "action_relation_loss": 0.00277303212787956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.056127221702527,
      "flow_matching_loss": 0.03657566159963608,
      "grad_norm": 0.12084712833166122,
      "learning_rate": 7.618834095040509e-05,
      "loss": 0.0398,
      "relation_loss": 0.003270865371450782,
      "selected_pairs": 26896.0,
      "state_query_norm": 1035.4467224121095,
      "state_relation_loss": 0.0004978332319296897,
      "step": 10750,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.003270865371450782,
      "weighted_relation_loss": 0.003270865371450782
    },
    {
      "action_query_norm": 585.2216125488281,
      "action_relation_loss": 0.0027343228459358214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.065481758652947,
      "flow_matching_loss": 0.03607484027743339,
      "grad_norm": 0.1104847639799118,
      "learning_rate": 7.614137413012e-05,
      "loss": 0.0393,
      "relation_loss": 0.003243199805729091,
      "selected_pairs": 27458.3,
      "state_query_norm": 1025.882794189453,
      "state_relation_loss": 0.0005088769568828866,
      "step": 10760,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.003243199805729091,
      "weighted_relation_loss": 0.003243199805729091
    },
    {
      "action_query_norm": 588.9391967773438,
      "action_relation_loss": 0.0027952064760029317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.074836295603367,
      "flow_matching_loss": 0.03966236114501953,
      "grad_norm": 0.12421277165412903,
      "learning_rate": 7.609437554560042e-05,
      "loss": 0.0429,
      "relation_loss": 0.0032866350840777157,
      "selected_pairs": 26650.3,
      "state_query_norm": 1035.1674926757812,
      "state_relation_loss": 0.0004914285673294216,
      "step": 10770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0032866350840777157,
      "weighted_relation_loss": 0.0032866350840777157
    },
    {
      "action_query_norm": 581.505712890625,
      "action_relation_loss": 0.002710125409066677,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.084190832553789,
      "flow_matching_loss": 0.036466007307171824,
      "grad_norm": 0.1137770414352417,
      "learning_rate": 7.604734525395398e-05,
      "loss": 0.0397,
      "relation_loss": 0.003212259500287473,
      "selected_pairs": 26591.3,
      "state_query_norm": 1037.4117431640625,
      "state_relation_loss": 0.0005021340330131352,
      "step": 10780,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.003212259500287473,
      "weighted_relation_loss": 0.003212259500287473
    },
    {
      "action_query_norm": 585.4696105957031,
      "action_relation_loss": 0.0027573848608881237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.09354536950421,
      "flow_matching_loss": 0.03689358793199062,
      "grad_norm": 0.10653228312730789,
      "learning_rate": 7.600028331232699e-05,
      "loss": 0.0401,
      "relation_loss": 0.003252403810620308,
      "selected_pairs": 26884.9,
      "state_query_norm": 1036.7572814941407,
      "state_relation_loss": 0.0004950189497321845,
      "step": 10790,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.003252403810620308,
      "weighted_relation_loss": 0.003252403810620308
    },
    {
      "action_query_norm": 578.3841369628906,
      "action_relation_loss": 0.0028409695718437434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.102899906454631,
      "flow_matching_loss": 0.040407586842775345,
      "grad_norm": 0.11423677951097488,
      "learning_rate": 7.595318977790408e-05,
      "loss": 0.0438,
      "relation_loss": 0.0033453973475843666,
      "selected_pairs": 27106.9,
      "state_query_norm": 1037.048272705078,
      "state_relation_loss": 0.00050442774372641,
      "step": 10800,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0033453973475843666,
      "weighted_relation_loss": 0.0033453973475843666
    },
    {
      "action_query_norm": 582.66171875,
      "action_relation_loss": 0.002778580063022673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.112254443405051,
      "flow_matching_loss": 0.03655520137399435,
      "grad_norm": 0.12528498470783234,
      "learning_rate": 7.590606470790836e-05,
      "loss": 0.0398,
      "relation_loss": 0.003275648131966591,
      "selected_pairs": 26969.6,
      "state_query_norm": 1030.4581787109375,
      "state_relation_loss": 0.0004970680456608534,
      "step": 10810,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003275648131966591,
      "weighted_relation_loss": 0.003275648131966591
    },
    {
      "action_query_norm": 597.58203125,
      "action_relation_loss": 0.002692090719938278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.121608980355472,
      "flow_matching_loss": 0.037885862216353414,
      "grad_norm": 0.1348525434732437,
      "learning_rate": 7.585890815960125e-05,
      "loss": 0.0411,
      "relation_loss": 0.003210669523105025,
      "selected_pairs": 27032.6,
      "state_query_norm": 1038.5738159179687,
      "state_relation_loss": 0.0005185788118978963,
      "step": 10820,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003210669523105025,
      "weighted_relation_loss": 0.003210669523105025
    },
    {
      "action_query_norm": 588.8954833984375,
      "action_relation_loss": 0.002723343879915774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.130963517305894,
      "flow_matching_loss": 0.03583159111440182,
      "grad_norm": 0.13521184027194977,
      "learning_rate": 7.581172019028238e-05,
      "loss": 0.0391,
      "relation_loss": 0.0032243737718090416,
      "selected_pairs": 27123.2,
      "state_query_norm": 1035.7806091308594,
      "state_relation_loss": 0.000501029915176332,
      "step": 10830,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0032243737718090416,
      "weighted_relation_loss": 0.0032243737718090416
    },
    {
      "action_query_norm": 579.6425354003907,
      "action_relation_loss": 0.002750232769176364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.140318054256314,
      "flow_matching_loss": 0.03874000236392021,
      "grad_norm": 0.1539447009563446,
      "learning_rate": 7.576450085728959e-05,
      "loss": 0.042,
      "relation_loss": 0.0032542522298172117,
      "selected_pairs": 27161.6,
      "state_query_norm": 1050.6592590332032,
      "state_relation_loss": 0.0005040194722823798,
      "step": 10840,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0032542522298172117,
      "weighted_relation_loss": 0.0032542522298172117
    },
    {
      "action_query_norm": 583.66650390625,
      "action_relation_loss": 0.002677595615386963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.149672591206736,
      "flow_matching_loss": 0.03758749812841415,
      "grad_norm": 0.13491035997867584,
      "learning_rate": 7.571725021799885e-05,
      "loss": 0.0408,
      "relation_loss": 0.0031657660845667126,
      "selected_pairs": 27010.4,
      "state_query_norm": 1030.7497802734374,
      "state_relation_loss": 0.0004881704691797495,
      "step": 10850,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0031657660845667126,
      "weighted_relation_loss": 0.0031657660845667126
    },
    {
      "action_query_norm": 580.7268920898438,
      "action_relation_loss": 0.002603101497516036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.159027128157156,
      "flow_matching_loss": 0.03686913065612316,
      "grad_norm": 0.13216601312160492,
      "learning_rate": 7.566996832982409e-05,
      "loss": 0.04,
      "relation_loss": 0.003111265250481665,
      "selected_pairs": 26663.3,
      "state_query_norm": 1036.6094909667968,
      "state_relation_loss": 0.000508163747144863,
      "step": 10860,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003111265250481665,
      "weighted_relation_loss": 0.003111265250481665
    },
    {
      "action_query_norm": 574.8493469238281,
      "action_relation_loss": 0.002696357434615493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.168381665107578,
      "flow_matching_loss": 0.03450994156301022,
      "grad_norm": 0.10641450434923172,
      "learning_rate": 7.56226552502173e-05,
      "loss": 0.0377,
      "relation_loss": 0.003188563697040081,
      "selected_pairs": 26916.8,
      "state_query_norm": 1049.3733642578125,
      "state_relation_loss": 0.0004922062245896087,
      "step": 10870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.003188563697040081,
      "weighted_relation_loss": 0.003188563697040081
    },
    {
      "action_query_norm": 579.1943725585937,
      "action_relation_loss": 0.002758481982164085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.177736202057998,
      "flow_matching_loss": 0.03655639216303826,
      "grad_norm": 0.10709607601165771,
      "learning_rate": 7.557531103666833e-05,
      "loss": 0.0398,
      "relation_loss": 0.0032741726841777562,
      "selected_pairs": 26753.4,
      "state_query_norm": 1036.8956970214845,
      "state_relation_loss": 0.0005156907165655867,
      "step": 10880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0032741726841777562,
      "weighted_relation_loss": 0.0032741726841777562
    },
    {
      "action_query_norm": 583.4462158203125,
      "action_relation_loss": 0.0027253355598077177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.187090739008418,
      "flow_matching_loss": 0.037668995559215546,
      "grad_norm": 0.09651674330234528,
      "learning_rate": 7.552793574670485e-05,
      "loss": 0.0409,
      "relation_loss": 0.003197696269489825,
      "selected_pairs": 27129.9,
      "state_query_norm": 1061.3956665039063,
      "state_relation_loss": 0.00047236070677172395,
      "step": 10890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.003197696269489825,
      "weighted_relation_loss": 0.003197696269489825
    },
    {
      "action_query_norm": 580.8753845214844,
      "action_relation_loss": 0.0026735124411061406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.19644527595884,
      "flow_matching_loss": 0.03948240205645561,
      "grad_norm": 0.13531629741191864,
      "learning_rate": 7.54805294378923e-05,
      "loss": 0.0426,
      "relation_loss": 0.003159721894189715,
      "selected_pairs": 27207.3,
      "state_query_norm": 1051.3353332519532,
      "state_relation_loss": 0.0004862094618147239,
      "step": 10900,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.003159721894189715,
      "weighted_relation_loss": 0.003159721894189715
    },
    {
      "action_query_norm": 572.4536560058593,
      "action_relation_loss": 0.002811158378608525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.20579981290926,
      "flow_matching_loss": 0.03891429528594017,
      "grad_norm": 0.12286977469921112,
      "learning_rate": 7.543309216783384e-05,
      "loss": 0.0422,
      "relation_loss": 0.0033168056281283496,
      "selected_pairs": 27387.4,
      "state_query_norm": 1060.329443359375,
      "state_relation_loss": 0.0005056472378782928,
      "step": 10910,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0033168056281283496,
      "weighted_relation_loss": 0.0033168056281283496
    },
    {
      "action_query_norm": 586.7894592285156,
      "action_relation_loss": 0.0026474435115233065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.215154349859683,
      "flow_matching_loss": 0.03678808100521565,
      "grad_norm": 0.11270102113485336,
      "learning_rate": 7.53856239941702e-05,
      "loss": 0.0399,
      "relation_loss": 0.003141956822946668,
      "selected_pairs": 26578.0,
      "state_query_norm": 1051.0745239257812,
      "state_relation_loss": 0.0004945132590364665,
      "step": 10920,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.003141956822946668,
      "weighted_relation_loss": 0.003141956822946668
    },
    {
      "action_query_norm": 592.5041320800781,
      "action_relation_loss": 0.0026128563797101377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.224508886810103,
      "flow_matching_loss": 0.036106215044856074,
      "grad_norm": 0.0926932618021965,
      "learning_rate": 7.533812497457972e-05,
      "loss": 0.0392,
      "relation_loss": 0.0031091361539438366,
      "selected_pairs": 26759.7,
      "state_query_norm": 1053.6265991210937,
      "state_relation_loss": 0.0004962797771440819,
      "step": 10930,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0031091361539438366,
      "weighted_relation_loss": 0.0031091361539438366
    },
    {
      "action_query_norm": 587.22626953125,
      "action_relation_loss": 0.0027054297272115944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.233863423760523,
      "flow_matching_loss": 0.0371135700494051,
      "grad_norm": 0.1027621254324913,
      "learning_rate": 7.529059516677814e-05,
      "loss": 0.0403,
      "relation_loss": 0.003206475032493472,
      "selected_pairs": 27034.5,
      "state_query_norm": 1047.1380615234375,
      "state_relation_loss": 0.0005010453081922605,
      "step": 10940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003206475032493472,
      "weighted_relation_loss": 0.003206475032493472
    },
    {
      "action_query_norm": 589.3006713867187,
      "action_relation_loss": 0.0026578346733003855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.243217960710945,
      "flow_matching_loss": 0.038081304356455806,
      "grad_norm": 0.10450845956802368,
      "learning_rate": 7.524303462851872e-05,
      "loss": 0.0412,
      "relation_loss": 0.0031505549093708398,
      "selected_pairs": 27127.5,
      "state_query_norm": 1049.5824829101562,
      "state_relation_loss": 0.0004927202418912203,
      "step": 10950,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0031505549093708398,
      "weighted_relation_loss": 0.0031505549093708398
    },
    {
      "action_query_norm": 587.2358154296875,
      "action_relation_loss": 0.00264399447478354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.252572497661365,
      "flow_matching_loss": 0.038633637689054014,
      "grad_norm": 0.13211247324943542,
      "learning_rate": 7.519544341759192e-05,
      "loss": 0.0418,
      "relation_loss": 0.003137459349818528,
      "selected_pairs": 27010.5,
      "state_query_norm": 1058.4788391113282,
      "state_relation_loss": 0.0004934648895869032,
      "step": 10960,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003137459349818528,
      "weighted_relation_loss": 0.003137459349818528
    },
    {
      "action_query_norm": 581.6018798828125,
      "action_relation_loss": 0.0026676300913095474,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.261927034611787,
      "flow_matching_loss": 0.038416876643896106,
      "grad_norm": 0.11652672290802002,
      "learning_rate": 7.514782159182562e-05,
      "loss": 0.0416,
      "relation_loss": 0.003168505569919944,
      "selected_pairs": 26504.1,
      "state_query_norm": 1045.4509887695312,
      "state_relation_loss": 0.0005008754407754168,
      "step": 10970,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.003168505569919944,
      "weighted_relation_loss": 0.003168505569919944
    },
    {
      "action_query_norm": 582.4691467285156,
      "action_relation_loss": 0.002751353173516691,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.271281571562207,
      "flow_matching_loss": 0.03895521946251392,
      "grad_norm": 0.1440751999616623,
      "learning_rate": 7.510016920908481e-05,
      "loss": 0.0422,
      "relation_loss": 0.0032454315572977067,
      "selected_pairs": 27034.4,
      "state_query_norm": 1057.1408325195312,
      "state_relation_loss": 0.0004940784128848463,
      "step": 10980,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0032454315572977067,
      "weighted_relation_loss": 0.0032454315572977067
    },
    {
      "action_query_norm": 582.5723571777344,
      "action_relation_loss": 0.002532741613686085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.28063610851263,
      "flow_matching_loss": 0.03694958817213774,
      "grad_norm": 0.14644689857959747,
      "learning_rate": 7.505248632727158e-05,
      "loss": 0.04,
      "relation_loss": 0.003047881764359772,
      "selected_pairs": 26257.4,
      "state_query_norm": 1060.359033203125,
      "state_relation_loss": 0.0005151401710463688,
      "step": 10990,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.003047881764359772,
      "weighted_relation_loss": 0.003047881764359772
    },
    {
      "action_query_norm": 580.3922241210937,
      "action_relation_loss": 0.0027310092467814685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.28999064546305,
      "flow_matching_loss": 0.03832645788788795,
      "grad_norm": 0.10622383654117584,
      "learning_rate": 7.50047730043252e-05,
      "loss": 0.0416,
      "relation_loss": 0.0032577496953308583,
      "selected_pairs": 26094.0,
      "state_query_norm": 1050.5703247070312,
      "state_relation_loss": 0.0005267404456390068,
      "step": 11000,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.0032577496953308583,
      "weighted_relation_loss": 0.0032577496953308583
    },
    {
      "action_query_norm": 587.675927734375,
      "action_relation_loss": 0.002684104023501277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.29934518241347,
      "flow_matching_loss": 0.036725346744060514,
      "grad_norm": 0.11181418597698212,
      "learning_rate": 7.495702929822183e-05,
      "loss": 0.0399,
      "relation_loss": 0.0031871391460299493,
      "selected_pairs": 26578.5,
      "state_query_norm": 1063.5290649414062,
      "state_relation_loss": 0.0005030350759625435,
      "step": 11010,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0031871391460299493,
      "weighted_relation_loss": 0.0031871391460299493
    },
    {
      "action_query_norm": 590.6367370605469,
      "action_relation_loss": 0.0027390907518565653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.308699719363892,
      "flow_matching_loss": 0.035589095391333106,
      "grad_norm": 0.10279650241136551,
      "learning_rate": 7.490925526697455e-05,
      "loss": 0.0388,
      "relation_loss": 0.0032324263593181968,
      "selected_pairs": 27011.2,
      "state_query_norm": 1059.5292724609376,
      "state_relation_loss": 0.0004933356161927804,
      "step": 11020,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0032324263593181968,
      "weighted_relation_loss": 0.0032324263593181968
    },
    {
      "action_query_norm": 586.4481262207031,
      "action_relation_loss": 0.002649734169244766,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.318054256314312,
      "flow_matching_loss": 0.038209404051303866,
      "grad_norm": 0.12120696157217026,
      "learning_rate": 7.486145096863334e-05,
      "loss": 0.0414,
      "relation_loss": 0.0031462957384064795,
      "selected_pairs": 27032.0,
      "state_query_norm": 1050.4921997070312,
      "state_relation_loss": 0.0004965616099070758,
      "step": 11030,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0031462957384064795,
      "weighted_relation_loss": 0.0031462957384064795
    },
    {
      "action_query_norm": 584.039306640625,
      "action_relation_loss": 0.0026501607848331333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.327408793264734,
      "flow_matching_loss": 0.0377998985350132,
      "grad_norm": 0.10279949754476547,
      "learning_rate": 7.481361646128491e-05,
      "loss": 0.0409,
      "relation_loss": 0.003144199424423277,
      "selected_pairs": 26665.8,
      "state_query_norm": 1044.5741088867187,
      "state_relation_loss": 0.0004940386512316764,
      "step": 11040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003144199424423277,
      "weighted_relation_loss": 0.003144199424423277
    },
    {
      "action_query_norm": 584.7494384765625,
      "action_relation_loss": 0.002581495442427695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.336763330215154,
      "flow_matching_loss": 0.03988162279129028,
      "grad_norm": 0.1190066933631897,
      "learning_rate": 7.476575180305271e-05,
      "loss": 0.0429,
      "relation_loss": 0.0030662832548841836,
      "selected_pairs": 27012.6,
      "state_query_norm": 1056.3375366210937,
      "state_relation_loss": 0.0004847878066357225,
      "step": 11050,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0030662832548841836,
      "weighted_relation_loss": 0.0030662832548841836
    },
    {
      "action_query_norm": 586.8612487792968,
      "action_relation_loss": 0.0027171278838068248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.346117867165574,
      "flow_matching_loss": 0.03620995953679085,
      "grad_norm": 0.14839765429496765,
      "learning_rate": 7.471785705209682e-05,
      "loss": 0.0394,
      "relation_loss": 0.003197473846375942,
      "selected_pairs": 26667.4,
      "state_query_norm": 1053.20771484375,
      "state_relation_loss": 0.00048034593055490404,
      "step": 11060,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003197473846375942,
      "weighted_relation_loss": 0.003197473846375942
    },
    {
      "action_query_norm": 577.2212463378906,
      "action_relation_loss": 0.002704055421054363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.355472404115996,
      "flow_matching_loss": 0.03970691245049238,
      "grad_norm": 0.14348705112934113,
      "learning_rate": 7.466993226661387e-05,
      "loss": 0.0429,
      "relation_loss": 0.0031905506737530233,
      "selected_pairs": 26899.9,
      "state_query_norm": 1061.2945922851563,
      "state_relation_loss": 0.0004864952847128734,
      "step": 11070,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0031905506737530233,
      "weighted_relation_loss": 0.0031905506737530233
    },
    {
      "action_query_norm": 588.8195190429688,
      "action_relation_loss": 0.0026418758323416116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.364826941066417,
      "flow_matching_loss": 0.037965621799230576,
      "grad_norm": 0.15257766842842102,
      "learning_rate": 7.462197750483699e-05,
      "loss": 0.0411,
      "relation_loss": 0.0031197125324979426,
      "selected_pairs": 27222.4,
      "state_query_norm": 1058.6712768554687,
      "state_relation_loss": 0.00047783668560441583,
      "step": 11080,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0031197125324979426,
      "weighted_relation_loss": 0.0031197125324979426
    },
    {
      "action_query_norm": 585.4237915039063,
      "action_relation_loss": 0.0025876139523461463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.374181478016839,
      "flow_matching_loss": 0.03530840519815683,
      "grad_norm": 0.11600776761770248,
      "learning_rate": 7.457399282503574e-05,
      "loss": 0.0384,
      "relation_loss": 0.0030827252892777325,
      "selected_pairs": 27004.6,
      "state_query_norm": 1034.3173217773438,
      "state_relation_loss": 0.0004951113660354167,
      "step": 11090,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0030827252892777325,
      "weighted_relation_loss": 0.0030827252892777325
    },
    {
      "action_query_norm": 589.0010681152344,
      "action_relation_loss": 0.0025619819760322573,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.383536014967259,
      "flow_matching_loss": 0.03916010111570358,
      "grad_norm": 0.14696431159973145,
      "learning_rate": 7.452597828551604e-05,
      "loss": 0.0422,
      "relation_loss": 0.003045777464285493,
      "selected_pairs": 27295.1,
      "state_query_norm": 1041.698260498047,
      "state_relation_loss": 0.0004837955260882154,
      "step": 11100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.003045777464285493,
      "weighted_relation_loss": 0.003045777464285493
    },
    {
      "action_query_norm": 585.7789733886718,
      "action_relation_loss": 0.0025529991136863826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.39289055191768,
      "flow_matching_loss": 0.03430365491658449,
      "grad_norm": 0.12975619733333588,
      "learning_rate": 7.447793394462006e-05,
      "loss": 0.0373,
      "relation_loss": 0.0030339936260133983,
      "selected_pairs": 26667.7,
      "state_query_norm": 1058.2552001953125,
      "state_relation_loss": 0.0004809945181477815,
      "step": 11110,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030339936260133983,
      "weighted_relation_loss": 0.0030339936260133983
    },
    {
      "action_query_norm": 583.7252746582031,
      "action_relation_loss": 0.002733589126728475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.402245088868101,
      "flow_matching_loss": 0.0381065059453249,
      "grad_norm": 0.09613016247749329,
      "learning_rate": 7.442985986072624e-05,
      "loss": 0.0413,
      "relation_loss": 0.0032416863599792124,
      "selected_pairs": 27247.6,
      "state_query_norm": 1055.2953063964844,
      "state_relation_loss": 0.0005080972478026524,
      "step": 11120,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0032416863599792124,
      "weighted_relation_loss": 0.0032416863599792124
    },
    {
      "action_query_norm": 582.7498474121094,
      "action_relation_loss": 0.0025532698491588235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.411599625818521,
      "flow_matching_loss": 0.03509034067392349,
      "grad_norm": 0.1333370804786682,
      "learning_rate": 7.438175609224908e-05,
      "loss": 0.0381,
      "relation_loss": 0.0030385993188247083,
      "selected_pairs": 26073.2,
      "state_query_norm": 1060.1438354492188,
      "state_relation_loss": 0.0004853294754866511,
      "step": 11130,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.0030385993188247083,
      "weighted_relation_loss": 0.0030385993188247083
    },
    {
      "action_query_norm": 594.3569274902344,
      "action_relation_loss": 0.002655711374245584,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.420954162768943,
      "flow_matching_loss": 0.03848962262272835,
      "grad_norm": 0.15950678288936615,
      "learning_rate": 7.433362269763924e-05,
      "loss": 0.0416,
      "relation_loss": 0.003126668161712587,
      "selected_pairs": 26759.0,
      "state_query_norm": 1054.782177734375,
      "state_relation_loss": 0.0004709567467216402,
      "step": 11140,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.003126668161712587,
      "weighted_relation_loss": 0.003126668161712587
    },
    {
      "action_query_norm": 588.3574829101562,
      "action_relation_loss": 0.002664898382499814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.430308699719363,
      "flow_matching_loss": 0.04031639769673347,
      "grad_norm": 0.1518390029668808,
      "learning_rate": 7.428545973538329e-05,
      "loss": 0.0435,
      "relation_loss": 0.0031459354562684894,
      "selected_pairs": 26951.3,
      "state_query_norm": 1054.7822143554688,
      "state_relation_loss": 0.0004810370795894414,
      "step": 11150,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0031459354562684894,
      "weighted_relation_loss": 0.0031459354562684894
    },
    {
      "action_query_norm": 585.0470947265625,
      "action_relation_loss": 0.0026591625064611435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.439663236669785,
      "flow_matching_loss": 0.03528047390282154,
      "grad_norm": 0.1510099470615387,
      "learning_rate": 7.42372672640038e-05,
      "loss": 0.0384,
      "relation_loss": 0.0031105814268812537,
      "selected_pairs": 26686.2,
      "state_query_norm": 1058.6668029785155,
      "state_relation_loss": 0.00045141893497202545,
      "step": 11160,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0031105814268812537,
      "weighted_relation_loss": 0.0031105814268812537
    },
    {
      "action_query_norm": 585.2159057617188,
      "action_relation_loss": 0.0025725222891196608,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.449017773620206,
      "flow_matching_loss": 0.03474451843649149,
      "grad_norm": 0.1124187782406807,
      "learning_rate": 7.418904534205917e-05,
      "loss": 0.0378,
      "relation_loss": 0.003063383512198925,
      "selected_pairs": 26969.3,
      "state_query_norm": 1041.8473999023438,
      "state_relation_loss": 0.0004908612085273489,
      "step": 11170,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003063383512198925,
      "weighted_relation_loss": 0.003063383512198925
    },
    {
      "action_query_norm": 575.8316101074219,
      "action_relation_loss": 0.002781274705193937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.458372310570628,
      "flow_matching_loss": 0.0381793100386858,
      "grad_norm": 0.10046274214982986,
      "learning_rate": 7.414079402814356e-05,
      "loss": 0.0414,
      "relation_loss": 0.00326370180118829,
      "selected_pairs": 26328.1,
      "state_query_norm": 1061.4695190429688,
      "state_relation_loss": 0.00048242706689052286,
      "step": 11180,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.00326370180118829,
      "weighted_relation_loss": 0.00326370180118829
    },
    {
      "action_query_norm": 586.9819519042969,
      "action_relation_loss": 0.002852989174425602,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.467726847521048,
      "flow_matching_loss": 0.039092781767249105,
      "grad_norm": 0.11420910805463791,
      "learning_rate": 7.409251338088687e-05,
      "loss": 0.0424,
      "relation_loss": 0.0033345210598781706,
      "selected_pairs": 26691.4,
      "state_query_norm": 1061.5492431640625,
      "state_relation_loss": 0.000481531917466782,
      "step": 11190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0033345210598781706,
      "weighted_relation_loss": 0.0033345210598781706
    },
    {
      "action_query_norm": 585.0539978027343,
      "action_relation_loss": 0.0026162200141698123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.477081384471468,
      "flow_matching_loss": 0.0352821184322238,
      "grad_norm": 0.12741856276988983,
      "learning_rate": 7.404420345895467e-05,
      "loss": 0.0384,
      "relation_loss": 0.003125476907007396,
      "selected_pairs": 26949.8,
      "state_query_norm": 1069.1484741210938,
      "state_relation_loss": 0.0005092569277621806,
      "step": 11200,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.003125476907007396,
      "weighted_relation_loss": 0.003125476907007396
    },
    {
      "action_query_norm": 580.3390380859375,
      "action_relation_loss": 0.0026865007355809213,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.48643592142189,
      "flow_matching_loss": 0.03715943600982428,
      "grad_norm": 0.10614420473575592,
      "learning_rate": 7.399586432104804e-05,
      "loss": 0.0403,
      "relation_loss": 0.0031851775711402297,
      "selected_pairs": 26594.2,
      "state_query_norm": 1064.6062866210937,
      "state_relation_loss": 0.0004986768239177763,
      "step": 11210,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0031851775711402297,
      "weighted_relation_loss": 0.0031851775711402297
    },
    {
      "action_query_norm": 582.9479248046875,
      "action_relation_loss": 0.0026283425744622947,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.49579045837231,
      "flow_matching_loss": 0.03633533008396626,
      "grad_norm": 0.12321925163269043,
      "learning_rate": 7.394749602590358e-05,
      "loss": 0.0395,
      "relation_loss": 0.003126429859548807,
      "selected_pairs": 27272.6,
      "state_query_norm": 1072.20087890625,
      "state_relation_loss": 0.0004980872879968956,
      "step": 11220,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003126429859548807,
      "weighted_relation_loss": 0.003126429859548807
    },
    {
      "action_query_norm": 583.5225402832032,
      "action_relation_loss": 0.00270166031550616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.505144995322732,
      "flow_matching_loss": 0.03536543324589729,
      "grad_norm": 0.15638367831707,
      "learning_rate": 7.389909863229336e-05,
      "loss": 0.0386,
      "relation_loss": 0.0032186443684622647,
      "selected_pairs": 27263.8,
      "state_query_norm": 1059.6887451171874,
      "state_relation_loss": 0.0005169840500457212,
      "step": 11230,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0032186443684622647,
      "weighted_relation_loss": 0.0032186443684622647
    },
    {
      "action_query_norm": 580.4442321777344,
      "action_relation_loss": 0.0026298139709979296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.514499532273152,
      "flow_matching_loss": 0.03796984814107418,
      "grad_norm": 0.12479949742555618,
      "learning_rate": 7.385067219902477e-05,
      "loss": 0.0411,
      "relation_loss": 0.003118559834547341,
      "selected_pairs": 27270.3,
      "state_query_norm": 1049.2471313476562,
      "state_relation_loss": 0.0004887458577286452,
      "step": 11240,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003118559834547341,
      "weighted_relation_loss": 0.003118559834547341
    },
    {
      "action_query_norm": 589.128955078125,
      "action_relation_loss": 0.002645980822853744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.523854069223573,
      "flow_matching_loss": 0.03557755295187235,
      "grad_norm": 0.09729206562042236,
      "learning_rate": 7.380221678494049e-05,
      "loss": 0.0387,
      "relation_loss": 0.003131220582872629,
      "selected_pairs": 27061.9,
      "state_query_norm": 1060.05087890625,
      "state_relation_loss": 0.0004852397600188851,
      "step": 11250,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003131220582872629,
      "weighted_relation_loss": 0.003131220582872629
    },
    {
      "action_query_norm": 579.4497375488281,
      "action_relation_loss": 0.002763229701668024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.533208606173995,
      "flow_matching_loss": 0.036382973194122314,
      "grad_norm": 0.09035037457942963,
      "learning_rate": 7.37537324489184e-05,
      "loss": 0.0397,
      "relation_loss": 0.0032671427354216574,
      "selected_pairs": 26575.1,
      "state_query_norm": 1069.3384399414062,
      "state_relation_loss": 0.0005039130308432504,
      "step": 11260,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0032671427354216574,
      "weighted_relation_loss": 0.0032671427354216574
    },
    {
      "action_query_norm": 581.757470703125,
      "action_relation_loss": 0.002629160275682807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.542563143124415,
      "flow_matching_loss": 0.036399119347333905,
      "grad_norm": 0.1131431981921196,
      "learning_rate": 7.370521924987155e-05,
      "loss": 0.0395,
      "relation_loss": 0.0031401999294757843,
      "selected_pairs": 26226.8,
      "state_query_norm": 1060.4366333007813,
      "state_relation_loss": 0.0005110396712552756,
      "step": 11270,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0031401999294757843,
      "weighted_relation_loss": 0.0031401999294757843
    },
    {
      "action_query_norm": 596.1833068847657,
      "action_relation_loss": 0.0026318372692912817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.551917680074837,
      "flow_matching_loss": 0.03630288355052471,
      "grad_norm": 0.0966014713048935,
      "learning_rate": 7.365667724674806e-05,
      "loss": 0.0395,
      "relation_loss": 0.003157167066819966,
      "selected_pairs": 26668.4,
      "state_query_norm": 1064.6223510742188,
      "state_relation_loss": 0.0005253298382740468,
      "step": 11280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003157167066819966,
      "weighted_relation_loss": 0.003157167066819966
    },
    {
      "action_query_norm": 586.6360534667969,
      "action_relation_loss": 0.0024975789710879326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.561272217025257,
      "flow_matching_loss": 0.03648414649069309,
      "grad_norm": 0.11697801947593689,
      "learning_rate": 7.360810649853105e-05,
      "loss": 0.0395,
      "relation_loss": 0.0029935281723737716,
      "selected_pairs": 26665.5,
      "state_query_norm": 1059.3166381835938,
      "state_relation_loss": 0.0004959491576300934,
      "step": 11290,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0029935281723737716,
      "weighted_relation_loss": 0.0029935281723737716
    },
    {
      "action_query_norm": 587.9284851074219,
      "action_relation_loss": 0.002578176581300795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.570626753975677,
      "flow_matching_loss": 0.036285719648003575,
      "grad_norm": 0.1122799813747406,
      "learning_rate": 7.355950706423853e-05,
      "loss": 0.0394,
      "relation_loss": 0.003078709589317441,
      "selected_pairs": 26735.5,
      "state_query_norm": 1043.677752685547,
      "state_relation_loss": 0.0005005330371204763,
      "step": 11300,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003078709589317441,
      "weighted_relation_loss": 0.003078709589317441
    },
    {
      "action_query_norm": 584.9015014648437,
      "action_relation_loss": 0.002707634144462645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.5799812909261,
      "flow_matching_loss": 0.03520261161029339,
      "grad_norm": 0.11502532660961151,
      "learning_rate": 7.351087900292342e-05,
      "loss": 0.0384,
      "relation_loss": 0.0031773028895258904,
      "selected_pairs": 26895.8,
      "state_query_norm": 1069.4310363769532,
      "state_relation_loss": 0.0004696687450632453,
      "step": 11310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0031773028895258904,
      "weighted_relation_loss": 0.0031773028895258904
    },
    {
      "action_query_norm": 581.6957702636719,
      "action_relation_loss": 0.002537318225950003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.58933582787652,
      "flow_matching_loss": 0.035300750844180585,
      "grad_norm": 0.11822690814733505,
      "learning_rate": 7.346222237367339e-05,
      "loss": 0.0383,
      "relation_loss": 0.0030241752974689007,
      "selected_pairs": 26762.2,
      "state_query_norm": 1072.9532104492187,
      "state_relation_loss": 0.00048685707442928103,
      "step": 11320,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0030241752974689007,
      "weighted_relation_loss": 0.0030241752974689007
    },
    {
      "action_query_norm": 586.1106750488282,
      "action_relation_loss": 0.002470770291984081,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.598690364826941,
      "flow_matching_loss": 0.03736238796263933,
      "grad_norm": 0.09836886078119278,
      "learning_rate": 7.341353723561081e-05,
      "loss": 0.0403,
      "relation_loss": 0.0029437160352244972,
      "selected_pairs": 27264.6,
      "state_query_norm": 1078.9829956054687,
      "state_relation_loss": 0.000472945740330033,
      "step": 11330,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0029437160352244972,
      "weighted_relation_loss": 0.0029437160352244972
    },
    {
      "action_query_norm": 591.7412719726562,
      "action_relation_loss": 0.0026822598185390232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.608044901777362,
      "flow_matching_loss": 0.03605849407613278,
      "grad_norm": 0.13165412843227386,
      "learning_rate": 7.336482364789277e-05,
      "loss": 0.0392,
      "relation_loss": 0.0031553914304822683,
      "selected_pairs": 26961.9,
      "state_query_norm": 1067.77373046875,
      "state_relation_loss": 0.00047313162358477714,
      "step": 11340,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0031553914304822683,
      "weighted_relation_loss": 0.0031553914304822683
    },
    {
      "action_query_norm": 587.4517456054688,
      "action_relation_loss": 0.0026119346264749767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.617399438727784,
      "flow_matching_loss": 0.03609959837049246,
      "grad_norm": 0.10471615195274353,
      "learning_rate": 7.331608166971082e-05,
      "loss": 0.0392,
      "relation_loss": 0.0030908955726772548,
      "selected_pairs": 26433.9,
      "state_query_norm": 1055.211279296875,
      "state_relation_loss": 0.0004789609258295968,
      "step": 11350,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0030908955726772548,
      "weighted_relation_loss": 0.0030908955726772548
    },
    {
      "action_query_norm": 578.214794921875,
      "action_relation_loss": 0.0026470049750059845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.626753975678204,
      "flow_matching_loss": 0.035767647251486775,
      "grad_norm": 0.10379671305418015,
      "learning_rate": 7.326731136029108e-05,
      "loss": 0.0389,
      "relation_loss": 0.0031281188363209366,
      "selected_pairs": 26737.2,
      "state_query_norm": 1072.4241821289063,
      "state_relation_loss": 0.00048111387877725065,
      "step": 11360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0031281188363209366,
      "weighted_relation_loss": 0.0031281188363209366
    },
    {
      "action_query_norm": 578.6315979003906,
      "action_relation_loss": 0.002703284821473062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.636108512628624,
      "flow_matching_loss": 0.036986421421170236,
      "grad_norm": 0.13111470639705658,
      "learning_rate": 7.321851277889408e-05,
      "loss": 0.0402,
      "relation_loss": 0.003183485311456025,
      "selected_pairs": 26780.4,
      "state_query_norm": 1063.5567626953125,
      "state_relation_loss": 0.000480200513266027,
      "step": 11370,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003183485311456025,
      "weighted_relation_loss": 0.003183485311456025
    },
    {
      "action_query_norm": 581.214892578125,
      "action_relation_loss": 0.002592381229624152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.645463049579046,
      "flow_matching_loss": 0.037468301877379415,
      "grad_norm": 0.10869243741035461,
      "learning_rate": 7.316968598481469e-05,
      "loss": 0.0405,
      "relation_loss": 0.0030786270275712014,
      "selected_pairs": 26923.5,
      "state_query_norm": 1057.0260864257812,
      "state_relation_loss": 0.00048624579212628303,
      "step": 11380,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0030786270275712014,
      "weighted_relation_loss": 0.0030786270275712014
    },
    {
      "action_query_norm": 586.5305847167969,
      "action_relation_loss": 0.002560661151073873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.654817586529466,
      "flow_matching_loss": 0.03590383045375347,
      "grad_norm": 0.12168978154659271,
      "learning_rate": 7.312083103738206e-05,
      "loss": 0.039,
      "relation_loss": 0.0030501357046887278,
      "selected_pairs": 26967.1,
      "state_query_norm": 1067.9098754882812,
      "state_relation_loss": 0.0004894745594356209,
      "step": 11390,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0030501357046887278,
      "weighted_relation_loss": 0.0030501357046887278
    },
    {
      "action_query_norm": 582.8883728027344,
      "action_relation_loss": 0.0026518150931224226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.664172123479888,
      "flow_matching_loss": 0.0381842702627182,
      "grad_norm": 0.11373430490493774,
      "learning_rate": 7.307194799595958e-05,
      "loss": 0.0413,
      "relation_loss": 0.0031556339003145696,
      "selected_pairs": 26484.3,
      "state_query_norm": 1055.1495239257813,
      "state_relation_loss": 0.0005038188479375094,
      "step": 11400,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0031556339003145696,
      "weighted_relation_loss": 0.0031556339003145696
    },
    {
      "action_query_norm": 590.2837219238281,
      "action_relation_loss": 0.0025009882170706987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.673526660430309,
      "flow_matching_loss": 0.035688852146267894,
      "grad_norm": 0.11801960319280624,
      "learning_rate": 7.302303691994473e-05,
      "loss": 0.0387,
      "relation_loss": 0.002988729253411293,
      "selected_pairs": 26506.0,
      "state_query_norm": 1064.8971435546875,
      "state_relation_loss": 0.00048774107417557386,
      "step": 11410,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002988729253411293,
      "weighted_relation_loss": 0.002988729253411293
    },
    {
      "action_query_norm": 584.1136962890625,
      "action_relation_loss": 0.0024703916860744357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.68288119738073,
      "flow_matching_loss": 0.03557984139770269,
      "grad_norm": 0.12251688539981842,
      "learning_rate": 7.29740978687691e-05,
      "loss": 0.0385,
      "relation_loss": 0.0029591221129521726,
      "selected_pairs": 26430.6,
      "state_query_norm": 1063.4026428222655,
      "state_relation_loss": 0.0004887304210569709,
      "step": 11420,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0029591221129521726,
      "weighted_relation_loss": 0.0029591221129521726
    },
    {
      "action_query_norm": 589.8452270507812,
      "action_relation_loss": 0.002736518019810319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.69223573433115,
      "flow_matching_loss": 0.03527127131819725,
      "grad_norm": 0.11459746956825256,
      "learning_rate": 7.292513090189825e-05,
      "loss": 0.0385,
      "relation_loss": 0.00323737608268857,
      "selected_pairs": 26664.2,
      "state_query_norm": 1051.8598327636719,
      "state_relation_loss": 0.0005008580657886342,
      "step": 11430,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00323737608268857,
      "weighted_relation_loss": 0.00323737608268857
    },
    {
      "action_query_norm": 588.5183898925782,
      "action_relation_loss": 0.002680023992434144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.701590271281571,
      "flow_matching_loss": 0.036907556280493733,
      "grad_norm": 0.12726792693138123,
      "learning_rate": 7.287613607883163e-05,
      "loss": 0.0401,
      "relation_loss": 0.0031828623032197356,
      "selected_pairs": 27060.1,
      "state_query_norm": 1055.6798217773437,
      "state_relation_loss": 0.0005028382700402289,
      "step": 11440,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0031828623032197356,
      "weighted_relation_loss": 0.0031828623032197356
    },
    {
      "action_query_norm": 589.5879760742188,
      "action_relation_loss": 0.002511670161038637,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.710944808231993,
      "flow_matching_loss": 0.03431951589882374,
      "grad_norm": 0.13058534264564514,
      "learning_rate": 7.282711345910263e-05,
      "loss": 0.0373,
      "relation_loss": 0.002973582781851292,
      "selected_pairs": 27407.9,
      "state_query_norm": 1063.7157165527344,
      "state_relation_loss": 0.00046191263536456975,
      "step": 11450,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.002973582781851292,
      "weighted_relation_loss": 0.002973582781851292
    },
    {
      "action_query_norm": 590.39189453125,
      "action_relation_loss": 0.0025335713056847454,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.720299345182413,
      "flow_matching_loss": 0.03579627387225628,
      "grad_norm": 0.1186022236943245,
      "learning_rate": 7.27780631022783e-05,
      "loss": 0.0388,
      "relation_loss": 0.0030116323847323655,
      "selected_pairs": 27029.1,
      "state_query_norm": 1067.6644165039063,
      "state_relation_loss": 0.0004780610819580033,
      "step": 11460,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0030116323847323655,
      "weighted_relation_loss": 0.0030116323847323655
    },
    {
      "action_query_norm": 586.5123840332031,
      "action_relation_loss": 0.002571548381820321,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.729653882132835,
      "flow_matching_loss": 0.03471260350197554,
      "grad_norm": 0.13757285475730896,
      "learning_rate": 7.272898506795948e-05,
      "loss": 0.0378,
      "relation_loss": 0.0030640542041510343,
      "selected_pairs": 26784.8,
      "state_query_norm": 1060.648291015625,
      "state_relation_loss": 0.0004925057961372659,
      "step": 11470,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0030640542041510343,
      "weighted_relation_loss": 0.0030640542041510343
    },
    {
      "action_query_norm": 589.6929870605469,
      "action_relation_loss": 0.002505569369532168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.739008419083255,
      "flow_matching_loss": 0.035290867649018766,
      "grad_norm": 0.13106298446655273,
      "learning_rate": 7.267987941578058e-05,
      "loss": 0.0383,
      "relation_loss": 0.002985844830982387,
      "selected_pairs": 26668.7,
      "state_query_norm": 1051.9680419921874,
      "state_relation_loss": 0.00048027546727098526,
      "step": 11480,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002985844830982387,
      "weighted_relation_loss": 0.002985844830982387
    },
    {
      "action_query_norm": 590.1081420898438,
      "action_relation_loss": 0.0025567005621269347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.748362956033676,
      "flow_matching_loss": 0.03692970778793096,
      "grad_norm": 0.13037215173244476,
      "learning_rate": 7.263074620540962e-05,
      "loss": 0.04,
      "relation_loss": 0.0030412626452744006,
      "selected_pairs": 27226.3,
      "state_query_norm": 1044.1935302734375,
      "state_relation_loss": 0.0004845621035201475,
      "step": 11490,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0030412626452744006,
      "weighted_relation_loss": 0.0030412626452744006
    },
    {
      "action_query_norm": 589.5554504394531,
      "action_relation_loss": 0.0025107275927439333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.757717492984098,
      "flow_matching_loss": 0.035880091786384585,
      "grad_norm": 0.1586795449256897,
      "learning_rate": 7.25815854965481e-05,
      "loss": 0.0389,
      "relation_loss": 0.00299070633482188,
      "selected_pairs": 26736.3,
      "state_query_norm": 1057.9346435546875,
      "state_relation_loss": 0.0004799786896910518,
      "step": 11500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00299070633482188,
      "weighted_relation_loss": 0.00299070633482188
    },
    {
      "action_query_norm": 583.9082336425781,
      "action_relation_loss": 0.002647430752404034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.767072029934518,
      "flow_matching_loss": 0.03703801967203617,
      "grad_norm": 0.1418064683675766,
      "learning_rate": 7.253239734893089e-05,
      "loss": 0.0402,
      "relation_loss": 0.003125940333120525,
      "selected_pairs": 26833.8,
      "state_query_norm": 1053.3589721679687,
      "state_relation_loss": 0.00047850956034380945,
      "step": 11510,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.003125940333120525,
      "weighted_relation_loss": 0.003125940333120525
    },
    {
      "action_query_norm": 583.5107788085937,
      "action_relation_loss": 0.0026290297973901033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.77642656688494,
      "flow_matching_loss": 0.03673680666834116,
      "grad_norm": 0.13489842414855957,
      "learning_rate": 7.248318182232623e-05,
      "loss": 0.0398,
      "relation_loss": 0.003095956868492067,
      "selected_pairs": 26438.2,
      "state_query_norm": 1085.5259399414062,
      "state_relation_loss": 0.0004669270594604313,
      "step": 11520,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.003095956868492067,
      "weighted_relation_loss": 0.003095956868492067
    },
    {
      "action_query_norm": 592.6989868164062,
      "action_relation_loss": 0.0025443785125389695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.78578110383536,
      "flow_matching_loss": 0.0348734550178051,
      "grad_norm": 0.15191678702831268,
      "learning_rate": 7.243393897653565e-05,
      "loss": 0.0379,
      "relation_loss": 0.0030372622422873974,
      "selected_pairs": 26876.3,
      "state_query_norm": 1067.0993896484374,
      "state_relation_loss": 0.0004928837093757466,
      "step": 11530,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0030372622422873974,
      "weighted_relation_loss": 0.0030372622422873974
    },
    {
      "action_query_norm": 585.8127990722656,
      "action_relation_loss": 0.002510261000134051,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.795135640785782,
      "flow_matching_loss": 0.035273030400276184,
      "grad_norm": 0.12034521996974945,
      "learning_rate": 7.23846688713938e-05,
      "loss": 0.0383,
      "relation_loss": 0.002989504113793373,
      "selected_pairs": 27482.0,
      "state_query_norm": 1061.4246826171875,
      "state_relation_loss": 0.0004792431223904714,
      "step": 11540,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.002989504113793373,
      "weighted_relation_loss": 0.002989504113793373
    },
    {
      "action_query_norm": 584.6291625976562,
      "action_relation_loss": 0.002632087608799338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.804490177736202,
      "flow_matching_loss": 0.036684670485556126,
      "grad_norm": 0.12929487228393555,
      "learning_rate": 7.233537156676854e-05,
      "loss": 0.0398,
      "relation_loss": 0.003105954360216856,
      "selected_pairs": 27172.1,
      "state_query_norm": 1061.6870544433593,
      "state_relation_loss": 0.0004738668038044125,
      "step": 11550,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.003105954360216856,
      "weighted_relation_loss": 0.003105954360216856
    },
    {
      "action_query_norm": 584.9003784179688,
      "action_relation_loss": 0.0024611776694655418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.813844714686622,
      "flow_matching_loss": 0.03490044269710779,
      "grad_norm": 0.12219370901584625,
      "learning_rate": 7.228604712256076e-05,
      "loss": 0.0378,
      "relation_loss": 0.0029117093654349446,
      "selected_pairs": 26418.3,
      "state_query_norm": 1072.561572265625,
      "state_relation_loss": 0.00045053169887978586,
      "step": 11560,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0029117093654349446,
      "weighted_relation_loss": 0.0029117093654349446
    },
    {
      "action_query_norm": 590.5695922851562,
      "action_relation_loss": 0.002519293245859444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.823199251637044,
      "flow_matching_loss": 0.03553153406828642,
      "grad_norm": 0.12839066982269287,
      "learning_rate": 7.223669559870426e-05,
      "loss": 0.0385,
      "relation_loss": 0.002991375746205449,
      "selected_pairs": 26900.5,
      "state_query_norm": 1082.8580200195313,
      "state_relation_loss": 0.00047208253818098456,
      "step": 11570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.002991375746205449,
      "weighted_relation_loss": 0.002991375746205449
    },
    {
      "action_query_norm": 585.5403564453125,
      "action_relation_loss": 0.0026399827096611262,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.832553788587465,
      "flow_matching_loss": 0.03463032841682434,
      "grad_norm": 0.10232599824666977,
      "learning_rate": 7.218731705516585e-05,
      "loss": 0.0377,
      "relation_loss": 0.0031171666691079737,
      "selected_pairs": 26733.7,
      "state_query_norm": 1068.8050048828125,
      "state_relation_loss": 0.00047718395362608137,
      "step": 11580,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0031171666691079737,
      "weighted_relation_loss": 0.0031171666691079737
    },
    {
      "action_query_norm": 590.7166442871094,
      "action_relation_loss": 0.0024743717862293125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.841908325537887,
      "flow_matching_loss": 0.035637783259153365,
      "grad_norm": 0.11835853010416031,
      "learning_rate": 7.21379115519451e-05,
      "loss": 0.0386,
      "relation_loss": 0.002934827678836882,
      "selected_pairs": 26987.5,
      "state_query_norm": 1066.411474609375,
      "state_relation_loss": 0.0004604559042491019,
      "step": 11590,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002934827678836882,
      "weighted_relation_loss": 0.002934827678836882
    },
    {
      "action_query_norm": 593.9475891113282,
      "action_relation_loss": 0.0025638817809522153,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.851262862488307,
      "flow_matching_loss": 0.035452688857913014,
      "grad_norm": 0.09994036704301834,
      "learning_rate": 7.20884791490743e-05,
      "loss": 0.0385,
      "relation_loss": 0.0030546547612175347,
      "selected_pairs": 27298.1,
      "state_query_norm": 1065.9439025878905,
      "state_relation_loss": 0.0004907730035483837,
      "step": 11600,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0030546547612175347,
      "weighted_relation_loss": 0.0030546547612175347
    },
    {
      "action_query_norm": 585.6124389648437,
      "action_relation_loss": 0.002464338974095881,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.860617399438727,
      "flow_matching_loss": 0.03610174544155598,
      "grad_norm": 0.1280621886253357,
      "learning_rate": 7.203901990661857e-05,
      "loss": 0.039,
      "relation_loss": 0.0029341007117182016,
      "selected_pairs": 26891.0,
      "state_query_norm": 1059.89384765625,
      "state_relation_loss": 0.0004697617026977241,
      "step": 11610,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0029341007117182016,
      "weighted_relation_loss": 0.0029341007117182016
    },
    {
      "action_query_norm": 592.6474182128907,
      "action_relation_loss": 0.0024717025691643356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.869971936389149,
      "flow_matching_loss": 0.035825255885720254,
      "grad_norm": 0.11518838256597519,
      "learning_rate": 7.198953388467549e-05,
      "loss": 0.0388,
      "relation_loss": 0.0029576390516012907,
      "selected_pairs": 26734.3,
      "state_query_norm": 1058.4962524414063,
      "state_relation_loss": 0.000485936485347338,
      "step": 11620,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0029576390516012907,
      "weighted_relation_loss": 0.0029576390516012907
    },
    {
      "action_query_norm": 592.1308776855469,
      "action_relation_loss": 0.00236443723551929,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.87932647333957,
      "flow_matching_loss": 0.037082156538963316,
      "grad_norm": 0.16187229752540588,
      "learning_rate": 7.194002114337526e-05,
      "loss": 0.0399,
      "relation_loss": 0.0028481390560045837,
      "selected_pairs": 27008.5,
      "state_query_norm": 1068.0429626464843,
      "state_relation_loss": 0.0004837018728721887,
      "step": 11630,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0028481390560045837,
      "weighted_relation_loss": 0.0028481390560045837
    },
    {
      "action_query_norm": 589.7568908691406,
      "action_relation_loss": 0.0025189480977132915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.888681010289991,
      "flow_matching_loss": 0.035815367847681044,
      "grad_norm": 0.14676006138324738,
      "learning_rate": 7.189048174288054e-05,
      "loss": 0.0388,
      "relation_loss": 0.003012753487564623,
      "selected_pairs": 26854.2,
      "state_query_norm": 1068.8401245117188,
      "state_relation_loss": 0.0004938053898513317,
      "step": 11640,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003012753487564623,
      "weighted_relation_loss": 0.003012753487564623
    },
    {
      "action_query_norm": 589.1761901855468,
      "action_relation_loss": 0.002603569929488003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.898035547240411,
      "flow_matching_loss": 0.03761522527784109,
      "grad_norm": 0.11075732111930847,
      "learning_rate": 7.184091574338636e-05,
      "loss": 0.0407,
      "relation_loss": 0.0030918648233637215,
      "selected_pairs": 26297.4,
      "state_query_norm": 1066.9061889648438,
      "state_relation_loss": 0.0004882949026068673,
      "step": 11650,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0030918648233637215,
      "weighted_relation_loss": 0.0030918648233637215
    },
    {
      "action_query_norm": 596.372509765625,
      "action_relation_loss": 0.00250860545784235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.907390084190833,
      "flow_matching_loss": 0.03376577366143465,
      "grad_norm": 0.13749423623085022,
      "learning_rate": 7.179132320512009e-05,
      "loss": 0.0367,
      "relation_loss": 0.0029741539154201748,
      "selected_pairs": 26598.3,
      "state_query_norm": 1067.0642700195312,
      "state_relation_loss": 0.00046554844302590934,
      "step": 11660,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0029741539154201748,
      "weighted_relation_loss": 0.0029741539154201748
    },
    {
      "action_query_norm": 591.0693176269531,
      "action_relation_loss": 0.0024242073530331256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.916744621141254,
      "flow_matching_loss": 0.034636716544628146,
      "grad_norm": 0.13060034811496735,
      "learning_rate": 7.174170418834134e-05,
      "loss": 0.0375,
      "relation_loss": 0.0029044917551800608,
      "selected_pairs": 27315.0,
      "state_query_norm": 1064.1577514648438,
      "state_relation_loss": 0.00048028439632616935,
      "step": 11670,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0029044917551800608,
      "weighted_relation_loss": 0.0029044917551800608
    },
    {
      "action_query_norm": 596.0936157226563,
      "action_relation_loss": 0.0024281847756356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.926099158091674,
      "flow_matching_loss": 0.03499734923243523,
      "grad_norm": 0.13961002230644226,
      "learning_rate": 7.169205875334189e-05,
      "loss": 0.0379,
      "relation_loss": 0.0029063345165923236,
      "selected_pairs": 27383.0,
      "state_query_norm": 1048.9538330078126,
      "state_relation_loss": 0.0004781497176736593,
      "step": 11680,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0029063345165923236,
      "weighted_relation_loss": 0.0029063345165923236
    },
    {
      "action_query_norm": 597.71318359375,
      "action_relation_loss": 0.0023781362222507598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.935453695042096,
      "flow_matching_loss": 0.03419851418584585,
      "grad_norm": 0.13318273425102234,
      "learning_rate": 7.164238696044562e-05,
      "loss": 0.037,
      "relation_loss": 0.0028453761944547294,
      "selected_pairs": 27036.3,
      "state_query_norm": 1069.6393188476563,
      "state_relation_loss": 0.0004672399692935869,
      "step": 11690,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0028453761944547294,
      "weighted_relation_loss": 0.0028453761944547294
    },
    {
      "action_query_norm": 595.4638061523438,
      "action_relation_loss": 0.0025412957184016705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.944808231992516,
      "flow_matching_loss": 0.03460884653031826,
      "grad_norm": 0.1115444004535675,
      "learning_rate": 7.159268887000846e-05,
      "loss": 0.0376,
      "relation_loss": 0.003024539118632674,
      "selected_pairs": 26338.3,
      "state_query_norm": 1061.3128784179687,
      "state_relation_loss": 0.0004832433973206207,
      "step": 11700,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.003024539118632674,
      "weighted_relation_loss": 0.003024539118632674
    },
    {
      "action_query_norm": 591.3639831542969,
      "action_relation_loss": 0.002383920317515731,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.954162768942938,
      "flow_matching_loss": 0.03547843638807535,
      "grad_norm": 0.10367591679096222,
      "learning_rate": 7.154296454241827e-05,
      "loss": 0.0383,
      "relation_loss": 0.002839660132303834,
      "selected_pairs": 27289.7,
      "state_query_norm": 1063.965966796875,
      "state_relation_loss": 0.00045573982642963526,
      "step": 11710,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.002839660132303834,
      "weighted_relation_loss": 0.002839660132303834
    },
    {
      "action_query_norm": 591.1381652832031,
      "action_relation_loss": 0.0024905495811253786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.963517305893358,
      "flow_matching_loss": 0.035350169986486435,
      "grad_norm": 0.12679985165596008,
      "learning_rate": 7.149321403809479e-05,
      "loss": 0.0383,
      "relation_loss": 0.0029584364499896766,
      "selected_pairs": 27196.9,
      "state_query_norm": 1053.8985473632813,
      "state_relation_loss": 0.0004678868514019996,
      "step": 11720,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0029584364499896766,
      "weighted_relation_loss": 0.0029584364499896766
    },
    {
      "action_query_norm": 585.9715026855469,
      "action_relation_loss": 0.002476440160535276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.97287184284378,
      "flow_matching_loss": 0.035276667773723604,
      "grad_norm": 0.11529547721147537,
      "learning_rate": 7.144343741748956e-05,
      "loss": 0.0382,
      "relation_loss": 0.002960105147212744,
      "selected_pairs": 27101.6,
      "state_query_norm": 1070.8758972167968,
      "state_relation_loss": 0.0004836649808567017,
      "step": 11730,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.002960105147212744,
      "weighted_relation_loss": 0.002960105147212744
    },
    {
      "action_query_norm": 579.6447570800781,
      "action_relation_loss": 0.002507360279560089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.9822263797942,
      "flow_matching_loss": 0.036755166575312616,
      "grad_norm": 0.14395292103290558,
      "learning_rate": 7.139363474108589e-05,
      "loss": 0.0397,
      "relation_loss": 0.002981861308217049,
      "selected_pairs": 26139.4,
      "state_query_norm": 1073.0966430664062,
      "state_relation_loss": 0.0004745010315673426,
      "step": 11740,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.002981861308217049,
      "weighted_relation_loss": 0.002981861308217049
    },
    {
      "action_query_norm": 581.8624084472656,
      "action_relation_loss": 0.002339055575430393,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.99158091674462,
      "flow_matching_loss": 0.035909166745841506,
      "grad_norm": 0.14598348736763,
      "learning_rate": 7.134380606939874e-05,
      "loss": 0.0387,
      "relation_loss": 0.0028268886730074883,
      "selected_pairs": 26686.4,
      "state_query_norm": 1079.2962890625,
      "state_relation_loss": 0.0004878330713836476,
      "step": 11750,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0028268886730074883,
      "weighted_relation_loss": 0.0028268886730074883
    },
    {
      "action_query_norm": 584.157373046875,
      "action_relation_loss": 0.0024050466949120165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.000935453695043,
      "flow_matching_loss": 0.036382117122411725,
      "grad_norm": 0.134908065199852,
      "learning_rate": 7.12939514629746e-05,
      "loss": 0.0393,
      "relation_loss": 0.0028701026923954485,
      "selected_pairs": 24618.5,
      "state_query_norm": 1087.6267456054688,
      "state_relation_loss": 0.0004650559625588357,
      "step": 11760,
      "valid_action_samples": 215.4,
      "weighted_auxiliary_loss": 0.0028701026923954485,
      "weighted_relation_loss": 0.0028701026923954485
    },
    {
      "action_query_norm": 587.4504821777343,
      "action_relation_loss": 0.0024404515977948906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.010289990645463,
      "flow_matching_loss": 0.03309692796319723,
      "grad_norm": 0.12229353934526443,
      "learning_rate": 7.124407098239155e-05,
      "loss": 0.036,
      "relation_loss": 0.0029093659948557617,
      "selected_pairs": 26585.4,
      "state_query_norm": 1074.0393310546874,
      "state_relation_loss": 0.00046891439124010505,
      "step": 11770,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0029093659948557617,
      "weighted_relation_loss": 0.0029093659948557617
    },
    {
      "action_query_norm": 586.6218505859375,
      "action_relation_loss": 0.002428990975022316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.019644527595885,
      "flow_matching_loss": 0.03390200901776552,
      "grad_norm": 0.10999208688735962,
      "learning_rate": 7.119416468825908e-05,
      "loss": 0.0368,
      "relation_loss": 0.00290277034509927,
      "selected_pairs": 26483.4,
      "state_query_norm": 1068.1063842773438,
      "state_relation_loss": 0.0004737793409731239,
      "step": 11780,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00290277034509927,
      "weighted_relation_loss": 0.00290277034509927
    },
    {
      "action_query_norm": 583.306640625,
      "action_relation_loss": 0.0024877307470887898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.028999064546305,
      "flow_matching_loss": 0.03548036515712738,
      "grad_norm": 0.12746691703796387,
      "learning_rate": 7.114423264121804e-05,
      "loss": 0.0384,
      "relation_loss": 0.002967275306582451,
      "selected_pairs": 26434.3,
      "state_query_norm": 1071.3084106445312,
      "state_relation_loss": 0.00047954457404557614,
      "step": 11790,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002967275306582451,
      "weighted_relation_loss": 0.002967275306582451
    },
    {
      "action_query_norm": 586.7407958984375,
      "action_relation_loss": 0.0023567414144054057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.038353601496725,
      "flow_matching_loss": 0.03446125276386738,
      "grad_norm": 0.10763181746006012,
      "learning_rate": 7.109427490194056e-05,
      "loss": 0.0373,
      "relation_loss": 0.0028372103115543723,
      "selected_pairs": 26965.4,
      "state_query_norm": 1076.6965942382812,
      "state_relation_loss": 0.00048046892625279724,
      "step": 11800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0028372103115543723,
      "weighted_relation_loss": 0.0028372103115543723
    },
    {
      "action_query_norm": 579.1288391113281,
      "action_relation_loss": 0.0024293618276715278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.047708138447147,
      "flow_matching_loss": 0.035299723595380784,
      "grad_norm": 0.11361213773488998,
      "learning_rate": 7.104429153113001e-05,
      "loss": 0.0382,
      "relation_loss": 0.002910447330214083,
      "selected_pairs": 27017.2,
      "state_query_norm": 1078.5135375976563,
      "state_relation_loss": 0.000481085543287918,
      "step": 11810,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002910447330214083,
      "weighted_relation_loss": 0.002910447330214083
    },
    {
      "action_query_norm": 580.851513671875,
      "action_relation_loss": 0.0024024541955441235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.057062675397567,
      "flow_matching_loss": 0.03550129607319832,
      "grad_norm": 0.0982966274023056,
      "learning_rate": 7.099428258952092e-05,
      "loss": 0.0384,
      "relation_loss": 0.0028790093958377836,
      "selected_pairs": 26448.5,
      "state_query_norm": 1070.3621704101563,
      "state_relation_loss": 0.00047655517410021274,
      "step": 11820,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0028790093958377836,
      "weighted_relation_loss": 0.0028790093958377836
    },
    {
      "action_query_norm": 587.0829284667968,
      "action_relation_loss": 0.0023662101943045855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.06641721234799,
      "flow_matching_loss": 0.03389930240809917,
      "grad_norm": 0.11741889268159866,
      "learning_rate": 7.094424813787883e-05,
      "loss": 0.0367,
      "relation_loss": 0.0028271151008084417,
      "selected_pairs": 27315.3,
      "state_query_norm": 1078.2565551757812,
      "state_relation_loss": 0.0004609048774000257,
      "step": 11830,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0028271151008084417,
      "weighted_relation_loss": 0.0028271151008084417
    },
    {
      "action_query_norm": 585.8682067871093,
      "action_relation_loss": 0.0025043666129931806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.07577174929841,
      "flow_matching_loss": 0.035245454497635366,
      "grad_norm": 0.13127994537353516,
      "learning_rate": 7.089418823700034e-05,
      "loss": 0.0382,
      "relation_loss": 0.002950376598164439,
      "selected_pairs": 27105.6,
      "state_query_norm": 1082.8592041015625,
      "state_relation_loss": 0.0004460099822608754,
      "step": 11840,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.002950376598164439,
      "weighted_relation_loss": 0.002950376598164439
    },
    {
      "action_query_norm": 580.9610168457032,
      "action_relation_loss": 0.0023553602397441862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.085126286248832,
      "flow_matching_loss": 0.03511770684272051,
      "grad_norm": 0.09924931079149246,
      "learning_rate": 7.084410294771298e-05,
      "loss": 0.0379,
      "relation_loss": 0.002822066587395966,
      "selected_pairs": 26826.0,
      "state_query_norm": 1077.9424194335938,
      "state_relation_loss": 0.00046670633601024746,
      "step": 11850,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002822066587395966,
      "weighted_relation_loss": 0.002822066587395966
    },
    {
      "action_query_norm": 589.986474609375,
      "action_relation_loss": 0.0023516528541222214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.094480823199252,
      "flow_matching_loss": 0.03376961182802916,
      "grad_norm": 0.12447118759155273,
      "learning_rate": 7.079399233087504e-05,
      "loss": 0.0366,
      "relation_loss": 0.0028075150214135645,
      "selected_pairs": 27038.3,
      "state_query_norm": 1079.7100708007813,
      "state_relation_loss": 0.0004558621294563636,
      "step": 11860,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0028075150214135645,
      "weighted_relation_loss": 0.0028075150214135645
    },
    {
      "action_query_norm": 592.3392272949219,
      "action_relation_loss": 0.0023667493602260946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.103835360149672,
      "flow_matching_loss": 0.034148591570556164,
      "grad_norm": 0.11680294573307037,
      "learning_rate": 7.074385644737568e-05,
      "loss": 0.037,
      "relation_loss": 0.0028126871213316916,
      "selected_pairs": 26963.6,
      "state_query_norm": 1070.0263916015624,
      "state_relation_loss": 0.0004459377640159801,
      "step": 11870,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0028126871213316916,
      "weighted_relation_loss": 0.0028126871213316916
    },
    {
      "action_query_norm": 592.4055419921875,
      "action_relation_loss": 0.0022860393160954117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.113189897100094,
      "flow_matching_loss": 0.03437246382236481,
      "grad_norm": 0.13868865370750427,
      "learning_rate": 7.069369535813473e-05,
      "loss": 0.0371,
      "relation_loss": 0.002771283336915076,
      "selected_pairs": 27737.2,
      "state_query_norm": 1062.9272827148438,
      "state_relation_loss": 0.00048524400626774876,
      "step": 11880,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.002771283336915076,
      "weighted_relation_loss": 0.002771283336915076
    },
    {
      "action_query_norm": 588.39501953125,
      "action_relation_loss": 0.0022801301442086697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.122544434050514,
      "flow_matching_loss": 0.03432848509401083,
      "grad_norm": 0.1414693146944046,
      "learning_rate": 7.06435091241026e-05,
      "loss": 0.0371,
      "relation_loss": 0.0027365521527826786,
      "selected_pairs": 26622.0,
      "state_query_norm": 1066.7010375976563,
      "state_relation_loss": 0.0004564220056636259,
      "step": 11890,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0027365521527826786,
      "weighted_relation_loss": 0.0027365521527826786
    },
    {
      "action_query_norm": 585.1667541503906,
      "action_relation_loss": 0.002360347262583673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.131898971000936,
      "flow_matching_loss": 0.03530493713915348,
      "grad_norm": 0.12662579119205475,
      "learning_rate": 7.059329780626034e-05,
      "loss": 0.0381,
      "relation_loss": 0.0028271037852391602,
      "selected_pairs": 26446.0,
      "state_query_norm": 1063.097265625,
      "state_relation_loss": 0.0004667565255658701,
      "step": 11900,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0028271037852391602,
      "weighted_relation_loss": 0.0028271037852391602
    },
    {
      "action_query_norm": 588.30859375,
      "action_relation_loss": 0.0023232737090438603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.141253507951356,
      "flow_matching_loss": 0.03531058132648468,
      "grad_norm": 0.11705803871154785,
      "learning_rate": 7.054306146561944e-05,
      "loss": 0.0381,
      "relation_loss": 0.002777530252933502,
      "selected_pairs": 26214.8,
      "state_query_norm": 1064.3648864746094,
      "state_relation_loss": 0.00045425656717270615,
      "step": 11910,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.002777530252933502,
      "weighted_relation_loss": 0.002777530252933502
    },
    {
      "action_query_norm": 592.9891357421875,
      "action_relation_loss": 0.0023937753401696684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.150608044901777,
      "flow_matching_loss": 0.03502142019569874,
      "grad_norm": 0.12548798322677612,
      "learning_rate": 7.049280016322178e-05,
      "loss": 0.0379,
      "relation_loss": 0.002856044890359044,
      "selected_pairs": 26755.9,
      "state_query_norm": 1077.5708251953124,
      "state_relation_loss": 0.00046226954145822674,
      "step": 11920,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.002856044890359044,
      "weighted_relation_loss": 0.002856044890359044
    },
    {
      "action_query_norm": 587.2282958984375,
      "action_relation_loss": 0.0023413091199472546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.159962581852199,
      "flow_matching_loss": 0.0338261291384697,
      "grad_norm": 0.13916553556919098,
      "learning_rate": 7.044251396013957e-05,
      "loss": 0.0366,
      "relation_loss": 0.002804011502303183,
      "selected_pairs": 27012.8,
      "state_query_norm": 1058.3178466796876,
      "state_relation_loss": 0.0004627023852663115,
      "step": 11930,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002804011502303183,
      "weighted_relation_loss": 0.002804011502303183
    },
    {
      "action_query_norm": 588.7974670410156,
      "action_relation_loss": 0.002337657962925732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.169317118802619,
      "flow_matching_loss": 0.03343935459852219,
      "grad_norm": 0.14773893356323242,
      "learning_rate": 7.039220291747528e-05,
      "loss": 0.0363,
      "relation_loss": 0.002829613955691457,
      "selected_pairs": 26807.0,
      "state_query_norm": 1072.8861206054687,
      "state_relation_loss": 0.0004919559927657246,
      "step": 11940,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002829613955691457,
      "weighted_relation_loss": 0.002829613955691457
    },
    {
      "action_query_norm": 590.0069946289062,
      "action_relation_loss": 0.002400082489475608,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.17867165575304,
      "flow_matching_loss": 0.034648696146905425,
      "grad_norm": 0.1615697741508484,
      "learning_rate": 7.034186709636159e-05,
      "loss": 0.0375,
      "relation_loss": 0.0028668553568422794,
      "selected_pairs": 26144.8,
      "state_query_norm": 1069.1374267578126,
      "state_relation_loss": 0.0004667729081120342,
      "step": 11950,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0028668553568422794,
      "weighted_relation_loss": 0.0028668553568422794
    },
    {
      "action_query_norm": 581.40244140625,
      "action_relation_loss": 0.0024409455712884665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.188026192703461,
      "flow_matching_loss": 0.03499126601964235,
      "grad_norm": 0.10809938609600067,
      "learning_rate": 7.029150655796129e-05,
      "loss": 0.0379,
      "relation_loss": 0.0029007041128352286,
      "selected_pairs": 26782.3,
      "state_query_norm": 1077.5200927734375,
      "state_relation_loss": 0.00045975857647135855,
      "step": 11960,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0029007041128352286,
      "weighted_relation_loss": 0.0029007041128352286
    },
    {
      "action_query_norm": 587.3994812011719,
      "action_relation_loss": 0.002376227080821991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.197380729653883,
      "flow_matching_loss": 0.03331143744289875,
      "grad_norm": 0.1167958527803421,
      "learning_rate": 7.024112136346712e-05,
      "loss": 0.0362,
      "relation_loss": 0.0028437295462936164,
      "selected_pairs": 26667.8,
      "state_query_norm": 1080.8598999023438,
      "state_relation_loss": 0.00046750251203775406,
      "step": 11970,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0028437295462936164,
      "weighted_relation_loss": 0.0028437295462936164
    },
    {
      "action_query_norm": 590.8161682128906,
      "action_relation_loss": 0.002489307406358421,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.206735266604303,
      "flow_matching_loss": 0.03534009326249361,
      "grad_norm": 0.14329619705677032,
      "learning_rate": 7.01907115741019e-05,
      "loss": 0.0383,
      "relation_loss": 0.002962442487478256,
      "selected_pairs": 26966.5,
      "state_query_norm": 1067.1979248046875,
      "state_relation_loss": 0.00047313509858213363,
      "step": 11980,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002962442487478256,
      "weighted_relation_loss": 0.002962442487478256
    },
    {
      "action_query_norm": 584.951806640625,
      "action_relation_loss": 0.002454768237657845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.216089803554723,
      "flow_matching_loss": 0.03619084358215332,
      "grad_norm": 0.1441434621810913,
      "learning_rate": 7.014027725111825e-05,
      "loss": 0.0391,
      "relation_loss": 0.0029327028896659613,
      "selected_pairs": 26689.1,
      "state_query_norm": 1077.4131225585938,
      "state_relation_loss": 0.00047793461999390274,
      "step": 11990,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0029327028896659613,
      "weighted_relation_loss": 0.0029327028896659613
    },
    {
      "action_query_norm": 586.4189514160156,
      "action_relation_loss": 0.002486987109296024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.225444340505145,
      "flow_matching_loss": 0.036281734704971313,
      "grad_norm": 0.13402040302753448,
      "learning_rate": 7.008981845579864e-05,
      "loss": 0.0392,
      "relation_loss": 0.0029664295492693783,
      "selected_pairs": 26622.4,
      "state_query_norm": 1059.9059631347657,
      "state_relation_loss": 0.00047944244870450347,
      "step": 12000,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0029664295492693783,
      "weighted_relation_loss": 0.0029664295492693783
    },
    {
      "action_query_norm": 583.4759826660156,
      "action_relation_loss": 0.002434703498147428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.234798877455566,
      "flow_matching_loss": 0.03354371711611748,
      "grad_norm": 0.14511500298976898,
      "learning_rate": 7.003933524945528e-05,
      "loss": 0.0364,
      "relation_loss": 0.0028964459197595716,
      "selected_pairs": 26753.4,
      "state_query_norm": 1075.9895751953125,
      "state_relation_loss": 0.00046174237795639785,
      "step": 12010,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0028964459197595716,
      "weighted_relation_loss": 0.0028964459197595716
    },
    {
      "action_query_norm": 591.7545715332031,
      "action_relation_loss": 0.002381231763865799,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.244153414405988,
      "flow_matching_loss": 0.03272958565503359,
      "grad_norm": 0.16543307900428772,
      "learning_rate": 6.998882769342998e-05,
      "loss": 0.0356,
      "relation_loss": 0.0028690960723906757,
      "selected_pairs": 26365.0,
      "state_query_norm": 1076.148486328125,
      "state_relation_loss": 0.000487864320166409,
      "step": 12020,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0028690960723906757,
      "weighted_relation_loss": 0.0028690960723906757
    },
    {
      "action_query_norm": 584.2999084472656,
      "action_relation_loss": 0.002405069605447352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.253507951356408,
      "flow_matching_loss": 0.033635391108691695,
      "grad_norm": 0.18845584988594055,
      "learning_rate": 6.993829584909423e-05,
      "loss": 0.0365,
      "relation_loss": 0.0028736147098243235,
      "selected_pairs": 27320.1,
      "state_query_norm": 1066.5842895507812,
      "state_relation_loss": 0.0004685451072873548,
      "step": 12030,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0028736147098243235,
      "weighted_relation_loss": 0.0028736147098243235
    },
    {
      "action_query_norm": 581.5526123046875,
      "action_relation_loss": 0.002409095480106771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.262862488306828,
      "flow_matching_loss": 0.034113811515271666,
      "grad_norm": 0.15289755165576935,
      "learning_rate": 6.988773977784895e-05,
      "loss": 0.037,
      "relation_loss": 0.0028666001278907062,
      "selected_pairs": 26660.9,
      "state_query_norm": 1073.0797973632812,
      "state_relation_loss": 0.0004575046361424029,
      "step": 12040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0028666001278907062,
      "weighted_relation_loss": 0.0028666001278907062
    },
    {
      "action_query_norm": 591.9750183105468,
      "action_relation_loss": 0.0024847593158483505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.27221702525725,
      "flow_matching_loss": 0.035207966528832915,
      "grad_norm": 0.14380104839801788,
      "learning_rate": 6.983715954112454e-05,
      "loss": 0.0382,
      "relation_loss": 0.0029439833015203476,
      "selected_pairs": 26876.1,
      "state_query_norm": 1088.5220458984375,
      "state_relation_loss": 0.00045922399440314623,
      "step": 12050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0029439833015203476,
      "weighted_relation_loss": 0.0029439833015203476
    },
    {
      "action_query_norm": 590.5081176757812,
      "action_relation_loss": 0.0022940762341022492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.28157156220767,
      "flow_matching_loss": 0.03451072089374065,
      "grad_norm": 0.13700228929519653,
      "learning_rate": 6.978655520038079e-05,
      "loss": 0.0373,
      "relation_loss": 0.0027701480081304907,
      "selected_pairs": 27069.4,
      "state_query_norm": 1066.6166442871095,
      "state_relation_loss": 0.0004760717856697738,
      "step": 12060,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0027701480081304907,
      "weighted_relation_loss": 0.0027701480081304907
    },
    {
      "action_query_norm": 588.1843688964843,
      "action_relation_loss": 0.002277453860733658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.290926099158092,
      "flow_matching_loss": 0.033747642487287524,
      "grad_norm": 0.1411544531583786,
      "learning_rate": 6.97359268171067e-05,
      "loss": 0.0365,
      "relation_loss": 0.002753961249254644,
      "selected_pairs": 27086.9,
      "state_query_norm": 1064.1397827148437,
      "state_relation_loss": 0.000476507397252135,
      "step": 12070,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002753961249254644,
      "weighted_relation_loss": 0.002753961249254644
    },
    {
      "action_query_norm": 594.4331237792969,
      "action_relation_loss": 0.0024342876626178623,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.300280636108512,
      "flow_matching_loss": 0.03625383265316486,
      "grad_norm": 0.16960503160953522,
      "learning_rate": 6.968527445282056e-05,
      "loss": 0.0392,
      "relation_loss": 0.0028994993306696417,
      "selected_pairs": 26880.7,
      "state_query_norm": 1076.3366821289062,
      "state_relation_loss": 0.00046521167387254537,
      "step": 12080,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0028994993306696417,
      "weighted_relation_loss": 0.0028994993306696417
    },
    {
      "action_query_norm": 598.6151611328125,
      "action_relation_loss": 0.0023561554960906507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.309635173058934,
      "flow_matching_loss": 0.03537815809249878,
      "grad_norm": 0.12753607332706451,
      "learning_rate": 6.963459816906977e-05,
      "loss": 0.0382,
      "relation_loss": 0.00281388727016747,
      "selected_pairs": 27314.1,
      "state_query_norm": 1093.7542114257812,
      "state_relation_loss": 0.00045773173333145677,
      "step": 12090,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.00281388727016747,
      "weighted_relation_loss": 0.00281388727016747
    },
    {
      "action_query_norm": 593.8742492675781,
      "action_relation_loss": 0.0023627069778740404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.318989710009355,
      "flow_matching_loss": 0.034651538357138634,
      "grad_norm": 0.13537205755710602,
      "learning_rate": 6.958389802743077e-05,
      "loss": 0.0375,
      "relation_loss": 0.0028137773741036655,
      "selected_pairs": 27246.4,
      "state_query_norm": 1084.982861328125,
      "state_relation_loss": 0.0004510703613050282,
      "step": 12100,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0028137773741036655,
      "weighted_relation_loss": 0.0028137773741036655
    },
    {
      "action_query_norm": 589.2226684570312,
      "action_relation_loss": 0.002385623403824866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.328344246959775,
      "flow_matching_loss": 0.03387575224041939,
      "grad_norm": 0.13691961765289307,
      "learning_rate": 6.953317408950903e-05,
      "loss": 0.0367,
      "relation_loss": 0.00284537966363132,
      "selected_pairs": 27183.5,
      "state_query_norm": 1071.962939453125,
      "state_relation_loss": 0.00045975627144798634,
      "step": 12110,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.00284537966363132,
      "weighted_relation_loss": 0.00284537966363132
    },
    {
      "action_query_norm": 592.0669860839844,
      "action_relation_loss": 0.002341410983353853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.337698783910197,
      "flow_matching_loss": 0.0334299499168992,
      "grad_norm": 0.10375971347093582,
      "learning_rate": 6.94824264169389e-05,
      "loss": 0.0362,
      "relation_loss": 0.002794922050088644,
      "selected_pairs": 26641.9,
      "state_query_norm": 1079.5413696289063,
      "state_relation_loss": 0.0004535110667347908,
      "step": 12120,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002794922050088644,
      "weighted_relation_loss": 0.002794922050088644
    },
    {
      "action_query_norm": 582.2088623046875,
      "action_relation_loss": 0.002406908106058836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.347053320860617,
      "flow_matching_loss": 0.034324743784964085,
      "grad_norm": 0.12164396047592163,
      "learning_rate": 6.94316550713836e-05,
      "loss": 0.0372,
      "relation_loss": 0.002866396773606539,
      "selected_pairs": 27012.4,
      "state_query_norm": 1075.493603515625,
      "state_relation_loss": 0.0004594886559061706,
      "step": 12130,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002866396773606539,
      "weighted_relation_loss": 0.002866396773606539
    },
    {
      "action_query_norm": 588.2107849121094,
      "action_relation_loss": 0.0023270251462236048,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.356407857811039,
      "flow_matching_loss": 0.03514351807534695,
      "grad_norm": 0.12773412466049194,
      "learning_rate": 6.938086011453513e-05,
      "loss": 0.038,
      "relation_loss": 0.002811067458242178,
      "selected_pairs": 25660.6,
      "state_query_norm": 1066.0833129882812,
      "state_relation_loss": 0.00048404230910819023,
      "step": 12140,
      "valid_action_samples": 227.0,
      "weighted_auxiliary_loss": 0.002811067458242178,
      "weighted_relation_loss": 0.002811067458242178
    },
    {
      "action_query_norm": 595.4145690917969,
      "action_relation_loss": 0.0023587157716974615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.36576239476146,
      "flow_matching_loss": 0.03508441168814898,
      "grad_norm": 0.10016600787639618,
      "learning_rate": 6.93300416081141e-05,
      "loss": 0.0379,
      "relation_loss": 0.0028211686760187147,
      "selected_pairs": 26953.9,
      "state_query_norm": 1076.1653442382812,
      "state_relation_loss": 0.00046245291596278546,
      "step": 12150,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0028211686760187147,
      "weighted_relation_loss": 0.0028211686760187147
    },
    {
      "action_query_norm": 591.9876525878906,
      "action_relation_loss": 0.0023643413791432978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.37511693171188,
      "flow_matching_loss": 0.03481542877852917,
      "grad_norm": 0.11799128353595734,
      "learning_rate": 6.927919961386984e-05,
      "loss": 0.0376,
      "relation_loss": 0.00283113659825176,
      "selected_pairs": 26287.6,
      "state_query_norm": 1068.6143859863282,
      "state_relation_loss": 0.000466795195825398,
      "step": 12160,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.00283113659825176,
      "weighted_relation_loss": 0.00283113659825176
    },
    {
      "action_query_norm": 589.1390808105468,
      "action_relation_loss": 0.0024097955552861094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.384471468662301,
      "flow_matching_loss": 0.03596591129899025,
      "grad_norm": 0.11458948999643326,
      "learning_rate": 6.922833419358013e-05,
      "loss": 0.0388,
      "relation_loss": 0.002865920262411237,
      "selected_pairs": 26914.2,
      "state_query_norm": 1065.0655395507813,
      "state_relation_loss": 0.0004561246983939782,
      "step": 12170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002865920262411237,
      "weighted_relation_loss": 0.002865920262411237
    },
    {
      "action_query_norm": 588.7036071777344,
      "action_relation_loss": 0.0023967577377334236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.393826005612722,
      "flow_matching_loss": 0.0327927578240633,
      "grad_norm": 0.12398968636989594,
      "learning_rate": 6.917744540905124e-05,
      "loss": 0.0356,
      "relation_loss": 0.0028389596845954657,
      "selected_pairs": 26782.5,
      "state_query_norm": 1077.3052001953124,
      "state_relation_loss": 0.00044220194686204193,
      "step": 12180,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0028389596845954657,
      "weighted_relation_loss": 0.0028389596845954657
    },
    {
      "action_query_norm": 594.2677062988281,
      "action_relation_loss": 0.0022594690788537266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.403180542563144,
      "flow_matching_loss": 0.03485199734568596,
      "grad_norm": 0.12060175091028214,
      "learning_rate": 6.912653332211787e-05,
      "loss": 0.0376,
      "relation_loss": 0.002718453761190176,
      "selected_pairs": 26827.0,
      "state_query_norm": 1074.6140869140625,
      "state_relation_loss": 0.00045898467651568354,
      "step": 12190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002718453761190176,
      "weighted_relation_loss": 0.002718453761190176
    },
    {
      "action_query_norm": 598.8714111328125,
      "action_relation_loss": 0.0023348458111286164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.412535079513564,
      "flow_matching_loss": 0.0349993048235774,
      "grad_norm": 0.11146512627601624,
      "learning_rate": 6.9075597994643e-05,
      "loss": 0.0378,
      "relation_loss": 0.002798677142709494,
      "selected_pairs": 27693.7,
      "state_query_norm": 1061.4846557617188,
      "state_relation_loss": 0.000463831314118579,
      "step": 12200,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.002798677142709494,
      "weighted_relation_loss": 0.002798677142709494
    },
    {
      "action_query_norm": 594.3979675292969,
      "action_relation_loss": 0.0023827782133594155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.421889616463986,
      "flow_matching_loss": 0.037187335826456544,
      "grad_norm": 0.11160606890916824,
      "learning_rate": 6.902463948851787e-05,
      "loss": 0.04,
      "relation_loss": 0.0028554487274959684,
      "selected_pairs": 26669.6,
      "state_query_norm": 1056.624072265625,
      "state_relation_loss": 0.000472670528688468,
      "step": 12210,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0028554487274959684,
      "weighted_relation_loss": 0.0028554487274959684
    },
    {
      "action_query_norm": 598.1979248046875,
      "action_relation_loss": 0.0023800351191312075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.431244153414406,
      "flow_matching_loss": 0.03275201190263033,
      "grad_norm": 0.1062617152929306,
      "learning_rate": 6.897365786566184e-05,
      "loss": 0.0356,
      "relation_loss": 0.0028236223151907325,
      "selected_pairs": 26777.7,
      "state_query_norm": 1082.2440307617187,
      "state_relation_loss": 0.00044358722225297243,
      "step": 12220,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0028236223151907325,
      "weighted_relation_loss": 0.0028236223151907325
    },
    {
      "action_query_norm": 600.8682678222656,
      "action_relation_loss": 0.0023283723974600436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.440598690364826,
      "flow_matching_loss": 0.033512264490127563,
      "grad_norm": 0.1279781311750412,
      "learning_rate": 6.892265318802242e-05,
      "loss": 0.0363,
      "relation_loss": 0.0027833838015794755,
      "selected_pairs": 27125.6,
      "state_query_norm": 1079.4659912109375,
      "state_relation_loss": 0.00045501140994019806,
      "step": 12230,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0027833838015794755,
      "weighted_relation_loss": 0.0027833838015794755
    },
    {
      "action_query_norm": 595.7564636230469,
      "action_relation_loss": 0.0024563393322750925,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.449953227315248,
      "flow_matching_loss": 0.0315348781645298,
      "grad_norm": 0.09675057232379913,
      "learning_rate": 6.887162551757507e-05,
      "loss": 0.0345,
      "relation_loss": 0.0029152984032407405,
      "selected_pairs": 26224.0,
      "state_query_norm": 1086.2616821289062,
      "state_relation_loss": 0.0004589590505929664,
      "step": 12240,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0029152984032407405,
      "weighted_relation_loss": 0.0029152984032407405
    },
    {
      "action_query_norm": 600.7739562988281,
      "action_relation_loss": 0.0024430318968370556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.459307764265668,
      "flow_matching_loss": 0.03195214793086052,
      "grad_norm": 0.1064244955778122,
      "learning_rate": 6.882057491632326e-05,
      "loss": 0.0349,
      "relation_loss": 0.00290464092977345,
      "selected_pairs": 26712.1,
      "state_query_norm": 1052.596875,
      "state_relation_loss": 0.0004616090445779264,
      "step": 12250,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00290464092977345,
      "weighted_relation_loss": 0.00290464092977345
    },
    {
      "action_query_norm": 597.5413146972656,
      "action_relation_loss": 0.002251963899470866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.46866230121609,
      "flow_matching_loss": 0.03529158476740122,
      "grad_norm": 0.14232762157917023,
      "learning_rate": 6.876950144629824e-05,
      "loss": 0.038,
      "relation_loss": 0.002709666336886585,
      "selected_pairs": 27131.5,
      "state_query_norm": 1082.1303466796876,
      "state_relation_loss": 0.00045770241704303773,
      "step": 12260,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.002709666336886585,
      "weighted_relation_loss": 0.002709666336886585
    },
    {
      "action_query_norm": 602.2042846679688,
      "action_relation_loss": 0.0022018693154677747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.47801683816651,
      "flow_matching_loss": 0.03457159698009491,
      "grad_norm": 0.12630358338356018,
      "learning_rate": 6.87184051695591e-05,
      "loss": 0.0372,
      "relation_loss": 0.002653695596382022,
      "selected_pairs": 27336.5,
      "state_query_norm": 1085.9982666015626,
      "state_relation_loss": 0.0004518262983765453,
      "step": 12270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.002653695596382022,
      "weighted_relation_loss": 0.002653695596382022
    },
    {
      "action_query_norm": 601.8292358398437,
      "action_relation_loss": 0.0024483122397214174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.487371375116933,
      "flow_matching_loss": 0.03492950089275837,
      "grad_norm": 0.11922919750213623,
      "learning_rate": 6.866728614819268e-05,
      "loss": 0.0378,
      "relation_loss": 0.00289951735176146,
      "selected_pairs": 26895.2,
      "state_query_norm": 1079.1410278320313,
      "state_relation_loss": 0.00045120508875697853,
      "step": 12280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00289951735176146,
      "weighted_relation_loss": 0.00289951735176146
    },
    {
      "action_query_norm": 595.1318786621093,
      "action_relation_loss": 0.002223655139096081,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.496725912067353,
      "flow_matching_loss": 0.03427658621221781,
      "grad_norm": 0.09766056388616562,
      "learning_rate": 6.861614444431337e-05,
      "loss": 0.037,
      "relation_loss": 0.0026822644285857677,
      "selected_pairs": 27101.7,
      "state_query_norm": 1084.7510131835938,
      "state_relation_loss": 0.0004586092953104526,
      "step": 12290,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0026822644285857677,
      "weighted_relation_loss": 0.0026822644285857677
    },
    {
      "action_query_norm": 594.3901428222656,
      "action_relation_loss": 0.002284350199624896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.506080449017773,
      "flow_matching_loss": 0.03390701077878475,
      "grad_norm": 0.10996246337890625,
      "learning_rate": 6.856498012006318e-05,
      "loss": 0.0367,
      "relation_loss": 0.0027571812504902484,
      "selected_pairs": 27151.1,
      "state_query_norm": 1083.6947875976562,
      "state_relation_loss": 0.0004728310013888404,
      "step": 12300,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0027571812504902484,
      "weighted_relation_loss": 0.0027571812504902484
    },
    {
      "action_query_norm": 595.6304077148437,
      "action_relation_loss": 0.0023261036025360227,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.515434985968195,
      "flow_matching_loss": 0.03637358564883471,
      "grad_norm": 0.1305568516254425,
      "learning_rate": 6.851379323761157e-05,
      "loss": 0.0392,
      "relation_loss": 0.0027775400085374713,
      "selected_pairs": 27295.4,
      "state_query_norm": 1082.2563354492188,
      "state_relation_loss": 0.00045143643801566214,
      "step": 12310,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0027775400085374713,
      "weighted_relation_loss": 0.0027775400085374713
    },
    {
      "action_query_norm": 601.6433715820312,
      "action_relation_loss": 0.0023431431502103804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.524789522918615,
      "flow_matching_loss": 0.033403844386339185,
      "grad_norm": 0.13653887808322906,
      "learning_rate": 6.846258385915545e-05,
      "loss": 0.0362,
      "relation_loss": 0.002804846107028425,
      "selected_pairs": 27245.4,
      "state_query_norm": 1074.7107055664062,
      "state_relation_loss": 0.0004617030004737899,
      "step": 12320,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.002804846107028425,
      "weighted_relation_loss": 0.002804846107028425
    },
    {
      "action_query_norm": 609.6140686035156,
      "action_relation_loss": 0.002206621947698295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.534144059869037,
      "flow_matching_loss": 0.0355842212215066,
      "grad_norm": 0.15272603929042816,
      "learning_rate": 6.841135204691902e-05,
      "loss": 0.0382,
      "relation_loss": 0.0026607936015352607,
      "selected_pairs": 26643.9,
      "state_query_norm": 1080.2798217773438,
      "state_relation_loss": 0.00045417165965773166,
      "step": 12330,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0026607936015352607,
      "weighted_relation_loss": 0.0026607936015352607
    },
    {
      "action_query_norm": 599.0171508789062,
      "action_relation_loss": 0.002498311665840447,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.543498596819457,
      "flow_matching_loss": 0.032686217501759526,
      "grad_norm": 0.11028816550970078,
      "learning_rate": 6.836009786315377e-05,
      "loss": 0.0356,
      "relation_loss": 0.0029544387012720107,
      "selected_pairs": 26457.6,
      "state_query_norm": 1111.0704956054688,
      "state_relation_loss": 0.0004561270558042452,
      "step": 12340,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0029544387012720107,
      "weighted_relation_loss": 0.0029544387012720107
    },
    {
      "action_query_norm": 604.4397338867187,
      "action_relation_loss": 0.0021576783852651714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.552853133769878,
      "flow_matching_loss": 0.03498399034142494,
      "grad_norm": 0.1338968724012375,
      "learning_rate": 6.830882137013839e-05,
      "loss": 0.0376,
      "relation_loss": 0.0026245305547490717,
      "selected_pairs": 26918.0,
      "state_query_norm": 1071.2074462890625,
      "state_relation_loss": 0.0004668521782150492,
      "step": 12350,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0026245305547490717,
      "weighted_relation_loss": 0.0026245305547490717
    },
    {
      "action_query_norm": 602.0271301269531,
      "action_relation_loss": 0.0023379871621727944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.5622076707203,
      "flow_matching_loss": 0.03243873957544565,
      "grad_norm": 0.1287156343460083,
      "learning_rate": 6.825752263017863e-05,
      "loss": 0.0352,
      "relation_loss": 0.0027832719031721355,
      "selected_pairs": 26665.1,
      "state_query_norm": 1075.53916015625,
      "state_relation_loss": 0.00044528476137202234,
      "step": 12360,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0027832719031721355,
      "weighted_relation_loss": 0.0027832719031721355
    },
    {
      "action_query_norm": 605.0549255371094,
      "action_relation_loss": 0.002223390876315534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.57156220767072,
      "flow_matching_loss": 0.033802751637995246,
      "grad_norm": 0.14084568619728088,
      "learning_rate": 6.820620170560731e-05,
      "loss": 0.0365,
      "relation_loss": 0.002672808850184083,
      "selected_pairs": 26782.0,
      "state_query_norm": 1083.8289184570312,
      "state_relation_loss": 0.00044941797968931494,
      "step": 12370,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002672808850184083,
      "weighted_relation_loss": 0.002672808850184083
    },
    {
      "action_query_norm": 596.9108947753906,
      "action_relation_loss": 0.002262460836209357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.580916744621142,
      "flow_matching_loss": 0.032972214929759505,
      "grad_norm": 0.10266339033842087,
      "learning_rate": 6.815485865878418e-05,
      "loss": 0.0357,
      "relation_loss": 0.0027101996121928094,
      "selected_pairs": 27293.4,
      "state_query_norm": 1078.470361328125,
      "state_relation_loss": 0.000447738790535368,
      "step": 12380,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0027101996121928094,
      "weighted_relation_loss": 0.0027101996121928094
    },
    {
      "action_query_norm": 595.9407165527343,
      "action_relation_loss": 0.0023667526664212348,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.590271281571562,
      "flow_matching_loss": 0.032048468664288524,
      "grad_norm": 0.09543468058109283,
      "learning_rate": 6.810349355209588e-05,
      "loss": 0.0349,
      "relation_loss": 0.0028316544834524394,
      "selected_pairs": 26920.8,
      "state_query_norm": 1082.5737182617188,
      "state_relation_loss": 0.00046490184031426906,
      "step": 12390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0028316544834524394,
      "weighted_relation_loss": 0.0028316544834524394
    },
    {
      "action_query_norm": 599.03798828125,
      "action_relation_loss": 0.0023857083870097995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.599625818521982,
      "flow_matching_loss": 0.03486110921949148,
      "grad_norm": 0.11121159791946411,
      "learning_rate": 6.805210644795588e-05,
      "loss": 0.0377,
      "relation_loss": 0.0028477939777076244,
      "selected_pairs": 27360.9,
      "state_query_norm": 1088.5473266601562,
      "state_relation_loss": 0.00046208557905629276,
      "step": 12400,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0028477939777076244,
      "weighted_relation_loss": 0.0028477939777076244
    },
    {
      "action_query_norm": 601.6607849121094,
      "action_relation_loss": 0.0022668806836009026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.608980355472404,
      "flow_matching_loss": 0.0342138385400176,
      "grad_norm": 0.12092415988445282,
      "learning_rate": 6.800069740880433e-05,
      "loss": 0.0369,
      "relation_loss": 0.0027186237974092366,
      "selected_pairs": 27626.9,
      "state_query_norm": 1076.7053466796874,
      "state_relation_loss": 0.0004517430759733543,
      "step": 12410,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0027186237974092366,
      "weighted_relation_loss": 0.0027186237974092366
    },
    {
      "action_query_norm": 599.9092407226562,
      "action_relation_loss": 0.002330519864335656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.618334892422824,
      "flow_matching_loss": 0.034175369329750536,
      "grad_norm": 0.11089842766523361,
      "learning_rate": 6.794926649710807e-05,
      "loss": 0.037,
      "relation_loss": 0.0027825551806017756,
      "selected_pairs": 26575.5,
      "state_query_norm": 1097.0464111328124,
      "state_relation_loss": 0.0004520353308180347,
      "step": 12420,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0027825551806017756,
      "weighted_relation_loss": 0.0027825551806017756
    },
    {
      "action_query_norm": 605.875146484375,
      "action_relation_loss": 0.0023400239180773495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.627689429373246,
      "flow_matching_loss": 0.03203383795917034,
      "grad_norm": 0.09485378116369247,
      "learning_rate": 6.789781377536052e-05,
      "loss": 0.0348,
      "relation_loss": 0.002780151693150401,
      "selected_pairs": 26898.8,
      "state_query_norm": 1097.8274169921874,
      "state_relation_loss": 0.00044012776925228536,
      "step": 12430,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.002780151693150401,
      "weighted_relation_loss": 0.002780151693150401
    },
    {
      "action_query_norm": 599.9137084960937,
      "action_relation_loss": 0.0021989618195220827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.637043966323667,
      "flow_matching_loss": 0.03255477324128151,
      "grad_norm": 0.11415298283100128,
      "learning_rate": 6.784633930608158e-05,
      "loss": 0.0352,
      "relation_loss": 0.0026605403050780295,
      "selected_pairs": 26990.0,
      "state_query_norm": 1092.7946166992188,
      "state_relation_loss": 0.000461578494287096,
      "step": 12440,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0026605403050780295,
      "weighted_relation_loss": 0.0026605403050780295
    },
    {
      "action_query_norm": 591.6056945800781,
      "action_relation_loss": 0.00250788563862443,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.646398503274089,
      "flow_matching_loss": 0.03454246073961258,
      "grad_norm": 0.11086437106132507,
      "learning_rate": 6.779484315181759e-05,
      "loss": 0.0375,
      "relation_loss": 0.0029654776910319923,
      "selected_pairs": 26112.4,
      "state_query_norm": 1099.440673828125,
      "state_relation_loss": 0.000457592008751817,
      "step": 12450,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.0029654776910319923,
      "weighted_relation_loss": 0.0029654776910319923
    },
    {
      "action_query_norm": 597.5889465332032,
      "action_relation_loss": 0.0024243003688752653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.655753040224509,
      "flow_matching_loss": 0.033630984090268615,
      "grad_norm": 0.11625493317842484,
      "learning_rate": 6.774332537514122e-05,
      "loss": 0.0365,
      "relation_loss": 0.0028762066271156073,
      "selected_pairs": 27047.1,
      "state_query_norm": 1085.55966796875,
      "state_relation_loss": 0.0004519062815234065,
      "step": 12460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0028762066271156073,
      "weighted_relation_loss": 0.0028762066271156073
    },
    {
      "action_query_norm": 591.1009033203125,
      "action_relation_loss": 0.002864615363068879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.665107577174929,
      "flow_matching_loss": 0.034513887763023374,
      "grad_norm": 0.11876765638589859,
      "learning_rate": 6.769178603865143e-05,
      "loss": 0.0378,
      "relation_loss": 0.0033185731386765837,
      "selected_pairs": 26392.8,
      "state_query_norm": 1080.7606506347656,
      "state_relation_loss": 0.00045395775232464074,
      "step": 12470,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0033185731386765837,
      "weighted_relation_loss": 0.0033185731386765837
    },
    {
      "action_query_norm": 595.9557189941406,
      "action_relation_loss": 0.0023171811131760477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.674462114125351,
      "flow_matching_loss": 0.03483507540076971,
      "grad_norm": 0.11510707437992096,
      "learning_rate": 6.764022520497337e-05,
      "loss": 0.0376,
      "relation_loss": 0.002777262218296528,
      "selected_pairs": 27253.1,
      "state_query_norm": 1089.2232177734375,
      "state_relation_loss": 0.0004600811196723953,
      "step": 12480,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.002777262218296528,
      "weighted_relation_loss": 0.002777262218296528
    },
    {
      "action_query_norm": 595.2483459472656,
      "action_relation_loss": 0.0024315960705280303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.683816651075771,
      "flow_matching_loss": 0.034066378697752954,
      "grad_norm": 0.11769759654998779,
      "learning_rate": 6.758864293675833e-05,
      "loss": 0.037,
      "relation_loss": 0.0028882227139547466,
      "selected_pairs": 26385.2,
      "state_query_norm": 1092.8181274414062,
      "state_relation_loss": 0.00045662666962016375,
      "step": 12490,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0028882227139547466,
      "weighted_relation_loss": 0.0028882227139547466
    },
    {
      "action_query_norm": 601.8346130371094,
      "action_relation_loss": 0.0023710794514045118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.693171188026193,
      "flow_matching_loss": 0.03469369858503342,
      "grad_norm": 0.11379878222942352,
      "learning_rate": 6.753703929668362e-05,
      "loss": 0.0375,
      "relation_loss": 0.0028296133736148477,
      "selected_pairs": 26874.4,
      "state_query_norm": 1094.1310791015626,
      "state_relation_loss": 0.0004585339513141662,
      "step": 12500,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0028296133736148477,
      "weighted_relation_loss": 0.0028296133736148477
    },
    {
      "action_query_norm": 601.1434875488281,
      "action_relation_loss": 0.002351343911141157,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.702525724976613,
      "flow_matching_loss": 0.03352461643517017,
      "grad_norm": 0.13618330657482147,
      "learning_rate": 6.748541434745253e-05,
      "loss": 0.0363,
      "relation_loss": 0.0027950541349127888,
      "selected_pairs": 26785.2,
      "state_query_norm": 1091.0864013671876,
      "state_relation_loss": 0.00044371019757818433,
      "step": 12510,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0027950541349127888,
      "weighted_relation_loss": 0.0027950541349127888
    },
    {
      "action_query_norm": 602.7100769042969,
      "action_relation_loss": 0.0023565758019685745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.711880261927035,
      "flow_matching_loss": 0.03386247679591179,
      "grad_norm": 0.140828937292099,
      "learning_rate": 6.743376815179424e-05,
      "loss": 0.0367,
      "relation_loss": 0.002815322717651725,
      "selected_pairs": 27300.0,
      "state_query_norm": 1095.0710083007812,
      "state_relation_loss": 0.00045874686911702157,
      "step": 12520,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.002815322717651725,
      "weighted_relation_loss": 0.002815322717651725
    },
    {
      "action_query_norm": 602.9900390625,
      "action_relation_loss": 0.00229814515914768,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.721234798877456,
      "flow_matching_loss": 0.03572481330484152,
      "grad_norm": 0.1025080606341362,
      "learning_rate": 6.738210077246376e-05,
      "loss": 0.0385,
      "relation_loss": 0.002755955373868346,
      "selected_pairs": 27669.9,
      "state_query_norm": 1088.22626953125,
      "state_relation_loss": 0.0004578102292725816,
      "step": 12530,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.002755955373868346,
      "weighted_relation_loss": 0.002755955373868346
    },
    {
      "action_query_norm": 603.2574096679688,
      "action_relation_loss": 0.0023172730347141624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.730589335827876,
      "flow_matching_loss": 0.03516951110213995,
      "grad_norm": 0.1051463708281517,
      "learning_rate": 6.733041227224181e-05,
      "loss": 0.038,
      "relation_loss": 0.002788258343935013,
      "selected_pairs": 26340.6,
      "state_query_norm": 1072.2880249023438,
      "state_relation_loss": 0.00047098531504161654,
      "step": 12540,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002788258343935013,
      "weighted_relation_loss": 0.002788258343935013
    },
    {
      "action_query_norm": 613.0945922851563,
      "action_relation_loss": 0.002395313838496804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.739943872778298,
      "flow_matching_loss": 0.03488606195896864,
      "grad_norm": 0.2010829746723175,
      "learning_rate": 6.72787027139348e-05,
      "loss": 0.0378,
      "relation_loss": 0.002869080868549645,
      "selected_pairs": 27527.7,
      "state_query_norm": 1093.482275390625,
      "state_relation_loss": 0.0004737670911708847,
      "step": 12550,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.002869080868549645,
      "weighted_relation_loss": 0.002869080868549645
    },
    {
      "action_query_norm": 601.8582336425782,
      "action_relation_loss": 0.0023055693949572743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.749298409728718,
      "flow_matching_loss": 0.03509730920195579,
      "grad_norm": 0.16568049788475037,
      "learning_rate": 6.72269721603747e-05,
      "loss": 0.0379,
      "relation_loss": 0.002761120768263936,
      "selected_pairs": 26340.6,
      "state_query_norm": 1091.9159912109376,
      "state_relation_loss": 0.00045555135293398055,
      "step": 12560,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002761120768263936,
      "weighted_relation_loss": 0.002761120768263936
    },
    {
      "action_query_norm": 605.8569519042969,
      "action_relation_loss": 0.0023553104838356376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.75865294667914,
      "flow_matching_loss": 0.03245116211473942,
      "grad_norm": 0.14318139851093292,
      "learning_rate": 6.717522067441904e-05,
      "loss": 0.0353,
      "relation_loss": 0.002804699703119695,
      "selected_pairs": 27059.0,
      "state_query_norm": 1088.0393676757812,
      "state_relation_loss": 0.0004493892251048237,
      "step": 12570,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002804699703119695,
      "weighted_relation_loss": 0.002804699703119695
    },
    {
      "action_query_norm": 612.7757080078125,
      "action_relation_loss": 0.0023029573960229753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.76800748362956,
      "flow_matching_loss": 0.0329938005656004,
      "grad_norm": 0.16696225106716156,
      "learning_rate": 6.712344831895074e-05,
      "loss": 0.0357,
      "relation_loss": 0.002748175454325974,
      "selected_pairs": 27197.7,
      "state_query_norm": 1105.1378295898437,
      "state_relation_loss": 0.0004452180612133816,
      "step": 12580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.002748175454325974,
      "weighted_relation_loss": 0.002748175454325974
    },
    {
      "action_query_norm": 611.3850280761719,
      "action_relation_loss": 0.002262839348986745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.77736202057998,
      "flow_matching_loss": 0.033697668462991714,
      "grad_norm": 0.17057105898857117,
      "learning_rate": 6.707165515687811e-05,
      "loss": 0.0364,
      "relation_loss": 0.0027138417353853583,
      "selected_pairs": 26733.8,
      "state_query_norm": 1089.802099609375,
      "state_relation_loss": 0.00045100242132321,
      "step": 12590,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0027138417353853583,
      "weighted_relation_loss": 0.0027138417353853583
    },
    {
      "action_query_norm": 602.2433532714844,
      "action_relation_loss": 0.0021525413496419786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.786716557530402,
      "flow_matching_loss": 0.030586194433271884,
      "grad_norm": 0.11631742119789124,
      "learning_rate": 6.70198412511347e-05,
      "loss": 0.0332,
      "relation_loss": 0.002606663550250232,
      "selected_pairs": 26995.4,
      "state_query_norm": 1082.1671752929688,
      "state_relation_loss": 0.00045412219769787044,
      "step": 12600,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002606663550250232,
      "weighted_relation_loss": 0.002606663550250232
    },
    {
      "action_query_norm": 598.9256896972656,
      "action_relation_loss": 0.002211976074613631,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.796071094480823,
      "flow_matching_loss": 0.033312239684164524,
      "grad_norm": 0.13085949420928955,
      "learning_rate": 6.69680066646793e-05,
      "loss": 0.036,
      "relation_loss": 0.0026614591479301453,
      "selected_pairs": 26547.1,
      "state_query_norm": 1098.0794555664063,
      "state_relation_loss": 0.0004494830674957484,
      "step": 12610,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0026614591479301453,
      "weighted_relation_loss": 0.0026614591479301453
    },
    {
      "action_query_norm": 600.27314453125,
      "action_relation_loss": 0.002364496421068907,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.805425631431245,
      "flow_matching_loss": 0.03442361168563366,
      "grad_norm": 0.10591136664152145,
      "learning_rate": 6.691615146049583e-05,
      "loss": 0.0372,
      "relation_loss": 0.0028023674385622144,
      "selected_pairs": 27256.1,
      "state_query_norm": 1107.3255249023437,
      "state_relation_loss": 0.0004378710204036906,
      "step": 12620,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0028023674385622144,
      "weighted_relation_loss": 0.0028023674385622144
    },
    {
      "action_query_norm": 602.3035217285156,
      "action_relation_loss": 0.002316775405779481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.814780168381665,
      "flow_matching_loss": 0.034630035236477855,
      "grad_norm": 0.11597934365272522,
      "learning_rate": 6.686427570159324e-05,
      "loss": 0.0374,
      "relation_loss": 0.0027642229106277227,
      "selected_pairs": 26552.8,
      "state_query_norm": 1102.5962280273438,
      "state_relation_loss": 0.000447447516489774,
      "step": 12630,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0027642229106277227,
      "weighted_relation_loss": 0.0027642229106277227
    },
    {
      "action_query_norm": 604.9446411132812,
      "action_relation_loss": 0.002385587221942842,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.824134705332087,
      "flow_matching_loss": 0.033129045739769936,
      "grad_norm": 0.11292941123247147,
      "learning_rate": 6.681237945100548e-05,
      "loss": 0.036,
      "relation_loss": 0.0028529898962005975,
      "selected_pairs": 26994.9,
      "state_query_norm": 1073.6228759765625,
      "state_relation_loss": 0.0004674026771681383,
      "step": 12640,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0028529898962005975,
      "weighted_relation_loss": 0.0028529898962005975
    },
    {
      "action_query_norm": 599.3655090332031,
      "action_relation_loss": 0.0022291664499789476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.833489242282507,
      "flow_matching_loss": 0.034449337236583236,
      "grad_norm": 0.12534639239311218,
      "learning_rate": 6.676046277179139e-05,
      "loss": 0.0371,
      "relation_loss": 0.0026905406499281525,
      "selected_pairs": 26810.6,
      "state_query_norm": 1096.9802978515625,
      "state_relation_loss": 0.00046137423196341843,
      "step": 12650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0026905406499281525,
      "weighted_relation_loss": 0.0026905406499281525
    },
    {
      "action_query_norm": 611.8672058105469,
      "action_relation_loss": 0.002330569364130497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.842843779232927,
      "flow_matching_loss": 0.03402473740279675,
      "grad_norm": 0.10803059488534927,
      "learning_rate": 6.670852572703462e-05,
      "loss": 0.0368,
      "relation_loss": 0.0027667622081935404,
      "selected_pairs": 27151.7,
      "state_query_norm": 1082.7875,
      "state_relation_loss": 0.0004361928586149588,
      "step": 12660,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0027667622081935404,
      "weighted_relation_loss": 0.0027667622081935404
    },
    {
      "action_query_norm": 607.9387573242187,
      "action_relation_loss": 0.0023497101152315735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.85219831618335,
      "flow_matching_loss": 0.03603212758898735,
      "grad_norm": 0.11840081959962845,
      "learning_rate": 6.665656837984359e-05,
      "loss": 0.0388,
      "relation_loss": 0.0027929908595979215,
      "selected_pairs": 26317.5,
      "state_query_norm": 1078.9505737304687,
      "state_relation_loss": 0.00044328075018711387,
      "step": 12670,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0027929908595979215,
      "weighted_relation_loss": 0.0027929908595979215
    },
    {
      "action_query_norm": 599.9649597167969,
      "action_relation_loss": 0.0022607383551076056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.86155285313377,
      "flow_matching_loss": 0.03307078424841166,
      "grad_norm": 0.15849047899246216,
      "learning_rate": 6.660459079335135e-05,
      "loss": 0.0358,
      "relation_loss": 0.0027233670232817532,
      "selected_pairs": 26664.8,
      "state_query_norm": 1081.885400390625,
      "state_relation_loss": 0.00046262869727797806,
      "step": 12680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0027233670232817532,
      "weighted_relation_loss": 0.0027233670232817532
    },
    {
      "action_query_norm": 602.7334411621093,
      "action_relation_loss": 0.002460913290269673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.870907390084191,
      "flow_matching_loss": 0.03333245422691107,
      "grad_norm": 0.12409272789955139,
      "learning_rate": 6.655259303071558e-05,
      "loss": 0.0362,
      "relation_loss": 0.0029142807703465222,
      "selected_pairs": 26989.2,
      "state_query_norm": 1083.0121215820313,
      "state_relation_loss": 0.00045336745388340204,
      "step": 12690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0029142807703465222,
      "weighted_relation_loss": 0.0029142807703465222
    },
    {
      "action_query_norm": 609.9051208496094,
      "action_relation_loss": 0.0021825019735842942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.880261927034612,
      "flow_matching_loss": 0.032720995508134364,
      "grad_norm": 0.12532974779605865,
      "learning_rate": 6.650057515511848e-05,
      "loss": 0.0353,
      "relation_loss": 0.0026272376300767062,
      "selected_pairs": 27255.2,
      "state_query_norm": 1102.5488159179688,
      "state_relation_loss": 0.0004447356652235612,
      "step": 12700,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0026272376300767062,
      "weighted_relation_loss": 0.0026272376300767062
    },
    {
      "action_query_norm": 594.6632751464844,
      "action_relation_loss": 0.0022826199419796467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.889616463985032,
      "flow_matching_loss": 0.034827405400574205,
      "grad_norm": 0.12099163234233856,
      "learning_rate": 6.644853722976667e-05,
      "loss": 0.0375,
      "relation_loss": 0.0027208998100832103,
      "selected_pairs": 26874.4,
      "state_query_norm": 1092.7980102539063,
      "state_relation_loss": 0.00043827984191011635,
      "step": 12710,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0027208998100832103,
      "weighted_relation_loss": 0.0027208998100832103
    },
    {
      "action_query_norm": 603.4538879394531,
      "action_relation_loss": 0.002250173897482455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.898971000935454,
      "flow_matching_loss": 0.03189084138721228,
      "grad_norm": 0.1448775976896286,
      "learning_rate": 6.639647931789113e-05,
      "loss": 0.0346,
      "relation_loss": 0.0026894503505900504,
      "selected_pairs": 26895.6,
      "state_query_norm": 1091.2690185546876,
      "state_relation_loss": 0.00043927648803219197,
      "step": 12720,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0026894503505900504,
      "weighted_relation_loss": 0.0026894503505900504
    },
    {
      "action_query_norm": 605.5553833007813,
      "action_relation_loss": 0.0022071030689403415,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.908325537885874,
      "flow_matching_loss": 0.03146015722304583,
      "grad_norm": 0.11967277526855469,
      "learning_rate": 6.634440148274713e-05,
      "loss": 0.0341,
      "relation_loss": 0.0026355101959779858,
      "selected_pairs": 27104.5,
      "state_query_norm": 1085.6001342773438,
      "state_relation_loss": 0.00042840708047151563,
      "step": 12730,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0026355101959779858,
      "weighted_relation_loss": 0.0026355101959779858
    },
    {
      "action_query_norm": 607.4049011230469,
      "action_relation_loss": 0.0022630233550444247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.917680074836296,
      "flow_matching_loss": 0.033396252617239955,
      "grad_norm": 0.11576607078313828,
      "learning_rate": 6.629230378761415e-05,
      "loss": 0.0361,
      "relation_loss": 0.0027025653282180427,
      "selected_pairs": 26848.5,
      "state_query_norm": 1107.7890991210938,
      "state_relation_loss": 0.0004395419964566827,
      "step": 12740,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0027025653282180427,
      "weighted_relation_loss": 0.0027025653282180427
    },
    {
      "action_query_norm": 606.1117736816407,
      "action_relation_loss": 0.0022774572018533943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.927034611786716,
      "flow_matching_loss": 0.033112581446766855,
      "grad_norm": 0.11869197338819504,
      "learning_rate": 6.624018629579581e-05,
      "loss": 0.0358,
      "relation_loss": 0.0027176304021850227,
      "selected_pairs": 26568.8,
      "state_query_norm": 1098.424365234375,
      "state_relation_loss": 0.0004401732061523944,
      "step": 12750,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0027176304021850227,
      "weighted_relation_loss": 0.0027176304021850227
    },
    {
      "action_query_norm": 614.5052856445312,
      "action_relation_loss": 0.0021867305389605463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.936389148737138,
      "flow_matching_loss": 0.03488588500767946,
      "grad_norm": 0.148989737033844,
      "learning_rate": 6.618804907061976e-05,
      "loss": 0.0375,
      "relation_loss": 0.002624740055762231,
      "selected_pairs": 26992.3,
      "state_query_norm": 1090.2285888671875,
      "state_relation_loss": 0.00043800949933938684,
      "step": 12760,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002624740055762231,
      "weighted_relation_loss": 0.002624740055762231
    },
    {
      "action_query_norm": 605.8968200683594,
      "action_relation_loss": 0.0022464474430307748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.945743685687559,
      "flow_matching_loss": 0.0332870626822114,
      "grad_norm": 0.14080913364887238,
      "learning_rate": 6.613589217543766e-05,
      "loss": 0.036,
      "relation_loss": 0.002687493385747075,
      "selected_pairs": 26709.7,
      "state_query_norm": 1092.7495727539062,
      "state_relation_loss": 0.0004410459514474496,
      "step": 12770,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.002687493385747075,
      "weighted_relation_loss": 0.002687493385747075
    },
    {
      "action_query_norm": 605.6385314941406,
      "action_relation_loss": 0.0022391112055629493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.955098222637979,
      "flow_matching_loss": 0.03364765029400587,
      "grad_norm": 0.11456132680177689,
      "learning_rate": 6.608371567362504e-05,
      "loss": 0.0363,
      "relation_loss": 0.0026634174166247247,
      "selected_pairs": 26477.0,
      "state_query_norm": 1089.9814086914062,
      "state_relation_loss": 0.00042430620815139266,
      "step": 12780,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0026634174166247247,
      "weighted_relation_loss": 0.0026634174166247247
    },
    {
      "action_query_norm": 602.8826477050782,
      "action_relation_loss": 0.0022859199438244105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.9644527595884,
      "flow_matching_loss": 0.034291849471628666,
      "grad_norm": 0.12740647792816162,
      "learning_rate": 6.60315196285813e-05,
      "loss": 0.037,
      "relation_loss": 0.0027030180674046276,
      "selected_pairs": 26526.1,
      "state_query_norm": 1094.6597900390625,
      "state_relation_loss": 0.00041709812066983434,
      "step": 12790,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0027030180674046276,
      "weighted_relation_loss": 0.0027030180674046276
    },
    {
      "action_query_norm": 600.7379760742188,
      "action_relation_loss": 0.0021738919313065708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.973807296538821,
      "flow_matching_loss": 0.03368379212915897,
      "grad_norm": 0.1372307389974594,
      "learning_rate": 6.597930410372954e-05,
      "loss": 0.0363,
      "relation_loss": 0.0026193806435912846,
      "selected_pairs": 26178.8,
      "state_query_norm": 1088.1764770507812,
      "state_relation_loss": 0.00044548872392624617,
      "step": 12800,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0026193806435912846,
      "weighted_relation_loss": 0.0026193806435912846
    },
    {
      "action_query_norm": 604.2021118164063,
      "action_relation_loss": 0.0022331704501993956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.983161833489243,
      "flow_matching_loss": 0.033027677051723006,
      "grad_norm": 0.10508355498313904,
      "learning_rate": 6.592706916251653e-05,
      "loss": 0.0357,
      "relation_loss": 0.002686502574943006,
      "selected_pairs": 26595.4,
      "state_query_norm": 1091.8400146484375,
      "state_relation_loss": 0.0004533321276539937,
      "step": 12810,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.002686502574943006,
      "weighted_relation_loss": 0.002686502574943006
    },
    {
      "action_query_norm": 610.5849609375,
      "action_relation_loss": 0.0022057359921745954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.992516370439663,
      "flow_matching_loss": 0.03352704979479313,
      "grad_norm": 0.12406573444604874,
      "learning_rate": 6.587481486841267e-05,
      "loss": 0.0362,
      "relation_loss": 0.0026507416740059854,
      "selected_pairs": 27595.8,
      "state_query_norm": 1088.8194458007813,
      "state_relation_loss": 0.00044500570802483705,
      "step": 12820,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.0026507416740059854,
      "weighted_relation_loss": 0.0026507416740059854
    },
    {
      "action_query_norm": 602.5350402832031,
      "action_relation_loss": 0.0022591177839785814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.001870907390083,
      "flow_matching_loss": 0.03355993404984474,
      "grad_norm": 0.12626861035823822,
      "learning_rate": 6.582254128491184e-05,
      "loss": 0.0363,
      "relation_loss": 0.0027027517091482876,
      "selected_pairs": 24547.5,
      "state_query_norm": 1085.203515625,
      "state_relation_loss": 0.00044363390479702504,
      "step": 12830,
      "valid_action_samples": 215.1,
      "weighted_auxiliary_loss": 0.0027027517091482876,
      "weighted_relation_loss": 0.0027027517091482876
    },
    {
      "action_query_norm": 601.0787414550781,
      "action_relation_loss": 0.0022487434791401027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.011225444340505,
      "flow_matching_loss": 0.03165528662502766,
      "grad_norm": 0.11085600405931473,
      "learning_rate": 6.577024847553139e-05,
      "loss": 0.0343,
      "relation_loss": 0.0026876211166381838,
      "selected_pairs": 26880.0,
      "state_query_norm": 1089.2202026367188,
      "state_relation_loss": 0.0004388776083942503,
      "step": 12840,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0026876211166381838,
      "weighted_relation_loss": 0.0026876211166381838
    },
    {
      "action_query_norm": 603.937744140625,
      "action_relation_loss": 0.002272061095573008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.020579981290926,
      "flow_matching_loss": 0.03143967427313328,
      "grad_norm": 0.11743800342082977,
      "learning_rate": 6.571793650381201e-05,
      "loss": 0.0341,
      "relation_loss": 0.0026985984994098546,
      "selected_pairs": 27223.7,
      "state_query_norm": 1102.2623291015625,
      "state_relation_loss": 0.00042653743585105983,
      "step": 12850,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0026985984994098546,
      "weighted_relation_loss": 0.0026985984994098546
    },
    {
      "action_query_norm": 597.3402099609375,
      "action_relation_loss": 0.0022256717318668963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.029934518241348,
      "flow_matching_loss": 0.032957667484879494,
      "grad_norm": 0.1278129369020462,
      "learning_rate": 6.566560543331766e-05,
      "loss": 0.0356,
      "relation_loss": 0.0026658405316993595,
      "selected_pairs": 26802.0,
      "state_query_norm": 1084.9295532226563,
      "state_relation_loss": 0.0004401687998324633,
      "step": 12860,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0026658405316993595,
      "weighted_relation_loss": 0.0026658405316993595
    },
    {
      "action_query_norm": 608.4322875976562,
      "action_relation_loss": 0.002124490647111088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.039289055191768,
      "flow_matching_loss": 0.031392197497189044,
      "grad_norm": 0.11687912046909332,
      "learning_rate": 6.561325532763554e-05,
      "loss": 0.034,
      "relation_loss": 0.002572151063941419,
      "selected_pairs": 26877.0,
      "state_query_norm": 1096.1437255859375,
      "state_relation_loss": 0.00044766041974071414,
      "step": 12870,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.002572151063941419,
      "weighted_relation_loss": 0.002572151063941419
    },
    {
      "action_query_norm": 601.8395263671875,
      "action_relation_loss": 0.0021200371789745986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.04864359214219,
      "flow_matching_loss": 0.033469361066818235,
      "grad_norm": 0.10824152082204819,
      "learning_rate": 6.556088625037597e-05,
      "loss": 0.036,
      "relation_loss": 0.0025672904914245008,
      "selected_pairs": 26921.9,
      "state_query_norm": 1090.596923828125,
      "state_relation_loss": 0.0004472533066291362,
      "step": 12880,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0025672904914245008,
      "weighted_relation_loss": 0.0025672904914245008
    },
    {
      "action_query_norm": 606.6485656738281,
      "action_relation_loss": 0.0020903979544527828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.05799812909261,
      "flow_matching_loss": 0.031311589665710925,
      "grad_norm": 0.14116831123828888,
      "learning_rate": 6.55084982651723e-05,
      "loss": 0.0338,
      "relation_loss": 0.0025205973768606784,
      "selected_pairs": 26688.2,
      "state_query_norm": 1082.2154052734375,
      "state_relation_loss": 0.00043019945151172577,
      "step": 12890,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0025205973768606784,
      "weighted_relation_loss": 0.0025205973768606784
    },
    {
      "action_query_norm": 603.1840881347656,
      "action_relation_loss": 0.0021656138240359725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.06735266604303,
      "flow_matching_loss": 0.032963427156209944,
      "grad_norm": 0.1307523250579834,
      "learning_rate": 6.54560914356809e-05,
      "loss": 0.0356,
      "relation_loss": 0.0025867816293612123,
      "selected_pairs": 27220.9,
      "state_query_norm": 1099.3361694335938,
      "state_relation_loss": 0.0004211678431602195,
      "step": 12900,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0025867816293612123,
      "weighted_relation_loss": 0.0025867816293612123
    },
    {
      "action_query_norm": 604.671875,
      "action_relation_loss": 0.0021160670439712703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.076707202993452,
      "flow_matching_loss": 0.03113273773342371,
      "grad_norm": 0.09293841570615768,
      "learning_rate": 6.540366582558097e-05,
      "loss": 0.0337,
      "relation_loss": 0.002553595951758325,
      "selected_pairs": 26476.0,
      "state_query_norm": 1084.8754760742188,
      "state_relation_loss": 0.0004375289281597361,
      "step": 12910,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.002553595951758325,
      "weighted_relation_loss": 0.002553595951758325
    },
    {
      "action_query_norm": 600.0126892089844,
      "action_relation_loss": 0.002268676506355405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.086061739943872,
      "flow_matching_loss": 0.03450290150940418,
      "grad_norm": 0.10434791445732117,
      "learning_rate": 6.53512214985746e-05,
      "loss": 0.0372,
      "relation_loss": 0.0027170491172000765,
      "selected_pairs": 26573.0,
      "state_query_norm": 1094.793115234375,
      "state_relation_loss": 0.0004483726545004174,
      "step": 12920,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0027170491172000765,
      "weighted_relation_loss": 0.0027170491172000765
    },
    {
      "action_query_norm": 608.8674987792969,
      "action_relation_loss": 0.002100022556260228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.095416276894294,
      "flow_matching_loss": 0.03171364292502403,
      "grad_norm": 0.11100537329912186,
      "learning_rate": 6.529875851838658e-05,
      "loss": 0.0342,
      "relation_loss": 0.002522237179800868,
      "selected_pairs": 27602.7,
      "state_query_norm": 1089.28134765625,
      "state_relation_loss": 0.00042221462936140595,
      "step": 12930,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.002522237179800868,
      "weighted_relation_loss": 0.002522237179800868
    },
    {
      "action_query_norm": 609.6849243164063,
      "action_relation_loss": 0.0021872769575566053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.104770813844715,
      "flow_matching_loss": 0.032893420197069645,
      "grad_norm": 0.08873893320560455,
      "learning_rate": 6.52462769487644e-05,
      "loss": 0.0355,
      "relation_loss": 0.002611237787641585,
      "selected_pairs": 27436.1,
      "state_query_norm": 1078.3434692382812,
      "state_relation_loss": 0.00042396083590574564,
      "step": 12940,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.002611237787641585,
      "weighted_relation_loss": 0.002611237787641585
    },
    {
      "action_query_norm": 612.4074645996094,
      "action_relation_loss": 0.002148902544286102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.114125350795135,
      "flow_matching_loss": 0.03173232804983854,
      "grad_norm": 0.1242876797914505,
      "learning_rate": 6.519377685347807e-05,
      "loss": 0.0343,
      "relation_loss": 0.0025808098958805203,
      "selected_pairs": 27340.1,
      "state_query_norm": 1111.221875,
      "state_relation_loss": 0.000431907331221737,
      "step": 12950,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0025808098958805203,
      "weighted_relation_loss": 0.0025808098958805203
    },
    {
      "action_query_norm": 596.7235229492187,
      "action_relation_loss": 0.002309044939465821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.123479887745557,
      "flow_matching_loss": 0.02980396933853626,
      "grad_norm": 0.11745313555002213,
      "learning_rate": 6.514125829632021e-05,
      "loss": 0.0326,
      "relation_loss": 0.0027547533391043544,
      "selected_pairs": 26253.3,
      "state_query_norm": 1102.4137329101563,
      "state_relation_loss": 0.0004457084025489166,
      "step": 12960,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0027547533391043544,
      "weighted_relation_loss": 0.0027547533391043544
    },
    {
      "action_query_norm": 610.2008728027344,
      "action_relation_loss": 0.001973303337581456,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.132834424695977,
      "flow_matching_loss": 0.030771730095148088,
      "grad_norm": 0.09621332585811615,
      "learning_rate": 6.508872134110577e-05,
      "loss": 0.0332,
      "relation_loss": 0.0024158000946044923,
      "selected_pairs": 26734.1,
      "state_query_norm": 1106.9825439453125,
      "state_relation_loss": 0.00044249677448533476,
      "step": 12970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0024158000946044923,
      "weighted_relation_loss": 0.0024158000946044923
    },
    {
      "action_query_norm": 602.0256713867187,
      "action_relation_loss": 0.002213268168270588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.142188961646399,
      "flow_matching_loss": 0.03311787024140358,
      "grad_norm": 0.12946631014347076,
      "learning_rate": 6.503616605167213e-05,
      "loss": 0.0358,
      "relation_loss": 0.0026470425305888056,
      "selected_pairs": 27276.7,
      "state_query_norm": 1100.1090087890625,
      "state_relation_loss": 0.00043377438851166517,
      "step": 12980,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0026470425305888056,
      "weighted_relation_loss": 0.0026470425305888056
    },
    {
      "action_query_norm": 608.9246887207031,
      "action_relation_loss": 0.002137751178815961,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.15154349859682,
      "flow_matching_loss": 0.03178839255124331,
      "grad_norm": 0.11695310473442078,
      "learning_rate": 6.498359249187893e-05,
      "loss": 0.0344,
      "relation_loss": 0.0025733225513249634,
      "selected_pairs": 26964.9,
      "state_query_norm": 1089.2505493164062,
      "state_relation_loss": 0.0004355713492259383,
      "step": 12990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0025733225513249634,
      "weighted_relation_loss": 0.0025733225513249634
    },
    {
      "action_query_norm": 600.6252014160157,
      "action_relation_loss": 0.0021900271996855734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.160898035547241,
      "flow_matching_loss": 0.031871323101222516,
      "grad_norm": 0.15664760768413544,
      "learning_rate": 6.493100072560799e-05,
      "loss": 0.0345,
      "relation_loss": 0.0026378074660897257,
      "selected_pairs": 27036.2,
      "state_query_norm": 1080.6196899414062,
      "state_relation_loss": 0.0004477802256587893,
      "step": 13000,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0026378074660897257,
      "weighted_relation_loss": 0.0026378074660897257
    },
    {
      "action_query_norm": 598.5071166992187,
      "action_relation_loss": 0.002169573574792594,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.170252572497661,
      "flow_matching_loss": 0.031176414340734482,
      "grad_norm": 0.13040763139724731,
      "learning_rate": 6.487839081676327e-05,
      "loss": 0.0338,
      "relation_loss": 0.0025997554650530217,
      "selected_pairs": 27311.5,
      "state_query_norm": 1079.9340698242188,
      "state_relation_loss": 0.0004301818960811943,
      "step": 13010,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0025997554650530217,
      "weighted_relation_loss": 0.0025997554650530217
    },
    {
      "action_query_norm": 604.1590515136719,
      "action_relation_loss": 0.0021736321737989782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.179607109448082,
      "flow_matching_loss": 0.030992406979203226,
      "grad_norm": 0.13288649916648865,
      "learning_rate": 6.482576282927076e-05,
      "loss": 0.0336,
      "relation_loss": 0.00262346335221082,
      "selected_pairs": 26485.3,
      "state_query_norm": 1103.0633178710937,
      "state_relation_loss": 0.00044983117550145835,
      "step": 13020,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00262346335221082,
      "weighted_relation_loss": 0.00262346335221082
    },
    {
      "action_query_norm": 605.16142578125,
      "action_relation_loss": 0.002039345051161945,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.188961646398504,
      "flow_matching_loss": 0.02986641861498356,
      "grad_norm": 0.12451792508363724,
      "learning_rate": 6.477311682707845e-05,
      "loss": 0.0323,
      "relation_loss": 0.0024712520418688653,
      "selected_pairs": 26778.8,
      "state_query_norm": 1102.9070190429688,
      "state_relation_loss": 0.00043190699943806976,
      "step": 13030,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0024712520418688653,
      "weighted_relation_loss": 0.0024712520418688653
    },
    {
      "action_query_norm": 605.106494140625,
      "action_relation_loss": 0.0021464433753862976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.198316183348924,
      "flow_matching_loss": 0.032043876871466634,
      "grad_norm": 0.11329551041126251,
      "learning_rate": 6.472045287415616e-05,
      "loss": 0.0346,
      "relation_loss": 0.0025726461550220846,
      "selected_pairs": 26755.5,
      "state_query_norm": 1101.772607421875,
      "state_relation_loss": 0.00042620276217348875,
      "step": 13040,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0025726461550220846,
      "weighted_relation_loss": 0.0025726461550220846
    },
    {
      "action_query_norm": 604.6617614746094,
      "action_relation_loss": 0.0020724707283079624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.207670720299346,
      "flow_matching_loss": 0.02893949709832668,
      "grad_norm": 0.10797414928674698,
      "learning_rate": 6.466777103449558e-05,
      "loss": 0.0315,
      "relation_loss": 0.002524193888530135,
      "selected_pairs": 27207.5,
      "state_query_norm": 1087.7265380859376,
      "state_relation_loss": 0.000451723177684471,
      "step": 13050,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.002524193888530135,
      "weighted_relation_loss": 0.002524193888530135
    },
    {
      "action_query_norm": 610.4782348632813,
      "action_relation_loss": 0.002050083223730326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.217025257249766,
      "flow_matching_loss": 0.03150821086019277,
      "grad_norm": 0.10437870770692825,
      "learning_rate": 6.461507137211012e-05,
      "loss": 0.034,
      "relation_loss": 0.002496363501995802,
      "selected_pairs": 27175.2,
      "state_query_norm": 1092.5959716796874,
      "state_relation_loss": 0.0004462803131900728,
      "step": 13060,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002496363501995802,
      "weighted_relation_loss": 0.002496363501995802
    },
    {
      "action_query_norm": 613.7081848144531,
      "action_relation_loss": 0.0021398361423052847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.226379794200188,
      "flow_matching_loss": 0.03263262249529362,
      "grad_norm": 0.10916348546743393,
      "learning_rate": 6.456235395103483e-05,
      "loss": 0.0352,
      "relation_loss": 0.002564799343235791,
      "selected_pairs": 27244.8,
      "state_query_norm": 1107.6657592773438,
      "state_relation_loss": 0.00042496321548242123,
      "step": 13070,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.002564799343235791,
      "weighted_relation_loss": 0.002564799343235791
    },
    {
      "action_query_norm": 605.9758544921875,
      "action_relation_loss": 0.002117685379926115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.235734331150608,
      "flow_matching_loss": 0.030719158053398133,
      "grad_norm": 0.12346835434436798,
      "learning_rate": 6.450961883532634e-05,
      "loss": 0.0333,
      "relation_loss": 0.0025565305491909384,
      "selected_pairs": 26650.8,
      "state_query_norm": 1097.8978881835938,
      "state_relation_loss": 0.0004388451314298436,
      "step": 13080,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0025565305491909384,
      "weighted_relation_loss": 0.0025565305491909384
    },
    {
      "action_query_norm": 602.7891296386719,
      "action_relation_loss": 0.002082657942082733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.245088868101028,
      "flow_matching_loss": 0.03252710457891226,
      "grad_norm": 0.1293681561946869,
      "learning_rate": 6.445686608906283e-05,
      "loss": 0.035,
      "relation_loss": 0.002519411803223193,
      "selected_pairs": 26325.9,
      "state_query_norm": 1090.5689575195313,
      "state_relation_loss": 0.00043675389024429024,
      "step": 13090,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.002519411803223193,
      "weighted_relation_loss": 0.002519411803223193
    },
    {
      "action_query_norm": 598.8243103027344,
      "action_relation_loss": 0.0021714982809498905,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.25444340505145,
      "flow_matching_loss": 0.03279796466231346,
      "grad_norm": 0.10355739295482635,
      "learning_rate": 6.44040957763438e-05,
      "loss": 0.0354,
      "relation_loss": 0.0026164171053096653,
      "selected_pairs": 27033.9,
      "state_query_norm": 1096.4259887695312,
      "state_relation_loss": 0.00044491881562862544,
      "step": 13100,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0026164171053096653,
      "weighted_relation_loss": 0.0026164171053096653
    },
    {
      "action_query_norm": 601.5588439941406,
      "action_relation_loss": 0.0021083674277178944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.26379794200187,
      "flow_matching_loss": 0.033075755834579466,
      "grad_norm": 0.16894939541816711,
      "learning_rate": 6.435130796129018e-05,
      "loss": 0.0356,
      "relation_loss": 0.002544720913283527,
      "selected_pairs": 26436.6,
      "state_query_norm": 1118.000390625,
      "state_relation_loss": 0.00043635347101371734,
      "step": 13110,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002544720913283527,
      "weighted_relation_loss": 0.002544720913283527
    },
    {
      "action_query_norm": 605.8345703125,
      "action_relation_loss": 0.002147778426297009,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.273152478952293,
      "flow_matching_loss": 0.03310327660292387,
      "grad_norm": 0.13023002445697784,
      "learning_rate": 6.429850270804416e-05,
      "loss": 0.0357,
      "relation_loss": 0.002591253980062902,
      "selected_pairs": 26513.1,
      "state_query_norm": 1111.0264038085938,
      "state_relation_loss": 0.00044347555085550995,
      "step": 13120,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002591253980062902,
      "weighted_relation_loss": 0.002591253980062902
    },
    {
      "action_query_norm": 612.6222351074218,
      "action_relation_loss": 0.0021718507166951896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.282507015902713,
      "flow_matching_loss": 0.0330713901668787,
      "grad_norm": 0.16061703860759735,
      "learning_rate": 6.424568008076909e-05,
      "loss": 0.0357,
      "relation_loss": 0.0026189177762717007,
      "selected_pairs": 27102.7,
      "state_query_norm": 1104.6815551757813,
      "state_relation_loss": 0.0004470670741284266,
      "step": 13130,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0026189177762717007,
      "weighted_relation_loss": 0.0026189177762717007
    },
    {
      "action_query_norm": 608.9313171386718,
      "action_relation_loss": 0.002111037028953433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.291861552853133,
      "flow_matching_loss": 0.030861887149512768,
      "grad_norm": 0.1310485154390335,
      "learning_rate": 6.419284014364943e-05,
      "loss": 0.0334,
      "relation_loss": 0.002562099159695208,
      "selected_pairs": 26779.5,
      "state_query_norm": 1094.1035400390624,
      "state_relation_loss": 0.0004510621336521581,
      "step": 13140,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002562099159695208,
      "weighted_relation_loss": 0.002562099159695208
    },
    {
      "action_query_norm": 609.2576293945312,
      "action_relation_loss": 0.00206926321843639,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.301216089803555,
      "flow_matching_loss": 0.030805003270506857,
      "grad_norm": 0.10945121943950653,
      "learning_rate": 6.41399829608907e-05,
      "loss": 0.0333,
      "relation_loss": 0.002495611645281315,
      "selected_pairs": 26763.6,
      "state_query_norm": 1109.8050537109375,
      "state_relation_loss": 0.000426348447217606,
      "step": 13150,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.002495611645281315,
      "weighted_relation_loss": 0.002495611645281315
    },
    {
      "action_query_norm": 605.194091796875,
      "action_relation_loss": 0.0021595486672595144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.310570626753975,
      "flow_matching_loss": 0.03245875649154186,
      "grad_norm": 0.12402195483446121,
      "learning_rate": 6.408710859671938e-05,
      "loss": 0.0351,
      "relation_loss": 0.002602544054389,
      "selected_pairs": 27382.1,
      "state_query_norm": 1104.5512084960938,
      "state_relation_loss": 0.0004429953929502517,
      "step": 13160,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.002602544054389,
      "weighted_relation_loss": 0.002602544054389
    },
    {
      "action_query_norm": 607.2007385253906,
      "action_relation_loss": 0.0021984797669574617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.319925163704397,
      "flow_matching_loss": 0.03303167354315519,
      "grad_norm": 0.13460001349449158,
      "learning_rate": 6.403421711538278e-05,
      "loss": 0.0357,
      "relation_loss": 0.0026227773865684866,
      "selected_pairs": 27035.4,
      "state_query_norm": 1099.7504150390625,
      "state_relation_loss": 0.0004242976399837062,
      "step": 13170,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0026227773865684866,
      "weighted_relation_loss": 0.0026227773865684866
    },
    {
      "action_query_norm": 603.3445251464843,
      "action_relation_loss": 0.002050523681100458,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.329279700654817,
      "flow_matching_loss": 0.03422455880790949,
      "grad_norm": 0.11029355973005295,
      "learning_rate": 6.398130858114903e-05,
      "loss": 0.0367,
      "relation_loss": 0.0025035129860043527,
      "selected_pairs": 26919.8,
      "state_query_norm": 1095.6237548828126,
      "state_relation_loss": 0.0004529892758000642,
      "step": 13180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0025035129860043527,
      "weighted_relation_loss": 0.0025035129860043527
    },
    {
      "action_query_norm": 610.5588806152343,
      "action_relation_loss": 0.002082572563085705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.33863423760524,
      "flow_matching_loss": 0.03239958826452494,
      "grad_norm": 0.12648087739944458,
      "learning_rate": 6.392838305830701e-05,
      "loss": 0.0349,
      "relation_loss": 0.002536133024841547,
      "selected_pairs": 26835.9,
      "state_query_norm": 1097.8651611328125,
      "state_relation_loss": 0.0004535604850389063,
      "step": 13190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002536133024841547,
      "weighted_relation_loss": 0.002536133024841547
    },
    {
      "action_query_norm": 608.2232360839844,
      "action_relation_loss": 0.002140212384983897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.34798877455566,
      "flow_matching_loss": 0.032102938555181024,
      "grad_norm": 0.11283863335847855,
      "learning_rate": 6.387544061116621e-05,
      "loss": 0.0347,
      "relation_loss": 0.0025723955594003202,
      "selected_pairs": 26901.6,
      "state_query_norm": 1103.8960083007812,
      "state_relation_loss": 0.00043218315695412455,
      "step": 13200,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0025723955594003202,
      "weighted_relation_loss": 0.0025723955594003202
    },
    {
      "action_query_norm": 593.672509765625,
      "action_relation_loss": 0.002202288038097322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.35734331150608,
      "flow_matching_loss": 0.031041776388883592,
      "grad_norm": 0.13761134445667267,
      "learning_rate": 6.382248130405671e-05,
      "loss": 0.0337,
      "relation_loss": 0.0026657531270757318,
      "selected_pairs": 26781.2,
      "state_query_norm": 1084.1928344726562,
      "state_relation_loss": 0.0004634650831576437,
      "step": 13210,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0026657531270757318,
      "weighted_relation_loss": 0.0026657531270757318
    },
    {
      "action_query_norm": 601.4317321777344,
      "action_relation_loss": 0.0022449620650149883,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.366697848456502,
      "flow_matching_loss": 0.03241296987980604,
      "grad_norm": 0.0985293760895729,
      "learning_rate": 6.376950520132905e-05,
      "loss": 0.0351,
      "relation_loss": 0.00268546671140939,
      "selected_pairs": 26045.0,
      "state_query_norm": 1098.5028564453125,
      "state_relation_loss": 0.00044050465221516787,
      "step": 13220,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.00268546671140939,
      "weighted_relation_loss": 0.00268546671140939
    },
    {
      "action_query_norm": 611.5924560546875,
      "action_relation_loss": 0.002060225419700146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.376052385406922,
      "flow_matching_loss": 0.03200349286198616,
      "grad_norm": 0.09995131194591522,
      "learning_rate": 6.371651236735418e-05,
      "loss": 0.0345,
      "relation_loss": 0.002498094271868467,
      "selected_pairs": 27433.1,
      "state_query_norm": 1076.776611328125,
      "state_relation_loss": 0.00043786882306449113,
      "step": 13230,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.002498094271868467,
      "weighted_relation_loss": 0.002498094271868467
    },
    {
      "action_query_norm": 613.3797485351563,
      "action_relation_loss": 0.0021810464677400887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.385406922357344,
      "flow_matching_loss": 0.03100866302847862,
      "grad_norm": 0.1224338710308075,
      "learning_rate": 6.366350286652341e-05,
      "loss": 0.0336,
      "relation_loss": 0.00262028609868139,
      "selected_pairs": 26117.2,
      "state_query_norm": 1103.1921020507812,
      "state_relation_loss": 0.0004392396367620677,
      "step": 13240,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.00262028609868139,
      "weighted_relation_loss": 0.00262028609868139
    },
    {
      "action_query_norm": 612.938427734375,
      "action_relation_loss": 0.002046879567205906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.394761459307764,
      "flow_matching_loss": 0.03126925453543663,
      "grad_norm": 0.09821475297212601,
      "learning_rate": 6.361047676324827e-05,
      "loss": 0.0338,
      "relation_loss": 0.0024950697319582106,
      "selected_pairs": 26666.4,
      "state_query_norm": 1105.6857055664063,
      "state_relation_loss": 0.0004481901996769011,
      "step": 13250,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0024950697319582106,
      "weighted_relation_loss": 0.0024950697319582106
    },
    {
      "action_query_norm": 605.6651794433594,
      "action_relation_loss": 0.0020179801736958324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.404115996258184,
      "flow_matching_loss": 0.029964292608201503,
      "grad_norm": 0.09964530915021896,
      "learning_rate": 6.355743412196047e-05,
      "loss": 0.0324,
      "relation_loss": 0.002445649611763656,
      "selected_pairs": 26467.0,
      "state_query_norm": 1089.9562377929688,
      "state_relation_loss": 0.0004276694089639932,
      "step": 13260,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002445649611763656,
      "weighted_relation_loss": 0.002445649611763656
    },
    {
      "action_query_norm": 604.3653747558594,
      "action_relation_loss": 0.0020856383023783566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.413470533208606,
      "flow_matching_loss": 0.03251039627939463,
      "grad_norm": 0.11595912277698517,
      "learning_rate": 6.350437500711184e-05,
      "loss": 0.035,
      "relation_loss": 0.0025058031314983964,
      "selected_pairs": 26553.9,
      "state_query_norm": 1090.909765625,
      "state_relation_loss": 0.00042016482912003993,
      "step": 13270,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0025058031314983964,
      "weighted_relation_loss": 0.0025058031314983964
    },
    {
      "action_query_norm": 606.3575866699218,
      "action_relation_loss": 0.002161407587118447,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.422825070159027,
      "flow_matching_loss": 0.03110039383172989,
      "grad_norm": 0.12341142445802689,
      "learning_rate": 6.345129948317418e-05,
      "loss": 0.0337,
      "relation_loss": 0.0026110158767551185,
      "selected_pairs": 27041.1,
      "state_query_norm": 1088.6707885742187,
      "state_relation_loss": 0.0004496083362028003,
      "step": 13280,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0026110158767551185,
      "weighted_relation_loss": 0.0026110158767551185
    },
    {
      "action_query_norm": 610.735205078125,
      "action_relation_loss": 0.002001738839317113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.432179607109449,
      "flow_matching_loss": 0.03209179323166609,
      "grad_norm": 0.09505422413349152,
      "learning_rate": 6.33982076146393e-05,
      "loss": 0.0345,
      "relation_loss": 0.0024279871955513953,
      "selected_pairs": 26554.9,
      "state_query_norm": 1092.309619140625,
      "state_relation_loss": 0.0004262483620550483,
      "step": 13290,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0024279871955513953,
      "weighted_relation_loss": 0.0024279871955513953
    },
    {
      "action_query_norm": 606.3759460449219,
      "action_relation_loss": 0.0021056412952020764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.441534144059869,
      "flow_matching_loss": 0.030439609847962856,
      "grad_norm": 0.09648609161376953,
      "learning_rate": 6.334509946601879e-05,
      "loss": 0.033,
      "relation_loss": 0.002537339320406318,
      "selected_pairs": 27034.5,
      "state_query_norm": 1090.0376342773438,
      "state_relation_loss": 0.00043169800483155996,
      "step": 13300,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.002537339320406318,
      "weighted_relation_loss": 0.002537339320406318
    },
    {
      "action_query_norm": 597.6647583007813,
      "action_relation_loss": 0.0020736352307721972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.45088868101029,
      "flow_matching_loss": 0.03197192661464214,
      "grad_norm": 0.10408243536949158,
      "learning_rate": 6.329197510184406e-05,
      "loss": 0.0345,
      "relation_loss": 0.00249721088912338,
      "selected_pairs": 26698.5,
      "state_query_norm": 1107.0596557617187,
      "state_relation_loss": 0.00042357565835118296,
      "step": 13310,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00249721088912338,
      "weighted_relation_loss": 0.00249721088912338
    },
    {
      "action_query_norm": 608.8142517089843,
      "action_relation_loss": 0.002055315603502095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.460243217960711,
      "flow_matching_loss": 0.031754080578684805,
      "grad_norm": 0.13243892788887024,
      "learning_rate": 6.323883458666624e-05,
      "loss": 0.0342,
      "relation_loss": 0.0024725200375542043,
      "selected_pairs": 26439.3,
      "state_query_norm": 1101.1417358398437,
      "state_relation_loss": 0.00041720442241057756,
      "step": 13320,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0024725200375542043,
      "weighted_relation_loss": 0.0024725200375542043
    },
    {
      "action_query_norm": 608.2134460449219,
      "action_relation_loss": 0.002014367934316397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.469597754911131,
      "flow_matching_loss": 0.031126100197434427,
      "grad_norm": 0.11440060287714005,
      "learning_rate": 6.318567798505605e-05,
      "loss": 0.0336,
      "relation_loss": 0.002444853214547038,
      "selected_pairs": 26685.4,
      "state_query_norm": 1118.444921875,
      "state_relation_loss": 0.0004304852860514075,
      "step": 13330,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002444853214547038,
      "weighted_relation_loss": 0.002444853214547038
    },
    {
      "action_query_norm": 614.5931396484375,
      "action_relation_loss": 0.0020871978253126145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.478952291861553,
      "flow_matching_loss": 0.03343084584921598,
      "grad_norm": 0.09991472959518433,
      "learning_rate": 6.313250536160378e-05,
      "loss": 0.0359,
      "relation_loss": 0.002518180012702942,
      "selected_pairs": 26621.0,
      "state_query_norm": 1102.1001220703124,
      "state_relation_loss": 0.0004309821728384122,
      "step": 13340,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002518180012702942,
      "weighted_relation_loss": 0.002518180012702942
    },
    {
      "action_query_norm": 612.7509216308594,
      "action_relation_loss": 0.001969197520520538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.488306828811973,
      "flow_matching_loss": 0.03062810394912958,
      "grad_norm": 0.10121588408946991,
      "learning_rate": 6.307931678091917e-05,
      "loss": 0.033,
      "relation_loss": 0.00239233432803303,
      "selected_pairs": 27183.3,
      "state_query_norm": 1091.9780883789062,
      "state_relation_loss": 0.00042313678422942755,
      "step": 13350,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.00239233432803303,
      "weighted_relation_loss": 0.00239233432803303
    },
    {
      "action_query_norm": 605.2272583007813,
      "action_relation_loss": 0.00211799843236804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.497661365762395,
      "flow_matching_loss": 0.0306370934471488,
      "grad_norm": 0.1314411759376526,
      "learning_rate": 6.302611230763138e-05,
      "loss": 0.0332,
      "relation_loss": 0.0025534558109939097,
      "selected_pairs": 26871.9,
      "state_query_norm": 1103.5982421875,
      "state_relation_loss": 0.00043545736698433755,
      "step": 13360,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0025534558109939097,
      "weighted_relation_loss": 0.0025534558109939097
    },
    {
      "action_query_norm": 605.4099548339843,
      "action_relation_loss": 0.001958121603820473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.507015902712816,
      "flow_matching_loss": 0.029899484105408193,
      "grad_norm": 0.10692600160837173,
      "learning_rate": 6.297289200638887e-05,
      "loss": 0.0323,
      "relation_loss": 0.002397816302254796,
      "selected_pairs": 26944.6,
      "state_query_norm": 1090.773779296875,
      "state_relation_loss": 0.00043969467224087567,
      "step": 13370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.002397816302254796,
      "weighted_relation_loss": 0.002397816302254796
    },
    {
      "action_query_norm": 605.2557189941406,
      "action_relation_loss": 0.0020937861991114914,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.516370439663238,
      "flow_matching_loss": 0.02842062134295702,
      "grad_norm": 0.10889061540365219,
      "learning_rate": 6.29196559418593e-05,
      "loss": 0.0309,
      "relation_loss": 0.0025163703365251424,
      "selected_pairs": 27368.6,
      "state_query_norm": 1067.02783203125,
      "state_relation_loss": 0.0004225841141305864,
      "step": 13380,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0025163703365251424,
      "weighted_relation_loss": 0.0025163703365251424
    },
    {
      "action_query_norm": 605.4597290039062,
      "action_relation_loss": 0.0020795216783881187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.525724976613658,
      "flow_matching_loss": 0.031689762510359286,
      "grad_norm": 0.12372899800539017,
      "learning_rate": 6.286640417872951e-05,
      "loss": 0.0342,
      "relation_loss": 0.0025077291531488298,
      "selected_pairs": 26754.9,
      "state_query_norm": 1088.6429443359375,
      "state_relation_loss": 0.0004282074747607112,
      "step": 13390,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0025077291531488298,
      "weighted_relation_loss": 0.0025077291531488298
    },
    {
      "action_query_norm": 608.4028259277344,
      "action_relation_loss": 0.0020470713498070836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.535079513564078,
      "flow_matching_loss": 0.03137514442205429,
      "grad_norm": 0.10255860537290573,
      "learning_rate": 6.281313678170542e-05,
      "loss": 0.0338,
      "relation_loss": 0.0024728438816964625,
      "selected_pairs": 26960.9,
      "state_query_norm": 1105.9679931640626,
      "state_relation_loss": 0.00042577253188937905,
      "step": 13400,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0024728438816964625,
      "weighted_relation_loss": 0.0024728438816964625
    },
    {
      "action_query_norm": 612.4325439453125,
      "action_relation_loss": 0.0020388708682730792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.5444340505145,
      "flow_matching_loss": 0.03181691747158766,
      "grad_norm": 0.11446128785610199,
      "learning_rate": 6.275985381551194e-05,
      "loss": 0.0343,
      "relation_loss": 0.0024620697367936373,
      "selected_pairs": 27107.3,
      "state_query_norm": 1106.2573852539062,
      "state_relation_loss": 0.0004231988685205579,
      "step": 13410,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0024620697367936373,
      "weighted_relation_loss": 0.0024620697367936373
    },
    {
      "action_query_norm": 608.5149597167969,
      "action_relation_loss": 0.002076579385902733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.55378858746492,
      "flow_matching_loss": 0.03196210879832506,
      "grad_norm": 0.13221623003482819,
      "learning_rate": 6.270655534489292e-05,
      "loss": 0.0345,
      "relation_loss": 0.0024905255064368247,
      "selected_pairs": 26902.8,
      "state_query_norm": 1105.5233764648438,
      "state_relation_loss": 0.00041394611762370916,
      "step": 13420,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0024905255064368247,
      "weighted_relation_loss": 0.0024905255064368247
    },
    {
      "action_query_norm": 602.6429321289063,
      "action_relation_loss": 0.0020481805317103864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.563143124415342,
      "flow_matching_loss": 0.03215092401951551,
      "grad_norm": 0.11999957263469696,
      "learning_rate": 6.265324143461098e-05,
      "loss": 0.0346,
      "relation_loss": 0.0024824495194479823,
      "selected_pairs": 27383.8,
      "state_query_norm": 1096.53994140625,
      "state_relation_loss": 0.0004342689906479791,
      "step": 13430,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0024824495194479823,
      "weighted_relation_loss": 0.0024824495194479823
    },
    {
      "action_query_norm": 605.9620178222656,
      "action_relation_loss": 0.002041645220015198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.572497661365762,
      "flow_matching_loss": 0.032141545228660105,
      "grad_norm": 0.10228121280670166,
      "learning_rate": 6.259991214944758e-05,
      "loss": 0.0346,
      "relation_loss": 0.0024609218584373594,
      "selected_pairs": 26664.5,
      "state_query_norm": 1115.4046142578125,
      "state_relation_loss": 0.00041927666461560875,
      "step": 13440,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0024609218584373594,
      "weighted_relation_loss": 0.0024609218584373594
    },
    {
      "action_query_norm": 605.6233459472656,
      "action_relation_loss": 0.0021741521428339182,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.581852198316183,
      "flow_matching_loss": 0.03177440222352743,
      "grad_norm": 0.12174858897924423,
      "learning_rate": 6.254656755420283e-05,
      "loss": 0.0344,
      "relation_loss": 0.0026020998833701015,
      "selected_pairs": 26691.5,
      "state_query_norm": 1107.1172973632813,
      "state_relation_loss": 0.00042794774926733223,
      "step": 13450,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0026020998833701015,
      "weighted_relation_loss": 0.0026020998833701015
    },
    {
      "action_query_norm": 618.3367492675782,
      "action_relation_loss": 0.002032770961523056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.591206735266605,
      "flow_matching_loss": 0.031700849533081055,
      "grad_norm": 0.1385974884033203,
      "learning_rate": 6.249320771369545e-05,
      "loss": 0.0342,
      "relation_loss": 0.002458686614409089,
      "selected_pairs": 26973.5,
      "state_query_norm": 1102.42392578125,
      "state_relation_loss": 0.00042591567325871437,
      "step": 13460,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002458686614409089,
      "weighted_relation_loss": 0.002458686614409089
    },
    {
      "action_query_norm": 614.5244384765625,
      "action_relation_loss": 0.0020447824499569833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.600561272217025,
      "flow_matching_loss": 0.029990683868527414,
      "grad_norm": 0.1258295625448227,
      "learning_rate": 6.243983269276263e-05,
      "loss": 0.0325,
      "relation_loss": 0.002483629365451634,
      "selected_pairs": 26165.8,
      "state_query_norm": 1106.9603271484375,
      "state_relation_loss": 0.0004388469213154167,
      "step": 13470,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.002483629365451634,
      "weighted_relation_loss": 0.002483629365451634
    },
    {
      "action_query_norm": 611.2072021484375,
      "action_relation_loss": 0.0020921544055454433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.609915809167447,
      "flow_matching_loss": 0.03173977453261614,
      "grad_norm": 0.11949027329683304,
      "learning_rate": 6.238644255626012e-05,
      "loss": 0.0343,
      "relation_loss": 0.0025333869038149715,
      "selected_pairs": 26592.1,
      "state_query_norm": 1104.3044311523438,
      "state_relation_loss": 0.000441232489538379,
      "step": 13480,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0025333869038149715,
      "weighted_relation_loss": 0.0025333869038149715
    },
    {
      "action_query_norm": 617.0270141601562,
      "action_relation_loss": 0.0020446377689950167,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.619270346117867,
      "flow_matching_loss": 0.032834955118596555,
      "grad_norm": 0.13953176140785217,
      "learning_rate": 6.233303736906193e-05,
      "loss": 0.0353,
      "relation_loss": 0.0024796420708298682,
      "selected_pairs": 27061.4,
      "state_query_norm": 1092.9013061523438,
      "state_relation_loss": 0.0004350042872829363,
      "step": 13490,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0024796420708298682,
      "weighted_relation_loss": 0.0024796420708298682
    },
    {
      "action_query_norm": 603.8503234863281,
      "action_relation_loss": 0.002187376283109188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.628624883068287,
      "flow_matching_loss": 0.03174741733819246,
      "grad_norm": 0.1228722482919693,
      "learning_rate": 6.227961719606039e-05,
      "loss": 0.0344,
      "relation_loss": 0.002603114419616759,
      "selected_pairs": 27171.5,
      "state_query_norm": 1104.5678344726562,
      "state_relation_loss": 0.00041573814232833686,
      "step": 13500,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002603114419616759,
      "weighted_relation_loss": 0.002603114419616759
    },
    {
      "action_query_norm": 612.6406494140625,
      "action_relation_loss": 0.002035864850040525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.63797942001871,
      "flow_matching_loss": 0.03416495453566313,
      "grad_norm": 0.10483758896589279,
      "learning_rate": 6.222618210216608e-05,
      "loss": 0.0366,
      "relation_loss": 0.002459517470560968,
      "selected_pairs": 26555.0,
      "state_query_norm": 1108.4627807617187,
      "state_relation_loss": 0.00042365260305814444,
      "step": 13510,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.002459517470560968,
      "weighted_relation_loss": 0.002459517470560968
    },
    {
      "action_query_norm": 616.601318359375,
      "action_relation_loss": 0.0020157050574198367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.64733395696913,
      "flow_matching_loss": 0.03163347486406565,
      "grad_norm": 0.13326387107372284,
      "learning_rate": 6.217273215230767e-05,
      "loss": 0.0341,
      "relation_loss": 0.002436502301134169,
      "selected_pairs": 26808.8,
      "state_query_norm": 1103.0445678710937,
      "state_relation_loss": 0.0004207972757285461,
      "step": 13520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002436502301134169,
      "weighted_relation_loss": 0.002436502301134169
    },
    {
      "action_query_norm": 610.2033813476562,
      "action_relation_loss": 0.002091418590862304,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.656688493919551,
      "flow_matching_loss": 0.0313785394653678,
      "grad_norm": 0.10633352398872375,
      "learning_rate": 6.211926741143188e-05,
      "loss": 0.0339,
      "relation_loss": 0.0025239855283871295,
      "selected_pairs": 27130.9,
      "state_query_norm": 1101.1747192382813,
      "state_relation_loss": 0.0004325669608078897,
      "step": 13530,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0025239855283871295,
      "weighted_relation_loss": 0.0025239855283871295
    },
    {
      "action_query_norm": 610.0370544433594,
      "action_relation_loss": 0.002069113589823246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.666043030869972,
      "flow_matching_loss": 0.0321712963283062,
      "grad_norm": 0.12847520411014557,
      "learning_rate": 6.20657879445034e-05,
      "loss": 0.0347,
      "relation_loss": 0.0025010933401063084,
      "selected_pairs": 26985.8,
      "state_query_norm": 1103.0277954101562,
      "state_relation_loss": 0.00043197976774536075,
      "step": 13540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0025010933401063084,
      "weighted_relation_loss": 0.0025010933401063084
    },
    {
      "action_query_norm": 602.9588073730469,
      "action_relation_loss": 0.002052867296151817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.675397567820394,
      "flow_matching_loss": 0.031611287407577036,
      "grad_norm": 0.12266813218593597,
      "learning_rate": 6.201229381650485e-05,
      "loss": 0.0341,
      "relation_loss": 0.002481533237732947,
      "selected_pairs": 26808.9,
      "state_query_norm": 1097.3873901367188,
      "state_relation_loss": 0.0004286659153876826,
      "step": 13550,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002481533237732947,
      "weighted_relation_loss": 0.002481533237732947
    },
    {
      "action_query_norm": 615.1520080566406,
      "action_relation_loss": 0.002001380024012178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.684752104770814,
      "flow_matching_loss": 0.03023350816220045,
      "grad_norm": 0.14927923679351807,
      "learning_rate": 6.195878509243661e-05,
      "loss": 0.0327,
      "relation_loss": 0.0024254602380096913,
      "selected_pairs": 26989.8,
      "state_query_norm": 1109.1601928710938,
      "state_relation_loss": 0.0004240802023559809,
      "step": 13560,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0024254602380096913,
      "weighted_relation_loss": 0.0024254602380096913
    },
    {
      "action_query_norm": 611.0423156738282,
      "action_relation_loss": 0.0020979218767024576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.694106641721234,
      "flow_matching_loss": 0.03241065721958876,
      "grad_norm": 0.12727384269237518,
      "learning_rate": 6.190526183731685e-05,
      "loss": 0.0349,
      "relation_loss": 0.0025159602519124746,
      "selected_pairs": 27036.0,
      "state_query_norm": 1105.6377319335938,
      "state_relation_loss": 0.00041803841304499656,
      "step": 13570,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0025159602519124746,
      "weighted_relation_loss": 0.0025159602519124746
    },
    {
      "action_query_norm": 609.6227294921875,
      "action_relation_loss": 0.002162112877704203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.703461178671656,
      "flow_matching_loss": 0.03171341586858034,
      "grad_norm": 0.1000877171754837,
      "learning_rate": 6.185172411618138e-05,
      "loss": 0.0343,
      "relation_loss": 0.0025808403501287104,
      "selected_pairs": 27200.5,
      "state_query_norm": 1109.6682373046874,
      "state_relation_loss": 0.0004187274957075715,
      "step": 13580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0025808403501287104,
      "weighted_relation_loss": 0.0025808403501287104
    },
    {
      "action_query_norm": 613.2097778320312,
      "action_relation_loss": 0.002124752209056169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.712815715622076,
      "flow_matching_loss": 0.03157674130052328,
      "grad_norm": 0.12466197460889816,
      "learning_rate": 6.179817199408355e-05,
      "loss": 0.0341,
      "relation_loss": 0.002551334514282644,
      "selected_pairs": 26666.7,
      "state_query_norm": 1108.2386840820313,
      "state_relation_loss": 0.0004265823255991563,
      "step": 13590,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002551334514282644,
      "weighted_relation_loss": 0.002551334514282644
    },
    {
      "action_query_norm": 596.88720703125,
      "action_relation_loss": 0.002118642465211451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.722170252572498,
      "flow_matching_loss": 0.0329906539991498,
      "grad_norm": 0.0978163629770279,
      "learning_rate": 6.174460553609426e-05,
      "loss": 0.0355,
      "relation_loss": 0.0025392209179699423,
      "selected_pairs": 26779.6,
      "state_query_norm": 1089.106298828125,
      "state_relation_loss": 0.00042057844693772496,
      "step": 13600,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0025392209179699423,
      "weighted_relation_loss": 0.0025392209179699423
    },
    {
      "action_query_norm": 607.9323486328125,
      "action_relation_loss": 0.0021146131795831023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.731524789522918,
      "flow_matching_loss": 0.030736659280955793,
      "grad_norm": 0.11656945198774338,
      "learning_rate": 6.16910248073018e-05,
      "loss": 0.0333,
      "relation_loss": 0.002536309836432338,
      "selected_pairs": 27351.9,
      "state_query_norm": 1102.571728515625,
      "state_relation_loss": 0.00042169665393885223,
      "step": 13610,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.002536309836432338,
      "weighted_relation_loss": 0.002536309836432338
    },
    {
      "action_query_norm": 606.7369506835937,
      "action_relation_loss": 0.0020193577161990105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.74087932647334,
      "flow_matching_loss": 0.032296201586723326,
      "grad_norm": 0.13692934811115265,
      "learning_rate": 6.16374298728118e-05,
      "loss": 0.0347,
      "relation_loss": 0.0024403920862823723,
      "selected_pairs": 27082.4,
      "state_query_norm": 1097.8725219726562,
      "state_relation_loss": 0.0004210344020975754,
      "step": 13620,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0024403920862823723,
      "weighted_relation_loss": 0.0024403920862823723
    },
    {
      "action_query_norm": 611.8899047851562,
      "action_relation_loss": 0.002064759097993374,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.75023386342376,
      "flow_matching_loss": 0.032121849805116655,
      "grad_norm": 0.11450935900211334,
      "learning_rate": 6.158382079774716e-05,
      "loss": 0.0346,
      "relation_loss": 0.0025055899750441313,
      "selected_pairs": 26715.2,
      "state_query_norm": 1106.2251708984375,
      "state_relation_loss": 0.0004408308770507574,
      "step": 13630,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0025055899750441313,
      "weighted_relation_loss": 0.0025055899750441313
    },
    {
      "action_query_norm": 612.9968872070312,
      "action_relation_loss": 0.0019459597882814706,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.75958840037418,
      "flow_matching_loss": 0.030618276447057724,
      "grad_norm": 0.11239399015903473,
      "learning_rate": 6.153019764724799e-05,
      "loss": 0.033,
      "relation_loss": 0.0023591057630255817,
      "selected_pairs": 26780.5,
      "state_query_norm": 1108.4164916992188,
      "state_relation_loss": 0.00041314598347526045,
      "step": 13640,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0023591057630255817,
      "weighted_relation_loss": 0.0023591057630255817
    },
    {
      "action_query_norm": 619.9793090820312,
      "action_relation_loss": 0.001975701656192541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.768942937324603,
      "flow_matching_loss": 0.03149980735033751,
      "grad_norm": 0.10226183384656906,
      "learning_rate": 6.147656048647144e-05,
      "loss": 0.0339,
      "relation_loss": 0.0023998017190024258,
      "selected_pairs": 27152.0,
      "state_query_norm": 1101.455419921875,
      "state_relation_loss": 0.000424100027885288,
      "step": 13650,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0023998017190024258,
      "weighted_relation_loss": 0.0023998017190024258
    },
    {
      "action_query_norm": 613.4735595703125,
      "action_relation_loss": 0.00206811495590955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.778297474275023,
      "flow_matching_loss": 0.030789589509367943,
      "grad_norm": 0.1329176425933838,
      "learning_rate": 6.142290938059173e-05,
      "loss": 0.0333,
      "relation_loss": 0.002495621400885284,
      "selected_pairs": 26666.9,
      "state_query_norm": 1119.7878295898438,
      "state_relation_loss": 0.00042750642460305245,
      "step": 13660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002495621400885284,
      "weighted_relation_loss": 0.002495621400885284
    },
    {
      "action_query_norm": 607.1142700195312,
      "action_relation_loss": 0.0020792130846530197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.787652011225445,
      "flow_matching_loss": 0.030728031508624555,
      "grad_norm": 0.12779313325881958,
      "learning_rate": 6.136924439480001e-05,
      "loss": 0.0332,
      "relation_loss": 0.002502301032654941,
      "selected_pairs": 26531.1,
      "state_query_norm": 1089.1298706054688,
      "state_relation_loss": 0.00042308794800192116,
      "step": 13670,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.002502301032654941,
      "weighted_relation_loss": 0.002502301032654941
    },
    {
      "action_query_norm": 605.7713073730469,
      "action_relation_loss": 0.0020677834283560513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.797006548175865,
      "flow_matching_loss": 0.03185160662978888,
      "grad_norm": 0.11917181313037872,
      "learning_rate": 6.13155655943043e-05,
      "loss": 0.0343,
      "relation_loss": 0.0024843988474458458,
      "selected_pairs": 26679.4,
      "state_query_norm": 1119.3037353515624,
      "state_relation_loss": 0.00041661544528324155,
      "step": 13680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0024843988474458458,
      "weighted_relation_loss": 0.0024843988474458458
    },
    {
      "action_query_norm": 606.14052734375,
      "action_relation_loss": 0.002059109474066645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.806361085126285,
      "flow_matching_loss": 0.02966428566724062,
      "grad_norm": 0.09640011936426163,
      "learning_rate": 6.126187304432941e-05,
      "loss": 0.0322,
      "relation_loss": 0.0024857440032064913,
      "selected_pairs": 26899.8,
      "state_query_norm": 1105.9299072265626,
      "state_relation_loss": 0.0004266345378709957,
      "step": 13690,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0024857440032064913,
      "weighted_relation_loss": 0.0024857440032064913
    },
    {
      "action_query_norm": 613.4591796875,
      "action_relation_loss": 0.0019395652576349675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.815715622076707,
      "flow_matching_loss": 0.031553853675723074,
      "grad_norm": 0.136774942278862,
      "learning_rate": 6.120816681011684e-05,
      "loss": 0.0339,
      "relation_loss": 0.00235938357654959,
      "selected_pairs": 26870.5,
      "state_query_norm": 1110.5543090820313,
      "state_relation_loss": 0.0004198183392873034,
      "step": 13700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00235938357654959,
      "weighted_relation_loss": 0.00235938357654959
    },
    {
      "action_query_norm": 607.3587097167969,
      "action_relation_loss": 0.001970143523067236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.825070159027128,
      "flow_matching_loss": 0.03103095982223749,
      "grad_norm": 0.11439710110425949,
      "learning_rate": 6.115444695692474e-05,
      "loss": 0.0334,
      "relation_loss": 0.0023956396849825977,
      "selected_pairs": 26735.2,
      "state_query_norm": 1114.6484008789062,
      "state_relation_loss": 0.0004254961444530636,
      "step": 13710,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0023956396849825977,
      "weighted_relation_loss": 0.0023956396849825977
    },
    {
      "action_query_norm": 613.2147094726563,
      "action_relation_loss": 0.002081019780598581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.83442469597755,
      "flow_matching_loss": 0.03187572602182627,
      "grad_norm": 0.11125581711530685,
      "learning_rate": 6.110071355002779e-05,
      "loss": 0.0344,
      "relation_loss": 0.002510720933787525,
      "selected_pairs": 27173.6,
      "state_query_norm": 1102.6076171875,
      "state_relation_loss": 0.00042970113281626254,
      "step": 13720,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002510720933787525,
      "weighted_relation_loss": 0.002510720933787525
    },
    {
      "action_query_norm": 612.6161560058594,
      "action_relation_loss": 0.0021733155590482056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.84377923292797,
      "flow_matching_loss": 0.031598724983632566,
      "grad_norm": 0.12083958834409714,
      "learning_rate": 6.104696665471714e-05,
      "loss": 0.0342,
      "relation_loss": 0.002588135050609708,
      "selected_pairs": 27318.0,
      "state_query_norm": 1090.658935546875,
      "state_relation_loss": 0.00041481947991997006,
      "step": 13730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.002588135050609708,
      "weighted_relation_loss": 0.002588135050609708
    },
    {
      "action_query_norm": 609.8914916992187,
      "action_relation_loss": 0.002006978972349316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.853133769878392,
      "flow_matching_loss": 0.03220851570367813,
      "grad_norm": 0.13948982954025269,
      "learning_rate": 6.099320633630036e-05,
      "loss": 0.0346,
      "relation_loss": 0.002427850058302283,
      "selected_pairs": 26666.7,
      "state_query_norm": 1104.9994018554687,
      "state_relation_loss": 0.0004208710946841165,
      "step": 13740,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002427850058302283,
      "weighted_relation_loss": 0.002427850058302283
    },
    {
      "action_query_norm": 614.0021789550781,
      "action_relation_loss": 0.0021076167235150934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.862488306828812,
      "flow_matching_loss": 0.0303935369476676,
      "grad_norm": 0.12975747883319855,
      "learning_rate": 6.093943266010128e-05,
      "loss": 0.0329,
      "relation_loss": 0.002513996697962284,
      "selected_pairs": 26918.8,
      "state_query_norm": 1100.8751586914063,
      "state_relation_loss": 0.00040637995407450945,
      "step": 13750,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002513996697962284,
      "weighted_relation_loss": 0.002513996697962284
    },
    {
      "action_query_norm": 615.2633850097657,
      "action_relation_loss": 0.001994344242848456,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.871842843779232,
      "flow_matching_loss": 0.03084080033004284,
      "grad_norm": 0.09782424569129944,
      "learning_rate": 6.088564569146e-05,
      "loss": 0.0333,
      "relation_loss": 0.0024191787000745536,
      "selected_pairs": 26756.2,
      "state_query_norm": 1103.4676147460937,
      "state_relation_loss": 0.00042483444849494845,
      "step": 13760,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0024191787000745536,
      "weighted_relation_loss": 0.0024191787000745536
    },
    {
      "action_query_norm": 616.2848449707031,
      "action_relation_loss": 0.002032150758896023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.881197380729654,
      "flow_matching_loss": 0.03172776885330677,
      "grad_norm": 0.11079451441764832,
      "learning_rate": 6.083184549573274e-05,
      "loss": 0.0342,
      "relation_loss": 0.0024428960168734193,
      "selected_pairs": 26640.2,
      "state_query_norm": 1112.7803588867187,
      "state_relation_loss": 0.00041074525797739624,
      "step": 13770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0024428960168734193,
      "weighted_relation_loss": 0.0024428960168734193
    },
    {
      "action_query_norm": 613.95224609375,
      "action_relation_loss": 0.0020355933345854284,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.890551917680074,
      "flow_matching_loss": 0.030874835699796675,
      "grad_norm": 0.12089426815509796,
      "learning_rate": 6.077803213829184e-05,
      "loss": 0.0333,
      "relation_loss": 0.002456818940117955,
      "selected_pairs": 26778.2,
      "state_query_norm": 1102.1223266601562,
      "state_relation_loss": 0.0004212256520986557,
      "step": 13780,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002456818940117955,
      "weighted_relation_loss": 0.002456818940117955
    },
    {
      "action_query_norm": 610.7011657714844,
      "action_relation_loss": 0.0020270192879252137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.899906454630496,
      "flow_matching_loss": 0.02998768538236618,
      "grad_norm": 0.12502452731132507,
      "learning_rate": 6.0724205684525594e-05,
      "loss": 0.0324,
      "relation_loss": 0.0024436434265226125,
      "selected_pairs": 26803.8,
      "state_query_norm": 1106.8047973632813,
      "state_relation_loss": 0.0004166241706116125,
      "step": 13790,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0024436434265226125,
      "weighted_relation_loss": 0.0024436434265226125
    },
    {
      "action_query_norm": 612.6514892578125,
      "action_relation_loss": 0.0020567184663377704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.909260991580917,
      "flow_matching_loss": 0.02915733903646469,
      "grad_norm": 0.11178167164325714,
      "learning_rate": 6.0670366199838215e-05,
      "loss": 0.0316,
      "relation_loss": 0.0024597265291959047,
      "selected_pairs": 26410.9,
      "state_query_norm": 1116.7618041992187,
      "state_relation_loss": 0.00040300801338162274,
      "step": 13800,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0024597265291959047,
      "weighted_relation_loss": 0.0024597265291959047
    },
    {
      "action_query_norm": 610.1201904296875,
      "action_relation_loss": 0.002033911959733814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.918615528531337,
      "flow_matching_loss": 0.03252036646008492,
      "grad_norm": 0.1225813776254654,
      "learning_rate": 6.061651374964974e-05,
      "loss": 0.035,
      "relation_loss": 0.002460620179772377,
      "selected_pairs": 27181.3,
      "state_query_norm": 1106.566943359375,
      "state_relation_loss": 0.0004267082258593291,
      "step": 13810,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002460620179772377,
      "weighted_relation_loss": 0.002460620179772377
    },
    {
      "action_query_norm": 610.0756896972656,
      "action_relation_loss": 0.0019377728109247982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.927970065481759,
      "flow_matching_loss": 0.02953320201486349,
      "grad_norm": 0.12945574522018433,
      "learning_rate": 6.0562648399396004e-05,
      "loss": 0.0319,
      "relation_loss": 0.002358162123709917,
      "selected_pairs": 26455.7,
      "state_query_norm": 1104.9534423828125,
      "state_relation_loss": 0.00042038935353048144,
      "step": 13820,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002358162123709917,
      "weighted_relation_loss": 0.002358162123709917
    },
    {
      "action_query_norm": 604.3768615722656,
      "action_relation_loss": 0.002028410916682333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.937324602432179,
      "flow_matching_loss": 0.03135566059499979,
      "grad_norm": 0.12685588002204895,
      "learning_rate": 6.050877021452845e-05,
      "loss": 0.0338,
      "relation_loss": 0.002425410784780979,
      "selected_pairs": 26668.8,
      "state_query_norm": 1119.445068359375,
      "state_relation_loss": 0.00039699984190519897,
      "step": 13830,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002425410784780979,
      "weighted_relation_loss": 0.002425410784780979
    },
    {
      "action_query_norm": 612.4060852050782,
      "action_relation_loss": 0.002029964071698487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.946679139382601,
      "flow_matching_loss": 0.02968762181699276,
      "grad_norm": 0.11066488176584244,
      "learning_rate": 6.04548792605142e-05,
      "loss": 0.0321,
      "relation_loss": 0.002434048382565379,
      "selected_pairs": 26260.4,
      "state_query_norm": 1101.9166748046875,
      "state_relation_loss": 0.00040408431668765844,
      "step": 13840,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.002434048382565379,
      "weighted_relation_loss": 0.002434048382565379
    },
    {
      "action_query_norm": 618.2074645996094,
      "action_relation_loss": 0.0019457227899692953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.956033676333021,
      "flow_matching_loss": 0.030530607886612417,
      "grad_norm": 0.12549692392349243,
      "learning_rate": 6.04009756028358e-05,
      "loss": 0.0329,
      "relation_loss": 0.0023437135154381395,
      "selected_pairs": 27220.0,
      "state_query_norm": 1106.0459106445312,
      "state_relation_loss": 0.0003979907080065459,
      "step": 13850,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0023437135154381395,
      "weighted_relation_loss": 0.0023437135154381395
    },
    {
      "action_query_norm": 608.4699890136719,
      "action_relation_loss": 0.0020947532961145044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.965388213283443,
      "flow_matching_loss": 0.03168937284499407,
      "grad_norm": 0.12444470077753067,
      "learning_rate": 6.0347059306991294e-05,
      "loss": 0.0342,
      "relation_loss": 0.0025150310480967163,
      "selected_pairs": 26502.8,
      "state_query_norm": 1120.54599609375,
      "state_relation_loss": 0.000420277746161446,
      "step": 13860,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0025150310480967163,
      "weighted_relation_loss": 0.0025150310480967163
    },
    {
      "action_query_norm": 613.0667602539063,
      "action_relation_loss": 0.001997241540811956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.974742750233863,
      "flow_matching_loss": 0.030742823518812655,
      "grad_norm": 0.1438678503036499,
      "learning_rate": 6.029313043849407e-05,
      "loss": 0.0332,
      "relation_loss": 0.0024103914853185414,
      "selected_pairs": 27290.1,
      "state_query_norm": 1107.0930297851562,
      "state_relation_loss": 0.00041314994159620256,
      "step": 13870,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0024103914853185414,
      "weighted_relation_loss": 0.0024103914853185414
    },
    {
      "action_query_norm": 611.0908569335937,
      "action_relation_loss": 0.0019994493923150003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.984097287184284,
      "flow_matching_loss": 0.029203851893544196,
      "grad_norm": 0.10762180387973785,
      "learning_rate": 6.023918906287279e-05,
      "loss": 0.0316,
      "relation_loss": 0.0024305614875629543,
      "selected_pairs": 26736.5,
      "state_query_norm": 1111.47578125,
      "state_relation_loss": 0.0004311121185310185,
      "step": 13880,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0024305614875629543,
      "weighted_relation_loss": 0.0024305614875629543
    },
    {
      "action_query_norm": 618.2615051269531,
      "action_relation_loss": 0.0020218776888214053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.993451824134706,
      "flow_matching_loss": 0.031079219840466975,
      "grad_norm": 0.15659520030021667,
      "learning_rate": 6.01852352456713e-05,
      "loss": 0.0335,
      "relation_loss": 0.0024207695852965115,
      "selected_pairs": 27106.3,
      "state_query_norm": 1101.5740112304688,
      "state_relation_loss": 0.00039889190811663865,
      "step": 13890,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0024207695852965115,
      "weighted_relation_loss": 0.0024207695852965115
    },
    {
      "action_query_norm": 606.9338195800781,
      "action_relation_loss": 0.002013747452292591,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.002806361085126,
      "flow_matching_loss": 0.029593771696090697,
      "grad_norm": 0.1145625188946724,
      "learning_rate": 6.013126905244858e-05,
      "loss": 0.032,
      "relation_loss": 0.00241764890961349,
      "selected_pairs": 24235.0,
      "state_query_norm": 1117.1219970703125,
      "state_relation_loss": 0.00040390141075477006,
      "step": 13900,
      "valid_action_samples": 213.9,
      "weighted_auxiliary_loss": 0.00241764890961349,
      "weighted_relation_loss": 0.00241764890961349
    },
    {
      "action_query_norm": 613.2712219238281,
      "action_relation_loss": 0.0019649578956887127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.012160898035548,
      "flow_matching_loss": 0.03005520086735487,
      "grad_norm": 0.10650354623794556,
      "learning_rate": 6.007729054877864e-05,
      "loss": 0.0324,
      "relation_loss": 0.002368938270956278,
      "selected_pairs": 26963.8,
      "state_query_norm": 1106.038134765625,
      "state_relation_loss": 0.0004039803898194805,
      "step": 13910,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002368938270956278,
      "weighted_relation_loss": 0.002368938270956278
    },
    {
      "action_query_norm": 614.9777099609375,
      "action_relation_loss": 0.0019418166484683752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.021515434985968,
      "flow_matching_loss": 0.03185571599751711,
      "grad_norm": 0.10851626098155975,
      "learning_rate": 6.0023299800250466e-05,
      "loss": 0.0342,
      "relation_loss": 0.0023678334895521403,
      "selected_pairs": 27438.0,
      "state_query_norm": 1110.925390625,
      "state_relation_loss": 0.00042601686727721243,
      "step": 13920,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0023678334895521403,
      "weighted_relation_loss": 0.0023678334895521403
    },
    {
      "action_query_norm": 610.6982604980469,
      "action_relation_loss": 0.0019774649175815285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.030869971936388,
      "flow_matching_loss": 0.03074074424803257,
      "grad_norm": 0.1592930108308792,
      "learning_rate": 5.9969296872467894e-05,
      "loss": 0.0331,
      "relation_loss": 0.0023830773774534465,
      "selected_pairs": 27668.2,
      "state_query_norm": 1113.0136352539062,
      "state_relation_loss": 0.0004056124424096197,
      "step": 13930,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0023830773774534465,
      "weighted_relation_loss": 0.0023830773774534465
    },
    {
      "action_query_norm": 609.1172668457032,
      "action_relation_loss": 0.0018926346790976822,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.04022450888681,
      "flow_matching_loss": 0.03074539266526699,
      "grad_norm": 0.10959851741790771,
      "learning_rate": 5.9915281831049585e-05,
      "loss": 0.033,
      "relation_loss": 0.0022992967860773206,
      "selected_pairs": 27129.6,
      "state_query_norm": 1104.5621215820313,
      "state_relation_loss": 0.00040666212153155356,
      "step": 13940,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0022992967860773206,
      "weighted_relation_loss": 0.0022992967860773206
    },
    {
      "action_query_norm": 614.1784423828125,
      "action_relation_loss": 0.0019309277064166962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.04957904583723,
      "flow_matching_loss": 0.03227977231144905,
      "grad_norm": 0.14309823513031006,
      "learning_rate": 5.98612547416289e-05,
      "loss": 0.0346,
      "relation_loss": 0.0023327125469222664,
      "selected_pairs": 27150.1,
      "state_query_norm": 1109.2003662109375,
      "state_relation_loss": 0.00040178485214710237,
      "step": 13950,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0023327125469222664,
      "weighted_relation_loss": 0.0023327125469222664
    },
    {
      "action_query_norm": 615.3860534667969,
      "action_relation_loss": 0.001974331587553024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.058933582787652,
      "flow_matching_loss": 0.03015286773443222,
      "grad_norm": 0.12312139570713043,
      "learning_rate": 5.9807215669853855e-05,
      "loss": 0.0325,
      "relation_loss": 0.0023860343964770435,
      "selected_pairs": 26945.5,
      "state_query_norm": 1101.4407348632812,
      "state_relation_loss": 0.0004117028438486159,
      "step": 13960,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0023860343964770435,
      "weighted_relation_loss": 0.0023860343964770435
    },
    {
      "action_query_norm": 616.1745300292969,
      "action_relation_loss": 0.0018804565304890275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.068288119738073,
      "flow_matching_loss": 0.029740168154239653,
      "grad_norm": 0.1138615682721138,
      "learning_rate": 5.9753164681387e-05,
      "loss": 0.032,
      "relation_loss": 0.0022871535271406173,
      "selected_pairs": 27126.6,
      "state_query_norm": 1114.8245849609375,
      "state_relation_loss": 0.00040669699374120685,
      "step": 13970,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0022871535271406173,
      "weighted_relation_loss": 0.0022871535271406173
    },
    {
      "action_query_norm": 615.1128112792969,
      "action_relation_loss": 0.0020236557233147324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.077642656688495,
      "flow_matching_loss": 0.030725904367864133,
      "grad_norm": 0.12580902874469757,
      "learning_rate": 5.969910184190539e-05,
      "loss": 0.0332,
      "relation_loss": 0.0024310487555339933,
      "selected_pairs": 26726.3,
      "state_query_norm": 1107.4114379882812,
      "state_relation_loss": 0.0004073930496815592,
      "step": 13980,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0024310487555339933,
      "weighted_relation_loss": 0.0024310487555339933
    },
    {
      "action_query_norm": 610.2749755859375,
      "action_relation_loss": 0.001945434871595353,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.086997193638915,
      "flow_matching_loss": 0.02985665313899517,
      "grad_norm": 0.11643525213003159,
      "learning_rate": 5.9645027217100477e-05,
      "loss": 0.0322,
      "relation_loss": 0.002360853459686041,
      "selected_pairs": 26406.5,
      "state_query_norm": 1100.6574340820312,
      "state_relation_loss": 0.0004154185648076236,
      "step": 13990,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.002360853459686041,
      "weighted_relation_loss": 0.002360853459686041
    },
    {
      "action_query_norm": 605.6408203125,
      "action_relation_loss": 0.001994777156505734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.096351730589335,
      "flow_matching_loss": 0.030628155171871185,
      "grad_norm": 0.12341855466365814,
      "learning_rate": 5.959094087267804e-05,
      "loss": 0.033,
      "relation_loss": 0.002405952219851315,
      "selected_pairs": 26987.7,
      "state_query_norm": 1092.0918579101562,
      "state_relation_loss": 0.0004111750633455813,
      "step": 14000,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002405952219851315,
      "weighted_relation_loss": 0.002405952219851315
    },
    {
      "action_query_norm": 610.976953125,
      "action_relation_loss": 0.0019164543133229016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.105706267539757,
      "flow_matching_loss": 0.029603262431919575,
      "grad_norm": 0.12398241460323334,
      "learning_rate": 5.9536842874358065e-05,
      "loss": 0.0319,
      "relation_loss": 0.002310370816849172,
      "selected_pairs": 26228.4,
      "state_query_norm": 1101.4017700195313,
      "state_relation_loss": 0.0003939165151678026,
      "step": 14010,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.002310370816849172,
      "weighted_relation_loss": 0.002310370816849172
    },
    {
      "action_query_norm": 617.8085998535156,
      "action_relation_loss": 0.001888045109808445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.115060804490177,
      "flow_matching_loss": 0.0308985885232687,
      "grad_norm": 0.11331036686897278,
      "learning_rate": 5.9482733287874734e-05,
      "loss": 0.0332,
      "relation_loss": 0.0023004617542028427,
      "selected_pairs": 27153.3,
      "state_query_norm": 1109.4321044921876,
      "state_relation_loss": 0.00041241662402171644,
      "step": 14020,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0023004617542028427,
      "weighted_relation_loss": 0.0023004617542028427
    },
    {
      "action_query_norm": 618.422216796875,
      "action_relation_loss": 0.0018894109060056507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.1244153414406,
      "flow_matching_loss": 0.030439439788460733,
      "grad_norm": 0.12359686195850372,
      "learning_rate": 5.942861217897631e-05,
      "loss": 0.0328,
      "relation_loss": 0.002310828072950244,
      "selected_pairs": 26967.4,
      "state_query_norm": 1114.0828979492187,
      "state_relation_loss": 0.00042141714366152885,
      "step": 14030,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002310828072950244,
      "weighted_relation_loss": 0.002310828072950244
    },
    {
      "action_query_norm": 614.8125183105469,
      "action_relation_loss": 0.0019683346152305605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.13376987839102,
      "flow_matching_loss": 0.03135220538824797,
      "grad_norm": 0.1072252094745636,
      "learning_rate": 5.9374479613425e-05,
      "loss": 0.0337,
      "relation_loss": 0.0023932095849886537,
      "selected_pairs": 26530.9,
      "state_query_norm": 1108.17568359375,
      "state_relation_loss": 0.0004248749726684764,
      "step": 14040,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0023932095849886537,
      "weighted_relation_loss": 0.0023932095849886537
    },
    {
      "action_query_norm": 615.6448181152343,
      "action_relation_loss": 0.0018626414472237228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.14312441534144,
      "flow_matching_loss": 0.030283860489726068,
      "grad_norm": 0.13526107370853424,
      "learning_rate": 5.932033565699704e-05,
      "loss": 0.0326,
      "relation_loss": 0.002287574694491923,
      "selected_pairs": 26044.5,
      "state_query_norm": 1104.7807495117188,
      "state_relation_loss": 0.00042493324144743383,
      "step": 14050,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.002287574694491923,
      "weighted_relation_loss": 0.002287574694491923
    },
    {
      "action_query_norm": 618.0840270996093,
      "action_relation_loss": 0.001919781684409827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.152478952291862,
      "flow_matching_loss": 0.030637688003480434,
      "grad_norm": 0.11681585013866425,
      "learning_rate": 5.926618037548237e-05,
      "loss": 0.033,
      "relation_loss": 0.002338333521038294,
      "selected_pairs": 26323.2,
      "state_query_norm": 1117.4982055664063,
      "state_relation_loss": 0.00041855183080770076,
      "step": 14060,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.002338333521038294,
      "weighted_relation_loss": 0.002338333521038294
    },
    {
      "action_query_norm": 609.6848510742187,
      "action_relation_loss": 0.0018567783175967635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.161833489242282,
      "flow_matching_loss": 0.030269716307520865,
      "grad_norm": 0.10159806162118912,
      "learning_rate": 5.9212013834684824e-05,
      "loss": 0.0326,
      "relation_loss": 0.00229067737236619,
      "selected_pairs": 26480.9,
      "state_query_norm": 1096.0356689453124,
      "state_relation_loss": 0.00043389902566559613,
      "step": 14070,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00229067737236619,
      "weighted_relation_loss": 0.00229067737236619
    },
    {
      "action_query_norm": 610.0884826660156,
      "action_relation_loss": 0.001961368543561548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.171188026192704,
      "flow_matching_loss": 0.030732950195670128,
      "grad_norm": 0.09805040061473846,
      "learning_rate": 5.9157836100421824e-05,
      "loss": 0.0331,
      "relation_loss": 0.0023761938558891416,
      "selected_pairs": 26952.2,
      "state_query_norm": 1117.1203002929688,
      "state_relation_loss": 0.0004148252861341462,
      "step": 14080,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0023761938558891416,
      "weighted_relation_loss": 0.0023761938558891416
    },
    {
      "action_query_norm": 618.3239135742188,
      "action_relation_loss": 0.0018952277023345232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.180542563143124,
      "flow_matching_loss": 0.03234921712428331,
      "grad_norm": 0.09491021186113358,
      "learning_rate": 5.910364723852444e-05,
      "loss": 0.0347,
      "relation_loss": 0.0023073290707543493,
      "selected_pairs": 26897.0,
      "state_query_norm": 1120.6195556640625,
      "state_relation_loss": 0.0004121014033444226,
      "step": 14090,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0023073290707543493,
      "weighted_relation_loss": 0.0023073290707543493
    },
    {
      "action_query_norm": 619.88974609375,
      "action_relation_loss": 0.0018606455181725324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.189897100093546,
      "flow_matching_loss": 0.03133759666234255,
      "grad_norm": 0.11058465391397476,
      "learning_rate": 5.904944731483723e-05,
      "loss": 0.0336,
      "relation_loss": 0.0022571809589862824,
      "selected_pairs": 26528.1,
      "state_query_norm": 1125.1566528320313,
      "state_relation_loss": 0.00039653542917221784,
      "step": 14100,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0022571809589862824,
      "weighted_relation_loss": 0.0022571809589862824
    },
    {
      "action_query_norm": 617.8861877441407,
      "action_relation_loss": 0.0018842300865799188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.199251637043966,
      "flow_matching_loss": 0.03067285120487213,
      "grad_norm": 0.11487755924463272,
      "learning_rate": 5.899523639521825e-05,
      "loss": 0.033,
      "relation_loss": 0.002283787471242249,
      "selected_pairs": 26693.0,
      "state_query_norm": 1119.3823364257812,
      "state_relation_loss": 0.00039955739921424536,
      "step": 14110,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002283787471242249,
      "weighted_relation_loss": 0.002283787471242249
    },
    {
      "action_query_norm": 615.0807983398438,
      "action_relation_loss": 0.001912475167773664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.208606173994387,
      "flow_matching_loss": 0.031119037233293057,
      "grad_norm": 0.10817881673574448,
      "learning_rate": 5.894101454553883e-05,
      "loss": 0.0334,
      "relation_loss": 0.0023306669434532525,
      "selected_pairs": 26299.5,
      "state_query_norm": 1100.3697143554687,
      "state_relation_loss": 0.0004181917727692053,
      "step": 14120,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0023306669434532525,
      "weighted_relation_loss": 0.0023306669434532525
    },
    {
      "action_query_norm": 615.8955871582032,
      "action_relation_loss": 0.001883356540929526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.217960710944809,
      "flow_matching_loss": 0.029717043787240983,
      "grad_norm": 0.11125041544437408,
      "learning_rate": 5.888678183168368e-05,
      "loss": 0.032,
      "relation_loss": 0.002294316375628114,
      "selected_pairs": 26503.4,
      "state_query_norm": 1103.323876953125,
      "state_relation_loss": 0.00041095984342973677,
      "step": 14130,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002294316375628114,
      "weighted_relation_loss": 0.002294316375628114
    },
    {
      "action_query_norm": 614.8782043457031,
      "action_relation_loss": 0.0018417217186652123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.227315247895229,
      "flow_matching_loss": 0.02890914026647806,
      "grad_norm": 0.11988715082406998,
      "learning_rate": 5.883253831955061e-05,
      "loss": 0.0311,
      "relation_loss": 0.0022293673362582924,
      "selected_pairs": 27409.9,
      "state_query_norm": 1107.656787109375,
      "state_relation_loss": 0.00038764561177231374,
      "step": 14140,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0022293673362582924,
      "weighted_relation_loss": 0.0022293673362582924
    },
    {
      "action_query_norm": 613.0435485839844,
      "action_relation_loss": 0.0019140962394885719,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.23666978484565,
      "flow_matching_loss": 0.03000825271010399,
      "grad_norm": 0.13994742929935455,
      "learning_rate": 5.8778284075050625e-05,
      "loss": 0.0323,
      "relation_loss": 0.0023163069505244493,
      "selected_pairs": 26997.3,
      "state_query_norm": 1120.7422241210938,
      "state_relation_loss": 0.00040221071685664356,
      "step": 14150,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0023163069505244493,
      "weighted_relation_loss": 0.0023163069505244493
    },
    {
      "action_query_norm": 610.112744140625,
      "action_relation_loss": 0.001991934038233012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.246024321796071,
      "flow_matching_loss": 0.02921759057790041,
      "grad_norm": 0.10708094388246536,
      "learning_rate": 5.872401916410777e-05,
      "loss": 0.0316,
      "relation_loss": 0.00240203607827425,
      "selected_pairs": 27043.7,
      "state_query_norm": 1083.2875366210938,
      "state_relation_loss": 0.0004101020545931533,
      "step": 14160,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00240203607827425,
      "weighted_relation_loss": 0.00240203607827425
    },
    {
      "action_query_norm": 617.2084228515625,
      "action_relation_loss": 0.0019973572343587875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.255378858746493,
      "flow_matching_loss": 0.029178240895271303,
      "grad_norm": 0.13301439583301544,
      "learning_rate": 5.8669743652659014e-05,
      "loss": 0.0316,
      "relation_loss": 0.002401136909611523,
      "selected_pairs": 27269.7,
      "state_query_norm": 1115.9919677734374,
      "state_relation_loss": 0.00040377970144618304,
      "step": 14170,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.002401136909611523,
      "weighted_relation_loss": 0.002401136909611523
    },
    {
      "action_query_norm": 617.6924194335937,
      "action_relation_loss": 0.0018981433007866145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.264733395696913,
      "flow_matching_loss": 0.029839301854372023,
      "grad_norm": 0.11714421212673187,
      "learning_rate": 5.86154576066542e-05,
      "loss": 0.0322,
      "relation_loss": 0.0023146089632064106,
      "selected_pairs": 26804.7,
      "state_query_norm": 1120.3883666992188,
      "state_relation_loss": 0.0004164656682405621,
      "step": 14180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0023146089632064106,
      "weighted_relation_loss": 0.0023146089632064106
    },
    {
      "action_query_norm": 615.4978454589843,
      "action_relation_loss": 0.0018772460520267487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.274087932647333,
      "flow_matching_loss": 0.028084767423570155,
      "grad_norm": 0.13565701246261597,
      "learning_rate": 5.856116109205602e-05,
      "loss": 0.0304,
      "relation_loss": 0.0022823008242994548,
      "selected_pairs": 26897.6,
      "state_query_norm": 1107.115576171875,
      "state_relation_loss": 0.0004050547955557704,
      "step": 14190,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0022823008242994548,
      "weighted_relation_loss": 0.0022823008242994548
    },
    {
      "action_query_norm": 610.0103149414062,
      "action_relation_loss": 0.0019791524042375385,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.283442469597755,
      "flow_matching_loss": 0.0294219009578228,
      "grad_norm": 0.1230047270655632,
      "learning_rate": 5.850685417483983e-05,
      "loss": 0.0318,
      "relation_loss": 0.002367057465016842,
      "selected_pairs": 27012.3,
      "state_query_norm": 1114.2216430664062,
      "state_relation_loss": 0.00038790508406236766,
      "step": 14200,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002367057465016842,
      "weighted_relation_loss": 0.002367057465016842
    },
    {
      "action_query_norm": 618.4105285644531,
      "action_relation_loss": 0.0018418807187117637,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.292797006548176,
      "flow_matching_loss": 0.02841061931103468,
      "grad_norm": 0.12060046941041946,
      "learning_rate": 5.8452536920993684e-05,
      "loss": 0.0307,
      "relation_loss": 0.0022521477891132235,
      "selected_pairs": 26422.8,
      "state_query_norm": 1107.296337890625,
      "state_relation_loss": 0.0004102670151041821,
      "step": 14210,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0022521477891132235,
      "weighted_relation_loss": 0.0022521477891132235
    },
    {
      "action_query_norm": 617.3025268554687,
      "action_relation_loss": 0.0019863061257638036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.302151543498598,
      "flow_matching_loss": 0.030417160503566264,
      "grad_norm": 0.11766045540571213,
      "learning_rate": 5.839820939651817e-05,
      "loss": 0.0328,
      "relation_loss": 0.0023867387790232897,
      "selected_pairs": 26709.5,
      "state_query_norm": 1116.32724609375,
      "state_relation_loss": 0.00040043265325948594,
      "step": 14220,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0023867387790232897,
      "weighted_relation_loss": 0.0023867387790232897
    },
    {
      "action_query_norm": 610.11904296875,
      "action_relation_loss": 0.001943525462411344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.311506080449018,
      "flow_matching_loss": 0.02979886680841446,
      "grad_norm": 0.1295304149389267,
      "learning_rate": 5.834387166742632e-05,
      "loss": 0.0322,
      "relation_loss": 0.00237071649171412,
      "selected_pairs": 26697.4,
      "state_query_norm": 1114.0481811523437,
      "state_relation_loss": 0.000427191038033925,
      "step": 14230,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00237071649171412,
      "weighted_relation_loss": 0.00237071649171412
    },
    {
      "action_query_norm": 613.67763671875,
      "action_relation_loss": 0.0019154811627231539,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.320860617399438,
      "flow_matching_loss": 0.028042724542319774,
      "grad_norm": 0.09723604470491409,
      "learning_rate": 5.828952379974364e-05,
      "loss": 0.0304,
      "relation_loss": 0.002315859030932188,
      "selected_pairs": 26657.5,
      "state_query_norm": 1115.2816772460938,
      "state_relation_loss": 0.0004003778478363529,
      "step": 14240,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002315859030932188,
      "weighted_relation_loss": 0.002315859030932188
    },
    {
      "action_query_norm": 617.5238525390625,
      "action_relation_loss": 0.0018643203540705145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.33021515434986,
      "flow_matching_loss": 0.029458560049533844,
      "grad_norm": 0.10504590719938278,
      "learning_rate": 5.8235165859507864e-05,
      "loss": 0.0317,
      "relation_loss": 0.002280399319715798,
      "selected_pairs": 27042.5,
      "state_query_norm": 1106.0156127929688,
      "state_relation_loss": 0.00041607895982451735,
      "step": 14250,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.002280399319715798,
      "weighted_relation_loss": 0.002280399319715798
    },
    {
      "action_query_norm": 618.0186889648437,
      "action_relation_loss": 0.0018697483348660171,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.33956969130028,
      "flow_matching_loss": 0.029971462115645407,
      "grad_norm": 0.10847321897745132,
      "learning_rate": 5.818079791276907e-05,
      "loss": 0.0322,
      "relation_loss": 0.0022575317649170755,
      "selected_pairs": 26709.1,
      "state_query_norm": 1116.7027099609375,
      "state_relation_loss": 0.0003877834475133568,
      "step": 14260,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0022575317649170755,
      "weighted_relation_loss": 0.0022575317649170755
    },
    {
      "action_query_norm": 613.3239013671875,
      "action_relation_loss": 0.0018074199091643096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.348924228250702,
      "flow_matching_loss": 0.02875293307006359,
      "grad_norm": 0.12630389630794525,
      "learning_rate": 5.8126420025589415e-05,
      "loss": 0.031,
      "relation_loss": 0.0022212137933820485,
      "selected_pairs": 26344.3,
      "state_query_norm": 1103.5518310546875,
      "state_relation_loss": 0.00041379391332156955,
      "step": 14270,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0022212137933820485,
      "weighted_relation_loss": 0.0022212137933820485
    },
    {
      "action_query_norm": 611.3182983398438,
      "action_relation_loss": 0.0019293913966976107,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.358278765201122,
      "flow_matching_loss": 0.02694256938993931,
      "grad_norm": 0.09177443385124207,
      "learning_rate": 5.807203226404313e-05,
      "loss": 0.0293,
      "relation_loss": 0.002339512319304049,
      "selected_pairs": 26672.0,
      "state_query_norm": 1100.2791748046875,
      "state_relation_loss": 0.0004101209255168214,
      "step": 14280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002339512319304049,
      "weighted_relation_loss": 0.002339512319304049
    },
    {
      "action_query_norm": 613.4046875,
      "action_relation_loss": 0.001894161489326507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.367633302151544,
      "flow_matching_loss": 0.03075902070850134,
      "grad_norm": 0.12677481770515442,
      "learning_rate": 5.801763469421652e-05,
      "loss": 0.0331,
      "relation_loss": 0.0023071260889992117,
      "selected_pairs": 26849.9,
      "state_query_norm": 1126.175244140625,
      "state_relation_loss": 0.0004129645763896406,
      "step": 14290,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0023071260889992117,
      "weighted_relation_loss": 0.0023071260889992117
    },
    {
      "action_query_norm": 623.474267578125,
      "action_relation_loss": 0.0018942357273772359,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.376987839101965,
      "flow_matching_loss": 0.028717939741909504,
      "grad_norm": 0.09315479546785355,
      "learning_rate": 5.796322738220774e-05,
      "loss": 0.031,
      "relation_loss": 0.002310704020783305,
      "selected_pairs": 27014.1,
      "state_query_norm": 1099.7802734375,
      "state_relation_loss": 0.0004164683079579845,
      "step": 14300,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002310704020783305,
      "weighted_relation_loss": 0.002310704020783305
    },
    {
      "action_query_norm": 614.906982421875,
      "action_relation_loss": 0.0019307801849208772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.386342376052385,
      "flow_matching_loss": 0.029262212850153447,
      "grad_norm": 0.1077602207660675,
      "learning_rate": 5.79088103941268e-05,
      "loss": 0.0316,
      "relation_loss": 0.0023276111809536816,
      "selected_pairs": 26665.7,
      "state_query_norm": 1127.1313110351562,
      "state_relation_loss": 0.0003968310047639534,
      "step": 14310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0023276111809536816,
      "weighted_relation_loss": 0.0023276111809536816
    },
    {
      "action_query_norm": 610.3041381835938,
      "action_relation_loss": 0.0020547380903735758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.395696913002807,
      "flow_matching_loss": 0.0312443308532238,
      "grad_norm": 0.1066901758313179,
      "learning_rate": 5.785438379609549e-05,
      "loss": 0.0337,
      "relation_loss": 0.0024613522458821536,
      "selected_pairs": 26482.2,
      "state_query_norm": 1103.5673828125,
      "state_relation_loss": 0.0004066141555085778,
      "step": 14320,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0024613522458821536,
      "weighted_relation_loss": 0.0024613522458821536
    },
    {
      "action_query_norm": 616.2492431640625,
      "action_relation_loss": 0.0019433804554864764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.405051449953227,
      "flow_matching_loss": 0.0323661845177412,
      "grad_norm": 0.12209751456975937,
      "learning_rate": 5.779994765424724e-05,
      "loss": 0.0347,
      "relation_loss": 0.0023590776952914895,
      "selected_pairs": 26630.1,
      "state_query_norm": 1121.7477172851563,
      "state_relation_loss": 0.0004156972659984604,
      "step": 14330,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0023590776952914895,
      "weighted_relation_loss": 0.0023590776952914895
    },
    {
      "action_query_norm": 622.133154296875,
      "action_relation_loss": 0.001974190049804747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.414405986903649,
      "flow_matching_loss": 0.030540736205875874,
      "grad_norm": 0.12973688542842865,
      "learning_rate": 5.7745502034727114e-05,
      "loss": 0.0329,
      "relation_loss": 0.0023833293234929442,
      "selected_pairs": 26548.8,
      "state_query_norm": 1105.2495239257812,
      "state_relation_loss": 0.00040913927368819715,
      "step": 14340,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0023833293234929442,
      "weighted_relation_loss": 0.0023833293234929442
    },
    {
      "action_query_norm": 612.3498840332031,
      "action_relation_loss": 0.0019255729392170906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.42376052385407,
      "flow_matching_loss": 0.03242077417671681,
      "grad_norm": 0.14265680313110352,
      "learning_rate": 5.769104700369164e-05,
      "loss": 0.0347,
      "relation_loss": 0.0023204771568998694,
      "selected_pairs": 26848.3,
      "state_query_norm": 1114.3839233398437,
      "state_relation_loss": 0.00039490422059316186,
      "step": 14350,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0023204771568998694,
      "weighted_relation_loss": 0.0023204771568998694
    },
    {
      "action_query_norm": 615.4003479003907,
      "action_relation_loss": 0.001971238083206117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.43311506080449,
      "flow_matching_loss": 0.030918745324015617,
      "grad_norm": 0.10651903599500656,
      "learning_rate": 5.7636582627308854e-05,
      "loss": 0.0333,
      "relation_loss": 0.0023754430934786797,
      "selected_pairs": 26644.8,
      "state_query_norm": 1117.8967651367188,
      "state_relation_loss": 0.0004042050277348608,
      "step": 14360,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0023754430934786797,
      "weighted_relation_loss": 0.0023754430934786797
    },
    {
      "action_query_norm": 619.6197814941406,
      "action_relation_loss": 0.0020348828402347863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.442469597754911,
      "flow_matching_loss": 0.029823882132768632,
      "grad_norm": 0.09553495794534683,
      "learning_rate": 5.7582108971758095e-05,
      "loss": 0.0323,
      "relation_loss": 0.0024392370600253344,
      "selected_pairs": 26184.2,
      "state_query_norm": 1118.3810546875,
      "state_relation_loss": 0.00040435427508782594,
      "step": 14370,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0024392370600253344,
      "weighted_relation_loss": 0.0024392370600253344
    },
    {
      "action_query_norm": 621.4936645507812,
      "action_relation_loss": 0.0019476562272757291,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.451824134705332,
      "flow_matching_loss": 0.030242850258946418,
      "grad_norm": 0.10701152682304382,
      "learning_rate": 5.752762610322999e-05,
      "loss": 0.0326,
      "relation_loss": 0.002347113797441125,
      "selected_pairs": 26801.9,
      "state_query_norm": 1113.2160400390626,
      "state_relation_loss": 0.0003994575905380771,
      "step": 14380,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002347113797441125,
      "weighted_relation_loss": 0.002347113797441125
    },
    {
      "action_query_norm": 616.8948791503906,
      "action_relation_loss": 0.0019505901378579437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.461178671655754,
      "flow_matching_loss": 0.02789765764027834,
      "grad_norm": 0.11716393381357193,
      "learning_rate": 5.747313408792636e-05,
      "loss": 0.0302,
      "relation_loss": 0.0023463014513254166,
      "selected_pairs": 26741.5,
      "state_query_norm": 1120.1716186523438,
      "state_relation_loss": 0.0003957113338401541,
      "step": 14390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0023463014513254166,
      "weighted_relation_loss": 0.0023463014513254166
    },
    {
      "action_query_norm": 618.1042602539062,
      "action_relation_loss": 0.0019285765825770794,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.470533208606174,
      "flow_matching_loss": 0.029486768878996372,
      "grad_norm": 0.17546270787715912,
      "learning_rate": 5.741863299206014e-05,
      "loss": 0.0318,
      "relation_loss": 0.002338908636011183,
      "selected_pairs": 26550.0,
      "state_query_norm": 1121.12314453125,
      "state_relation_loss": 0.00041033207962755115,
      "step": 14400,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.002338908636011183,
      "weighted_relation_loss": 0.002338908636011183
    },
    {
      "action_query_norm": 622.5403503417969,
      "action_relation_loss": 0.0019479834591038525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.479887745556596,
      "flow_matching_loss": 0.027754241600632668,
      "grad_norm": 0.12372638285160065,
      "learning_rate": 5.73641228818553e-05,
      "loss": 0.0301,
      "relation_loss": 0.002352802362293005,
      "selected_pairs": 26754.5,
      "state_query_norm": 1107.2307373046874,
      "state_relation_loss": 0.0004048189060995355,
      "step": 14410,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.002352802362293005,
      "weighted_relation_loss": 0.002352802362293005
    },
    {
      "action_query_norm": 619.5184875488281,
      "action_relation_loss": 0.001981292408891022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.489242282507016,
      "flow_matching_loss": 0.029791616089642047,
      "grad_norm": 0.10311663895845413,
      "learning_rate": 5.730960382354676e-05,
      "loss": 0.0322,
      "relation_loss": 0.0023844497045502067,
      "selected_pairs": 27454.1,
      "state_query_norm": 1113.5806030273438,
      "state_relation_loss": 0.00040315729565918443,
      "step": 14420,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0023844497045502067,
      "weighted_relation_loss": 0.0023844497045502067
    },
    {
      "action_query_norm": 618.1000549316407,
      "action_relation_loss": 0.001876900834031403,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.498596819457436,
      "flow_matching_loss": 0.0283260703086853,
      "grad_norm": 0.11964543163776398,
      "learning_rate": 5.725507588338035e-05,
      "loss": 0.0306,
      "relation_loss": 0.00228628555778414,
      "selected_pairs": 27368.8,
      "state_query_norm": 1105.748046875,
      "state_relation_loss": 0.000409384720842354,
      "step": 14430,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.00228628555778414,
      "weighted_relation_loss": 0.00228628555778414
    },
    {
      "action_query_norm": 614.1468505859375,
      "action_relation_loss": 0.0019167510909028351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.507951356407858,
      "flow_matching_loss": 0.02832592912018299,
      "grad_norm": 0.10863686352968216,
      "learning_rate": 5.7200539127612604e-05,
      "loss": 0.0306,
      "relation_loss": 0.0023186935344710947,
      "selected_pairs": 27182.5,
      "state_query_norm": 1107.8052856445313,
      "state_relation_loss": 0.0004019424173748121,
      "step": 14440,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0023186935344710947,
      "weighted_relation_loss": 0.0023186935344710947
    },
    {
      "action_query_norm": 617.0953552246094,
      "action_relation_loss": 0.0018806851003319025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.517305893358278,
      "flow_matching_loss": 0.030577800795435907,
      "grad_norm": 0.09443915635347366,
      "learning_rate": 5.714599362251087e-05,
      "loss": 0.0329,
      "relation_loss": 0.002284926502034068,
      "selected_pairs": 27060.6,
      "state_query_norm": 1115.9973022460938,
      "state_relation_loss": 0.00040424139297101647,
      "step": 14450,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002284926502034068,
      "weighted_relation_loss": 0.002284926502034068
    },
    {
      "action_query_norm": 619.6279357910156,
      "action_relation_loss": 0.0018612534389831127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.5266604303087,
      "flow_matching_loss": 0.02995318565517664,
      "grad_norm": 0.09843060374259949,
      "learning_rate": 5.709143943435307e-05,
      "loss": 0.0322,
      "relation_loss": 0.002283513592556119,
      "selected_pairs": 27106.4,
      "state_query_norm": 1125.6650268554688,
      "state_relation_loss": 0.00042226013611070814,
      "step": 14460,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.002283513592556119,
      "weighted_relation_loss": 0.002283513592556119
    },
    {
      "action_query_norm": 616.1406188964844,
      "action_relation_loss": 0.001837534341029823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.53601496725912,
      "flow_matching_loss": 0.027688279189169406,
      "grad_norm": 0.11692646145820618,
      "learning_rate": 5.7036876629427646e-05,
      "loss": 0.0299,
      "relation_loss": 0.002234307152684778,
      "selected_pairs": 26642.5,
      "state_query_norm": 1122.2831665039062,
      "state_relation_loss": 0.0003967728087445721,
      "step": 14470,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002234307152684778,
      "weighted_relation_loss": 0.002234307152684778
    },
    {
      "action_query_norm": 616.156884765625,
      "action_relation_loss": 0.00182338961167261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.545369504209543,
      "flow_matching_loss": 0.02838930282741785,
      "grad_norm": 0.10490459203720093,
      "learning_rate": 5.698230527403361e-05,
      "loss": 0.0306,
      "relation_loss": 0.00224103539949283,
      "selected_pairs": 26872.1,
      "state_query_norm": 1103.1048706054687,
      "state_relation_loss": 0.0004176457761786878,
      "step": 14480,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00224103539949283,
      "weighted_relation_loss": 0.00224103539949283
    },
    {
      "action_query_norm": 611.4579162597656,
      "action_relation_loss": 0.0018081653281114996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.554724041159963,
      "flow_matching_loss": 0.029434585385024546,
      "grad_norm": 0.12224786728620529,
      "learning_rate": 5.6927725434480274e-05,
      "loss": 0.0316,
      "relation_loss": 0.0022121192887425423,
      "selected_pairs": 26482.2,
      "state_query_norm": 1110.5199584960938,
      "state_relation_loss": 0.00040395395189989357,
      "step": 14490,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0022121192887425423,
      "weighted_relation_loss": 0.0022121192887425423
    },
    {
      "action_query_norm": 617.0694702148437,
      "action_relation_loss": 0.001807462889701128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.564078578110383,
      "flow_matching_loss": 0.028043194115161894,
      "grad_norm": 0.10802370309829712,
      "learning_rate": 5.687313717708728e-05,
      "loss": 0.0302,
      "relation_loss": 0.002202321821823716,
      "selected_pairs": 26522.0,
      "state_query_norm": 1123.3422119140625,
      "state_relation_loss": 0.0003948589175706729,
      "step": 14500,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.002202321821823716,
      "weighted_relation_loss": 0.002202321821823716
    },
    {
      "action_query_norm": 615.2350891113281,
      "action_relation_loss": 0.0018560104537755251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.573433115060805,
      "flow_matching_loss": 0.029166086576879025,
      "grad_norm": 0.1075657457113266,
      "learning_rate": 5.681854056818453e-05,
      "loss": 0.0314,
      "relation_loss": 0.0022498297272250055,
      "selected_pairs": 26919.3,
      "state_query_norm": 1119.0283325195312,
      "state_relation_loss": 0.0003938192647183314,
      "step": 14510,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0022498297272250055,
      "weighted_relation_loss": 0.0022498297272250055
    },
    {
      "action_query_norm": 618.6439819335938,
      "action_relation_loss": 0.0018115906161256135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.582787652011225,
      "flow_matching_loss": 0.028687292523682118,
      "grad_norm": 0.08649096637964249,
      "learning_rate": 5.676393567411205e-05,
      "loss": 0.0309,
      "relation_loss": 0.0022257055854424833,
      "selected_pairs": 27484.0,
      "state_query_norm": 1103.2334838867187,
      "state_relation_loss": 0.00041411495476495475,
      "step": 14520,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0022257055854424833,
      "weighted_relation_loss": 0.0022257055854424833
    },
    {
      "action_query_norm": 627.0387756347657,
      "action_relation_loss": 0.0018307991325855256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.592142188961647,
      "flow_matching_loss": 0.030281986482441424,
      "grad_norm": 0.12674197554588318,
      "learning_rate": 5.670932256121991e-05,
      "loss": 0.0325,
      "relation_loss": 0.0022270164219662545,
      "selected_pairs": 26943.1,
      "state_query_norm": 1125.3438842773437,
      "state_relation_loss": 0.0003962173097534105,
      "step": 14530,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0022270164219662545,
      "weighted_relation_loss": 0.0022270164219662545
    },
    {
      "action_query_norm": 627.2548278808594,
      "action_relation_loss": 0.0018695791950449347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.601496725912067,
      "flow_matching_loss": 0.02927444875240326,
      "grad_norm": 0.1059761568903923,
      "learning_rate": 5.6654701295868216e-05,
      "loss": 0.0315,
      "relation_loss": 0.0022704561008140443,
      "selected_pairs": 26988.0,
      "state_query_norm": 1106.124853515625,
      "state_relation_loss": 0.00040087694360408933,
      "step": 14540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0022704561008140443,
      "weighted_relation_loss": 0.0022704561008140443
    },
    {
      "action_query_norm": 622.126513671875,
      "action_relation_loss": 0.0018388534081168473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.610851262862488,
      "flow_matching_loss": 0.029422486014664174,
      "grad_norm": 0.122929647564888,
      "learning_rate": 5.660007194442697e-05,
      "loss": 0.0317,
      "relation_loss": 0.0022520846920087933,
      "selected_pairs": 26924.0,
      "state_query_norm": 1132.1655029296876,
      "state_relation_loss": 0.0004132312664296478,
      "step": 14550,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0022520846920087933,
      "weighted_relation_loss": 0.0022520846920087933
    },
    {
      "action_query_norm": 622.1047729492187,
      "action_relation_loss": 0.0018621860654093325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.62020579981291,
      "flow_matching_loss": 0.030593937262892723,
      "grad_norm": 0.11325706541538239,
      "learning_rate": 5.6545434573275946e-05,
      "loss": 0.0329,
      "relation_loss": 0.002274283836595714,
      "selected_pairs": 27361.7,
      "state_query_norm": 1122.1684692382812,
      "state_relation_loss": 0.00041209773626178504,
      "step": 14560,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.002274283836595714,
      "weighted_relation_loss": 0.002274283836595714
    },
    {
      "action_query_norm": 624.1442321777344,
      "action_relation_loss": 0.0018329121521674097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.62956033676333,
      "flow_matching_loss": 0.030083341151475908,
      "grad_norm": 0.10894030332565308,
      "learning_rate": 5.649078924880472e-05,
      "loss": 0.0323,
      "relation_loss": 0.0022486978909000755,
      "selected_pairs": 27062.5,
      "state_query_norm": 1114.5566284179688,
      "state_relation_loss": 0.0004157857008976862,
      "step": 14570,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0022486978909000755,
      "weighted_relation_loss": 0.0022486978909000755
    },
    {
      "action_query_norm": 624.6346374511719,
      "action_relation_loss": 0.0018560312921181322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.638914873713752,
      "flow_matching_loss": 0.02712752092629671,
      "grad_norm": 0.10203156620264053,
      "learning_rate": 5.643613603741252e-05,
      "loss": 0.0294,
      "relation_loss": 0.002269064192660153,
      "selected_pairs": 26802.0,
      "state_query_norm": 1126.9818237304687,
      "state_relation_loss": 0.0004130328947212547,
      "step": 14580,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002269064192660153,
      "weighted_relation_loss": 0.002269064192660153
    },
    {
      "action_query_norm": 622.7877014160156,
      "action_relation_loss": 0.0018450746894814074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.648269410664172,
      "flow_matching_loss": 0.030729846842586995,
      "grad_norm": 0.09695184975862503,
      "learning_rate": 5.6381475005508154e-05,
      "loss": 0.033,
      "relation_loss": 0.002254356583580375,
      "selected_pairs": 26687.4,
      "state_query_norm": 1105.31455078125,
      "state_relation_loss": 0.0004092819115612656,
      "step": 14590,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002254356583580375,
      "weighted_relation_loss": 0.002254356583580375
    },
    {
      "action_query_norm": 614.0925354003906,
      "action_relation_loss": 0.0019103326834738255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.657623947614592,
      "flow_matching_loss": 0.0287424236536026,
      "grad_norm": 0.10015853494405746,
      "learning_rate": 5.632680621950992e-05,
      "loss": 0.031,
      "relation_loss": 0.0023056821431964637,
      "selected_pairs": 26598.0,
      "state_query_norm": 1116.9695434570312,
      "state_relation_loss": 0.0003953495004680008,
      "step": 14600,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0023056821431964637,
      "weighted_relation_loss": 0.0023056821431964637
    },
    {
      "action_query_norm": 622.8455810546875,
      "action_relation_loss": 0.0018525675870478154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.666978484565014,
      "flow_matching_loss": 0.02860135044902563,
      "grad_norm": 0.0849653035402298,
      "learning_rate": 5.627212974584555e-05,
      "loss": 0.0309,
      "relation_loss": 0.002258973941206932,
      "selected_pairs": 26434.8,
      "state_query_norm": 1121.6057373046874,
      "state_relation_loss": 0.0004064063250552863,
      "step": 14610,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002258973941206932,
      "weighted_relation_loss": 0.002258973941206932
    },
    {
      "action_query_norm": 612.3893859863281,
      "action_relation_loss": 0.00196973686106503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.676333021515434,
      "flow_matching_loss": 0.028177518211305142,
      "grad_norm": 0.11840510368347168,
      "learning_rate": 5.62174456509521e-05,
      "loss": 0.0306,
      "relation_loss": 0.002378599648363888,
      "selected_pairs": 26718.9,
      "state_query_norm": 1095.1586303710938,
      "state_relation_loss": 0.0004088627669261768,
      "step": 14620,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.002378599648363888,
      "weighted_relation_loss": 0.002378599648363888
    },
    {
      "action_query_norm": 619.1358520507813,
      "action_relation_loss": 0.0018271825276315213,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.685687558465856,
      "flow_matching_loss": 0.028564704582095146,
      "grad_norm": 0.10310970991849899,
      "learning_rate": 5.616275400127594e-05,
      "loss": 0.0308,
      "relation_loss": 0.0022363685537129642,
      "selected_pairs": 26794.1,
      "state_query_norm": 1120.9847534179687,
      "state_relation_loss": 0.0004091860173502937,
      "step": 14630,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0022363685537129642,
      "weighted_relation_loss": 0.0022363685537129642
    },
    {
      "action_query_norm": 623.456591796875,
      "action_relation_loss": 0.001882710983045399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.695042095416277,
      "flow_matching_loss": 0.02854135390371084,
      "grad_norm": 0.11015990376472473,
      "learning_rate": 5.610805486327254e-05,
      "loss": 0.0308,
      "relation_loss": 0.0022839857265353203,
      "selected_pairs": 27032.2,
      "state_query_norm": 1126.0265747070312,
      "state_relation_loss": 0.00040127475513145326,
      "step": 14640,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0022839857265353203,
      "weighted_relation_loss": 0.0022839857265353203
    },
    {
      "action_query_norm": 628.4025634765625,
      "action_relation_loss": 0.001960241422057152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.704396632366699,
      "flow_matching_loss": 0.029805420711636544,
      "grad_norm": 0.11579767614603043,
      "learning_rate": 5.6053348303406536e-05,
      "loss": 0.0322,
      "relation_loss": 0.002354924427345395,
      "selected_pairs": 27504.2,
      "state_query_norm": 1109.6822021484375,
      "state_relation_loss": 0.0003946829994674772,
      "step": 14650,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.002354924427345395,
      "weighted_relation_loss": 0.002354924427345395
    },
    {
      "action_query_norm": 614.6289978027344,
      "action_relation_loss": 0.001790717919357121,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.713751169317119,
      "flow_matching_loss": 0.02939808592200279,
      "grad_norm": 0.114793561398983,
      "learning_rate": 5.599863438815156e-05,
      "loss": 0.0316,
      "relation_loss": 0.0021886070724576713,
      "selected_pairs": 26939.2,
      "state_query_norm": 1099.0651489257812,
      "state_relation_loss": 0.00039788917056284847,
      "step": 14660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0021886070724576713,
      "weighted_relation_loss": 0.0021886070724576713
    },
    {
      "action_query_norm": 619.0191284179688,
      "action_relation_loss": 0.0018363341107033194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.723105706267539,
      "flow_matching_loss": 0.02736089564859867,
      "grad_norm": 0.08526190370321274,
      "learning_rate": 5.594391318399017e-05,
      "loss": 0.0296,
      "relation_loss": 0.002250136062502861,
      "selected_pairs": 26504.1,
      "state_query_norm": 1123.7766479492188,
      "state_relation_loss": 0.00041380197508260607,
      "step": 14670,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002250136062502861,
      "weighted_relation_loss": 0.002250136062502861
    },
    {
      "action_query_norm": 623.5259582519532,
      "action_relation_loss": 0.0018659020308405161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.732460243217961,
      "flow_matching_loss": 0.028656185790896415,
      "grad_norm": 0.09119591861963272,
      "learning_rate": 5.588918475741377e-05,
      "loss": 0.0309,
      "relation_loss": 0.0022638971218839287,
      "selected_pairs": 27316.3,
      "state_query_norm": 1114.7911254882813,
      "state_relation_loss": 0.0003979950619395822,
      "step": 14680,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0022638971218839287,
      "weighted_relation_loss": 0.0022638971218839287
    },
    {
      "action_query_norm": 621.9359741210938,
      "action_relation_loss": 0.0018465889850631356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.741814780168381,
      "flow_matching_loss": 0.028034813702106476,
      "grad_norm": 0.09863277524709702,
      "learning_rate": 5.5834449174922586e-05,
      "loss": 0.0303,
      "relation_loss": 0.0022460441221483054,
      "selected_pairs": 26320.7,
      "state_query_norm": 1110.7236694335938,
      "state_relation_loss": 0.00039945516036823394,
      "step": 14690,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0022460441221483054,
      "weighted_relation_loss": 0.0022460441221483054
    },
    {
      "action_query_norm": 625.1769165039062,
      "action_relation_loss": 0.0017964780679903924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.751169317118803,
      "flow_matching_loss": 0.028003212809562684,
      "grad_norm": 0.1317838579416275,
      "learning_rate": 5.57797065030255e-05,
      "loss": 0.0302,
      "relation_loss": 0.0021906205685809253,
      "selected_pairs": 26871.1,
      "state_query_norm": 1112.3349731445312,
      "state_relation_loss": 0.00039414248312823473,
      "step": 14700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0021906205685809253,
      "weighted_relation_loss": 0.0021906205685809253
    },
    {
      "action_query_norm": 616.8056213378907,
      "action_relation_loss": 0.0018269201857037843,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.760523854069223,
      "flow_matching_loss": 0.03016786277294159,
      "grad_norm": 0.1082088053226471,
      "learning_rate": 5.5724956808240016e-05,
      "loss": 0.0324,
      "relation_loss": 0.0022234687814489005,
      "selected_pairs": 27149.8,
      "state_query_norm": 1110.0286254882812,
      "state_relation_loss": 0.0003965486423112452,
      "step": 14710,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0022234687814489005,
      "weighted_relation_loss": 0.0022234687814489005
    },
    {
      "action_query_norm": 623.5685546875,
      "action_relation_loss": 0.0017712807981297373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.769878391019645,
      "flow_matching_loss": 0.029329702630639075,
      "grad_norm": 0.0900372639298439,
      "learning_rate": 5.567020015709219e-05,
      "loss": 0.0315,
      "relation_loss": 0.002166861412115395,
      "selected_pairs": 27228.4,
      "state_query_norm": 1112.5580932617188,
      "state_relation_loss": 0.0003955806139856577,
      "step": 14720,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.002166861412115395,
      "weighted_relation_loss": 0.002166861412115395
    },
    {
      "action_query_norm": 629.89970703125,
      "action_relation_loss": 0.001831173652317375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.779232927970066,
      "flow_matching_loss": 0.02717868648469448,
      "grad_norm": 0.11402464658021927,
      "learning_rate": 5.561543661611649e-05,
      "loss": 0.0294,
      "relation_loss": 0.0022400822723284363,
      "selected_pairs": 27686.0,
      "state_query_norm": 1113.7951293945312,
      "state_relation_loss": 0.00040890861419029534,
      "step": 14730,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.0022400822723284363,
      "weighted_relation_loss": 0.0022400822723284363
    },
    {
      "action_query_norm": 624.4393249511719,
      "action_relation_loss": 0.0018238820368424058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.788587464920486,
      "flow_matching_loss": 0.02851703520864248,
      "grad_norm": 0.09780272841453552,
      "learning_rate": 5.556066625185583e-05,
      "loss": 0.0308,
      "relation_loss": 0.002239390346221626,
      "selected_pairs": 27458.2,
      "state_query_norm": 1110.8658325195313,
      "state_relation_loss": 0.0004155083384830505,
      "step": 14740,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.002239390346221626,
      "weighted_relation_loss": 0.002239390346221626
    },
    {
      "action_query_norm": 621.1777587890625,
      "action_relation_loss": 0.0018527892418205739,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.797942001870908,
      "flow_matching_loss": 0.030324486084282398,
      "grad_norm": 0.12399090081453323,
      "learning_rate": 5.550588913086131e-05,
      "loss": 0.0326,
      "relation_loss": 0.002263914328068495,
      "selected_pairs": 27069.3,
      "state_query_norm": 1123.1398803710938,
      "state_relation_loss": 0.00041112510371021926,
      "step": 14750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002263914328068495,
      "weighted_relation_loss": 0.002263914328068495
    },
    {
      "action_query_norm": 628.8425415039062,
      "action_relation_loss": 0.0017337126191705465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.807296538821328,
      "flow_matching_loss": 0.02721866834908724,
      "grad_norm": 0.08281102776527405,
      "learning_rate": 5.545110531969234e-05,
      "loss": 0.0294,
      "relation_loss": 0.0021317928447388114,
      "selected_pairs": 27461.8,
      "state_query_norm": 1130.0709228515625,
      "state_relation_loss": 0.0003980802197474986,
      "step": 14760,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0021317928447388114,
      "weighted_relation_loss": 0.0021317928447388114
    },
    {
      "action_query_norm": 623.7360290527344,
      "action_relation_loss": 0.0019086814019829035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.81665107577175,
      "flow_matching_loss": 0.028907349146902562,
      "grad_norm": 0.09480992704629898,
      "learning_rate": 5.539631488491641e-05,
      "loss": 0.0312,
      "relation_loss": 0.0023143682396039368,
      "selected_pairs": 26993.7,
      "state_query_norm": 1119.4577270507812,
      "state_relation_loss": 0.0004056868376210332,
      "step": 14770,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0023143682396039368,
      "weighted_relation_loss": 0.0023143682396039368
    },
    {
      "action_query_norm": 629.3742614746094,
      "action_relation_loss": 0.0017395581002347172,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.82600561272217,
      "flow_matching_loss": 0.029902536422014236,
      "grad_norm": 0.11889626830816269,
      "learning_rate": 5.534151789310904e-05,
      "loss": 0.032,
      "relation_loss": 0.0021371558541432023,
      "selected_pairs": 26510.3,
      "state_query_norm": 1114.1978759765625,
      "state_relation_loss": 0.0003975977684604004,
      "step": 14780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0021371558541432023,
      "weighted_relation_loss": 0.0021371558541432023
    },
    {
      "action_query_norm": 634.8352722167969,
      "action_relation_loss": 0.0018151244148612022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.83536014967259,
      "flow_matching_loss": 0.02647628653794527,
      "grad_norm": 0.09736645966768265,
      "learning_rate": 5.528671441085376e-05,
      "loss": 0.0287,
      "relation_loss": 0.002211522264406085,
      "selected_pairs": 26941.6,
      "state_query_norm": 1113.83818359375,
      "state_relation_loss": 0.00039639785536564885,
      "step": 14790,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.002211522264406085,
      "weighted_relation_loss": 0.002211522264406085
    },
    {
      "action_query_norm": 625.3669311523438,
      "action_relation_loss": 0.0018576714908704161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.844714686623012,
      "flow_matching_loss": 0.0292761005461216,
      "grad_norm": 0.11357519775629044,
      "learning_rate": 5.523190450474197e-05,
      "loss": 0.0315,
      "relation_loss": 0.0022566650761291383,
      "selected_pairs": 26224.5,
      "state_query_norm": 1115.93828125,
      "state_relation_loss": 0.00039899356197565795,
      "step": 14800,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0022566650761291383,
      "weighted_relation_loss": 0.0022566650761291383
    },
    {
      "action_query_norm": 625.2283996582031,
      "action_relation_loss": 0.0019041598658077418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.854069223573433,
      "flow_matching_loss": 0.03312547020614147,
      "grad_norm": 0.14697672426700592,
      "learning_rate": 5.5177088241372865e-05,
      "loss": 0.0354,
      "relation_loss": 0.002311827056109905,
      "selected_pairs": 27159.1,
      "state_query_norm": 1089.1082397460937,
      "state_relation_loss": 0.0004076671932125464,
      "step": 14810,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.002311827056109905,
      "weighted_relation_loss": 0.002311827056109905
    },
    {
      "action_query_norm": 630.2235534667968,
      "action_relation_loss": 0.00174265056848526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.863423760523855,
      "flow_matching_loss": 0.028166293166577816,
      "grad_norm": 0.08069299161434174,
      "learning_rate": 5.512226568735338e-05,
      "loss": 0.0303,
      "relation_loss": 0.0021544085582718255,
      "selected_pairs": 26596.8,
      "state_query_norm": 1126.4794067382813,
      "state_relation_loss": 0.0004117580276215449,
      "step": 14820,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0021544085582718255,
      "weighted_relation_loss": 0.0021544085582718255
    },
    {
      "action_query_norm": 627.7113098144531,
      "action_relation_loss": 0.0018180469633080065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.872778297474275,
      "flow_matching_loss": 0.028947499766945838,
      "grad_norm": 0.11116442084312439,
      "learning_rate": 5.506743690929809e-05,
      "loss": 0.0312,
      "relation_loss": 0.0022161291679367425,
      "selected_pairs": 26601.5,
      "state_query_norm": 1118.4955932617188,
      "state_relation_loss": 0.00039808220171835276,
      "step": 14830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0022161291679367425,
      "weighted_relation_loss": 0.0022161291679367425
    },
    {
      "action_query_norm": 627.8976013183594,
      "action_relation_loss": 0.0018586278893053532,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.882132834424697,
      "flow_matching_loss": 0.02921781353652477,
      "grad_norm": 0.11757132411003113,
      "learning_rate": 5.501260197382912e-05,
      "loss": 0.0315,
      "relation_loss": 0.002263743872754276,
      "selected_pairs": 26644.9,
      "state_query_norm": 1109.0406860351563,
      "state_relation_loss": 0.0004051159834489226,
      "step": 14840,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002263743872754276,
      "weighted_relation_loss": 0.002263743872754276
    },
    {
      "action_query_norm": 628.4133483886719,
      "action_relation_loss": 0.0017378049320541321,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.891487371375117,
      "flow_matching_loss": 0.029453433863818646,
      "grad_norm": 0.13645519316196442,
      "learning_rate": 5.49577609475761e-05,
      "loss": 0.0316,
      "relation_loss": 0.002139756572432816,
      "selected_pairs": 26811.4,
      "state_query_norm": 1121.7781494140625,
      "state_relation_loss": 0.0004019516316475347,
      "step": 14850,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002139756572432816,
      "weighted_relation_loss": 0.002139756572432816
    },
    {
      "action_query_norm": 623.07275390625,
      "action_relation_loss": 0.0019085479667410254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.900841908325537,
      "flow_matching_loss": 0.029095147922635078,
      "grad_norm": 0.11184033006429672,
      "learning_rate": 5.490291389717603e-05,
      "loss": 0.0314,
      "relation_loss": 0.0023105868371203543,
      "selected_pairs": 26966.9,
      "state_query_norm": 1124.278076171875,
      "state_relation_loss": 0.00040203883836511525,
      "step": 14860,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0023105868371203543,
      "weighted_relation_loss": 0.0023105868371203543
    },
    {
      "action_query_norm": 626.7860778808594,
      "action_relation_loss": 0.0017563773784786464,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.91019644527596,
      "flow_matching_loss": 0.03025128524750471,
      "grad_norm": 0.096256785094738,
      "learning_rate": 5.484806088927329e-05,
      "loss": 0.0324,
      "relation_loss": 0.0021630197064951063,
      "selected_pairs": 27011.7,
      "state_query_norm": 1126.43486328125,
      "state_relation_loss": 0.00040664229600224645,
      "step": 14870,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0021630197064951063,
      "weighted_relation_loss": 0.0021630197064951063
    },
    {
      "action_query_norm": 620.5123596191406,
      "action_relation_loss": 0.0017683890066109598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.91955098222638,
      "flow_matching_loss": 0.028450832702219486,
      "grad_norm": 0.12026532739400864,
      "learning_rate": 5.479320199051942e-05,
      "loss": 0.0306,
      "relation_loss": 0.0021651645191013815,
      "selected_pairs": 26893.4,
      "state_query_norm": 1124.1158081054687,
      "state_relation_loss": 0.00039677550666965543,
      "step": 14880,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0021651645191013815,
      "weighted_relation_loss": 0.0021651645191013815
    },
    {
      "action_query_norm": 624.4130981445312,
      "action_relation_loss": 0.0017435744171962141,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.928905519176801,
      "flow_matching_loss": 0.029515615850687026,
      "grad_norm": 0.09023069590330124,
      "learning_rate": 5.473833726757314e-05,
      "loss": 0.0316,
      "relation_loss": 0.002127430937252939,
      "selected_pairs": 27177.7,
      "state_query_norm": 1117.3755615234375,
      "state_relation_loss": 0.0003838565025944263,
      "step": 14890,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002127430937252939,
      "weighted_relation_loss": 0.002127430937252939
    },
    {
      "action_query_norm": 616.5120666503906,
      "action_relation_loss": 0.0018340879585593938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.938260056127222,
      "flow_matching_loss": 0.029250368289649487,
      "grad_norm": 0.14339840412139893,
      "learning_rate": 5.4683466787100304e-05,
      "loss": 0.0315,
      "relation_loss": 0.002235823904629797,
      "selected_pairs": 26752.6,
      "state_query_norm": 1113.2816284179687,
      "state_relation_loss": 0.0004017359053250402,
      "step": 14900,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.002235823904629797,
      "weighted_relation_loss": 0.002235823904629797
    },
    {
      "action_query_norm": 618.1259460449219,
      "action_relation_loss": 0.0017932335962541402,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.947614593077642,
      "flow_matching_loss": 0.028307039104402066,
      "grad_norm": 0.10129605978727341,
      "learning_rate": 5.462859061577369e-05,
      "loss": 0.0305,
      "relation_loss": 0.0021964569576084616,
      "selected_pairs": 26453.7,
      "state_query_norm": 1128.8147583007812,
      "state_relation_loss": 0.00040322334389202296,
      "step": 14910,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0021964569576084616,
      "weighted_relation_loss": 0.0021964569576084616
    },
    {
      "action_query_norm": 625.2147033691406,
      "action_relation_loss": 0.0018225950887426734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.956969130028064,
      "flow_matching_loss": 0.028455728106200694,
      "grad_norm": 0.09795064479112625,
      "learning_rate": 5.4573708820273026e-05,
      "loss": 0.0307,
      "relation_loss": 0.0022097124252468347,
      "selected_pairs": 27289.4,
      "state_query_norm": 1123.4912353515624,
      "state_relation_loss": 0.00038711733359377833,
      "step": 14920,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0022097124252468347,
      "weighted_relation_loss": 0.0022097124252468347
    },
    {
      "action_query_norm": 617.4290161132812,
      "action_relation_loss": 0.0017569551593624056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.966323666978484,
      "flow_matching_loss": 0.02726615983992815,
      "grad_norm": 0.1031920462846756,
      "learning_rate": 5.451882146728488e-05,
      "loss": 0.0294,
      "relation_loss": 0.0021627815440297127,
      "selected_pairs": 27504.0,
      "state_query_norm": 1115.75703125,
      "state_relation_loss": 0.0004058263875776902,
      "step": 14930,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0021627815440297127,
      "weighted_relation_loss": 0.0021627815440297127
    },
    {
      "action_query_norm": 621.4861083984375,
      "action_relation_loss": 0.0018720333464443684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.975678203928906,
      "flow_matching_loss": 0.028287597000598907,
      "grad_norm": 0.15837955474853516,
      "learning_rate": 5.446392862350255e-05,
      "loss": 0.0305,
      "relation_loss": 0.002261831075884402,
      "selected_pairs": 26686.5,
      "state_query_norm": 1127.85283203125,
      "state_relation_loss": 0.000389797700336203,
      "step": 14940,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002261831075884402,
      "weighted_relation_loss": 0.002261831075884402
    },
    {
      "action_query_norm": 617.9078735351562,
      "action_relation_loss": 0.001810689817648381,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.985032740879326,
      "flow_matching_loss": 0.027821847610175608,
      "grad_norm": 0.13895410299301147,
      "learning_rate": 5.440903035562603e-05,
      "loss": 0.03,
      "relation_loss": 0.002191773324739188,
      "selected_pairs": 27024.4,
      "state_query_norm": 1131.1088012695313,
      "state_relation_loss": 0.0003810835012700409,
      "step": 14950,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002191773324739188,
      "weighted_relation_loss": 0.002191773324739188
    },
    {
      "action_query_norm": 618.7152160644531,
      "action_relation_loss": 0.0017739286529831588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.994387277829748,
      "flow_matching_loss": 0.028401616029441357,
      "grad_norm": 0.10644802451133728,
      "learning_rate": 5.435412673036188e-05,
      "loss": 0.0306,
      "relation_loss": 0.0021615814068354664,
      "selected_pairs": 26136.1,
      "state_query_norm": 1123.3252075195312,
      "state_relation_loss": 0.00038765273930039255,
      "step": 14960,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0021615814068354664,
      "weighted_relation_loss": 0.0021615814068354664
    },
    {
      "action_query_norm": 623.2882568359375,
      "action_relation_loss": 0.0017463035997934639,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.003741814780168,
      "flow_matching_loss": 0.031643435917794706,
      "grad_norm": 0.13035160303115845,
      "learning_rate": 5.429921781442318e-05,
      "loss": 0.0338,
      "relation_loss": 0.002133293473161757,
      "selected_pairs": 24327.9,
      "state_query_norm": 1127.1051147460937,
      "state_relation_loss": 0.0003869898762786761,
      "step": 14970,
      "valid_action_samples": 214.3,
      "weighted_auxiliary_loss": 0.002133293473161757,
      "weighted_relation_loss": 0.002133293473161757
    },
    {
      "action_query_norm": 628.5813659667969,
      "action_relation_loss": 0.0017350451205857098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.013096351730589,
      "flow_matching_loss": 0.027803649194538593,
      "grad_norm": 0.11860576272010803,
      "learning_rate": 5.424430367452945e-05,
      "loss": 0.0299,
      "relation_loss": 0.0021230728249065577,
      "selected_pairs": 27086.2,
      "state_query_norm": 1115.5170654296876,
      "state_relation_loss": 0.00038802768976893277,
      "step": 14980,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0021230728249065577,
      "weighted_relation_loss": 0.0021230728249065577
    },
    {
      "action_query_norm": 628.4975769042969,
      "action_relation_loss": 0.0018015764537267387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.02245088868101,
      "flow_matching_loss": 0.029266617074608803,
      "grad_norm": 0.11384425312280655,
      "learning_rate": 5.418938437740655e-05,
      "loss": 0.0315,
      "relation_loss": 0.0021906936657615007,
      "selected_pairs": 26500.3,
      "state_query_norm": 1126.4060668945312,
      "state_relation_loss": 0.00038911722949706016,
      "step": 14990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0021906936657615007,
      "weighted_relation_loss": 0.0021906936657615007
    },
    {
      "action_query_norm": 629.9838806152344,
      "action_relation_loss": 0.0017704084399156272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.03180542563143,
      "flow_matching_loss": 0.02844274491071701,
      "grad_norm": 0.09740863740444183,
      "learning_rate": 5.413445998978658e-05,
      "loss": 0.0306,
      "relation_loss": 0.0021701092133298515,
      "selected_pairs": 27195.5,
      "state_query_norm": 1133.8551391601563,
      "state_relation_loss": 0.0003997007414000109,
      "step": 15000,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0021701092133298515,
      "weighted_relation_loss": 0.0021701092133298515
    },
    {
      "action_query_norm": 627.4828369140625,
      "action_relation_loss": 0.001746539247687906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.041159962581853,
      "flow_matching_loss": 0.028877121955156328,
      "grad_norm": 0.11404439061880112,
      "learning_rate": 5.407953057840789e-05,
      "loss": 0.031,
      "relation_loss": 0.002139397826977074,
      "selected_pairs": 26386.7,
      "state_query_norm": 1135.524267578125,
      "state_relation_loss": 0.0003928585851099342,
      "step": 15010,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.002139397826977074,
      "weighted_relation_loss": 0.002139397826977074
    },
    {
      "action_query_norm": 627.4115478515625,
      "action_relation_loss": 0.0017909504822455347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.050514499532273,
      "flow_matching_loss": 0.028866639360785484,
      "grad_norm": 0.10793838649988174,
      "learning_rate": 5.4024596210014853e-05,
      "loss": 0.031,
      "relation_loss": 0.002176979184150696,
      "selected_pairs": 27927.0,
      "state_query_norm": 1096.1875610351562,
      "state_relation_loss": 0.0003860287135466933,
      "step": 15020,
      "valid_action_samples": 236.8,
      "weighted_auxiliary_loss": 0.002176979184150696,
      "weighted_relation_loss": 0.002176979184150696
    },
    {
      "action_query_norm": 625.6189636230469,
      "action_relation_loss": 0.0018818069715052844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.059869036482693,
      "flow_matching_loss": 0.028477902710437774,
      "grad_norm": 0.10461729019880295,
      "learning_rate": 5.3969656951357935e-05,
      "loss": 0.0307,
      "relation_loss": 0.0022709135431796313,
      "selected_pairs": 26389.8,
      "state_query_norm": 1127.1004028320312,
      "state_relation_loss": 0.0003891065687639639,
      "step": 15030,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0022709135431796313,
      "weighted_relation_loss": 0.0022709135431796313
    },
    {
      "action_query_norm": 628.6683837890625,
      "action_relation_loss": 0.001688324159476906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.069223573433115,
      "flow_matching_loss": 0.027272957004606725,
      "grad_norm": 0.08749915659427643,
      "learning_rate": 5.3914712869193504e-05,
      "loss": 0.0293,
      "relation_loss": 0.002063361613545567,
      "selected_pairs": 26648.8,
      "state_query_norm": 1137.6388549804688,
      "state_relation_loss": 0.0003750374220544472,
      "step": 15040,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002063361613545567,
      "weighted_relation_loss": 0.002063361613545567
    },
    {
      "action_query_norm": 624.2876281738281,
      "action_relation_loss": 0.0017606787267141044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.078578110383535,
      "flow_matching_loss": 0.0282131003215909,
      "grad_norm": 0.10867772251367569,
      "learning_rate": 5.385976403028381e-05,
      "loss": 0.0304,
      "relation_loss": 0.0021512745413929226,
      "selected_pairs": 26926.5,
      "state_query_norm": 1112.619091796875,
      "state_relation_loss": 0.00039059577975422146,
      "step": 15050,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0021512745413929226,
      "weighted_relation_loss": 0.0021512745413929226
    },
    {
      "action_query_norm": 621.73955078125,
      "action_relation_loss": 0.0017588715767487884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.087932647333957,
      "flow_matching_loss": 0.027540707401931284,
      "grad_norm": 0.11116350442171097,
      "learning_rate": 5.3804810501396875e-05,
      "loss": 0.0297,
      "relation_loss": 0.002163459011353552,
      "selected_pairs": 26620.7,
      "state_query_norm": 1109.5130126953125,
      "state_relation_loss": 0.0004045874462462962,
      "step": 15060,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002163459011353552,
      "weighted_relation_loss": 0.002163459011353552
    },
    {
      "action_query_norm": 626.0446228027344,
      "action_relation_loss": 0.0017391129513271153,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.097287184284378,
      "flow_matching_loss": 0.029271696880459785,
      "grad_norm": 0.10936081409454346,
      "learning_rate": 5.374985234930639e-05,
      "loss": 0.0314,
      "relation_loss": 0.002136750076897442,
      "selected_pairs": 26509.2,
      "state_query_norm": 1112.6642456054688,
      "state_relation_loss": 0.00039763712557032703,
      "step": 15070,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002136750076897442,
      "weighted_relation_loss": 0.002136750076897442
    },
    {
      "action_query_norm": 635.5058532714844,
      "action_relation_loss": 0.001757193123921752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.1066417212348,
      "flow_matching_loss": 0.029582038708031177,
      "grad_norm": 0.11532894521951675,
      "learning_rate": 5.3694889640791724e-05,
      "loss": 0.0317,
      "relation_loss": 0.0021520106587558986,
      "selected_pairs": 26578.4,
      "state_query_norm": 1134.7656982421875,
      "state_relation_loss": 0.0003948175406549126,
      "step": 15080,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0021520106587558986,
      "weighted_relation_loss": 0.0021520106587558986
    },
    {
      "action_query_norm": 632.1077453613282,
      "action_relation_loss": 0.0017856918158940971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.11599625818522,
      "flow_matching_loss": 0.027459671534597874,
      "grad_norm": 0.15217038989067078,
      "learning_rate": 5.3639922442637736e-05,
      "loss": 0.0296,
      "relation_loss": 0.0021647050976753234,
      "selected_pairs": 27047.6,
      "state_query_norm": 1131.8103393554688,
      "state_relation_loss": 0.0003790132643189281,
      "step": 15090,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0021647050976753234,
      "weighted_relation_loss": 0.0021647050976753234
    },
    {
      "action_query_norm": 627.4204406738281,
      "action_relation_loss": 0.0017471662373282014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.12535079513564,
      "flow_matching_loss": 0.030082263797521592,
      "grad_norm": 0.1527486890554428,
      "learning_rate": 5.358495082163476e-05,
      "loss": 0.0322,
      "relation_loss": 0.0021342078456655146,
      "selected_pairs": 27266.9,
      "state_query_norm": 1137.517626953125,
      "state_relation_loss": 0.0003870415879646316,
      "step": 15100,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0021342078456655146,
      "weighted_relation_loss": 0.0021342078456655146
    },
    {
      "action_query_norm": 625.6316955566406,
      "action_relation_loss": 0.0018281336175277828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.134705332086062,
      "flow_matching_loss": 0.030436256155371665,
      "grad_norm": 0.11679722368717194,
      "learning_rate": 5.35299748445785e-05,
      "loss": 0.0327,
      "relation_loss": 0.002241733088158071,
      "selected_pairs": 26875.8,
      "state_query_norm": 1124.3785766601563,
      "state_relation_loss": 0.0004135994822718203,
      "step": 15110,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.002241733088158071,
      "weighted_relation_loss": 0.002241733088158071
    },
    {
      "action_query_norm": 626.6874633789063,
      "action_relation_loss": 0.0017615727498196065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.144059869036482,
      "flow_matching_loss": 0.028697358444333076,
      "grad_norm": 0.11410337686538696,
      "learning_rate": 5.3474994578269944e-05,
      "loss": 0.0309,
      "relation_loss": 0.002164740953594446,
      "selected_pairs": 27082.0,
      "state_query_norm": 1118.0714599609375,
      "state_relation_loss": 0.0004031681513879448,
      "step": 15120,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002164740953594446,
      "weighted_relation_loss": 0.002164740953594446
    },
    {
      "action_query_norm": 621.9441528320312,
      "action_relation_loss": 0.0018080141744576395,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.153414405986904,
      "flow_matching_loss": 0.02822024691849947,
      "grad_norm": 0.11914952844381332,
      "learning_rate": 5.34200100895153e-05,
      "loss": 0.0304,
      "relation_loss": 0.0021834929240867494,
      "selected_pairs": 27270.8,
      "state_query_norm": 1122.1042358398438,
      "state_relation_loss": 0.00037547872343566267,
      "step": 15130,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0021834929240867494,
      "weighted_relation_loss": 0.0021834929240867494
    },
    {
      "action_query_norm": 618.6084655761719,
      "action_relation_loss": 0.001857341022696346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.162768942937324,
      "flow_matching_loss": 0.029762431606650354,
      "grad_norm": 0.10917265713214874,
      "learning_rate": 5.3365021445125916e-05,
      "loss": 0.032,
      "relation_loss": 0.002238792832940817,
      "selected_pairs": 26434.4,
      "state_query_norm": 1120.0093383789062,
      "state_relation_loss": 0.0003814517956925556,
      "step": 15140,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002238792832940817,
      "weighted_relation_loss": 0.002238792832940817
    },
    {
      "action_query_norm": 614.4473266601562,
      "action_relation_loss": 0.00174916508840397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.172123479887745,
      "flow_matching_loss": 0.027199242636561392,
      "grad_norm": 0.1079409122467041,
      "learning_rate": 5.331002871191817e-05,
      "loss": 0.0293,
      "relation_loss": 0.002137687325011939,
      "selected_pairs": 27185.7,
      "state_query_norm": 1120.6801879882812,
      "state_relation_loss": 0.0003885222569806501,
      "step": 15150,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002137687325011939,
      "weighted_relation_loss": 0.002137687325011939
    },
    {
      "action_query_norm": 627.0008605957031,
      "action_relation_loss": 0.0017397383810020984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.181478016838167,
      "flow_matching_loss": 0.02914438359439373,
      "grad_norm": 0.12236383557319641,
      "learning_rate": 5.3255031956713443e-05,
      "loss": 0.0313,
      "relation_loss": 0.002130930032581091,
      "selected_pairs": 26733.3,
      "state_query_norm": 1130.8628784179687,
      "state_relation_loss": 0.0003911916719516739,
      "step": 15160,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.002130930032581091,
      "weighted_relation_loss": 0.002130930032581091
    },
    {
      "action_query_norm": 620.6963012695312,
      "action_relation_loss": 0.0017044657724909484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.190832553788587,
      "flow_matching_loss": 0.02918246854096651,
      "grad_norm": 0.12668476998806,
      "learning_rate": 5.320003124633795e-05,
      "loss": 0.0313,
      "relation_loss": 0.002092265523970127,
      "selected_pairs": 26667.5,
      "state_query_norm": 1122.872314453125,
      "state_relation_loss": 0.00038779976312071086,
      "step": 15170,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002092265523970127,
      "weighted_relation_loss": 0.002092265523970127
    },
    {
      "action_query_norm": 624.4303466796875,
      "action_relation_loss": 0.0017608521040529013,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.200187090739009,
      "flow_matching_loss": 0.0281412273645401,
      "grad_norm": 0.13874420523643494,
      "learning_rate": 5.314502664762274e-05,
      "loss": 0.0303,
      "relation_loss": 0.0021553965285420417,
      "selected_pairs": 27406.4,
      "state_query_norm": 1120.6531494140625,
      "state_relation_loss": 0.0003945444186683744,
      "step": 15180,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0021553965285420417,
      "weighted_relation_loss": 0.0021553965285420417
    },
    {
      "action_query_norm": 622.0891906738282,
      "action_relation_loss": 0.0017974121728911996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.209541627689429,
      "flow_matching_loss": 0.027436206676065923,
      "grad_norm": 0.09594906866550446,
      "learning_rate": 5.3090018227403604e-05,
      "loss": 0.0296,
      "relation_loss": 0.0021826557931490242,
      "selected_pairs": 26570.3,
      "state_query_norm": 1131.0157592773437,
      "state_relation_loss": 0.0003852436493616551,
      "step": 15190,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0021826557931490242,
      "weighted_relation_loss": 0.0021826557931490242
    },
    {
      "action_query_norm": 629.7858642578125,
      "action_relation_loss": 0.0017834702623076737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.218896164639851,
      "flow_matching_loss": 0.02980749923735857,
      "grad_norm": 0.1059081107378006,
      "learning_rate": 5.3035006052520955e-05,
      "loss": 0.032,
      "relation_loss": 0.00217205083463341,
      "selected_pairs": 27736.8,
      "state_query_norm": 1111.0849609375,
      "state_relation_loss": 0.00038858058396726845,
      "step": 15200,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.00217205083463341,
      "weighted_relation_loss": 0.00217205083463341
    },
    {
      "action_query_norm": 619.1208557128906,
      "action_relation_loss": 0.0018052169238217176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.228250701590271,
      "flow_matching_loss": 0.02905846871435642,
      "grad_norm": 0.10839883983135223,
      "learning_rate": 5.297999018981976e-05,
      "loss": 0.0313,
      "relation_loss": 0.002193428389728069,
      "selected_pairs": 27057.3,
      "state_query_norm": 1118.2329956054687,
      "state_relation_loss": 0.0003882115037413314,
      "step": 15210,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002193428389728069,
      "weighted_relation_loss": 0.002193428389728069
    },
    {
      "action_query_norm": 619.7973754882812,
      "action_relation_loss": 0.0017714710789732634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.237605238540691,
      "flow_matching_loss": 0.028823412768542767,
      "grad_norm": 0.11303922533988953,
      "learning_rate": 5.2924970706149505e-05,
      "loss": 0.031,
      "relation_loss": 0.002147031202912331,
      "selected_pairs": 26092.1,
      "state_query_norm": 1132.2634033203126,
      "state_relation_loss": 0.0003755601122975349,
      "step": 15220,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.002147031202912331,
      "weighted_relation_loss": 0.002147031202912331
    },
    {
      "action_query_norm": 629.1995422363282,
      "action_relation_loss": 0.0018032017396762967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.246959775491113,
      "flow_matching_loss": 0.02676177993416786,
      "grad_norm": 0.12107568979263306,
      "learning_rate": 5.2869947668364014e-05,
      "loss": 0.029,
      "relation_loss": 0.0021906167035922407,
      "selected_pairs": 26551.4,
      "state_query_norm": 1124.17138671875,
      "state_relation_loss": 0.00038741496100556103,
      "step": 15230,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0021906167035922407,
      "weighted_relation_loss": 0.0021906167035922407
    },
    {
      "action_query_norm": 632.0542846679688,
      "action_relation_loss": 0.0017063878243789078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.256314312441534,
      "flow_matching_loss": 0.028459927067160606,
      "grad_norm": 0.10409455001354218,
      "learning_rate": 5.2814921143321504e-05,
      "loss": 0.0306,
      "relation_loss": 0.0020905000623315574,
      "selected_pairs": 26454.3,
      "state_query_norm": 1126.0098510742187,
      "state_relation_loss": 0.00038411224377341566,
      "step": 15240,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0020905000623315574,
      "weighted_relation_loss": 0.0020905000623315574
    },
    {
      "action_query_norm": 624.7665222167968,
      "action_relation_loss": 0.0017860225052572787,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.265668849391956,
      "flow_matching_loss": 0.029215283691883087,
      "grad_norm": 0.1264689415693283,
      "learning_rate": 5.275989119788436e-05,
      "loss": 0.0314,
      "relation_loss": 0.0021665381966158746,
      "selected_pairs": 26371.2,
      "state_query_norm": 1123.5981567382812,
      "state_relation_loss": 0.00038051572337280957,
      "step": 15250,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0021665381966158746,
      "weighted_relation_loss": 0.0021665381966158746
    },
    {
      "action_query_norm": 626.472265625,
      "action_relation_loss": 0.0017652189475484193,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.275023386342376,
      "flow_matching_loss": 0.026269120536744593,
      "grad_norm": 0.09765785932540894,
      "learning_rate": 5.270485789891919e-05,
      "loss": 0.0284,
      "relation_loss": 0.0021558049018494787,
      "selected_pairs": 26929.4,
      "state_query_norm": 1131.4402221679688,
      "state_relation_loss": 0.0003905859455699101,
      "step": 15260,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0021558049018494787,
      "weighted_relation_loss": 0.0021558049018494787
    },
    {
      "action_query_norm": 623.5114013671875,
      "action_relation_loss": 0.001657309348229319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.284377923292798,
      "flow_matching_loss": 0.029351463168859483,
      "grad_norm": 0.10345542430877686,
      "learning_rate": 5.264982131329661e-05,
      "loss": 0.0314,
      "relation_loss": 0.0020531982416287066,
      "selected_pairs": 27337.0,
      "state_query_norm": 1126.9833862304688,
      "state_relation_loss": 0.00039588889630977065,
      "step": 15270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0020531982416287066,
      "weighted_relation_loss": 0.0020531982416287066
    },
    {
      "action_query_norm": 626.4777893066406,
      "action_relation_loss": 0.001637500512879342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.293732460243218,
      "flow_matching_loss": 0.026488932222127913,
      "grad_norm": 0.1148206889629364,
      "learning_rate": 5.259478150789128e-05,
      "loss": 0.0285,
      "relation_loss": 0.0020295187714509666,
      "selected_pairs": 26663.3,
      "state_query_norm": 1130.9758178710938,
      "state_relation_loss": 0.0003920182818546891,
      "step": 15280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0020295187714509666,
      "weighted_relation_loss": 0.0020295187714509666
    },
    {
      "action_query_norm": 623.9811218261718,
      "action_relation_loss": 0.0016872649430297315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.303086997193638,
      "flow_matching_loss": 0.02735109571367502,
      "grad_norm": 0.11988107860088348,
      "learning_rate": 5.253973854958173e-05,
      "loss": 0.0294,
      "relation_loss": 0.002073128684423864,
      "selected_pairs": 27337.3,
      "state_query_norm": 1124.1665283203124,
      "state_relation_loss": 0.0003858637559460476,
      "step": 15290,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.002073128684423864,
      "weighted_relation_loss": 0.002073128684423864
    },
    {
      "action_query_norm": 619.882861328125,
      "action_relation_loss": 0.0017089322325773537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.31244153414406,
      "flow_matching_loss": 0.028065636940300464,
      "grad_norm": 0.08830269426107407,
      "learning_rate": 5.2484692505250376e-05,
      "loss": 0.0302,
      "relation_loss": 0.0020917669986374675,
      "selected_pairs": 26891.5,
      "state_query_norm": 1129.75185546875,
      "state_relation_loss": 0.0003828347631497309,
      "step": 15300,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0020917669986374675,
      "weighted_relation_loss": 0.0020917669986374675
    },
    {
      "action_query_norm": 618.6648559570312,
      "action_relation_loss": 0.0017486059805378318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.32179607109448,
      "flow_matching_loss": 0.02785332463681698,
      "grad_norm": 0.10233472287654877,
      "learning_rate": 5.242964344178333e-05,
      "loss": 0.03,
      "relation_loss": 0.002133768377825618,
      "selected_pairs": 27132.1,
      "state_query_norm": 1126.3875366210937,
      "state_relation_loss": 0.0003851624031085521,
      "step": 15310,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.002133768377825618,
      "weighted_relation_loss": 0.002133768377825618
    },
    {
      "action_query_norm": 616.726025390625,
      "action_relation_loss": 0.0017937256256118418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.331150608044902,
      "flow_matching_loss": 0.02978628911077976,
      "grad_norm": 0.16085603833198547,
      "learning_rate": 5.237459142607041e-05,
      "loss": 0.032,
      "relation_loss": 0.0021901191445067524,
      "selected_pairs": 27429.6,
      "state_query_norm": 1123.046337890625,
      "state_relation_loss": 0.0003963934985222295,
      "step": 15320,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0021901191445067524,
      "weighted_relation_loss": 0.0021901191445067524
    },
    {
      "action_query_norm": 619.2390258789062,
      "action_relation_loss": 0.0017415418056771158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.340505144995323,
      "flow_matching_loss": 0.027046588622033595,
      "grad_norm": 0.1105339452624321,
      "learning_rate": 5.231953652500498e-05,
      "loss": 0.0292,
      "relation_loss": 0.002118058386258781,
      "selected_pairs": 26352.8,
      "state_query_norm": 1127.00419921875,
      "state_relation_loss": 0.00037651654274668546,
      "step": 15330,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002118058386258781,
      "weighted_relation_loss": 0.002118058386258781
    },
    {
      "action_query_norm": 624.9403137207031,
      "action_relation_loss": 0.0017390179098583757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.349859681945743,
      "flow_matching_loss": 0.027348433062434195,
      "grad_norm": 0.10215991735458374,
      "learning_rate": 5.2264478805483976e-05,
      "loss": 0.0295,
      "relation_loss": 0.0021177919930778443,
      "selected_pairs": 26985.5,
      "state_query_norm": 1123.4233276367188,
      "state_relation_loss": 0.0003787740948610008,
      "step": 15340,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0021177919930778443,
      "weighted_relation_loss": 0.0021177919930778443
    },
    {
      "action_query_norm": 623.4714172363281,
      "action_relation_loss": 0.0017904439475387335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.359214218896165,
      "flow_matching_loss": 0.02870432324707508,
      "grad_norm": 0.10342670977115631,
      "learning_rate": 5.220941833440768e-05,
      "loss": 0.0309,
      "relation_loss": 0.0021542908973060547,
      "selected_pairs": 26387.7,
      "state_query_norm": 1134.6235107421876,
      "state_relation_loss": 0.00036384696140885354,
      "step": 15350,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0021542908973060547,
      "weighted_relation_loss": 0.0021542908973060547
    },
    {
      "action_query_norm": 629.1272644042969,
      "action_relation_loss": 0.001686745323240757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.368568755846585,
      "flow_matching_loss": 0.02599481139332056,
      "grad_norm": 0.13308453559875488,
      "learning_rate": 5.215435517867977e-05,
      "loss": 0.0281,
      "relation_loss": 0.002067377313505858,
      "selected_pairs": 26984.3,
      "state_query_norm": 1130.5451904296874,
      "state_relation_loss": 0.0003806319698924199,
      "step": 15360,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002067377313505858,
      "weighted_relation_loss": 0.002067377313505858
    },
    {
      "action_query_norm": 622.0490356445313,
      "action_relation_loss": 0.0017776539316400886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.377923292797007,
      "flow_matching_loss": 0.027835555374622345,
      "grad_norm": 0.10479810833930969,
      "learning_rate": 5.209928940520719e-05,
      "loss": 0.03,
      "relation_loss": 0.0021622739266604187,
      "selected_pairs": 26897.5,
      "state_query_norm": 1128.5338134765625,
      "state_relation_loss": 0.00038461999502032996,
      "step": 15370,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0021622739266604187,
      "weighted_relation_loss": 0.0021622739266604187
    },
    {
      "action_query_norm": 615.8819763183594,
      "action_relation_loss": 0.0017074559931643308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.387277829747427,
      "flow_matching_loss": 0.0274257967248559,
      "grad_norm": 0.11874513328075409,
      "learning_rate": 5.204422108090004e-05,
      "loss": 0.0295,
      "relation_loss": 0.0021068678004667165,
      "selected_pairs": 26852.0,
      "state_query_norm": 1106.3755859375,
      "state_relation_loss": 0.0003994118393165991,
      "step": 15380,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0021068678004667165,
      "weighted_relation_loss": 0.0021068678004667165
    },
    {
      "action_query_norm": 625.8303344726562,
      "action_relation_loss": 0.0016865385812707245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.39663236669785,
      "flow_matching_loss": 0.02653376292437315,
      "grad_norm": 0.12805286049842834,
      "learning_rate": 5.19891502726715e-05,
      "loss": 0.0286,
      "relation_loss": 0.0020756810670718552,
      "selected_pairs": 27217.9,
      "state_query_norm": 1124.0930053710938,
      "state_relation_loss": 0.0003891424828907475,
      "step": 15390,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0020756810670718552,
      "weighted_relation_loss": 0.0020756810670718552
    },
    {
      "action_query_norm": 623.8283264160157,
      "action_relation_loss": 0.001765322347637266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.40598690364827,
      "flow_matching_loss": 0.02953724768012762,
      "grad_norm": 0.10802459716796875,
      "learning_rate": 5.193407704743782e-05,
      "loss": 0.0317,
      "relation_loss": 0.0021659241290763022,
      "selected_pairs": 27246.0,
      "state_query_norm": 1126.88369140625,
      "state_relation_loss": 0.0004006017552455887,
      "step": 15400,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0021659241290763022,
      "weighted_relation_loss": 0.0021659241290763022
    },
    {
      "action_query_norm": 622.0457885742187,
      "action_relation_loss": 0.0017180586815811693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.41534144059869,
      "flow_matching_loss": 0.026207932271063328,
      "grad_norm": 0.11875951290130615,
      "learning_rate": 5.187900147211815e-05,
      "loss": 0.0283,
      "relation_loss": 0.0021073520183563232,
      "selected_pairs": 26906.2,
      "state_query_norm": 1123.6178833007812,
      "state_relation_loss": 0.0003892933309543878,
      "step": 15410,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0021073520183563232,
      "weighted_relation_loss": 0.0021073520183563232
    },
    {
      "action_query_norm": 622.963330078125,
      "action_relation_loss": 0.0017335624084807932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.424695977549112,
      "flow_matching_loss": 0.025184255838394166,
      "grad_norm": 0.08777198195457458,
      "learning_rate": 5.182392361363453e-05,
      "loss": 0.0273,
      "relation_loss": 0.0021077893441542985,
      "selected_pairs": 26757.4,
      "state_query_norm": 1129.7186157226563,
      "state_relation_loss": 0.0003742269444046542,
      "step": 15420,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0021077893441542985,
      "weighted_relation_loss": 0.0021077893441542985
    },
    {
      "action_query_norm": 622.2206787109375,
      "action_relation_loss": 0.0017516916152089835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.434050514499532,
      "flow_matching_loss": 0.027138050831854343,
      "grad_norm": 0.1187640130519867,
      "learning_rate": 5.176884353891172e-05,
      "loss": 0.0293,
      "relation_loss": 0.0021491956431418657,
      "selected_pairs": 26598.6,
      "state_query_norm": 1131.4867919921876,
      "state_relation_loss": 0.00039750399882905186,
      "step": 15430,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0021491956431418657,
      "weighted_relation_loss": 0.0021491956431418657
    },
    {
      "action_query_norm": 625.318896484375,
      "action_relation_loss": 0.0016532171051949263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.443405051449954,
      "flow_matching_loss": 0.025782280415296555,
      "grad_norm": 0.13005509972572327,
      "learning_rate": 5.171376131487722e-05,
      "loss": 0.0278,
      "relation_loss": 0.0020304299308918417,
      "selected_pairs": 26835.8,
      "state_query_norm": 1135.6353271484375,
      "state_relation_loss": 0.00037721281114500016,
      "step": 15440,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0020304299308918417,
      "weighted_relation_loss": 0.0020304299308918417
    },
    {
      "action_query_norm": 620.0118469238281,
      "action_relation_loss": 0.0018365666037425399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.452759588400374,
      "flow_matching_loss": 0.028251747600734234,
      "grad_norm": 0.09912361949682236,
      "learning_rate": 5.165867700846113e-05,
      "loss": 0.0305,
      "relation_loss": 0.0022288011270575224,
      "selected_pairs": 26554.4,
      "state_query_norm": 1131.552392578125,
      "state_relation_loss": 0.00039223454659804703,
      "step": 15450,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0022288011270575224,
      "weighted_relation_loss": 0.0022288011270575224
    },
    {
      "action_query_norm": 621.7583251953125,
      "action_relation_loss": 0.0018468191381543874,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.462114125350794,
      "flow_matching_loss": 0.026119608804583548,
      "grad_norm": 0.10380896925926208,
      "learning_rate": 5.160359068659606e-05,
      "loss": 0.0284,
      "relation_loss": 0.002238493086770177,
      "selected_pairs": 27078.4,
      "state_query_norm": 1121.5931030273437,
      "state_relation_loss": 0.0003916739777196199,
      "step": 15460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002238493086770177,
      "weighted_relation_loss": 0.002238493086770177
    },
    {
      "action_query_norm": 627.3385009765625,
      "action_relation_loss": 0.0016952500329352916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.471468662301216,
      "flow_matching_loss": 0.030725820176303386,
      "grad_norm": 0.10910364985466003,
      "learning_rate": 5.154850241621712e-05,
      "loss": 0.0328,
      "relation_loss": 0.0020782514475286007,
      "selected_pairs": 26806.7,
      "state_query_norm": 1131.5580932617188,
      "state_relation_loss": 0.0003830013942206278,
      "step": 15470,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0020782514475286007,
      "weighted_relation_loss": 0.0020782514475286007
    },
    {
      "action_query_norm": 618.8203491210937,
      "action_relation_loss": 0.0017415968235582114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.480823199251637,
      "flow_matching_loss": 0.028665022924542427,
      "grad_norm": 0.10155536234378815,
      "learning_rate": 5.149341226426172e-05,
      "loss": 0.0308,
      "relation_loss": 0.002144236769527197,
      "selected_pairs": 26925.2,
      "state_query_norm": 1117.9406616210938,
      "state_relation_loss": 0.0004026399314170703,
      "step": 15480,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002144236769527197,
      "weighted_relation_loss": 0.002144236769527197
    },
    {
      "action_query_norm": 621.847998046875,
      "action_relation_loss": 0.0016609603771939874,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.490177736202059,
      "flow_matching_loss": 0.026469864882528782,
      "grad_norm": 0.10919497162103653,
      "learning_rate": 5.14383202976696e-05,
      "loss": 0.0285,
      "relation_loss": 0.002050447219517082,
      "selected_pairs": 27174.2,
      "state_query_norm": 1120.0966796875,
      "state_relation_loss": 0.00038948685105424377,
      "step": 15490,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002050447219517082,
      "weighted_relation_loss": 0.002050447219517082
    },
    {
      "action_query_norm": 612.6265258789062,
      "action_relation_loss": 0.0016965292394161223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.499532273152479,
      "flow_matching_loss": 0.027549635991454123,
      "grad_norm": 0.09899937361478806,
      "learning_rate": 5.1383226583382684e-05,
      "loss": 0.0296,
      "relation_loss": 0.002079967933241278,
      "selected_pairs": 26833.9,
      "state_query_norm": 1123.359521484375,
      "state_relation_loss": 0.0003834387083770707,
      "step": 15500,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002079967933241278,
      "weighted_relation_loss": 0.002079967933241278
    },
    {
      "action_query_norm": 625.573046875,
      "action_relation_loss": 0.0017084855469875037,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.5088868101029,
      "flow_matching_loss": 0.02640279121696949,
      "grad_norm": 0.1065623015165329,
      "learning_rate": 5.132813118834504e-05,
      "loss": 0.0285,
      "relation_loss": 0.00209066605893895,
      "selected_pairs": 26923.3,
      "state_query_norm": 1129.4595458984375,
      "state_relation_loss": 0.0003821804915787652,
      "step": 15510,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00209066605893895,
      "weighted_relation_loss": 0.00209066605893895
    },
    {
      "action_query_norm": 624.9024658203125,
      "action_relation_loss": 0.0016665623639710247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.518241347053321,
      "flow_matching_loss": 0.026121359318494797,
      "grad_norm": 0.09548677504062653,
      "learning_rate": 5.1273034179502777e-05,
      "loss": 0.0282,
      "relation_loss": 0.0020385423675179483,
      "selected_pairs": 26917.3,
      "state_query_norm": 1111.9525756835938,
      "state_relation_loss": 0.00037198001227807255,
      "step": 15520,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0020385423675179483,
      "weighted_relation_loss": 0.0020385423675179483
    },
    {
      "action_query_norm": 623.135595703125,
      "action_relation_loss": 0.0016961069777607918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.527595884003741,
      "flow_matching_loss": 0.028341253101825715,
      "grad_norm": 0.1075843796133995,
      "learning_rate": 5.1217935623803945e-05,
      "loss": 0.0304,
      "relation_loss": 0.002076835627667606,
      "selected_pairs": 27210.1,
      "state_query_norm": 1128.7025756835938,
      "state_relation_loss": 0.0003807286586379632,
      "step": 15530,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.002076835627667606,
      "weighted_relation_loss": 0.002076835627667606
    },
    {
      "action_query_norm": 623.7458862304687,
      "action_relation_loss": 0.0016111420933157205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.536950420954163,
      "flow_matching_loss": 0.025793830677866936,
      "grad_norm": 0.13224001228809357,
      "learning_rate": 5.1162835588198476e-05,
      "loss": 0.0278,
      "relation_loss": 0.0019968711072579025,
      "selected_pairs": 26991.9,
      "state_query_norm": 1140.9344360351563,
      "state_relation_loss": 0.0003857290168525651,
      "step": 15540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0019968711072579025,
      "weighted_relation_loss": 0.0019968711072579025
    },
    {
      "action_query_norm": 623.8096862792969,
      "action_relation_loss": 0.0016307523008435964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.546304957904583,
      "flow_matching_loss": 0.026538093015551566,
      "grad_norm": 0.1136549562215805,
      "learning_rate": 5.110773413963813e-05,
      "loss": 0.0286,
      "relation_loss": 0.002021892217453569,
      "selected_pairs": 26785.5,
      "state_query_norm": 1127.97099609375,
      "state_relation_loss": 0.000391139934072271,
      "step": 15550,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002021892217453569,
      "weighted_relation_loss": 0.002021892217453569
    },
    {
      "action_query_norm": 623.3687622070313,
      "action_relation_loss": 0.001729639561381191,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.555659494855005,
      "flow_matching_loss": 0.02656414471566677,
      "grad_norm": 0.10271530598402023,
      "learning_rate": 5.105263134507636e-05,
      "loss": 0.0287,
      "relation_loss": 0.002113901195116341,
      "selected_pairs": 26803.6,
      "state_query_norm": 1124.7424560546874,
      "state_relation_loss": 0.0003842616046313196,
      "step": 15560,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002113901195116341,
      "weighted_relation_loss": 0.002113901195116341
    },
    {
      "action_query_norm": 622.5367126464844,
      "action_relation_loss": 0.001643038634210825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.565014031805426,
      "flow_matching_loss": 0.027010437101125717,
      "grad_norm": 0.10285048931837082,
      "learning_rate": 5.099752727146824e-05,
      "loss": 0.029,
      "relation_loss": 0.002023639669641852,
      "selected_pairs": 26345.7,
      "state_query_norm": 1126.3880249023437,
      "state_relation_loss": 0.00038060103834141045,
      "step": 15570,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002023639669641852,
      "weighted_relation_loss": 0.002023639669641852
    },
    {
      "action_query_norm": 624.6343017578125,
      "action_relation_loss": 0.0016371653182432055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.574368568755848,
      "flow_matching_loss": 0.028603004291653633,
      "grad_norm": 0.11505335569381714,
      "learning_rate": 5.094242198577042e-05,
      "loss": 0.0306,
      "relation_loss": 0.002031586365774274,
      "selected_pairs": 26460.5,
      "state_query_norm": 1116.3065185546875,
      "state_relation_loss": 0.0003944210970075801,
      "step": 15580,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002031586365774274,
      "weighted_relation_loss": 0.002031586365774274
    },
    {
      "action_query_norm": 620.7366027832031,
      "action_relation_loss": 0.0016148037277162075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.583723105706268,
      "flow_matching_loss": 0.026175782456994055,
      "grad_norm": 0.108498215675354,
      "learning_rate": 5.088731555494102e-05,
      "loss": 0.0282,
      "relation_loss": 0.001983357931021601,
      "selected_pairs": 26974.0,
      "state_query_norm": 1128.5065795898438,
      "state_relation_loss": 0.0003685542120365426,
      "step": 15590,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001983357931021601,
      "weighted_relation_loss": 0.001983357931021601
    },
    {
      "action_query_norm": 625.3209350585937,
      "action_relation_loss": 0.001752396149095148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.593077642656688,
      "flow_matching_loss": 0.02863877937197685,
      "grad_norm": 0.1368701159954071,
      "learning_rate": 5.0832208045939556e-05,
      "loss": 0.0308,
      "relation_loss": 0.0021356699871830642,
      "selected_pairs": 25839.1,
      "state_query_norm": 1124.290185546875,
      "state_relation_loss": 0.0003832738351775333,
      "step": 15600,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.0021356699871830642,
      "weighted_relation_loss": 0.0021356699871830642
    },
    {
      "action_query_norm": 620.9181457519531,
      "action_relation_loss": 0.001684986415784806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.60243217960711,
      "flow_matching_loss": 0.02738232407718897,
      "grad_norm": 0.09896720945835114,
      "learning_rate": 5.0777099525726844e-05,
      "loss": 0.0295,
      "relation_loss": 0.002069907134864479,
      "selected_pairs": 26663.7,
      "state_query_norm": 1103.7194702148438,
      "state_relation_loss": 0.0003849206899758428,
      "step": 15610,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002069907134864479,
      "weighted_relation_loss": 0.002069907134864479
    },
    {
      "action_query_norm": 621.3631469726563,
      "action_relation_loss": 0.0016339031280949712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.61178671655753,
      "flow_matching_loss": 0.028047130070626737,
      "grad_norm": 0.11217331886291504,
      "learning_rate": 5.072199006126494e-05,
      "loss": 0.0301,
      "relation_loss": 0.0020169793046079577,
      "selected_pairs": 26808.0,
      "state_query_norm": 1124.3470458984375,
      "state_relation_loss": 0.0003830761241260916,
      "step": 15620,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0020169793046079577,
      "weighted_relation_loss": 0.0020169793046079577
    },
    {
      "action_query_norm": 620.2515563964844,
      "action_relation_loss": 0.001764670619741082,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.621141253507952,
      "flow_matching_loss": 0.028081906773149968,
      "grad_norm": 0.10294175893068314,
      "learning_rate": 5.0666879719517026e-05,
      "loss": 0.0302,
      "relation_loss": 0.0021664592204615474,
      "selected_pairs": 27228.9,
      "state_query_norm": 1132.5919677734375,
      "state_relation_loss": 0.00040178858616854993,
      "step": 15630,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0021664592204615474,
      "weighted_relation_loss": 0.0021664592204615474
    },
    {
      "action_query_norm": 622.6493469238281,
      "action_relation_loss": 0.0016641952097415924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.630495790458372,
      "flow_matching_loss": 0.027982152253389358,
      "grad_norm": 0.10680864751338959,
      "learning_rate": 5.061176856744737e-05,
      "loss": 0.03,
      "relation_loss": 0.002038838586304337,
      "selected_pairs": 26619.6,
      "state_query_norm": 1125.1105224609375,
      "state_relation_loss": 0.0003746433911146596,
      "step": 15640,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002038838586304337,
      "weighted_relation_loss": 0.002038838586304337
    },
    {
      "action_query_norm": 626.0131225585938,
      "action_relation_loss": 0.0016893923981115221,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.639850327408793,
      "flow_matching_loss": 0.027785910479724407,
      "grad_norm": 0.136728435754776,
      "learning_rate": 5.0556656672021205e-05,
      "loss": 0.0299,
      "relation_loss": 0.0020706490031443536,
      "selected_pairs": 27014.6,
      "state_query_norm": 1133.1600219726563,
      "state_relation_loss": 0.0003812566224951297,
      "step": 15650,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0020706490031443536,
      "weighted_relation_loss": 0.0020706490031443536
    },
    {
      "action_query_norm": 625.5955688476563,
      "action_relation_loss": 0.001677228230983019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.649204864359215,
      "flow_matching_loss": 0.025671720504760742,
      "grad_norm": 0.0864543467760086,
      "learning_rate": 5.050154410020472e-05,
      "loss": 0.0277,
      "relation_loss": 0.0020620651077479122,
      "selected_pairs": 26594.3,
      "state_query_norm": 1112.0438110351563,
      "state_relation_loss": 0.0003848369145998731,
      "step": 15660,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0020620651077479122,
      "weighted_relation_loss": 0.0020620651077479122
    },
    {
      "action_query_norm": 623.5420532226562,
      "action_relation_loss": 0.0017051746370270848,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.658559401309635,
      "flow_matching_loss": 0.02839454784989357,
      "grad_norm": 0.10509996116161346,
      "learning_rate": 5.044643091896485e-05,
      "loss": 0.0305,
      "relation_loss": 0.0020786124398000537,
      "selected_pairs": 26709.2,
      "state_query_norm": 1130.8127319335938,
      "state_relation_loss": 0.0003734378056833521,
      "step": 15670,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0020786124398000537,
      "weighted_relation_loss": 0.0020786124398000537
    },
    {
      "action_query_norm": 627.1909240722656,
      "action_relation_loss": 0.0017439107643440366,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.667913938260057,
      "flow_matching_loss": 0.027511977404356004,
      "grad_norm": 0.10055168718099594,
      "learning_rate": 5.039131719526932e-05,
      "loss": 0.0296,
      "relation_loss": 0.0021170211024582387,
      "selected_pairs": 26670.7,
      "state_query_norm": 1123.0984008789062,
      "state_relation_loss": 0.000373110335203819,
      "step": 15680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0021170211024582387,
      "weighted_relation_loss": 0.0021170211024582387
    },
    {
      "action_query_norm": 626.5107971191406,
      "action_relation_loss": 0.0016114995931275188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.677268475210477,
      "flow_matching_loss": 0.02695784214884043,
      "grad_norm": 0.13437630236148834,
      "learning_rate": 5.03362029960865e-05,
      "loss": 0.0289,
      "relation_loss": 0.001982290332671255,
      "selected_pairs": 26666.3,
      "state_query_norm": 1122.4627319335937,
      "state_relation_loss": 0.00037079074536450207,
      "step": 15690,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001982290332671255,
      "weighted_relation_loss": 0.001982290332671255
    },
    {
      "action_query_norm": 625.867333984375,
      "action_relation_loss": 0.0016997266793623567,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.686623012160897,
      "flow_matching_loss": 0.02835283726453781,
      "grad_norm": 0.13958513736724854,
      "learning_rate": 5.028108838838533e-05,
      "loss": 0.0304,
      "relation_loss": 0.0020779571612365546,
      "selected_pairs": 27431.7,
      "state_query_norm": 1121.303369140625,
      "state_relation_loss": 0.0003782304673222825,
      "step": 15700,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0020779571612365546,
      "weighted_relation_loss": 0.0020779571612365546
    },
    {
      "action_query_norm": 626.5257507324219,
      "action_relation_loss": 0.0016024089767597617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.69597754911132,
      "flow_matching_loss": 0.02692341860383749,
      "grad_norm": 0.11101843416690826,
      "learning_rate": 5.0225973439135276e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019882017746567725,
      "selected_pairs": 26760.9,
      "state_query_norm": 1131.78642578125,
      "state_relation_loss": 0.00038579278043471277,
      "step": 15710,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0019882017746567725,
      "weighted_relation_loss": 0.0019882017746567725
    },
    {
      "action_query_norm": 626.275439453125,
      "action_relation_loss": 0.0016568927909247576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.70533208606174,
      "flow_matching_loss": 0.027591524459421635,
      "grad_norm": 0.10929621756076813,
      "learning_rate": 5.017085821530617e-05,
      "loss": 0.0296,
      "relation_loss": 0.0020160626037977634,
      "selected_pairs": 27088.1,
      "state_query_norm": 1134.1593994140626,
      "state_relation_loss": 0.00035916982451453803,
      "step": 15720,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0020160626037977634,
      "weighted_relation_loss": 0.0020160626037977634
    },
    {
      "action_query_norm": 621.8812622070312,
      "action_relation_loss": 0.001687335455790162,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.714686623012161,
      "flow_matching_loss": 0.026254747807979584,
      "grad_norm": 0.10851947218179703,
      "learning_rate": 5.011574278386822e-05,
      "loss": 0.0283,
      "relation_loss": 0.002059533924330026,
      "selected_pairs": 26668.7,
      "state_query_norm": 1126.4807861328125,
      "state_relation_loss": 0.0003721984481671825,
      "step": 15730,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002059533924330026,
      "weighted_relation_loss": 0.002059533924330026
    },
    {
      "action_query_norm": 631.4829650878906,
      "action_relation_loss": 0.0016222814098000526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.724041159962582,
      "flow_matching_loss": 0.02765645645558834,
      "grad_norm": 0.11682838201522827,
      "learning_rate": 5.006062721179189e-05,
      "loss": 0.0297,
      "relation_loss": 0.0019950557732954623,
      "selected_pairs": 26573.9,
      "state_query_norm": 1103.1544067382813,
      "state_relation_loss": 0.0003727743693161756,
      "step": 15740,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0019950557732954623,
      "weighted_relation_loss": 0.0019950557732954623
    },
    {
      "action_query_norm": 621.623779296875,
      "action_relation_loss": 0.0016087675001472234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.733395696913004,
      "flow_matching_loss": 0.025978796556591986,
      "grad_norm": 0.11113345623016357,
      "learning_rate": 5.000551156604777e-05,
      "loss": 0.028,
      "relation_loss": 0.001982046267949045,
      "selected_pairs": 27247.2,
      "state_query_norm": 1122.6827880859375,
      "state_relation_loss": 0.00037327878817450255,
      "step": 15750,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001982046267949045,
      "weighted_relation_loss": 0.001982046267949045
    },
    {
      "action_query_norm": 620.4340942382812,
      "action_relation_loss": 0.001747726695612073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.742750233863424,
      "flow_matching_loss": 0.027785309590399265,
      "grad_norm": 0.12101084738969803,
      "learning_rate": 4.9950395913606594e-05,
      "loss": 0.0299,
      "relation_loss": 0.002131075004581362,
      "selected_pairs": 26918.3,
      "state_query_norm": 1108.1689086914062,
      "state_relation_loss": 0.0003833483177004382,
      "step": 15760,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002131075004581362,
      "weighted_relation_loss": 0.002131075004581362
    },
    {
      "action_query_norm": 619.1311767578125,
      "action_relation_loss": 0.0016699984553270042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.752104770813844,
      "flow_matching_loss": 0.02814734000712633,
      "grad_norm": 0.11996661871671677,
      "learning_rate": 4.989528032143903e-05,
      "loss": 0.0302,
      "relation_loss": 0.00204642474418506,
      "selected_pairs": 26343.4,
      "state_query_norm": 1119.924462890625,
      "state_relation_loss": 0.0003764263092307374,
      "step": 15770,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.00204642474418506,
      "weighted_relation_loss": 0.00204642474418506
    },
    {
      "action_query_norm": 625.0689208984375,
      "action_relation_loss": 0.0017063894774764776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.761459307764266,
      "flow_matching_loss": 0.027247336879372596,
      "grad_norm": 0.12914879620075226,
      "learning_rate": 4.984016485651578e-05,
      "loss": 0.0293,
      "relation_loss": 0.002086783014237881,
      "selected_pairs": 26507.2,
      "state_query_norm": 1127.5381469726562,
      "state_relation_loss": 0.00038039351929910483,
      "step": 15780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002086783014237881,
      "weighted_relation_loss": 0.002086783014237881
    },
    {
      "action_query_norm": 634.1330017089844,
      "action_relation_loss": 0.0016320960712619126,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.770813844714686,
      "flow_matching_loss": 0.027475863136351107,
      "grad_norm": 0.13809673488140106,
      "learning_rate": 4.9785049585807274e-05,
      "loss": 0.0295,
      "relation_loss": 0.001998427929356694,
      "selected_pairs": 27568.6,
      "state_query_norm": 1129.0468872070312,
      "state_relation_loss": 0.00036633186391554773,
      "step": 15790,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.001998427929356694,
      "weighted_relation_loss": 0.001998427929356694
    },
    {
      "action_query_norm": 623.1620910644531,
      "action_relation_loss": 0.001639030349906534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.780168381665108,
      "flow_matching_loss": 0.02754662185907364,
      "grad_norm": 0.10950028896331787,
      "learning_rate": 4.972993457628381e-05,
      "loss": 0.0296,
      "relation_loss": 0.002016910817474127,
      "selected_pairs": 26918.4,
      "state_query_norm": 1139.5301635742187,
      "state_relation_loss": 0.0003778804501052946,
      "step": 15800,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002016910817474127,
      "weighted_relation_loss": 0.002016910817474127
    },
    {
      "action_query_norm": 627.3099060058594,
      "action_relation_loss": 0.0016463637934066356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.789522918615528,
      "flow_matching_loss": 0.025808721035718917,
      "grad_norm": 0.13073626160621643,
      "learning_rate": 4.9674819894915306e-05,
      "loss": 0.0278,
      "relation_loss": 0.00200928122503683,
      "selected_pairs": 26868.7,
      "state_query_norm": 1129.431494140625,
      "state_relation_loss": 0.0003629174316301942,
      "step": 15810,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00200928122503683,
      "weighted_relation_loss": 0.00200928122503683
    },
    {
      "action_query_norm": 628.2115417480469,
      "action_relation_loss": 0.0015510722063481809,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.79887745556595,
      "flow_matching_loss": 0.027486870251595973,
      "grad_norm": 0.11203888058662415,
      "learning_rate": 4.9619705608671265e-05,
      "loss": 0.0294,
      "relation_loss": 0.0019364379579201341,
      "selected_pairs": 26462.3,
      "state_query_norm": 1125.4620483398437,
      "state_relation_loss": 0.00038536576903425157,
      "step": 15820,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0019364379579201341,
      "weighted_relation_loss": 0.0019364379579201341
    },
    {
      "action_query_norm": 629.9962463378906,
      "action_relation_loss": 0.0017455235356464982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.80823199251637,
      "flow_matching_loss": 0.026660351455211638,
      "grad_norm": 0.12386709451675415,
      "learning_rate": 4.956459178452079e-05,
      "loss": 0.0288,
      "relation_loss": 0.0021170400781556965,
      "selected_pairs": 26648.6,
      "state_query_norm": 1121.1895874023437,
      "state_relation_loss": 0.0003715165046742186,
      "step": 15830,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0021170400781556965,
      "weighted_relation_loss": 0.0021170400781556965
    },
    {
      "action_query_norm": 630.9769653320312,
      "action_relation_loss": 0.0016272477805614471,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.81758652946679,
      "flow_matching_loss": 0.027308629266917704,
      "grad_norm": 0.11472542583942413,
      "learning_rate": 4.950947848943235e-05,
      "loss": 0.0293,
      "relation_loss": 0.001994752383325249,
      "selected_pairs": 27200.2,
      "state_query_norm": 1130.6555786132812,
      "state_relation_loss": 0.00036750459403265265,
      "step": 15840,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001994752383325249,
      "weighted_relation_loss": 0.001994752383325249
    },
    {
      "action_query_norm": 627.6667785644531,
      "action_relation_loss": 0.0016732057672925293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.826941066417213,
      "flow_matching_loss": 0.02620470281690359,
      "grad_norm": 0.10805530101060867,
      "learning_rate": 4.9454365790373805e-05,
      "loss": 0.0283,
      "relation_loss": 0.0020515933982096614,
      "selected_pairs": 26966.3,
      "state_query_norm": 1111.2751098632812,
      "state_relation_loss": 0.00037838761345483364,
      "step": 15850,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0020515933982096614,
      "weighted_relation_loss": 0.0020515933982096614
    },
    {
      "action_query_norm": 623.1230651855469,
      "action_relation_loss": 0.0016993471304886042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.836295603367633,
      "flow_matching_loss": 0.02845376282930374,
      "grad_norm": 0.10453832894563675,
      "learning_rate": 4.939925375431226e-05,
      "loss": 0.0305,
      "relation_loss": 0.0020742548047564925,
      "selected_pairs": 26754.3,
      "state_query_norm": 1126.3907348632813,
      "state_relation_loss": 0.00037490766553673895,
      "step": 15860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0020742548047564925,
      "weighted_relation_loss": 0.0020742548047564925
    },
    {
      "action_query_norm": 624.9580932617188,
      "action_relation_loss": 0.00166855234419927,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.845650140318055,
      "flow_matching_loss": 0.02752523384988308,
      "grad_norm": 0.13593310117721558,
      "learning_rate": 4.9344142448214046e-05,
      "loss": 0.0296,
      "relation_loss": 0.0020261498750187455,
      "selected_pairs": 27386.8,
      "state_query_norm": 1127.2355834960938,
      "state_relation_loss": 0.0003575975395506248,
      "step": 15870,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0020261498750187455,
      "weighted_relation_loss": 0.0020261498750187455
    },
    {
      "action_query_norm": 620.7245178222656,
      "action_relation_loss": 0.0016398979350924492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.855004677268475,
      "flow_matching_loss": 0.027025093138217927,
      "grad_norm": 0.11950736492872238,
      "learning_rate": 4.928903193904461e-05,
      "loss": 0.029,
      "relation_loss": 0.00202066384954378,
      "selected_pairs": 27322.8,
      "state_query_norm": 1120.3169555664062,
      "state_relation_loss": 0.0003807658940786496,
      "step": 15880,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.00202066384954378,
      "weighted_relation_loss": 0.00202066384954378
    },
    {
      "action_query_norm": 625.1150085449219,
      "action_relation_loss": 0.0016186235821805894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.864359214218895,
      "flow_matching_loss": 0.026667487062513828,
      "grad_norm": 0.11453278362751007,
      "learning_rate": 4.923392229376842e-05,
      "loss": 0.0287,
      "relation_loss": 0.0020085974712856112,
      "selected_pairs": 27087.7,
      "state_query_norm": 1110.8816162109374,
      "state_relation_loss": 0.00038997391529846934,
      "step": 15890,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0020085974712856112,
      "weighted_relation_loss": 0.0020085974712856112
    },
    {
      "action_query_norm": 629.4188293457031,
      "action_relation_loss": 0.0016330578131601215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.873713751169317,
      "flow_matching_loss": 0.02670887764543295,
      "grad_norm": 0.11896637827157974,
      "learning_rate": 4.9178813579348865e-05,
      "loss": 0.0287,
      "relation_loss": 0.002031813969369978,
      "selected_pairs": 26689.0,
      "state_query_norm": 1113.2253173828126,
      "state_relation_loss": 0.0003987561707617715,
      "step": 15900,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002031813969369978,
      "weighted_relation_loss": 0.002031813969369978
    },
    {
      "action_query_norm": 628.6178955078125,
      "action_relation_loss": 0.0016178541700355709,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.883068288119738,
      "flow_matching_loss": 0.026776174642145635,
      "grad_norm": 0.13230100274085999,
      "learning_rate": 4.9123705862748253e-05,
      "loss": 0.0288,
      "relation_loss": 0.0020065529039129614,
      "selected_pairs": 27201.5,
      "state_query_norm": 1122.5560913085938,
      "state_relation_loss": 0.00038869873096700756,
      "step": 15910,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0020065529039129614,
      "weighted_relation_loss": 0.0020065529039129614
    },
    {
      "action_query_norm": 621.0710327148438,
      "action_relation_loss": 0.0016392994089983403,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.89242282507016,
      "flow_matching_loss": 0.02652240954339504,
      "grad_norm": 0.10874374210834503,
      "learning_rate": 4.9068599210927627e-05,
      "loss": 0.0285,
      "relation_loss": 0.0020160139654763045,
      "selected_pairs": 26781.0,
      "state_query_norm": 1134.451318359375,
      "state_relation_loss": 0.00037671453901566565,
      "step": 15920,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0020160139654763045,
      "weighted_relation_loss": 0.0020160139654763045
    },
    {
      "action_query_norm": 633.8908020019531,
      "action_relation_loss": 0.001718008692841977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.90177736202058,
      "flow_matching_loss": 0.027834450639784337,
      "grad_norm": 0.12722325325012207,
      "learning_rate": 4.901349369084681e-05,
      "loss": 0.0299,
      "relation_loss": 0.0020952452439814808,
      "selected_pairs": 26867.7,
      "state_query_norm": 1124.9158081054688,
      "state_relation_loss": 0.0003772365744225681,
      "step": 15930,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0020952452439814808,
      "weighted_relation_loss": 0.0020952452439814808
    },
    {
      "action_query_norm": 631.4693054199219,
      "action_relation_loss": 0.001645052875392139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.911131898971,
      "flow_matching_loss": 0.026809929125010967,
      "grad_norm": 0.09732016175985336,
      "learning_rate": 4.8958389369464156e-05,
      "loss": 0.0288,
      "relation_loss": 0.0020327748847194017,
      "selected_pairs": 27218.8,
      "state_query_norm": 1131.18623046875,
      "state_relation_loss": 0.000387722032610327,
      "step": 15940,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0020327748847194017,
      "weighted_relation_loss": 0.0020327748847194017
    },
    {
      "action_query_norm": 626.6702758789063,
      "action_relation_loss": 0.0015818468760699033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.920486435921422,
      "flow_matching_loss": 0.028685099072754384,
      "grad_norm": 0.13491947948932648,
      "learning_rate": 4.890328631373666e-05,
      "loss": 0.0307,
      "relation_loss": 0.0019697115174494685,
      "selected_pairs": 26825.5,
      "state_query_norm": 1119.44541015625,
      "state_relation_loss": 0.0003878646646626294,
      "step": 15950,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0019697115174494685,
      "weighted_relation_loss": 0.0019697115174494685
    },
    {
      "action_query_norm": 621.2768005371094,
      "action_relation_loss": 0.0017065526102669538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.929840972871842,
      "flow_matching_loss": 0.026897223293781282,
      "grad_norm": 0.10040416568517685,
      "learning_rate": 4.88481845906197e-05,
      "loss": 0.029,
      "relation_loss": 0.002073369163554162,
      "selected_pairs": 26560.2,
      "state_query_norm": 1131.1301391601562,
      "state_relation_loss": 0.0003668165532872081,
      "step": 15960,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.002073369163554162,
      "weighted_relation_loss": 0.002073369163554162
    },
    {
      "action_query_norm": 624.2103698730468,
      "action_relation_loss": 0.0016873687854968011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.939195509822264,
      "flow_matching_loss": 0.027449983917176724,
      "grad_norm": 0.13292613625526428,
      "learning_rate": 4.879308426706707e-05,
      "loss": 0.0295,
      "relation_loss": 0.0020635339198634027,
      "selected_pairs": 26880.1,
      "state_query_norm": 1115.851513671875,
      "state_relation_loss": 0.00037616515182889996,
      "step": 15970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0020635339198634027,
      "weighted_relation_loss": 0.0020635339198634027
    },
    {
      "action_query_norm": 624.5323913574218,
      "action_relation_loss": 0.0016756347380578519,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.948550046772684,
      "flow_matching_loss": 0.027122826501727103,
      "grad_norm": 0.14025025069713593,
      "learning_rate": 4.873798541003084e-05,
      "loss": 0.0292,
      "relation_loss": 0.0020309505285695194,
      "selected_pairs": 27573.8,
      "state_query_norm": 1128.494384765625,
      "state_relation_loss": 0.0003553157876012847,
      "step": 15980,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0020309505285695194,
      "weighted_relation_loss": 0.0020309505285695194
    },
    {
      "action_query_norm": 620.9393859863281,
      "action_relation_loss": 0.0016219743760302664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.957904583723106,
      "flow_matching_loss": 0.02835852950811386,
      "grad_norm": 0.13595661520957947,
      "learning_rate": 4.868288808646136e-05,
      "loss": 0.0303,
      "relation_loss": 0.001983897783793509,
      "selected_pairs": 26729.0,
      "state_query_norm": 1123.9261596679687,
      "state_relation_loss": 0.0003619234048528597,
      "step": 15990,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001983897783793509,
      "weighted_relation_loss": 0.001983897783793509
    },
    {
      "action_query_norm": 622.726953125,
      "action_relation_loss": 0.0016092009958811105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.967259120673527,
      "flow_matching_loss": 0.02609137035906315,
      "grad_norm": 0.16284571588039398,
      "learning_rate": 4.862779236330705e-05,
      "loss": 0.0281,
      "relation_loss": 0.0019772577797994018,
      "selected_pairs": 26467.5,
      "state_query_norm": 1150.9600708007813,
      "state_relation_loss": 0.00036805674899369477,
      "step": 16000,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0019772577797994018,
      "weighted_relation_loss": 0.0019772577797994018
    },
    {
      "action_query_norm": 621.4840454101562,
      "action_relation_loss": 0.0016129648662172257,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.976613657623947,
      "flow_matching_loss": 0.027714090794324873,
      "grad_norm": 0.11232663691043854,
      "learning_rate": 4.8572698307514395e-05,
      "loss": 0.0297,
      "relation_loss": 0.001966431981418282,
      "selected_pairs": 26179.4,
      "state_query_norm": 1116.61455078125,
      "state_relation_loss": 0.0003534671268425882,
      "step": 16010,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.001966431981418282,
      "weighted_relation_loss": 0.001966431981418282
    },
    {
      "action_query_norm": 627.9599975585937,
      "action_relation_loss": 0.0016852839267812668,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.985968194574369,
      "flow_matching_loss": 0.02541534658521414,
      "grad_norm": 0.10442496091127396,
      "learning_rate": 4.85176059860279e-05,
      "loss": 0.0275,
      "relation_loss": 0.0020629703998565675,
      "selected_pairs": 26940.3,
      "state_query_norm": 1129.05087890625,
      "state_relation_loss": 0.00037768646143376827,
      "step": 16020,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0020629703998565675,
      "weighted_relation_loss": 0.0020629703998565675
    },
    {
      "action_query_norm": 623.8820739746094,
      "action_relation_loss": 0.0017097967211157084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.995322731524789,
      "flow_matching_loss": 0.02877891678363085,
      "grad_norm": 0.10227034240961075,
      "learning_rate": 4.846251546578989e-05,
      "loss": 0.0308,
      "relation_loss": 0.0020647029858082535,
      "selected_pairs": 26806.9,
      "state_query_norm": 1122.1776489257813,
      "state_relation_loss": 0.0003549062501406297,
      "step": 16030,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0020647029858082535,
      "weighted_relation_loss": 0.0020647029858082535
    },
    {
      "action_query_norm": 624.6905456542969,
      "action_relation_loss": 0.0016130803618580104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.004677268475211,
      "flow_matching_loss": 0.027535683661699294,
      "grad_norm": 0.11719998717308044,
      "learning_rate": 4.840742681374058e-05,
      "loss": 0.0295,
      "relation_loss": 0.001976225257385522,
      "selected_pairs": 23953.5,
      "state_query_norm": 1130.1068359375,
      "state_relation_loss": 0.0003631448751548305,
      "step": 16040,
      "valid_action_samples": 212.3,
      "weighted_auxiliary_loss": 0.001976225257385522,
      "weighted_relation_loss": 0.001976225257385522
    },
    {
      "action_query_norm": 620.6281311035157,
      "action_relation_loss": 0.0015841954038478435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.014031805425631,
      "flow_matching_loss": 0.02620066422969103,
      "grad_norm": 0.12431429326534271,
      "learning_rate": 4.835234009681787e-05,
      "loss": 0.0282,
      "relation_loss": 0.001957004703581333,
      "selected_pairs": 26436.1,
      "state_query_norm": 1110.7277954101562,
      "state_relation_loss": 0.00037280929973348975,
      "step": 16050,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001957004703581333,
      "weighted_relation_loss": 0.001957004703581333
    },
    {
      "action_query_norm": 630.1491455078125,
      "action_relation_loss": 0.0015611118404194712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.023386342376053,
      "flow_matching_loss": 0.027565689571201802,
      "grad_norm": 0.14590027928352356,
      "learning_rate": 4.8297255381957296e-05,
      "loss": 0.0295,
      "relation_loss": 0.0019333130563609301,
      "selected_pairs": 27079.7,
      "state_query_norm": 1121.5427368164062,
      "state_relation_loss": 0.0003722012072103098,
      "step": 16060,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0019333130563609301,
      "weighted_relation_loss": 0.0019333130563609301
    },
    {
      "action_query_norm": 634.2204406738281,
      "action_relation_loss": 0.0016182180610485376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.032740879326473,
      "flow_matching_loss": 0.026944924145936966,
      "grad_norm": 0.08978911489248276,
      "learning_rate": 4.824217273609198e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019841894158162177,
      "selected_pairs": 26614.3,
      "state_query_norm": 1136.0058837890624,
      "state_relation_loss": 0.00036597137222997843,
      "step": 16070,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0019841894158162177,
      "weighted_relation_loss": 0.0019841894158162177
    },
    {
      "action_query_norm": 619.2404968261719,
      "action_relation_loss": 0.0017266930895857512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.042095416276894,
      "flow_matching_loss": 0.025306074507534505,
      "grad_norm": 0.092035211622715,
      "learning_rate": 4.8187092226152555e-05,
      "loss": 0.0274,
      "relation_loss": 0.0020942109869793056,
      "selected_pairs": 26833.3,
      "state_query_norm": 1104.0885620117188,
      "state_relation_loss": 0.0003675179119454697,
      "step": 16080,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0020942109869793056,
      "weighted_relation_loss": 0.0020942109869793056
    },
    {
      "action_query_norm": 630.2549621582032,
      "action_relation_loss": 0.0016381166642531753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.051449953227316,
      "flow_matching_loss": 0.026977106183767318,
      "grad_norm": 0.0820254310965538,
      "learning_rate": 4.8132013919067016e-05,
      "loss": 0.029,
      "relation_loss": 0.0020105949253775178,
      "selected_pairs": 26782.4,
      "state_query_norm": 1133.8788696289062,
      "state_relation_loss": 0.0003724782553035766,
      "step": 16090,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0020105949253775178,
      "weighted_relation_loss": 0.0020105949253775178
    },
    {
      "action_query_norm": 626.3678100585937,
      "action_relation_loss": 0.0016432834090664983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.060804490177736,
      "flow_matching_loss": 0.02611529137939215,
      "grad_norm": 0.09583177417516708,
      "learning_rate": 4.807693788176071e-05,
      "loss": 0.0281,
      "relation_loss": 0.0020128866308368742,
      "selected_pairs": 27039.5,
      "state_query_norm": 1118.0296020507812,
      "state_relation_loss": 0.0003696032130392268,
      "step": 16100,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0020128866308368742,
      "weighted_relation_loss": 0.0020128866308368742
    },
    {
      "action_query_norm": 632.4380554199219,
      "action_relation_loss": 0.0015670856228098273,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.070159027128158,
      "flow_matching_loss": 0.02649920266121626,
      "grad_norm": 0.13948778808116913,
      "learning_rate": 4.8021864181156214e-05,
      "loss": 0.0284,
      "relation_loss": 0.0019354368327185512,
      "selected_pairs": 26386.3,
      "state_query_norm": 1127.79091796875,
      "state_relation_loss": 0.0003683512128191069,
      "step": 16110,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0019354368327185512,
      "weighted_relation_loss": 0.0019354368327185512
    },
    {
      "action_query_norm": 620.4984558105468,
      "action_relation_loss": 0.0016178725636564195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.079513564078578,
      "flow_matching_loss": 0.026781858690083028,
      "grad_norm": 0.15315331518650055,
      "learning_rate": 4.796679288417326e-05,
      "loss": 0.0288,
      "relation_loss": 0.0019784471136517824,
      "selected_pairs": 26623.9,
      "state_query_norm": 1133.575341796875,
      "state_relation_loss": 0.00036057454126421363,
      "step": 16120,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0019784471136517824,
      "weighted_relation_loss": 0.0019784471136517824
    },
    {
      "action_query_norm": 624.9953491210938,
      "action_relation_loss": 0.0016080867033451795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.088868101028998,
      "flow_matching_loss": 0.024871737137436865,
      "grad_norm": 0.0906754806637764,
      "learning_rate": 4.791172405772866e-05,
      "loss": 0.0268,
      "relation_loss": 0.001969220826867968,
      "selected_pairs": 26922.5,
      "state_query_norm": 1127.349658203125,
      "state_relation_loss": 0.00036113410897087306,
      "step": 16130,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001969220826867968,
      "weighted_relation_loss": 0.001969220826867968
    },
    {
      "action_query_norm": 628.929052734375,
      "action_relation_loss": 0.00158376123290509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.09822263797942,
      "flow_matching_loss": 0.024168499186635016,
      "grad_norm": 0.09759001433849335,
      "learning_rate": 4.785665776873625e-05,
      "loss": 0.0261,
      "relation_loss": 0.001956107746809721,
      "selected_pairs": 26484.2,
      "state_query_norm": 1131.56083984375,
      "state_relation_loss": 0.00037234650226309897,
      "step": 16140,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001956107746809721,
      "weighted_relation_loss": 0.001956107746809721
    },
    {
      "action_query_norm": 626.7797241210938,
      "action_relation_loss": 0.0015832387376576661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.10757717492984,
      "flow_matching_loss": 0.027638241276144983,
      "grad_norm": 0.12823764979839325,
      "learning_rate": 4.7801594084106763e-05,
      "loss": 0.0296,
      "relation_loss": 0.0019502874813042582,
      "selected_pairs": 26597.9,
      "state_query_norm": 1136.8799682617187,
      "state_relation_loss": 0.00036704873200505973,
      "step": 16150,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0019502874813042582,
      "weighted_relation_loss": 0.0019502874813042582
    },
    {
      "action_query_norm": 626.7561889648438,
      "action_relation_loss": 0.0015186599688604475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.116931711880262,
      "flow_matching_loss": 0.026371847465634347,
      "grad_norm": 0.1089237630367279,
      "learning_rate": 4.774653307074775e-05,
      "loss": 0.0283,
      "relation_loss": 0.001895172824151814,
      "selected_pairs": 27200.5,
      "state_query_norm": 1130.7589599609375,
      "state_relation_loss": 0.00037651284656021745,
      "step": 16160,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001895172824151814,
      "weighted_relation_loss": 0.001895172824151814
    },
    {
      "action_query_norm": 632.407666015625,
      "action_relation_loss": 0.0015812457771971823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.126286248830683,
      "flow_matching_loss": 0.024479520879685877,
      "grad_norm": 0.0957636684179306,
      "learning_rate": 4.769147479556355e-05,
      "loss": 0.0264,
      "relation_loss": 0.00195554222445935,
      "selected_pairs": 26597.5,
      "state_query_norm": 1119.7794311523437,
      "state_relation_loss": 0.0003742964559933171,
      "step": 16170,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.00195554222445935,
      "weighted_relation_loss": 0.00195554222445935
    },
    {
      "action_query_norm": 621.0833129882812,
      "action_relation_loss": 0.0016941819223575294,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.135640785781105,
      "flow_matching_loss": 0.02605066429823637,
      "grad_norm": 0.11270596832036972,
      "learning_rate": 4.763641932545515e-05,
      "loss": 0.0281,
      "relation_loss": 0.0020636595436371863,
      "selected_pairs": 26625.3,
      "state_query_norm": 1115.16220703125,
      "state_relation_loss": 0.00036947762419003993,
      "step": 16180,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0020636595436371863,
      "weighted_relation_loss": 0.0020636595436371863
    },
    {
      "action_query_norm": 624.1210998535156,
      "action_relation_loss": 0.0015763712115585804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.144995322731525,
      "flow_matching_loss": 0.027619779482483862,
      "grad_norm": 0.1249212995171547,
      "learning_rate": 4.758136672732013e-05,
      "loss": 0.0296,
      "relation_loss": 0.0019439661991782486,
      "selected_pairs": 26684.1,
      "state_query_norm": 1117.4650024414063,
      "state_relation_loss": 0.00036759499053005127,
      "step": 16190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0019439661991782486,
      "weighted_relation_loss": 0.0019439661991782486
    },
    {
      "action_query_norm": 624.8042419433593,
      "action_relation_loss": 0.0015496711945161223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.154349859681945,
      "flow_matching_loss": 0.025060240738093854,
      "grad_norm": 0.1142757385969162,
      "learning_rate": 4.752631706805261e-05,
      "loss": 0.027,
      "relation_loss": 0.0019094669609330595,
      "selected_pairs": 26397.4,
      "state_query_norm": 1106.9645751953126,
      "state_relation_loss": 0.000359795760596171,
      "step": 16200,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0019094669609330595,
      "weighted_relation_loss": 0.0019094669609330595
    },
    {
      "action_query_norm": 634.4117980957031,
      "action_relation_loss": 0.001636657875496894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.163704396632367,
      "flow_matching_loss": 0.02619122825562954,
      "grad_norm": 0.11116888374090195,
      "learning_rate": 4.7471270414543114e-05,
      "loss": 0.0282,
      "relation_loss": 0.0020083505078218877,
      "selected_pairs": 26830.7,
      "state_query_norm": 1126.2279418945313,
      "state_relation_loss": 0.000371692655608058,
      "step": 16210,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0020083505078218877,
      "weighted_relation_loss": 0.0020083505078218877
    },
    {
      "action_query_norm": 631.8070434570312,
      "action_relation_loss": 0.001563053741119802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.173058933582787,
      "flow_matching_loss": 0.026697220280766487,
      "grad_norm": 0.09042052179574966,
      "learning_rate": 4.741622683367849e-05,
      "loss": 0.0286,
      "relation_loss": 0.0019363146158866585,
      "selected_pairs": 26804.1,
      "state_query_norm": 1128.3189819335937,
      "state_relation_loss": 0.00037326087476685643,
      "step": 16220,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0019363146158866585,
      "weighted_relation_loss": 0.0019363146158866585
    },
    {
      "action_query_norm": 632.2247436523437,
      "action_relation_loss": 0.001621763885486871,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.18241347053321,
      "flow_matching_loss": 0.026974067464470862,
      "grad_norm": 0.0973503440618515,
      "learning_rate": 4.736118639234191e-05,
      "loss": 0.029,
      "relation_loss": 0.0019781658076681195,
      "selected_pairs": 26479.2,
      "state_query_norm": 1118.8176147460938,
      "state_relation_loss": 0.00035640191345009953,
      "step": 16230,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0019781658076681195,
      "weighted_relation_loss": 0.0019781658076681195
    },
    {
      "action_query_norm": 628.4793334960938,
      "action_relation_loss": 0.0016528998385183513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.19176800748363,
      "flow_matching_loss": 0.021927941218018533,
      "grad_norm": 0.10144861787557602,
      "learning_rate": 4.730614915741267e-05,
      "loss": 0.0239,
      "relation_loss": 0.002018413844052702,
      "selected_pairs": 26667.5,
      "state_query_norm": 1128.3576049804688,
      "state_relation_loss": 0.0003655140259070322,
      "step": 16240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002018413844052702,
      "weighted_relation_loss": 0.002018413844052702
    },
    {
      "action_query_norm": 625.3318786621094,
      "action_relation_loss": 0.0015918155782856046,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.20112254443405,
      "flow_matching_loss": 0.025928672030568124,
      "grad_norm": 0.10293277353048325,
      "learning_rate": 4.7251115195766234e-05,
      "loss": 0.0279,
      "relation_loss": 0.001956288004294038,
      "selected_pairs": 27042.2,
      "state_query_norm": 1126.7050903320312,
      "state_relation_loss": 0.0003644724405603483,
      "step": 16250,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001956288004294038,
      "weighted_relation_loss": 0.001956288004294038
    },
    {
      "action_query_norm": 629.6076293945313,
      "action_relation_loss": 0.0015643802587874233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.210477081384472,
      "flow_matching_loss": 0.02678563166409731,
      "grad_norm": 0.08754877746105194,
      "learning_rate": 4.719608457427404e-05,
      "loss": 0.0287,
      "relation_loss": 0.001934516499750316,
      "selected_pairs": 26597.8,
      "state_query_norm": 1132.6132446289062,
      "state_relation_loss": 0.0003701362496940419,
      "step": 16260,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001934516499750316,
      "weighted_relation_loss": 0.001934516499750316
    },
    {
      "action_query_norm": 620.8857482910156,
      "action_relation_loss": 0.0015499804867431521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.219831618334892,
      "flow_matching_loss": 0.027220231667160987,
      "grad_norm": 0.11769066005945206,
      "learning_rate": 4.714105735980347e-05,
      "loss": 0.0291,
      "relation_loss": 0.0019074822892434895,
      "selected_pairs": 26482.7,
      "state_query_norm": 1119.2084838867188,
      "state_relation_loss": 0.0003575018170522526,
      "step": 16270,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0019074822892434895,
      "weighted_relation_loss": 0.0019074822892434895
    },
    {
      "action_query_norm": 620.2823181152344,
      "action_relation_loss": 0.0015298082260414958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.229186155285314,
      "flow_matching_loss": 0.024839270301163195,
      "grad_norm": 0.0999031588435173,
      "learning_rate": 4.70860336192178e-05,
      "loss": 0.0267,
      "relation_loss": 0.001899429364129901,
      "selected_pairs": 26388.8,
      "state_query_norm": 1139.6842407226563,
      "state_relation_loss": 0.0003696211468195543,
      "step": 16280,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001899429364129901,
      "weighted_relation_loss": 0.001899429364129901
    },
    {
      "action_query_norm": 622.3513000488281,
      "action_relation_loss": 0.001541177835315466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.238540692235734,
      "flow_matching_loss": 0.02535307630896568,
      "grad_norm": 0.10173380374908447,
      "learning_rate": 4.7031013419376034e-05,
      "loss": 0.0273,
      "relation_loss": 0.001898067886941135,
      "selected_pairs": 26894.5,
      "state_query_norm": 1128.8804077148438,
      "state_relation_loss": 0.0003568900458049029,
      "step": 16290,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001898067886941135,
      "weighted_relation_loss": 0.001898067886941135
    },
    {
      "action_query_norm": 626.7258056640625,
      "action_relation_loss": 0.0015723908320069313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.247895229186156,
      "flow_matching_loss": 0.024470621719956398,
      "grad_norm": 0.1049855574965477,
      "learning_rate": 4.6975996827132926e-05,
      "loss": 0.0264,
      "relation_loss": 0.0019148992374539374,
      "selected_pairs": 27060.2,
      "state_query_norm": 1133.4994873046876,
      "state_relation_loss": 0.0003425084229093045,
      "step": 16300,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0019148992374539374,
      "weighted_relation_loss": 0.0019148992374539374
    },
    {
      "action_query_norm": 626.1377197265625,
      "action_relation_loss": 0.0015465395990759134,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.257249766136576,
      "flow_matching_loss": 0.025998195074498652,
      "grad_norm": 0.127326101064682,
      "learning_rate": 4.692098390933883e-05,
      "loss": 0.0279,
      "relation_loss": 0.0019049342023208737,
      "selected_pairs": 26761.0,
      "state_query_norm": 1120.9019775390625,
      "state_relation_loss": 0.00035839460324496033,
      "step": 16310,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0019049342023208737,
      "weighted_relation_loss": 0.0019049342023208737
    },
    {
      "action_query_norm": 622.0345275878906,
      "action_relation_loss": 0.0015628237277269363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.266604303086996,
      "flow_matching_loss": 0.02706646006554365,
      "grad_norm": 0.10727079212665558,
      "learning_rate": 4.6865974732839615e-05,
      "loss": 0.029,
      "relation_loss": 0.0019523674622178079,
      "selected_pairs": 27195.6,
      "state_query_norm": 1126.3361328125,
      "state_relation_loss": 0.0003895437461324036,
      "step": 16320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0019523674622178079,
      "weighted_relation_loss": 0.0019523674622178079
    },
    {
      "action_query_norm": 620.7231811523437,
      "action_relation_loss": 0.0015682146418839692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.275958840037418,
      "flow_matching_loss": 0.029090935736894606,
      "grad_norm": 0.1071614995598793,
      "learning_rate": 4.6810969364476615e-05,
      "loss": 0.031,
      "relation_loss": 0.0019423435674980282,
      "selected_pairs": 26384.0,
      "state_query_norm": 1112.589794921875,
      "state_relation_loss": 0.0003741289518075064,
      "step": 16330,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0019423435674980282,
      "weighted_relation_loss": 0.0019423435674980282
    },
    {
      "action_query_norm": 626.2655456542968,
      "action_relation_loss": 0.001590803381986916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.285313376987839,
      "flow_matching_loss": 0.027532605454325676,
      "grad_norm": 0.11169344931840897,
      "learning_rate": 4.675596787108653e-05,
      "loss": 0.0295,
      "relation_loss": 0.0019526611664332449,
      "selected_pairs": 27460.7,
      "state_query_norm": 1124.2099487304688,
      "state_relation_loss": 0.0003618577844463289,
      "step": 16340,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0019526611664332449,
      "weighted_relation_loss": 0.0019526611664332449
    },
    {
      "action_query_norm": 618.9455139160157,
      "action_relation_loss": 0.0015979592106305062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.29466791393826,
      "flow_matching_loss": 0.026559126749634742,
      "grad_norm": 0.10024402290582657,
      "learning_rate": 4.670097031950139e-05,
      "loss": 0.0285,
      "relation_loss": 0.0019661238766275346,
      "selected_pairs": 27337.5,
      "state_query_norm": 1120.8885498046875,
      "state_relation_loss": 0.0003681646514451131,
      "step": 16350,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0019661238766275346,
      "weighted_relation_loss": 0.0019661238766275346
    },
    {
      "action_query_norm": 626.730419921875,
      "action_relation_loss": 0.001575932139530778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.30402245088868,
      "flow_matching_loss": 0.02547118254005909,
      "grad_norm": 0.12526877224445343,
      "learning_rate": 4.664597677654838e-05,
      "loss": 0.0274,
      "relation_loss": 0.001931438664905727,
      "selected_pairs": 27567.8,
      "state_query_norm": 1123.8461547851562,
      "state_relation_loss": 0.0003555065137334168,
      "step": 16360,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.001931438664905727,
      "weighted_relation_loss": 0.001931438664905727
    },
    {
      "action_query_norm": 626.1830993652344,
      "action_relation_loss": 0.0016614752705208958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.313376987839103,
      "flow_matching_loss": 0.027572025172412397,
      "grad_norm": 0.09159506857395172,
      "learning_rate": 4.659098730904986e-05,
      "loss": 0.0296,
      "relation_loss": 0.002027089975308627,
      "selected_pairs": 27596.4,
      "state_query_norm": 1140.610693359375,
      "state_relation_loss": 0.0003656147193396464,
      "step": 16370,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.002027089975308627,
      "weighted_relation_loss": 0.002027089975308627
    },
    {
      "action_query_norm": 628.5176879882813,
      "action_relation_loss": 0.0015784797491505743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.322731524789523,
      "flow_matching_loss": 0.026709439419209956,
      "grad_norm": 0.09429533779621124,
      "learning_rate": 4.6536001983823204e-05,
      "loss": 0.0287,
      "relation_loss": 0.0019498276174999773,
      "selected_pairs": 26575.6,
      "state_query_norm": 1145.3838500976562,
      "state_relation_loss": 0.0003713478799909353,
      "step": 16380,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0019498276174999773,
      "weighted_relation_loss": 0.0019498276174999773
    },
    {
      "action_query_norm": 621.5538452148437,
      "action_relation_loss": 0.0015413679415360094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.332086061739943,
      "flow_matching_loss": 0.024008403532207012,
      "grad_norm": 0.09322100132703781,
      "learning_rate": 4.648102086768078e-05,
      "loss": 0.0259,
      "relation_loss": 0.001904879731591791,
      "selected_pairs": 26801.4,
      "state_query_norm": 1120.6650634765624,
      "state_relation_loss": 0.0003635118016973138,
      "step": 16390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001904879731591791,
      "weighted_relation_loss": 0.001904879731591791
    },
    {
      "action_query_norm": 628.535498046875,
      "action_relation_loss": 0.0016361695947125555,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.341440598690365,
      "flow_matching_loss": 0.025566173158586024,
      "grad_norm": 0.09982341527938843,
      "learning_rate": 4.6426044027429786e-05,
      "loss": 0.0276,
      "relation_loss": 0.0019936724216677248,
      "selected_pairs": 26965.0,
      "state_query_norm": 1139.768603515625,
      "state_relation_loss": 0.0003575028444174677,
      "step": 16400,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0019936724216677248,
      "weighted_relation_loss": 0.0019936724216677248
    },
    {
      "action_query_norm": 620.710791015625,
      "action_relation_loss": 0.0016265622340142726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.350795135640785,
      "flow_matching_loss": 0.02520234677940607,
      "grad_norm": 0.13561324775218964,
      "learning_rate": 4.637107152987234e-05,
      "loss": 0.0272,
      "relation_loss": 0.0019884583889506757,
      "selected_pairs": 26121.3,
      "state_query_norm": 1119.9289184570312,
      "state_relation_loss": 0.00036189616948831824,
      "step": 16410,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.0019884583889506757,
      "weighted_relation_loss": 0.0019884583889506757
    },
    {
      "action_query_norm": 610.7047241210937,
      "action_relation_loss": 0.0016359666478820145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.360149672591207,
      "flow_matching_loss": 0.02675354778766632,
      "grad_norm": 0.08836651593446732,
      "learning_rate": 4.631610344180515e-05,
      "loss": 0.0287,
      "relation_loss": 0.001995504670776427,
      "selected_pairs": 27503.0,
      "state_query_norm": 1122.9711303710938,
      "state_relation_loss": 0.00035953802871517836,
      "step": 16420,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.001995504670776427,
      "weighted_relation_loss": 0.001995504670776427
    },
    {
      "action_query_norm": 625.5868774414063,
      "action_relation_loss": 0.0015391318127512931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.369504209541628,
      "flow_matching_loss": 0.025223356485366822,
      "grad_norm": 0.10757286846637726,
      "learning_rate": 4.626113983001965e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018940577516332268,
      "selected_pairs": 26832.9,
      "state_query_norm": 1131.6400756835938,
      "state_relation_loss": 0.00035492591850925235,
      "step": 16430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0018940577516332268,
      "weighted_relation_loss": 0.0018940577516332268
    },
    {
      "action_query_norm": 623.6839233398438,
      "action_relation_loss": 0.001513382326811552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.378858746492048,
      "flow_matching_loss": 0.025895045883953572,
      "grad_norm": 0.09073302149772644,
      "learning_rate": 4.620618076130182e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018704088870435953,
      "selected_pairs": 26994.1,
      "state_query_norm": 1131.7782958984376,
      "state_relation_loss": 0.0003570265806047246,
      "step": 16440,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0018704088870435953,
      "weighted_relation_loss": 0.0018704088870435953
    },
    {
      "action_query_norm": 624.872607421875,
      "action_relation_loss": 0.0015200524241663516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.38821328344247,
      "flow_matching_loss": 0.02730059251189232,
      "grad_norm": 0.11874212324619293,
      "learning_rate": 4.6151226302432074e-05,
      "loss": 0.0292,
      "relation_loss": 0.0018766639404930174,
      "selected_pairs": 26627.0,
      "state_query_norm": 1125.8142456054688,
      "state_relation_loss": 0.0003566115425201133,
      "step": 16450,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0018766639404930174,
      "weighted_relation_loss": 0.0018766639404930174
    },
    {
      "action_query_norm": 637.5346984863281,
      "action_relation_loss": 0.0015479772118851542,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.39756782039289,
      "flow_matching_loss": 0.02790445201098919,
      "grad_norm": 0.08789270371198654,
      "learning_rate": 4.6096276520185304e-05,
      "loss": 0.0298,
      "relation_loss": 0.0019055222161114215,
      "selected_pairs": 26763.7,
      "state_query_norm": 1145.6175537109375,
      "state_relation_loss": 0.00035754504497162996,
      "step": 16460,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0019055222161114215,
      "weighted_relation_loss": 0.0019055222161114215
    },
    {
      "action_query_norm": 623.4695068359375,
      "action_relation_loss": 0.001597966509871185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.406922357343312,
      "flow_matching_loss": 0.026031441055238247,
      "grad_norm": 0.08122924715280533,
      "learning_rate": 4.604133148133066e-05,
      "loss": 0.028,
      "relation_loss": 0.0019814942381344734,
      "selected_pairs": 27572.1,
      "state_query_norm": 1129.9868408203124,
      "state_relation_loss": 0.0003835277253529057,
      "step": 16470,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0019814942381344734,
      "weighted_relation_loss": 0.0019814942381344734
    },
    {
      "action_query_norm": 626.1858581542969,
      "action_relation_loss": 0.001512642775196582,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.416276894293732,
      "flow_matching_loss": 0.02577501367777586,
      "grad_norm": 0.09441741555929184,
      "learning_rate": 4.598639125263155e-05,
      "loss": 0.0277,
      "relation_loss": 0.0018803323269821703,
      "selected_pairs": 26531.1,
      "state_query_norm": 1124.2031616210938,
      "state_relation_loss": 0.00036768953723367306,
      "step": 16480,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0018803323269821703,
      "weighted_relation_loss": 0.0018803323269821703
    },
    {
      "action_query_norm": 620.90693359375,
      "action_relation_loss": 0.0015763798845000565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.425631431244154,
      "flow_matching_loss": 0.02677941247820854,
      "grad_norm": 0.10743671655654907,
      "learning_rate": 4.593145590084553e-05,
      "loss": 0.0287,
      "relation_loss": 0.0019382692058570682,
      "selected_pairs": 26943.6,
      "state_query_norm": 1119.1980590820312,
      "state_relation_loss": 0.00036188931844662877,
      "step": 16490,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0019382692058570682,
      "weighted_relation_loss": 0.0019382692058570682
    },
    {
      "action_query_norm": 631.9946350097656,
      "action_relation_loss": 0.001599931705277413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.434985968194574,
      "flow_matching_loss": 0.02594454735517502,
      "grad_norm": 0.12022215873003006,
      "learning_rate": 4.58765254927242e-05,
      "loss": 0.0279,
      "relation_loss": 0.0019648677087388934,
      "selected_pairs": 26988.8,
      "state_query_norm": 1132.8777221679688,
      "state_relation_loss": 0.00036493599181994794,
      "step": 16500,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0019648677087388934,
      "weighted_relation_loss": 0.0019648677087388934
    },
    {
      "action_query_norm": 629.9858520507812,
      "action_relation_loss": 0.001493324723560363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.444340505144995,
      "flow_matching_loss": 0.02510344758629799,
      "grad_norm": 0.0994594395160675,
      "learning_rate": 4.582160009501323e-05,
      "loss": 0.027,
      "relation_loss": 0.0018621390801854432,
      "selected_pairs": 26527.6,
      "state_query_norm": 1144.9559814453125,
      "state_relation_loss": 0.00036881434498354794,
      "step": 16510,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0018621390801854432,
      "weighted_relation_loss": 0.0018621390801854432
    },
    {
      "action_query_norm": 628.4996459960937,
      "action_relation_loss": 0.0015952857094816864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.453695042095417,
      "flow_matching_loss": 0.025022411346435548,
      "grad_norm": 0.09837498515844345,
      "learning_rate": 4.5766679774452143e-05,
      "loss": 0.027,
      "relation_loss": 0.0019520204979926348,
      "selected_pairs": 27430.7,
      "state_query_norm": 1119.1908569335938,
      "state_relation_loss": 0.0003567347739590332,
      "step": 16520,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0019520204979926348,
      "weighted_relation_loss": 0.0019520204979926348
    },
    {
      "action_query_norm": 626.489892578125,
      "action_relation_loss": 0.001537933829240501,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.463049579045837,
      "flow_matching_loss": 0.025819884054362773,
      "grad_norm": 0.10020947456359863,
      "learning_rate": 4.571176459777431e-05,
      "loss": 0.0277,
      "relation_loss": 0.0018978757667355239,
      "selected_pairs": 26241.3,
      "state_query_norm": 1130.6900634765625,
      "state_relation_loss": 0.0003599419316742569,
      "step": 16530,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0018978757667355239,
      "weighted_relation_loss": 0.0018978757667355239
    },
    {
      "action_query_norm": 627.5049621582032,
      "action_relation_loss": 0.001582346623763442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.472404115996259,
      "flow_matching_loss": 0.025579925253987314,
      "grad_norm": 0.10833552479743958,
      "learning_rate": 4.5656854631706846e-05,
      "loss": 0.0275,
      "relation_loss": 0.0019620620179921387,
      "selected_pairs": 26967.1,
      "state_query_norm": 1131.0971801757812,
      "state_relation_loss": 0.00037971537676639855,
      "step": 16540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0019620620179921387,
      "weighted_relation_loss": 0.0019620620179921387
    },
    {
      "action_query_norm": 624.5053283691407,
      "action_relation_loss": 0.001591278926935047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.481758652946679,
      "flow_matching_loss": 0.026416333578526973,
      "grad_norm": 0.10622455179691315,
      "learning_rate": 4.560194994297054e-05,
      "loss": 0.0284,
      "relation_loss": 0.001956625247839838,
      "selected_pairs": 27369.5,
      "state_query_norm": 1132.7264282226563,
      "state_relation_loss": 0.00036534630053211006,
      "step": 16550,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001956625247839838,
      "weighted_relation_loss": 0.001956625247839838
    },
    {
      "action_query_norm": 627.7834045410157,
      "action_relation_loss": 0.0015383372316136955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.4911131898971,
      "flow_matching_loss": 0.024737625755369662,
      "grad_norm": 0.11214304715394974,
      "learning_rate": 4.554705059827974e-05,
      "loss": 0.0266,
      "relation_loss": 0.0019058818463236094,
      "selected_pairs": 26457.3,
      "state_query_norm": 1135.2302368164062,
      "state_relation_loss": 0.00036754460306838156,
      "step": 16560,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0019058818463236094,
      "weighted_relation_loss": 0.0019058818463236094
    },
    {
      "action_query_norm": 619.731103515625,
      "action_relation_loss": 0.0015166389872319995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.500467726847521,
      "flow_matching_loss": 0.02572413217276335,
      "grad_norm": 0.10656771808862686,
      "learning_rate": 4.5492156664342365e-05,
      "loss": 0.0276,
      "relation_loss": 0.0018837381154298781,
      "selected_pairs": 26831.7,
      "state_query_norm": 1125.2879760742187,
      "state_relation_loss": 0.00036709912528749556,
      "step": 16570,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0018837381154298781,
      "weighted_relation_loss": 0.0018837381154298781
    },
    {
      "action_query_norm": 621.3878784179688,
      "action_relation_loss": 0.0016158051788806916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.509822263797941,
      "flow_matching_loss": 0.02737119197845459,
      "grad_norm": 0.11745055764913559,
      "learning_rate": 4.5437268207859695e-05,
      "loss": 0.0293,
      "relation_loss": 0.001975344796665013,
      "selected_pairs": 26573.1,
      "state_query_norm": 1132.64365234375,
      "state_relation_loss": 0.00035953959450125695,
      "step": 16580,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.001975344796665013,
      "weighted_relation_loss": 0.001975344796665013
    },
    {
      "action_query_norm": 621.5456604003906,
      "action_relation_loss": 0.00159146124497056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.519176800748363,
      "flow_matching_loss": 0.02623899970203638,
      "grad_norm": 0.09522007405757904,
      "learning_rate": 4.538238529552641e-05,
      "loss": 0.0282,
      "relation_loss": 0.0019473023596219719,
      "selected_pairs": 27269.3,
      "state_query_norm": 1132.862353515625,
      "state_relation_loss": 0.00035584110009949653,
      "step": 16590,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0019473023596219719,
      "weighted_relation_loss": 0.0019473023596219719
    },
    {
      "action_query_norm": 626.8728698730469,
      "action_relation_loss": 0.001578290993347764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.528531337698784,
      "flow_matching_loss": 0.02567635178565979,
      "grad_norm": 0.09458235651254654,
      "learning_rate": 4.5327507994030394e-05,
      "loss": 0.0276,
      "relation_loss": 0.0019479617243632675,
      "selected_pairs": 27010.5,
      "state_query_norm": 1132.248974609375,
      "state_relation_loss": 0.00036967073974665255,
      "step": 16600,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0019479617243632675,
      "weighted_relation_loss": 0.0019479617243632675
    },
    {
      "action_query_norm": 631.2817199707031,
      "action_relation_loss": 0.0014784990926273167,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.537885874649206,
      "flow_matching_loss": 0.024956371635198593,
      "grad_norm": 0.1075054183602333,
      "learning_rate": 4.527263637005273e-05,
      "loss": 0.0268,
      "relation_loss": 0.0018367061275057495,
      "selected_pairs": 26805.8,
      "state_query_norm": 1123.4932983398437,
      "state_relation_loss": 0.0003582070261472836,
      "step": 16610,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0018367061275057495,
      "weighted_relation_loss": 0.0018367061275057495
    },
    {
      "action_query_norm": 624.644091796875,
      "action_relation_loss": 0.001591205142904073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.547240411599626,
      "flow_matching_loss": 0.025078275054693223,
      "grad_norm": 0.12738652527332306,
      "learning_rate": 4.521777049026766e-05,
      "loss": 0.027,
      "relation_loss": 0.0019544584560208023,
      "selected_pairs": 26711.9,
      "state_query_norm": 1123.7181274414063,
      "state_relation_loss": 0.00036325331893749535,
      "step": 16620,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0019544584560208023,
      "weighted_relation_loss": 0.0019544584560208023
    },
    {
      "action_query_norm": 628.5033386230468,
      "action_relation_loss": 0.0015099524753168225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.556594948550046,
      "flow_matching_loss": 0.024483725801110266,
      "grad_norm": 0.08999163657426834,
      "learning_rate": 4.5162910421342386e-05,
      "loss": 0.0263,
      "relation_loss": 0.0018558173207566143,
      "selected_pairs": 26051.0,
      "state_query_norm": 1140.8663330078125,
      "state_relation_loss": 0.00034586484543979167,
      "step": 16630,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0018558173207566143,
      "weighted_relation_loss": 0.0018558173207566143
    },
    {
      "action_query_norm": 631.0990539550781,
      "action_relation_loss": 0.001538702892139554,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.565949485500468,
      "flow_matching_loss": 0.02482412029057741,
      "grad_norm": 0.10065293312072754,
      "learning_rate": 4.510805622993706e-05,
      "loss": 0.0267,
      "relation_loss": 0.0018952071317471565,
      "selected_pairs": 26943.8,
      "state_query_norm": 1133.026220703125,
      "state_relation_loss": 0.0003565042541595176,
      "step": 16640,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0018952071317471565,
      "weighted_relation_loss": 0.0018952071317471565
    },
    {
      "action_query_norm": 628.3453125,
      "action_relation_loss": 0.00154403435299173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.575304022450888,
      "flow_matching_loss": 0.027004349045455456,
      "grad_norm": 0.10283947736024857,
      "learning_rate": 4.505320798270467e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019162246491760016,
      "selected_pairs": 26361.5,
      "state_query_norm": 1124.1769653320312,
      "state_relation_loss": 0.0003721903049154207,
      "step": 16650,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0019162246491760016,
      "weighted_relation_loss": 0.0019162246491760016
    },
    {
      "action_query_norm": 630.1740295410157,
      "action_relation_loss": 0.0014920109300874175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.58465855940131,
      "flow_matching_loss": 0.02456688415259123,
      "grad_norm": 0.09535275399684906,
      "learning_rate": 4.499836574629105e-05,
      "loss": 0.0264,
      "relation_loss": 0.0018474504118785261,
      "selected_pairs": 27128.7,
      "state_query_norm": 1134.1923217773438,
      "state_relation_loss": 0.0003554394992534071,
      "step": 16660,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0018474504118785261,
      "weighted_relation_loss": 0.0018474504118785261
    },
    {
      "action_query_norm": 633.8884643554687,
      "action_relation_loss": 0.0015264556161127985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.59401309635173,
      "flow_matching_loss": 0.025273698568344116,
      "grad_norm": 0.11440540850162506,
      "learning_rate": 4.494352958733465e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018739054794423282,
      "selected_pairs": 27134.4,
      "state_query_norm": 1131.3175659179688,
      "state_relation_loss": 0.00034744988079182804,
      "step": 16670,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0018739054794423282,
      "weighted_relation_loss": 0.0018739054794423282
    },
    {
      "action_query_norm": 629.4229797363281,
      "action_relation_loss": 0.0015267012640833855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.603367633302152,
      "flow_matching_loss": 0.02619821224361658,
      "grad_norm": 0.09910322725772858,
      "learning_rate": 4.488869957246663e-05,
      "loss": 0.0281,
      "relation_loss": 0.0018821483477950095,
      "selected_pairs": 26765.7,
      "state_query_norm": 1120.75654296875,
      "state_relation_loss": 0.0003554470982635394,
      "step": 16680,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0018821483477950095,
      "weighted_relation_loss": 0.0018821483477950095
    },
    {
      "action_query_norm": 622.1153564453125,
      "action_relation_loss": 0.0015380572876892984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.612722170252573,
      "flow_matching_loss": 0.024503070674836636,
      "grad_norm": 0.12915261089801788,
      "learning_rate": 4.4833875768310575e-05,
      "loss": 0.0264,
      "relation_loss": 0.0019005959504283965,
      "selected_pairs": 26068.2,
      "state_query_norm": 1138.4086669921876,
      "state_relation_loss": 0.0003625386802013963,
      "step": 16690,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.0019005959504283965,
      "weighted_relation_loss": 0.0019005959504283965
    },
    {
      "action_query_norm": 627.2720458984375,
      "action_relation_loss": 0.0016157184611074627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.622076707202993,
      "flow_matching_loss": 0.026109476573765278,
      "grad_norm": 0.10308003425598145,
      "learning_rate": 4.47790582414826e-05,
      "loss": 0.0281,
      "relation_loss": 0.001975568139459938,
      "selected_pairs": 26926.0,
      "state_query_norm": 1114.0981079101562,
      "state_relation_loss": 0.0003598497103666887,
      "step": 16700,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001975568139459938,
      "weighted_relation_loss": 0.001975568139459938
    },
    {
      "action_query_norm": 626.6058532714844,
      "action_relation_loss": 0.0014879630878567695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.631431244153415,
      "flow_matching_loss": 0.026071978360414506,
      "grad_norm": 0.10410739481449127,
      "learning_rate": 4.472424705859114e-05,
      "loss": 0.0279,
      "relation_loss": 0.0018464389024302364,
      "selected_pairs": 27500.1,
      "state_query_norm": 1125.9116088867188,
      "state_relation_loss": 0.00035847582621499895,
      "step": 16710,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0018464389024302364,
      "weighted_relation_loss": 0.0018464389024302364
    },
    {
      "action_query_norm": 625.3874572753906,
      "action_relation_loss": 0.0016036443528719246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.640785781103835,
      "flow_matching_loss": 0.027159220725297927,
      "grad_norm": 0.09385878592729568,
      "learning_rate": 4.466944228623701e-05,
      "loss": 0.0291,
      "relation_loss": 0.001966170233208686,
      "selected_pairs": 26386.0,
      "state_query_norm": 1138.5539428710938,
      "state_relation_loss": 0.00036252589779905975,
      "step": 16720,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001966170233208686,
      "weighted_relation_loss": 0.001966170233208686
    },
    {
      "action_query_norm": 632.5817626953125,
      "action_relation_loss": 0.0014744100510142744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.650140318054257,
      "flow_matching_loss": 0.02577545288950205,
      "grad_norm": 0.12061510980129242,
      "learning_rate": 4.461464399101312e-05,
      "loss": 0.0276,
      "relation_loss": 0.00184024820337072,
      "selected_pairs": 26295.9,
      "state_query_norm": 1145.7684814453125,
      "state_relation_loss": 0.0003658381465356797,
      "step": 16730,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.00184024820337072,
      "weighted_relation_loss": 0.00184024820337072
    },
    {
      "action_query_norm": 625.5850219726562,
      "action_relation_loss": 0.0015645507257431746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.659494855004677,
      "flow_matching_loss": 0.026434784382581712,
      "grad_norm": 0.09992697834968567,
      "learning_rate": 4.45598522395046e-05,
      "loss": 0.0284,
      "relation_loss": 0.0019369565416127443,
      "selected_pairs": 27226.0,
      "state_query_norm": 1130.827978515625,
      "state_relation_loss": 0.0003724058246007189,
      "step": 16740,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0019369565416127443,
      "weighted_relation_loss": 0.0019369565416127443
    },
    {
      "action_query_norm": 624.3329711914063,
      "action_relation_loss": 0.0015353940427303315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.668849391955098,
      "flow_matching_loss": 0.02612975426018238,
      "grad_norm": 0.09817628562450409,
      "learning_rate": 4.4505067098288577e-05,
      "loss": 0.028,
      "relation_loss": 0.001905721437651664,
      "selected_pairs": 26974.2,
      "state_query_norm": 1123.2290161132812,
      "state_relation_loss": 0.0003703274152940139,
      "step": 16750,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001905721437651664,
      "weighted_relation_loss": 0.001905721437651664
    },
    {
      "action_query_norm": 626.7249267578125,
      "action_relation_loss": 0.0015258036437444389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.67820392890552,
      "flow_matching_loss": 0.02630047108978033,
      "grad_norm": 0.11174760013818741,
      "learning_rate": 4.445028863393417e-05,
      "loss": 0.0282,
      "relation_loss": 0.0018891829997301102,
      "selected_pairs": 26320.4,
      "state_query_norm": 1138.076220703125,
      "state_relation_loss": 0.0003633793501649052,
      "step": 16760,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0018891829997301102,
      "weighted_relation_loss": 0.0018891829997301102
    },
    {
      "action_query_norm": 625.84140625,
      "action_relation_loss": 0.00161369169363752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.68755846585594,
      "flow_matching_loss": 0.024614785425364973,
      "grad_norm": 0.10397052764892578,
      "learning_rate": 4.439551691300236e-05,
      "loss": 0.0266,
      "relation_loss": 0.0019669550587423147,
      "selected_pairs": 26663.1,
      "state_query_norm": 1124.423388671875,
      "state_relation_loss": 0.0003532633825670928,
      "step": 16770,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0019669550587423147,
      "weighted_relation_loss": 0.0019669550587423147
    },
    {
      "action_query_norm": 618.1975463867187,
      "action_relation_loss": 0.0015443003037944436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.696913002806362,
      "flow_matching_loss": 0.02322445474565029,
      "grad_norm": 0.10045225918292999,
      "learning_rate": 4.434075200204599e-05,
      "loss": 0.0251,
      "relation_loss": 0.0019069677800871432,
      "selected_pairs": 27221.9,
      "state_query_norm": 1113.723583984375,
      "state_relation_loss": 0.00036266747338231653,
      "step": 16780,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0019069677800871432,
      "weighted_relation_loss": 0.0019069677800871432
    },
    {
      "action_query_norm": 635.1909851074219,
      "action_relation_loss": 0.0014698377810418606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.706267539756782,
      "flow_matching_loss": 0.02385468650609255,
      "grad_norm": 0.08929239958524704,
      "learning_rate": 4.428599396760956e-05,
      "loss": 0.0257,
      "relation_loss": 0.001823655073530972,
      "selected_pairs": 26919.6,
      "state_query_norm": 1116.459228515625,
      "state_relation_loss": 0.00035381729539949445,
      "step": 16790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001823655073530972,
      "weighted_relation_loss": 0.001823655073530972
    },
    {
      "action_query_norm": 631.4894592285157,
      "action_relation_loss": 0.0015112732420675457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.715622076707202,
      "flow_matching_loss": 0.02578398697078228,
      "grad_norm": 0.1042541116476059,
      "learning_rate": 4.423124287622926e-05,
      "loss": 0.0276,
      "relation_loss": 0.0018642762675881385,
      "selected_pairs": 26828.6,
      "state_query_norm": 1131.9073486328125,
      "state_relation_loss": 0.00035300304007250815,
      "step": 16800,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0018642762675881385,
      "weighted_relation_loss": 0.0018642762675881385
    },
    {
      "action_query_norm": 627.8949890136719,
      "action_relation_loss": 0.0015055846306495368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.724976613657624,
      "flow_matching_loss": 0.026580004766583443,
      "grad_norm": 0.11177132278680801,
      "learning_rate": 4.417649879443282e-05,
      "loss": 0.0284,
      "relation_loss": 0.0018606019904837013,
      "selected_pairs": 27083.9,
      "state_query_norm": 1128.4944458007812,
      "state_relation_loss": 0.0003550173481926322,
      "step": 16810,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0018606019904837013,
      "weighted_relation_loss": 0.0018606019904837013
    },
    {
      "action_query_norm": 625.4736389160156,
      "action_relation_loss": 0.001482822117395699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.734331150608044,
      "flow_matching_loss": 0.025877521932125093,
      "grad_norm": 0.09365591406822205,
      "learning_rate": 4.4121761788739444e-05,
      "loss": 0.0277,
      "relation_loss": 0.0018488413305021822,
      "selected_pairs": 26829.6,
      "state_query_norm": 1124.7687622070312,
      "state_relation_loss": 0.00036601922765839846,
      "step": 16820,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0018488413305021822,
      "weighted_relation_loss": 0.0018488413305021822
    },
    {
      "action_query_norm": 622.3325866699219,
      "action_relation_loss": 0.0015171439386904239,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.743685687558466,
      "flow_matching_loss": 0.02463831678032875,
      "grad_norm": 0.09634888917207718,
      "learning_rate": 4.4067031925659805e-05,
      "loss": 0.0265,
      "relation_loss": 0.0018770483671687543,
      "selected_pairs": 27243.5,
      "state_query_norm": 1118.30693359375,
      "state_relation_loss": 0.00035990444885101167,
      "step": 16830,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0018770483671687543,
      "weighted_relation_loss": 0.0018770483671687543
    },
    {
      "action_query_norm": 627.9553344726562,
      "action_relation_loss": 0.001554221147671342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.753040224508887,
      "flow_matching_loss": 0.02441183440387249,
      "grad_norm": 0.10124845802783966,
      "learning_rate": 4.4012309271695816e-05,
      "loss": 0.0263,
      "relation_loss": 0.0019204277661629022,
      "selected_pairs": 27393.9,
      "state_query_norm": 1121.0324829101562,
      "state_relation_loss": 0.00036620664759539067,
      "step": 16840,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0019204277661629022,
      "weighted_relation_loss": 0.0019204277661629022
    },
    {
      "action_query_norm": 623.2980590820313,
      "action_relation_loss": 0.0015770749538205563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.762394761459309,
      "flow_matching_loss": 0.02455354705452919,
      "grad_norm": 0.086788609623909,
      "learning_rate": 4.395759389334067e-05,
      "loss": 0.0265,
      "relation_loss": 0.0019433696754276753,
      "selected_pairs": 27478.0,
      "state_query_norm": 1118.3865478515625,
      "state_relation_loss": 0.00036629472160711887,
      "step": 16850,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0019433696754276753,
      "weighted_relation_loss": 0.0019433696754276753
    },
    {
      "action_query_norm": 632.3189819335937,
      "action_relation_loss": 0.0015528293675743043,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.771749298409729,
      "flow_matching_loss": 0.024390616826713084,
      "grad_norm": 0.12390420585870743,
      "learning_rate": 4.3902885857078685e-05,
      "loss": 0.0263,
      "relation_loss": 0.0019227726268582046,
      "selected_pairs": 27087.4,
      "state_query_norm": 1121.3360717773437,
      "state_relation_loss": 0.00036994328256696464,
      "step": 16860,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0019227726268582046,
      "weighted_relation_loss": 0.0019227726268582046
    },
    {
      "action_query_norm": 630.9704223632813,
      "action_relation_loss": 0.0015004855464212596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.781103835360149,
      "flow_matching_loss": 0.025931896269321443,
      "grad_norm": 0.11358642578125,
      "learning_rate": 4.384818522938531e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018664350151084363,
      "selected_pairs": 26917.1,
      "state_query_norm": 1127.3982177734374,
      "state_relation_loss": 0.00036594946577679367,
      "step": 16870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0018664350151084363,
      "weighted_relation_loss": 0.0018664350151084363
    },
    {
      "action_query_norm": 635.15263671875,
      "action_relation_loss": 0.0015015369979664683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.790458372310571,
      "flow_matching_loss": 0.026873179897665977,
      "grad_norm": 0.11881125718355179,
      "learning_rate": 4.379349207672695e-05,
      "loss": 0.0287,
      "relation_loss": 0.0018572867847979069,
      "selected_pairs": 27385.7,
      "state_query_norm": 1130.9727783203125,
      "state_relation_loss": 0.00035574978101067246,
      "step": 16880,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0018572867847979069,
      "weighted_relation_loss": 0.0018572867847979069
    },
    {
      "action_query_norm": 632.1536315917969,
      "action_relation_loss": 0.001571458124089986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.799812909260991,
      "flow_matching_loss": 0.024284368939697744,
      "grad_norm": 0.106810562312603,
      "learning_rate": 4.373880646556098e-05,
      "loss": 0.0262,
      "relation_loss": 0.0019204598269425333,
      "selected_pairs": 27061.1,
      "state_query_norm": 1119.655810546875,
      "state_relation_loss": 0.00034900172031484546,
      "step": 16890,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0019204598269425333,
      "weighted_relation_loss": 0.0019204598269425333
    },
    {
      "action_query_norm": 633.8614562988281,
      "action_relation_loss": 0.0015161697869189084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.809167446211413,
      "flow_matching_loss": 0.025972824543714523,
      "grad_norm": 0.10052992403507233,
      "learning_rate": 4.3684128462335533e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018751577939838172,
      "selected_pairs": 26692.7,
      "state_query_norm": 1145.6838745117188,
      "state_relation_loss": 0.00035898802743759004,
      "step": 16900,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0018751577939838172,
      "weighted_relation_loss": 0.0018751577939838172
    },
    {
      "action_query_norm": 634.9003723144531,
      "action_relation_loss": 0.0015034512849524616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.818521983161833,
      "flow_matching_loss": 0.02528705671429634,
      "grad_norm": 0.09832936525344849,
      "learning_rate": 4.362945813348955e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018610829836688935,
      "selected_pairs": 26625.6,
      "state_query_norm": 1151.9931030273438,
      "state_relation_loss": 0.0003576316812541336,
      "step": 16910,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0018610829836688935,
      "weighted_relation_loss": 0.0018610829836688935
    },
    {
      "action_query_norm": 625.731884765625,
      "action_relation_loss": 0.001569874060805887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.827876520112255,
      "flow_matching_loss": 0.0252943504601717,
      "grad_norm": 0.1038731038570404,
      "learning_rate": 4.357479554545262e-05,
      "loss": 0.0272,
      "relation_loss": 0.0019186183344572783,
      "selected_pairs": 27200.8,
      "state_query_norm": 1122.6795043945312,
      "state_relation_loss": 0.00034874427947215734,
      "step": 16920,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0019186183344572783,
      "weighted_relation_loss": 0.0019186183344572783
    },
    {
      "action_query_norm": 615.6919189453125,
      "action_relation_loss": 0.001572922372724861,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.837231057062676,
      "flow_matching_loss": 0.024941625259816646,
      "grad_norm": 0.10648837685585022,
      "learning_rate": 4.352014076464499e-05,
      "loss": 0.0269,
      "relation_loss": 0.0019322292762808501,
      "selected_pairs": 26894.3,
      "state_query_norm": 1134.5861694335938,
      "state_relation_loss": 0.000359306886093691,
      "step": 16930,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0019322292762808501,
      "weighted_relation_loss": 0.0019322292762808501
    },
    {
      "action_query_norm": 630.6182495117188,
      "action_relation_loss": 0.001505148725118488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.846585594013096,
      "flow_matching_loss": 0.02489984892308712,
      "grad_norm": 0.12369513511657715,
      "learning_rate": 4.346549385747734e-05,
      "loss": 0.0268,
      "relation_loss": 0.0018636594177223742,
      "selected_pairs": 27008.5,
      "state_query_norm": 1123.06044921875,
      "state_relation_loss": 0.00035851069842465224,
      "step": 16940,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0018636594177223742,
      "weighted_relation_loss": 0.0018636594177223742
    },
    {
      "action_query_norm": 625.1272521972656,
      "action_relation_loss": 0.001475949224550277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.855940130963518,
      "flow_matching_loss": 0.025745330564677715,
      "grad_norm": 0.09939449280500412,
      "learning_rate": 4.34108548903508e-05,
      "loss": 0.0276,
      "relation_loss": 0.0018138080835342408,
      "selected_pairs": 27034.0,
      "state_query_norm": 1130.6500610351563,
      "state_relation_loss": 0.0003378588677151129,
      "step": 16950,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0018138080835342408,
      "weighted_relation_loss": 0.0018138080835342408
    },
    {
      "action_query_norm": 630.2885925292969,
      "action_relation_loss": 0.001518925663549453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.865294667913938,
      "flow_matching_loss": 0.025416882522404195,
      "grad_norm": 0.1385190635919571,
      "learning_rate": 4.335622392965689e-05,
      "loss": 0.0273,
      "relation_loss": 0.0018650857266038657,
      "selected_pairs": 26692.1,
      "state_query_norm": 1136.8312866210938,
      "state_relation_loss": 0.0003461600805167109,
      "step": 16960,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0018650857266038657,
      "weighted_relation_loss": 0.0018650857266038657
    },
    {
      "action_query_norm": 631.9147399902344,
      "action_relation_loss": 0.001551300345454365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.87464920486436,
      "flow_matching_loss": 0.027336221002042294,
      "grad_norm": 0.1253580003976822,
      "learning_rate": 4.330160104177738e-05,
      "loss": 0.0292,
      "relation_loss": 0.0018967015203088522,
      "selected_pairs": 27527.9,
      "state_query_norm": 1133.61044921875,
      "state_relation_loss": 0.0003454011864960194,
      "step": 16970,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0018967015203088522,
      "weighted_relation_loss": 0.0018967015203088522
    },
    {
      "action_query_norm": 633.7779968261718,
      "action_relation_loss": 0.0014978108345530927,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.88400374181478,
      "flow_matching_loss": 0.02482215892523527,
      "grad_norm": 0.11735403537750244,
      "learning_rate": 4.32469862930842e-05,
      "loss": 0.0267,
      "relation_loss": 0.0018413423444144429,
      "selected_pairs": 26807.5,
      "state_query_norm": 1130.490966796875,
      "state_relation_loss": 0.00034353151568211615,
      "step": 16980,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0018413423444144429,
      "weighted_relation_loss": 0.0018413423444144429
    },
    {
      "action_query_norm": 629.5772827148437,
      "action_relation_loss": 0.0015152710839174688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.8933582787652,
      "flow_matching_loss": 0.02616587094962597,
      "grad_norm": 0.11672481149435043,
      "learning_rate": 4.319237974993946e-05,
      "loss": 0.028,
      "relation_loss": 0.0018733486765995621,
      "selected_pairs": 27216.5,
      "state_query_norm": 1139.6330322265626,
      "state_relation_loss": 0.0003580776130547747,
      "step": 16990,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0018733486765995621,
      "weighted_relation_loss": 0.0018733486765995621
    },
    {
      "action_query_norm": 624.3539184570312,
      "action_relation_loss": 0.0014727527275681495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.902712815715622,
      "flow_matching_loss": 0.026569704711437225,
      "grad_norm": 0.09417860954999924,
      "learning_rate": 4.3137781478695236e-05,
      "loss": 0.0284,
      "relation_loss": 0.001816335308831185,
      "selected_pairs": 26617.3,
      "state_query_norm": 1135.1985107421874,
      "state_relation_loss": 0.0003435825812630355,
      "step": 17000,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001816335308831185,
      "weighted_relation_loss": 0.001816335308831185
    },
    {
      "action_query_norm": 629.3379699707032,
      "action_relation_loss": 0.0014892600709572434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.912067352666043,
      "flow_matching_loss": 0.026833930425345896,
      "grad_norm": 0.09349124878644943,
      "learning_rate": 4.308319154569358e-05,
      "loss": 0.0287,
      "relation_loss": 0.001845556974876672,
      "selected_pairs": 26991.5,
      "state_query_norm": 1132.9124877929687,
      "state_relation_loss": 0.00035629688645713033,
      "step": 17010,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001845556974876672,
      "weighted_relation_loss": 0.001845556974876672
    },
    {
      "action_query_norm": 628.7028930664062,
      "action_relation_loss": 0.0014591217506676913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.921421889616465,
      "flow_matching_loss": 0.024476057663559914,
      "grad_norm": 0.09374343603849411,
      "learning_rate": 4.302861001726641e-05,
      "loss": 0.0263,
      "relation_loss": 0.001809841999784112,
      "selected_pairs": 27675.7,
      "state_query_norm": 1135.75009765625,
      "state_relation_loss": 0.000350720266578719,
      "step": 17020,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.001809841999784112,
      "weighted_relation_loss": 0.001809841999784112
    },
    {
      "action_query_norm": 628.195166015625,
      "action_relation_loss": 0.0014858555397950113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.930776426566885,
      "flow_matching_loss": 0.02527585793286562,
      "grad_norm": 0.09122785925865173,
      "learning_rate": 4.2974036959735415e-05,
      "loss": 0.0271,
      "relation_loss": 0.00184954033466056,
      "selected_pairs": 27014.2,
      "state_query_norm": 1127.6829833984375,
      "state_relation_loss": 0.00036368479777593166,
      "step": 17030,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.00184954033466056,
      "weighted_relation_loss": 0.00184954033466056
    },
    {
      "action_query_norm": 633.9422241210938,
      "action_relation_loss": 0.0015190039179287852,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.940130963517305,
      "flow_matching_loss": 0.026631225273013114,
      "grad_norm": 0.09550797939300537,
      "learning_rate": 4.2919472439412034e-05,
      "loss": 0.0285,
      "relation_loss": 0.001865877618547529,
      "selected_pairs": 26272.7,
      "state_query_norm": 1129.2965942382812,
      "state_relation_loss": 0.0003468737122602761,
      "step": 17040,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.001865877618547529,
      "weighted_relation_loss": 0.001865877618547529
    },
    {
      "action_query_norm": 625.0002258300781,
      "action_relation_loss": 0.0014971922733820974,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.949485500467727,
      "flow_matching_loss": 0.02483872715383768,
      "grad_norm": 0.1177506074309349,
      "learning_rate": 4.286491652259729e-05,
      "loss": 0.0267,
      "relation_loss": 0.0018561528995633125,
      "selected_pairs": 26462.8,
      "state_query_norm": 1114.4864868164063,
      "state_relation_loss": 0.00035896060871891676,
      "step": 17050,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0018561528995633125,
      "weighted_relation_loss": 0.0018561528995633125
    },
    {
      "action_query_norm": 632.3008117675781,
      "action_relation_loss": 0.0014793930924497545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.958840037418147,
      "flow_matching_loss": 0.0283960098400712,
      "grad_norm": 0.10731944441795349,
      "learning_rate": 4.281036927558178e-05,
      "loss": 0.0302,
      "relation_loss": 0.0018402787391096354,
      "selected_pairs": 27129.1,
      "state_query_norm": 1135.4752685546875,
      "state_relation_loss": 0.00036088566412217915,
      "step": 17060,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0018402787391096354,
      "weighted_relation_loss": 0.0018402787391096354
    },
    {
      "action_query_norm": 632.4580078125,
      "action_relation_loss": 0.0014745552442036569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.96819457436857,
      "flow_matching_loss": 0.02493268232792616,
      "grad_norm": 0.08601221442222595,
      "learning_rate": 4.275583076464552e-05,
      "loss": 0.0267,
      "relation_loss": 0.0018160585663281382,
      "selected_pairs": 26433.5,
      "state_query_norm": 1137.503173828125,
      "state_relation_loss": 0.00034150334540754556,
      "step": 17070,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0018160585663281382,
      "weighted_relation_loss": 0.0018160585663281382
    },
    {
      "action_query_norm": 633.6856140136719,
      "action_relation_loss": 0.0014675089507363737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.97754911131899,
      "flow_matching_loss": 0.02357028629630804,
      "grad_norm": 0.09947951883077621,
      "learning_rate": 4.270130105605794e-05,
      "loss": 0.0254,
      "relation_loss": 0.0018230952322483064,
      "selected_pairs": 26940.6,
      "state_query_norm": 1117.8167236328125,
      "state_relation_loss": 0.0003555862727807835,
      "step": 17080,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0018230952322483064,
      "weighted_relation_loss": 0.0018230952322483064
    },
    {
      "action_query_norm": 629.6553588867188,
      "action_relation_loss": 0.0014943877700716256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.986903648269411,
      "flow_matching_loss": 0.025986503809690475,
      "grad_norm": 0.11113102734088898,
      "learning_rate": 4.264678021607782e-05,
      "loss": 0.0278,
      "relation_loss": 0.001848696218803525,
      "selected_pairs": 26878.4,
      "state_query_norm": 1132.0679565429687,
      "state_relation_loss": 0.0003543084691045806,
      "step": 17090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001848696218803525,
      "weighted_relation_loss": 0.001848696218803525
    },
    {
      "action_query_norm": 627.6485168457032,
      "action_relation_loss": 0.0014821584685705603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.996258185219832,
      "flow_matching_loss": 0.0256192147731781,
      "grad_norm": 0.12319190055131912,
      "learning_rate": 4.25922683109531e-05,
      "loss": 0.0275,
      "relation_loss": 0.0018429570249281825,
      "selected_pairs": 26776.8,
      "state_query_norm": 1135.9253295898438,
      "state_relation_loss": 0.0003607985592680052,
      "step": 17100,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0018429570249281825,
      "weighted_relation_loss": 0.0018429570249281825
    },
    {
      "action_query_norm": 631.1104675292969,
      "action_relation_loss": 0.0013478718348778785,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.00561272217025,
      "flow_matching_loss": 0.025425304658710957,
      "grad_norm": 0.1005810871720314,
      "learning_rate": 4.25377654069209e-05,
      "loss": 0.0271,
      "relation_loss": 0.001710706518497318,
      "selected_pairs": 24332.7,
      "state_query_norm": 1137.558819580078,
      "state_relation_loss": 0.00036283468070905653,
      "step": 17110,
      "valid_action_samples": 214.4,
      "weighted_auxiliary_loss": 0.001710706518497318,
      "weighted_relation_loss": 0.001710706518497318
    },
    {
      "action_query_norm": 625.5044311523437,
      "action_relation_loss": 0.001465614594053477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.014967259120674,
      "flow_matching_loss": 0.027458901703357696,
      "grad_norm": 0.10482347756624222,
      "learning_rate": 4.248327157020737e-05,
      "loss": 0.0293,
      "relation_loss": 0.0018125255475752056,
      "selected_pairs": 26858.2,
      "state_query_norm": 1137.0857299804688,
      "state_relation_loss": 0.0003469109651632607,
      "step": 17120,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0018125255475752056,
      "weighted_relation_loss": 0.0018125255475752056
    },
    {
      "action_query_norm": 627.8787963867187,
      "action_relation_loss": 0.0015171958482824266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.024321796071096,
      "flow_matching_loss": 0.024508091621100903,
      "grad_norm": 0.12448190152645111,
      "learning_rate": 4.242878686702763e-05,
      "loss": 0.0264,
      "relation_loss": 0.0018772599287331104,
      "selected_pairs": 27454.4,
      "state_query_norm": 1131.5074584960937,
      "state_relation_loss": 0.00036006407462991775,
      "step": 17130,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0018772599287331104,
      "weighted_relation_loss": 0.0018772599287331104
    },
    {
      "action_query_norm": 626.3085693359375,
      "action_relation_loss": 0.0014868807862512766,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.033676333021514,
      "flow_matching_loss": 0.02409947607666254,
      "grad_norm": 0.0859685018658638,
      "learning_rate": 4.23743113635858e-05,
      "loss": 0.0259,
      "relation_loss": 0.0018481410457752644,
      "selected_pairs": 26645.1,
      "state_query_norm": 1123.7521362304688,
      "state_relation_loss": 0.0003612602537032217,
      "step": 17140,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0018481410457752644,
      "weighted_relation_loss": 0.0018481410457752644
    },
    {
      "action_query_norm": 633.0797607421875,
      "action_relation_loss": 0.0013923182734288276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.043030869971936,
      "flow_matching_loss": 0.025503003410995005,
      "grad_norm": 0.11549417674541473,
      "learning_rate": 4.2319845126074705e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017335889046080411,
      "selected_pairs": 26788.6,
      "state_query_norm": 1132.5068237304688,
      "state_relation_loss": 0.00034127062535844744,
      "step": 17150,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0017335889046080411,
      "weighted_relation_loss": 0.0017335889046080411
    },
    {
      "action_query_norm": 635.0056457519531,
      "action_relation_loss": 0.0013514479156583547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.052385406922358,
      "flow_matching_loss": 0.022994223795831204,
      "grad_norm": 0.10552725940942764,
      "learning_rate": 4.226538822067598e-05,
      "loss": 0.0247,
      "relation_loss": 0.001694978028535843,
      "selected_pairs": 26862.8,
      "state_query_norm": 1152.3385620117188,
      "state_relation_loss": 0.0003435301274294034,
      "step": 17160,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001694978028535843,
      "weighted_relation_loss": 0.001694978028535843
    },
    {
      "action_query_norm": 626.9070556640625,
      "action_relation_loss": 0.0014810907770879567,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.061739943872777,
      "flow_matching_loss": 0.024867543391883372,
      "grad_norm": 0.09908952564001083,
      "learning_rate": 4.2210940713559896e-05,
      "loss": 0.0267,
      "relation_loss": 0.0018377029686234892,
      "selected_pairs": 26880.3,
      "state_query_norm": 1129.0476806640625,
      "state_relation_loss": 0.00035661220317706466,
      "step": 17170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0018377029686234892,
      "weighted_relation_loss": 0.0018377029686234892
    },
    {
      "action_query_norm": 630.0281677246094,
      "action_relation_loss": 0.0013882347964681685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.0710944808232,
      "flow_matching_loss": 0.024461602978408336,
      "grad_norm": 0.0837559923529625,
      "learning_rate": 4.21565026708853e-05,
      "loss": 0.0262,
      "relation_loss": 0.0017442678101360798,
      "selected_pairs": 26399.7,
      "state_query_norm": 1131.666162109375,
      "state_relation_loss": 0.0003560330194886774,
      "step": 17180,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0017442678101360798,
      "weighted_relation_loss": 0.0017442678101360798
    },
    {
      "action_query_norm": 630.134716796875,
      "action_relation_loss": 0.0014512979774735867,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.08044901777362,
      "flow_matching_loss": 0.02331931795924902,
      "grad_norm": 0.09390153735876083,
      "learning_rate": 4.2102074158799535e-05,
      "loss": 0.0251,
      "relation_loss": 0.0017910460010170937,
      "selected_pairs": 26803.0,
      "state_query_norm": 1130.6574462890626,
      "state_relation_loss": 0.0003397480119019747,
      "step": 17190,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0017910460010170937,
      "weighted_relation_loss": 0.0017910460010170937
    },
    {
      "action_query_norm": 625.0996948242188,
      "action_relation_loss": 0.0014204252627678215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.089803554724043,
      "flow_matching_loss": 0.025327707454562188,
      "grad_norm": 0.09825015813112259,
      "learning_rate": 4.204765524343841e-05,
      "loss": 0.0271,
      "relation_loss": 0.0017633653129450976,
      "selected_pairs": 26847.7,
      "state_query_norm": 1136.7139526367187,
      "state_relation_loss": 0.0003429400676395744,
      "step": 17200,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0017633653129450976,
      "weighted_relation_loss": 0.0017633653129450976
    },
    {
      "action_query_norm": 623.6362548828125,
      "action_relation_loss": 0.00145823759958148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.09915809167446,
      "flow_matching_loss": 0.023959286324679852,
      "grad_norm": 0.09148769080638885,
      "learning_rate": 4.1993245990926036e-05,
      "loss": 0.0258,
      "relation_loss": 0.0018110418925061822,
      "selected_pairs": 26280.7,
      "state_query_norm": 1119.3360229492187,
      "state_relation_loss": 0.0003528042783727869,
      "step": 17210,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0018110418925061822,
      "weighted_relation_loss": 0.0018110418925061822
    },
    {
      "action_query_norm": 622.4918518066406,
      "action_relation_loss": 0.0014209454413503409,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.108512628624883,
      "flow_matching_loss": 0.02528599984943867,
      "grad_norm": 0.10991528630256653,
      "learning_rate": 4.193884646737474e-05,
      "loss": 0.027,
      "relation_loss": 0.0017608635709621013,
      "selected_pairs": 26422.3,
      "state_query_norm": 1142.7638793945312,
      "state_relation_loss": 0.0003399181237909943,
      "step": 17220,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0017608635709621013,
      "weighted_relation_loss": 0.0017608635709621013
    },
    {
      "action_query_norm": 628.6698669433594,
      "action_relation_loss": 0.0014342518523335456,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.117867165575305,
      "flow_matching_loss": 0.024976877495646477,
      "grad_norm": 0.1437229961156845,
      "learning_rate": 4.1884456738885124e-05,
      "loss": 0.0268,
      "relation_loss": 0.0017845946247689426,
      "selected_pairs": 26233.6,
      "state_query_norm": 1132.8969360351562,
      "state_relation_loss": 0.000350342778256163,
      "step": 17230,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0017845946247689426,
      "weighted_relation_loss": 0.0017845946247689426
    },
    {
      "action_query_norm": 629.9466247558594,
      "action_relation_loss": 0.0013837552396580577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.127221702525723,
      "flow_matching_loss": 0.025514999777078627,
      "grad_norm": 0.10673125088214874,
      "learning_rate": 4.1830076871545795e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017288802191615105,
      "selected_pairs": 27080.3,
      "state_query_norm": 1117.6691650390626,
      "state_relation_loss": 0.00034512498823460194,
      "step": 17240,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0017288802191615105,
      "weighted_relation_loss": 0.0017288802191615105
    },
    {
      "action_query_norm": 629.828759765625,
      "action_relation_loss": 0.001381454325746745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.136576239476145,
      "flow_matching_loss": 0.025629484094679356,
      "grad_norm": 0.11404038965702057,
      "learning_rate": 4.177570693143346e-05,
      "loss": 0.0274,
      "relation_loss": 0.0017317982157692314,
      "selected_pairs": 27015.2,
      "state_query_norm": 1123.7308471679687,
      "state_relation_loss": 0.0003503438812913373,
      "step": 17250,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0017317982157692314,
      "weighted_relation_loss": 0.0017317982157692314
    },
    {
      "action_query_norm": 625.3996032714844,
      "action_relation_loss": 0.0015174246393144131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.145930776426567,
      "flow_matching_loss": 0.02501455321907997,
      "grad_norm": 0.11740010231733322,
      "learning_rate": 4.172134698461271e-05,
      "loss": 0.0269,
      "relation_loss": 0.001861076313070953,
      "selected_pairs": 26547.1,
      "state_query_norm": 1115.1449096679687,
      "state_relation_loss": 0.0003436516708461568,
      "step": 17260,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001861076313070953,
      "weighted_relation_loss": 0.001861076313070953
    },
    {
      "action_query_norm": 630.9160583496093,
      "action_relation_loss": 0.0014335865853354335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.15528531337699,
      "flow_matching_loss": 0.024806414358317853,
      "grad_norm": 0.10388179868459702,
      "learning_rate": 4.166699709713599e-05,
      "loss": 0.0266,
      "relation_loss": 0.0017850895877927543,
      "selected_pairs": 27318.2,
      "state_query_norm": 1134.239013671875,
      "state_relation_loss": 0.0003515029733534902,
      "step": 17270,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0017850895877927543,
      "weighted_relation_loss": 0.0017850895877927543
    },
    {
      "action_query_norm": 629.9865173339844,
      "action_relation_loss": 0.001403050997760147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.164639850327408,
      "flow_matching_loss": 0.025677903927862643,
      "grad_norm": 0.09059610962867737,
      "learning_rate": 4.1612657335043555e-05,
      "loss": 0.0274,
      "relation_loss": 0.0017425641650334,
      "selected_pairs": 26962.5,
      "state_query_norm": 1128.8040649414063,
      "state_relation_loss": 0.0003395131701836362,
      "step": 17280,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0017425641650334,
      "weighted_relation_loss": 0.0017425641650334
    },
    {
      "action_query_norm": 632.6601867675781,
      "action_relation_loss": 0.00146835392806679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.17399438727783,
      "flow_matching_loss": 0.02389972675591707,
      "grad_norm": 0.09049773216247559,
      "learning_rate": 4.155832776436331e-05,
      "loss": 0.0257,
      "relation_loss": 0.0018119669402949512,
      "selected_pairs": 26946.3,
      "state_query_norm": 1116.715966796875,
      "state_relation_loss": 0.00034361302095931023,
      "step": 17290,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0018119669402949512,
      "weighted_relation_loss": 0.0018119669402949512
    },
    {
      "action_query_norm": 628.0065673828125,
      "action_relation_loss": 0.001465037907473743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.18334892422825,
      "flow_matching_loss": 0.02591324057430029,
      "grad_norm": 0.10445056855678558,
      "learning_rate": 4.1504008451110845e-05,
      "loss": 0.0277,
      "relation_loss": 0.001818221027497202,
      "selected_pairs": 26807.3,
      "state_query_norm": 1119.9040405273438,
      "state_relation_loss": 0.0003531831142026931,
      "step": 17300,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001818221027497202,
      "weighted_relation_loss": 0.001818221027497202
    },
    {
      "action_query_norm": 626.4406860351562,
      "action_relation_loss": 0.0014268389670178295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.19270346117867,
      "flow_matching_loss": 0.02532276790589094,
      "grad_norm": 0.09617273509502411,
      "learning_rate": 4.1449699461289225e-05,
      "loss": 0.0271,
      "relation_loss": 0.001778743544127792,
      "selected_pairs": 26669.4,
      "state_query_norm": 1131.9732055664062,
      "state_relation_loss": 0.0003519045712891966,
      "step": 17310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001778743544127792,
      "weighted_relation_loss": 0.001778743544127792
    },
    {
      "action_query_norm": 628.7758239746094,
      "action_relation_loss": 0.0014365065726451576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.202057998129092,
      "flow_matching_loss": 0.025992877781391144,
      "grad_norm": 0.10186158865690231,
      "learning_rate": 4.139540086088901e-05,
      "loss": 0.0278,
      "relation_loss": 0.0017934271367266775,
      "selected_pairs": 26893.1,
      "state_query_norm": 1117.051123046875,
      "state_relation_loss": 0.0003569205553503707,
      "step": 17320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0017934271367266775,
      "weighted_relation_loss": 0.0017934271367266775
    },
    {
      "action_query_norm": 630.4550537109375,
      "action_relation_loss": 0.001472942065447569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.211412535079514,
      "flow_matching_loss": 0.025761521607637405,
      "grad_norm": 0.11415793001651764,
      "learning_rate": 4.1341112715888106e-05,
      "loss": 0.0276,
      "relation_loss": 0.001820465107448399,
      "selected_pairs": 26734.6,
      "state_query_norm": 1135.291650390625,
      "state_relation_loss": 0.0003475230390904471,
      "step": 17330,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001820465107448399,
      "weighted_relation_loss": 0.001820465107448399
    },
    {
      "action_query_norm": 625.5845886230469,
      "action_relation_loss": 0.0014849157305434346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.220767072029936,
      "flow_matching_loss": 0.0232160285115242,
      "grad_norm": 0.0794898197054863,
      "learning_rate": 4.128683509225172e-05,
      "loss": 0.025,
      "relation_loss": 0.001833927514962852,
      "selected_pairs": 26870.3,
      "state_query_norm": 1125.4399536132812,
      "state_relation_loss": 0.00034901179606094956,
      "step": 17340,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001833927514962852,
      "weighted_relation_loss": 0.001833927514962852
    },
    {
      "action_query_norm": 633.3877502441406,
      "action_relation_loss": 0.0014465697575360536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.230121608980355,
      "flow_matching_loss": 0.025693484768271446,
      "grad_norm": 0.0919424518942833,
      "learning_rate": 4.123256805593231e-05,
      "loss": 0.0275,
      "relation_loss": 0.001792518387082964,
      "selected_pairs": 27598.8,
      "state_query_norm": 1133.9181274414063,
      "state_relation_loss": 0.0003459486237261444,
      "step": 17350,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.001792518387082964,
      "weighted_relation_loss": 0.001792518387082964
    },
    {
      "action_query_norm": 625.1755920410156,
      "action_relation_loss": 0.0014009283506311477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.239476145930777,
      "flow_matching_loss": 0.024029051512479783,
      "grad_norm": 0.08202370256185532,
      "learning_rate": 4.117831167286943e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017404368845745921,
      "selected_pairs": 26575.5,
      "state_query_norm": 1127.3122802734374,
      "state_relation_loss": 0.00033950853685382754,
      "step": 17360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0017404368845745921,
      "weighted_relation_loss": 0.0017404368845745921
    },
    {
      "action_query_norm": 629.8281005859375,
      "action_relation_loss": 0.0014661679160781205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.2488306828812,
      "flow_matching_loss": 0.025822816789150237,
      "grad_norm": 0.0919056236743927,
      "learning_rate": 4.1124066008989685e-05,
      "loss": 0.0276,
      "relation_loss": 0.0018185873050242662,
      "selected_pairs": 26571.2,
      "state_query_norm": 1122.4070190429688,
      "state_relation_loss": 0.00035241937439423054,
      "step": 17370,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0018185873050242662,
      "weighted_relation_loss": 0.0018185873050242662
    },
    {
      "action_query_norm": 628.3415344238281,
      "action_relation_loss": 0.0014301735442131757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.258185219831617,
      "flow_matching_loss": 0.025127665884792805,
      "grad_norm": 0.1114618256688118,
      "learning_rate": 4.106983113020669e-05,
      "loss": 0.0269,
      "relation_loss": 0.0017756173736415803,
      "selected_pairs": 27059.1,
      "state_query_norm": 1134.4179565429688,
      "state_relation_loss": 0.0003454438265180215,
      "step": 17380,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0017756173736415803,
      "weighted_relation_loss": 0.0017756173736415803
    },
    {
      "action_query_norm": 628.6689880371093,
      "action_relation_loss": 0.001365733821876347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.26753975678204,
      "flow_matching_loss": 0.024293013103306293,
      "grad_norm": 0.09034299850463867,
      "learning_rate": 4.101560710242094e-05,
      "loss": 0.026,
      "relation_loss": 0.0017076535150408745,
      "selected_pairs": 27393.7,
      "state_query_norm": 1129.4306884765624,
      "state_relation_loss": 0.0003419196931645274,
      "step": 17390,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0017076535150408745,
      "weighted_relation_loss": 0.0017076535150408745
    },
    {
      "action_query_norm": 625.0274047851562,
      "action_relation_loss": 0.0014196483767591416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.27689429373246,
      "flow_matching_loss": 0.02624839637428522,
      "grad_norm": 0.11785527318716049,
      "learning_rate": 4.096139399151971e-05,
      "loss": 0.028,
      "relation_loss": 0.001760771474801004,
      "selected_pairs": 26617.4,
      "state_query_norm": 1147.9718627929688,
      "state_relation_loss": 0.00034112310386262833,
      "step": 17400,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001760771474801004,
      "weighted_relation_loss": 0.001760771474801004
    },
    {
      "action_query_norm": 626.6857055664062,
      "action_relation_loss": 0.0014622599934227764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.28624883068288,
      "flow_matching_loss": 0.024015611782670022,
      "grad_norm": 0.10208452492952347,
      "learning_rate": 4.090719186337709e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017923471285030245,
      "selected_pairs": 26439.1,
      "state_query_norm": 1137.9687133789062,
      "state_relation_loss": 0.0003300871147075668,
      "step": 17410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0017923471285030245,
      "weighted_relation_loss": 0.0017923471285030245
    },
    {
      "action_query_norm": 631.6742980957031,
      "action_relation_loss": 0.0014139490318484603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.2956033676333,
      "flow_matching_loss": 0.023802801221609115,
      "grad_norm": 0.0870119109749794,
      "learning_rate": 4.0853000783853746e-05,
      "loss": 0.0256,
      "relation_loss": 0.0017688071937300264,
      "selected_pairs": 27312.6,
      "state_query_norm": 1129.9989624023438,
      "state_relation_loss": 0.00035485816479194907,
      "step": 17420,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0017688071937300264,
      "weighted_relation_loss": 0.0017688071937300264
    },
    {
      "action_query_norm": 631.986865234375,
      "action_relation_loss": 0.0014160693739540876,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.304957904583723,
      "flow_matching_loss": 0.02404635064303875,
      "grad_norm": 0.1534673124551773,
      "learning_rate": 4.0798820818796956e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017587202368304134,
      "selected_pairs": 26836.7,
      "state_query_norm": 1142.9654541015625,
      "state_relation_loss": 0.0003426508686970919,
      "step": 17430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0017587202368304134,
      "weighted_relation_loss": 0.0017587202368304134
    },
    {
      "action_query_norm": 629.2829345703125,
      "action_relation_loss": 0.0014266188256442547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.314312441534145,
      "flow_matching_loss": 0.0241565078496933,
      "grad_norm": 0.10618575662374496,
      "learning_rate": 4.074465203404048e-05,
      "loss": 0.0259,
      "relation_loss": 0.0017673959606327116,
      "selected_pairs": 26873.4,
      "state_query_norm": 1133.53603515625,
      "state_relation_loss": 0.0003407771233469248,
      "step": 17440,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0017673959606327116,
      "weighted_relation_loss": 0.0017673959606327116
    },
    {
      "action_query_norm": 636.5069702148437,
      "action_relation_loss": 0.0013905040104873478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.323666978484564,
      "flow_matching_loss": 0.0255298076197505,
      "grad_norm": 0.09197545796632767,
      "learning_rate": 4.0690494495404495e-05,
      "loss": 0.0273,
      "relation_loss": 0.001752125658094883,
      "selected_pairs": 27151.7,
      "state_query_norm": 1141.6140625,
      "state_relation_loss": 0.0003616216592490673,
      "step": 17450,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001752125658094883,
      "weighted_relation_loss": 0.001752125658094883
    },
    {
      "action_query_norm": 632.8671325683594,
      "action_relation_loss": 0.0014854963636025786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.333021515434986,
      "flow_matching_loss": 0.023376107402145862,
      "grad_norm": 0.10372950881719589,
      "learning_rate": 4.0636348268695535e-05,
      "loss": 0.0252,
      "relation_loss": 0.0018276911228895187,
      "selected_pairs": 27220.3,
      "state_query_norm": 1123.6967163085938,
      "state_relation_loss": 0.0003421947330934927,
      "step": 17460,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0018276911228895187,
      "weighted_relation_loss": 0.0018276911228895187
    },
    {
      "action_query_norm": 627.723095703125,
      "action_relation_loss": 0.0013594830874353648,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.342376052385408,
      "flow_matching_loss": 0.025170144625008106,
      "grad_norm": 0.1095455065369606,
      "learning_rate": 4.058221341970637e-05,
      "loss": 0.0269,
      "relation_loss": 0.0016966637107543648,
      "selected_pairs": 27046.9,
      "state_query_norm": 1140.58505859375,
      "state_relation_loss": 0.0003371806349605322,
      "step": 17470,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0016966637107543648,
      "weighted_relation_loss": 0.0016966637107543648
    },
    {
      "action_query_norm": 629.4708801269531,
      "action_relation_loss": 0.0014055301202461124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.351730589335826,
      "flow_matching_loss": 0.02570490222424269,
      "grad_norm": 0.09495473653078079,
      "learning_rate": 4.0528090014215945e-05,
      "loss": 0.0275,
      "relation_loss": 0.0017489620833657682,
      "selected_pairs": 26392.7,
      "state_query_norm": 1139.4818115234375,
      "state_relation_loss": 0.00034343194856774063,
      "step": 17480,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0017489620833657682,
      "weighted_relation_loss": 0.0017489620833657682
    },
    {
      "action_query_norm": 630.9863525390625,
      "action_relation_loss": 0.0014066944597288967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.36108512628625,
      "flow_matching_loss": 0.02436126209795475,
      "grad_norm": 0.11314830183982849,
      "learning_rate": 4.047397811798929e-05,
      "loss": 0.0261,
      "relation_loss": 0.0017611637245863675,
      "selected_pairs": 27994.1,
      "state_query_norm": 1142.3618041992188,
      "state_relation_loss": 0.00035446927649900316,
      "step": 17490,
      "valid_action_samples": 237.1,
      "weighted_auxiliary_loss": 0.0017611637245863675,
      "weighted_relation_loss": 0.0017611637245863675
    },
    {
      "action_query_norm": 629.8797241210938,
      "action_relation_loss": 0.0014543715748004614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.37043966323667,
      "flow_matching_loss": 0.024250103905797005,
      "grad_norm": 0.08857724815607071,
      "learning_rate": 4.041987779677745e-05,
      "loss": 0.0261,
      "relation_loss": 0.0018085983349010349,
      "selected_pairs": 27208.8,
      "state_query_norm": 1121.3976928710938,
      "state_relation_loss": 0.00035422674845904113,
      "step": 17500,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0018085983349010349,
      "weighted_relation_loss": 0.0018085983349010349
    },
    {
      "action_query_norm": 627.1235534667969,
      "action_relation_loss": 0.001415088609792292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.379794200187092,
      "flow_matching_loss": 0.025656377710402013,
      "grad_norm": 0.10012554377317429,
      "learning_rate": 4.036578911631745e-05,
      "loss": 0.0274,
      "relation_loss": 0.0017642273451201618,
      "selected_pairs": 27363.2,
      "state_query_norm": 1121.860302734375,
      "state_relation_loss": 0.00034913873532786964,
      "step": 17510,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0017642273451201618,
      "weighted_relation_loss": 0.0017642273451201618
    },
    {
      "action_query_norm": 631.4826232910157,
      "action_relation_loss": 0.0014043655362911522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.38914873713751,
      "flow_matching_loss": 0.023865377716720104,
      "grad_norm": 0.0949656069278717,
      "learning_rate": 4.031171214233211e-05,
      "loss": 0.0256,
      "relation_loss": 0.0017346261534839868,
      "selected_pairs": 26322.3,
      "state_query_norm": 1138.9481201171875,
      "state_relation_loss": 0.00033026062010321766,
      "step": 17520,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0017346261534839868,
      "weighted_relation_loss": 0.0017346261534839868
    },
    {
      "action_query_norm": 630.8325744628906,
      "action_relation_loss": 0.001431349467020482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.398503274087933,
      "flow_matching_loss": 0.022628765553236008,
      "grad_norm": 0.08624167740345001,
      "learning_rate": 4.025764694053008e-05,
      "loss": 0.0244,
      "relation_loss": 0.0017787953140214086,
      "selected_pairs": 27383.6,
      "state_query_norm": 1126.12822265625,
      "state_relation_loss": 0.00034744584700092674,
      "step": 17530,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0017787953140214086,
      "weighted_relation_loss": 0.0017787953140214086
    },
    {
      "action_query_norm": 631.9399658203125,
      "action_relation_loss": 0.0013753070845268667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.407857811038355,
      "flow_matching_loss": 0.023178032226860525,
      "grad_norm": 0.11626897752285004,
      "learning_rate": 4.020359357660566e-05,
      "loss": 0.0249,
      "relation_loss": 0.0017240802408196032,
      "selected_pairs": 26552.7,
      "state_query_norm": 1122.5228881835938,
      "state_relation_loss": 0.0003487731533823535,
      "step": 17540,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0017240802408196032,
      "weighted_relation_loss": 0.0017240802408196032
    },
    {
      "action_query_norm": 631.5181091308593,
      "action_relation_loss": 0.0014054623548872768,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.417212347988773,
      "flow_matching_loss": 0.024194907397031784,
      "grad_norm": 0.12173433601856232,
      "learning_rate": 4.014955211623875e-05,
      "loss": 0.0259,
      "relation_loss": 0.001746524847112596,
      "selected_pairs": 27151.7,
      "state_query_norm": 1124.6806640625,
      "state_relation_loss": 0.0003410624980460852,
      "step": 17550,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001746524847112596,
      "weighted_relation_loss": 0.001746524847112596
    },
    {
      "action_query_norm": 637.9412902832031,
      "action_relation_loss": 0.001385501795448363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.426566884939195,
      "flow_matching_loss": 0.023309722542762756,
      "grad_norm": 0.10085251182317734,
      "learning_rate": 4.0095522625094895e-05,
      "loss": 0.025,
      "relation_loss": 0.001717034913599491,
      "selected_pairs": 26967.4,
      "state_query_norm": 1127.453515625,
      "state_relation_loss": 0.0003315331239718944,
      "step": 17560,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001717034913599491,
      "weighted_relation_loss": 0.001717034913599491
    },
    {
      "action_query_norm": 634.6944213867188,
      "action_relation_loss": 0.001432291278615594,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.435921421889617,
      "flow_matching_loss": 0.024493459984660147,
      "grad_norm": 0.10496089607477188,
      "learning_rate": 4.004150516882497e-05,
      "loss": 0.0263,
      "relation_loss": 0.0017821540939621627,
      "selected_pairs": 26732.0,
      "state_query_norm": 1122.752099609375,
      "state_relation_loss": 0.0003498628298984841,
      "step": 17570,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0017821540939621627,
      "weighted_relation_loss": 0.0017821540939621627
    },
    {
      "action_query_norm": 631.4994812011719,
      "action_relation_loss": 0.001379613415338099,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.44527595884004,
      "flow_matching_loss": 0.025255551002919673,
      "grad_norm": 0.09882257133722305,
      "learning_rate": 3.99874998130653e-05,
      "loss": 0.027,
      "relation_loss": 0.0017168115242384373,
      "selected_pairs": 26926.2,
      "state_query_norm": 1124.3557739257812,
      "state_relation_loss": 0.00033719813218340277,
      "step": 17580,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0017168115242384373,
      "weighted_relation_loss": 0.0017168115242384373
    },
    {
      "action_query_norm": 634.0878662109375,
      "action_relation_loss": 0.0013716797810047864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.454630495790457,
      "flow_matching_loss": 0.024982335045933723,
      "grad_norm": 0.13351452350616455,
      "learning_rate": 3.9933506623437455e-05,
      "loss": 0.0267,
      "relation_loss": 0.001715329138096422,
      "selected_pairs": 27055.6,
      "state_query_norm": 1131.1023803710937,
      "state_relation_loss": 0.00034364935709163547,
      "step": 17590,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001715329138096422,
      "weighted_relation_loss": 0.001715329138096422
    },
    {
      "action_query_norm": 631.5165222167968,
      "action_relation_loss": 0.0013942432589828968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.46398503274088,
      "flow_matching_loss": 0.02413430605083704,
      "grad_norm": 0.09957394003868103,
      "learning_rate": 3.9879525665548276e-05,
      "loss": 0.0259,
      "relation_loss": 0.0017417774302884935,
      "selected_pairs": 26026.3,
      "state_query_norm": 1131.176025390625,
      "state_relation_loss": 0.0003475341567536816,
      "step": 17600,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.0017417774302884935,
      "weighted_relation_loss": 0.0017417774302884935
    },
    {
      "action_query_norm": 635.1592834472656,
      "action_relation_loss": 0.0013769960263743997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.4733395696913,
      "flow_matching_loss": 0.02376259006559849,
      "grad_norm": 0.09330791980028152,
      "learning_rate": 3.9825557004989704e-05,
      "loss": 0.0255,
      "relation_loss": 0.0017261144239455462,
      "selected_pairs": 27162.2,
      "state_query_norm": 1133.262841796875,
      "state_relation_loss": 0.00034911840048152954,
      "step": 17610,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0017261144239455462,
      "weighted_relation_loss": 0.0017261144239455462
    },
    {
      "action_query_norm": 641.7688415527343,
      "action_relation_loss": 0.0013733540661633014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.48269410664172,
      "flow_matching_loss": 0.02378924246877432,
      "grad_norm": 0.07328592240810394,
      "learning_rate": 3.977160070733877e-05,
      "loss": 0.0255,
      "relation_loss": 0.0017095563001930715,
      "selected_pairs": 26940.3,
      "state_query_norm": 1144.8695068359375,
      "state_relation_loss": 0.00033620222529862076,
      "step": 17620,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017095563001930715,
      "weighted_relation_loss": 0.0017095563001930715
    },
    {
      "action_query_norm": 642.9761047363281,
      "action_relation_loss": 0.001290616241749376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.492048643592142,
      "flow_matching_loss": 0.02661747857928276,
      "grad_norm": 0.08578212559223175,
      "learning_rate": 3.971765683815746e-05,
      "loss": 0.0283,
      "relation_loss": 0.0016387037467211485,
      "selected_pairs": 27475.4,
      "state_query_norm": 1133.1842407226563,
      "state_relation_loss": 0.0003480875166133046,
      "step": 17630,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0016387037467211485,
      "weighted_relation_loss": 0.0016387037467211485
    },
    {
      "action_query_norm": 630.2209045410157,
      "action_relation_loss": 0.0014678919338621198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.501403180542564,
      "flow_matching_loss": 0.024003583006560802,
      "grad_norm": 0.09917329996824265,
      "learning_rate": 3.966372546299265e-05,
      "loss": 0.0258,
      "relation_loss": 0.0018092213431373239,
      "selected_pairs": 26815.7,
      "state_query_norm": 1136.1727416992187,
      "state_relation_loss": 0.0003413293947232887,
      "step": 17640,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0018092213431373239,
      "weighted_relation_loss": 0.0018092213431373239
    },
    {
      "action_query_norm": 625.1163513183594,
      "action_relation_loss": 0.0014119972940534354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.510757717492986,
      "flow_matching_loss": 0.023401861637830736,
      "grad_norm": 0.08943401277065277,
      "learning_rate": 3.960980664737604e-05,
      "loss": 0.0251,
      "relation_loss": 0.0017388844513334333,
      "selected_pairs": 26658.6,
      "state_query_norm": 1133.2038940429688,
      "state_relation_loss": 0.00032688714854884894,
      "step": 17650,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0017388844513334333,
      "weighted_relation_loss": 0.0017388844513334333
    },
    {
      "action_query_norm": 638.3549865722656,
      "action_relation_loss": 0.0013942192192189396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.520112254443404,
      "flow_matching_loss": 0.02424431387335062,
      "grad_norm": 0.12531214952468872,
      "learning_rate": 3.955590045682408e-05,
      "loss": 0.026,
      "relation_loss": 0.0017344186548143626,
      "selected_pairs": 26832.6,
      "state_query_norm": 1137.2062744140626,
      "state_relation_loss": 0.0003401994239538908,
      "step": 17660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0017344186548143626,
      "weighted_relation_loss": 0.0017344186548143626
    },
    {
      "action_query_norm": 637.9476745605468,
      "action_relation_loss": 0.0013799473992548883,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.529466791393826,
      "flow_matching_loss": 0.0234805541113019,
      "grad_norm": 0.10403989255428314,
      "learning_rate": 3.950200695683788e-05,
      "loss": 0.0252,
      "relation_loss": 0.001714442193042487,
      "selected_pairs": 27194.1,
      "state_query_norm": 1111.35205078125,
      "state_relation_loss": 0.00033449480251874776,
      "step": 17670,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001714442193042487,
      "weighted_relation_loss": 0.001714442193042487
    },
    {
      "action_query_norm": 627.9995239257812,
      "action_relation_loss": 0.001320689870044589,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.53882132834425,
      "flow_matching_loss": 0.023381470143795012,
      "grad_norm": 0.08942461013793945,
      "learning_rate": 3.944812621290314e-05,
      "loss": 0.025,
      "relation_loss": 0.0016630062367767096,
      "selected_pairs": 27108.0,
      "state_query_norm": 1116.8515991210938,
      "state_relation_loss": 0.00034231636382173745,
      "step": 17680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0016630062367767096,
      "weighted_relation_loss": 0.0016630062367767096
    },
    {
      "action_query_norm": 628.8501159667969,
      "action_relation_loss": 0.0013948126812465488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.548175865294667,
      "flow_matching_loss": 0.024704702757298948,
      "grad_norm": 0.11103122681379318,
      "learning_rate": 3.939425829049002e-05,
      "loss": 0.0264,
      "relation_loss": 0.0017369788023643196,
      "selected_pairs": 27337.5,
      "state_query_norm": 1124.1025024414062,
      "state_relation_loss": 0.0003421661123866215,
      "step": 17690,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0017369788023643196,
      "weighted_relation_loss": 0.0017369788023643196
    },
    {
      "action_query_norm": 631.8985656738281,
      "action_relation_loss": 0.001449936267454177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.55753040224509,
      "flow_matching_loss": 0.025335188768804074,
      "grad_norm": 0.08941829204559326,
      "learning_rate": 3.9340403255053135e-05,
      "loss": 0.0271,
      "relation_loss": 0.0017869172152131796,
      "selected_pairs": 26460.6,
      "state_query_norm": 1136.0185302734376,
      "state_relation_loss": 0.0003369809477590024,
      "step": 17700,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0017869172152131796,
      "weighted_relation_loss": 0.0017869172152131796
    },
    {
      "action_query_norm": 635.7083435058594,
      "action_relation_loss": 0.0014728520531207324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.56688493919551,
      "flow_matching_loss": 0.0224959596991539,
      "grad_norm": 0.09231492877006531,
      "learning_rate": 3.9286561172031407e-05,
      "loss": 0.0243,
      "relation_loss": 0.001812659320421517,
      "selected_pairs": 26971.5,
      "state_query_norm": 1111.8738037109374,
      "state_relation_loss": 0.00033980729058384894,
      "step": 17710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001812659320421517,
      "weighted_relation_loss": 0.001812659320421517
    },
    {
      "action_query_norm": 622.4057495117188,
      "action_relation_loss": 0.001386623806320131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.57623947614593,
      "flow_matching_loss": 0.02402661833912134,
      "grad_norm": 0.09965422749519348,
      "learning_rate": 3.923273210684809e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017286509624682366,
      "selected_pairs": 26762.8,
      "state_query_norm": 1122.5870483398437,
      "state_relation_loss": 0.000342027164879255,
      "step": 17720,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0017286509624682366,
      "weighted_relation_loss": 0.0017286509624682366
    },
    {
      "action_query_norm": 624.3699951171875,
      "action_relation_loss": 0.00143054110230878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.58559401309635,
      "flow_matching_loss": 0.024599607288837432,
      "grad_norm": 0.11326348781585693,
      "learning_rate": 3.9178916124910555e-05,
      "loss": 0.0264,
      "relation_loss": 0.0017672194284386934,
      "selected_pairs": 26547.8,
      "state_query_norm": 1149.6079711914062,
      "state_relation_loss": 0.0003366783232195303,
      "step": 17730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0017672194284386934,
      "weighted_relation_loss": 0.0017672194284386934
    },
    {
      "action_query_norm": 630.2910827636719,
      "action_relation_loss": 0.0013231260469183325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.594948550046773,
      "flow_matching_loss": 0.02331945560872555,
      "grad_norm": 0.09529493004083633,
      "learning_rate": 3.9125113291610274e-05,
      "loss": 0.025,
      "relation_loss": 0.0016558436444029212,
      "selected_pairs": 27244.7,
      "state_query_norm": 1135.5809692382813,
      "state_relation_loss": 0.0003327176149468869,
      "step": 17740,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0016558436444029212,
      "weighted_relation_loss": 0.0016558436444029212
    },
    {
      "action_query_norm": 628.0523620605469,
      "action_relation_loss": 0.0014521989389322699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.604303086997195,
      "flow_matching_loss": 0.024582487158477308,
      "grad_norm": 0.08804705739021301,
      "learning_rate": 3.907132367232278e-05,
      "loss": 0.0264,
      "relation_loss": 0.0017975121620111168,
      "selected_pairs": 26736.8,
      "state_query_norm": 1140.6007568359375,
      "state_relation_loss": 0.00034531323763076216,
      "step": 17750,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0017975121620111168,
      "weighted_relation_loss": 0.0017975121620111168
    },
    {
      "action_query_norm": 622.3732238769531,
      "action_relation_loss": 0.0014323581825010478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.613657623947613,
      "flow_matching_loss": 0.024578607641160488,
      "grad_norm": 0.10274101793766022,
      "learning_rate": 3.901754733240752e-05,
      "loss": 0.0264,
      "relation_loss": 0.001780480076558888,
      "selected_pairs": 26618.2,
      "state_query_norm": 1120.644189453125,
      "state_relation_loss": 0.0003481219173409045,
      "step": 17760,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001780480076558888,
      "weighted_relation_loss": 0.001780480076558888
    },
    {
      "action_query_norm": 626.23505859375,
      "action_relation_loss": 0.0013883326668292285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.623012160898035,
      "flow_matching_loss": 0.023160865530371666,
      "grad_norm": 0.0882866308093071,
      "learning_rate": 3.896378433720785e-05,
      "loss": 0.0249,
      "relation_loss": 0.0017243764363229276,
      "selected_pairs": 26640.1,
      "state_query_norm": 1112.4766967773437,
      "state_relation_loss": 0.000336043763672933,
      "step": 17770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0017243764363229276,
      "weighted_relation_loss": 0.0017243764363229276
    },
    {
      "action_query_norm": 622.7406494140625,
      "action_relation_loss": 0.001423204888124019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.632366697848457,
      "flow_matching_loss": 0.024018500931560993,
      "grad_norm": 0.1019035205245018,
      "learning_rate": 3.891003475205086e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017602089792490005,
      "selected_pairs": 27084.0,
      "state_query_norm": 1138.593603515625,
      "state_relation_loss": 0.0003370040736626834,
      "step": 17780,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0017602089792490005,
      "weighted_relation_loss": 0.0017602089792490005
    },
    {
      "action_query_norm": 625.6791076660156,
      "action_relation_loss": 0.0014043089700862764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.641721234798876,
      "flow_matching_loss": 0.025306364707648753,
      "grad_norm": 0.1135186031460762,
      "learning_rate": 3.885629864224736e-05,
      "loss": 0.0271,
      "relation_loss": 0.0017440598574467003,
      "selected_pairs": 26903.4,
      "state_query_norm": 1136.0224975585938,
      "state_relation_loss": 0.0003397508757188916,
      "step": 17790,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0017440598574467003,
      "weighted_relation_loss": 0.0017440598574467003
    },
    {
      "action_query_norm": 632.3656005859375,
      "action_relation_loss": 0.0014063861686736344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.651075771749298,
      "flow_matching_loss": 0.02409601379185915,
      "grad_norm": 0.12262585014104843,
      "learning_rate": 3.880257607309178e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017487897304818035,
      "selected_pairs": 26876.4,
      "state_query_norm": 1119.4163696289063,
      "state_relation_loss": 0.0003424035472562537,
      "step": 17800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0017487897304818035,
      "weighted_relation_loss": 0.0017487897304818035
    },
    {
      "action_query_norm": 625.2025817871094,
      "action_relation_loss": 0.0014260744908824563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.66043030869972,
      "flow_matching_loss": 0.02416005786508322,
      "grad_norm": 0.09318706393241882,
      "learning_rate": 3.874886710986213e-05,
      "loss": 0.0259,
      "relation_loss": 0.0017779360758140683,
      "selected_pairs": 27157.5,
      "state_query_norm": 1133.8794677734375,
      "state_relation_loss": 0.00035186157911084595,
      "step": 17810,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0017779360758140683,
      "weighted_relation_loss": 0.0017779360758140683
    },
    {
      "action_query_norm": 628.8109191894531,
      "action_relation_loss": 0.0013989522005431354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.669784845650142,
      "flow_matching_loss": 0.024018285796046257,
      "grad_norm": 0.11261710524559021,
      "learning_rate": 3.8695171817819834e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017476157401688397,
      "selected_pairs": 26962.6,
      "state_query_norm": 1115.8667724609375,
      "state_relation_loss": 0.000348663522163406,
      "step": 17820,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0017476157401688397,
      "weighted_relation_loss": 0.0017476157401688397
    },
    {
      "action_query_norm": 623.2121948242187,
      "action_relation_loss": 0.0014059275854378939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.67913938260056,
      "flow_matching_loss": 0.023314406536519526,
      "grad_norm": 0.09789609909057617,
      "learning_rate": 3.8641490262209774e-05,
      "loss": 0.0251,
      "relation_loss": 0.0017447848920710384,
      "selected_pairs": 26708.2,
      "state_query_norm": 1123.710595703125,
      "state_relation_loss": 0.0003388573037227616,
      "step": 17830,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0017447848920710384,
      "weighted_relation_loss": 0.0017447848920710384
    },
    {
      "action_query_norm": 626.0501770019531,
      "action_relation_loss": 0.0013382513774558903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.688493919550982,
      "flow_matching_loss": 0.02262104246765375,
      "grad_norm": 0.11057166755199432,
      "learning_rate": 3.8587822508260084e-05,
      "loss": 0.0243,
      "relation_loss": 0.0016766512184403838,
      "selected_pairs": 26896.2,
      "state_query_norm": 1128.3775634765625,
      "state_relation_loss": 0.00033839984680525956,
      "step": 17840,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0016766512184403838,
      "weighted_relation_loss": 0.0016766512184403838
    },
    {
      "action_query_norm": 625.5229370117188,
      "action_relation_loss": 0.0014444043044932187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.697848456501404,
      "flow_matching_loss": 0.02363350447267294,
      "grad_norm": 0.13117556273937225,
      "learning_rate": 3.853416862118214e-05,
      "loss": 0.0254,
      "relation_loss": 0.0017809685552492738,
      "selected_pairs": 26569.2,
      "state_query_norm": 1134.1491455078126,
      "state_relation_loss": 0.00033656425366643814,
      "step": 17850,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0017809685552492738,
      "weighted_relation_loss": 0.0017809685552492738
    },
    {
      "action_query_norm": 626.4808898925781,
      "action_relation_loss": 0.001384394569322467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.707202993451823,
      "flow_matching_loss": 0.024205374531447887,
      "grad_norm": 0.12528696656227112,
      "learning_rate": 3.848052866617049e-05,
      "loss": 0.0259,
      "relation_loss": 0.0017225703806616366,
      "selected_pairs": 26688.0,
      "state_query_norm": 1133.5839111328125,
      "state_relation_loss": 0.00033817579969763756,
      "step": 17860,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0017225703806616366,
      "weighted_relation_loss": 0.0017225703806616366
    },
    {
      "action_query_norm": 633.6660827636719,
      "action_relation_loss": 0.0013680904405191541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.716557530402245,
      "flow_matching_loss": 0.02571813389658928,
      "grad_norm": 0.10819315910339355,
      "learning_rate": 3.84269027084027e-05,
      "loss": 0.0274,
      "relation_loss": 0.0017115275491960346,
      "selected_pairs": 27124.5,
      "state_query_norm": 1132.9638305664062,
      "state_relation_loss": 0.0003434371086768806,
      "step": 17870,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0017115275491960346,
      "weighted_relation_loss": 0.0017115275491960346
    },
    {
      "action_query_norm": 630.21982421875,
      "action_relation_loss": 0.0014922643546015023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.725912067352667,
      "flow_matching_loss": 0.022778710909187795,
      "grad_norm": 0.10457289963960648,
      "learning_rate": 3.83732908130394e-05,
      "loss": 0.0246,
      "relation_loss": 0.0018247179104946554,
      "selected_pairs": 26163.1,
      "state_query_norm": 1135.6621215820312,
      "state_relation_loss": 0.0003324535675346851,
      "step": 17880,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0018247179104946554,
      "weighted_relation_loss": 0.0018247179104946554
    },
    {
      "action_query_norm": 636.4406188964844,
      "action_relation_loss": 0.0013851250521838665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.73526660430309,
      "flow_matching_loss": 0.02510254308581352,
      "grad_norm": 0.10970592498779297,
      "learning_rate": 3.831969304522409e-05,
      "loss": 0.0268,
      "relation_loss": 0.001741020381450653,
      "selected_pairs": 27135.7,
      "state_query_norm": 1131.3619140625,
      "state_relation_loss": 0.0003558953409083188,
      "step": 17890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.001741020381450653,
      "weighted_relation_loss": 0.001741020381450653
    },
    {
      "action_query_norm": 639.5964904785156,
      "action_relation_loss": 0.0013846558285877109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.744621141253507,
      "flow_matching_loss": 0.022400877624750137,
      "grad_norm": 0.07450105249881744,
      "learning_rate": 3.826610947008312e-05,
      "loss": 0.0241,
      "relation_loss": 0.0017329945461824537,
      "selected_pairs": 26800.8,
      "state_query_norm": 1128.5311279296875,
      "state_relation_loss": 0.0003483387117739767,
      "step": 17900,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0017329945461824537,
      "weighted_relation_loss": 0.0017329945461824537
    },
    {
      "action_query_norm": 625.0863647460938,
      "action_relation_loss": 0.0014248988940380515,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.75397567820393,
      "flow_matching_loss": 0.023634997941553593,
      "grad_norm": 0.09457673877477646,
      "learning_rate": 3.821254015272559e-05,
      "loss": 0.0254,
      "relation_loss": 0.0017726084915921092,
      "selected_pairs": 27391.1,
      "state_query_norm": 1131.4465087890626,
      "state_relation_loss": 0.00034770958300214263,
      "step": 17910,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0017726084915921092,
      "weighted_relation_loss": 0.0017726084915921092
    },
    {
      "action_query_norm": 634.3853332519532,
      "action_relation_loss": 0.00136662432923913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.76333021515435,
      "flow_matching_loss": 0.021509294770658015,
      "grad_norm": 0.09210275858640671,
      "learning_rate": 3.8158985158243214e-05,
      "loss": 0.0232,
      "relation_loss": 0.001720433379523456,
      "selected_pairs": 27115.2,
      "state_query_norm": 1126.1149169921875,
      "state_relation_loss": 0.00035380904446355996,
      "step": 17920,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001720433379523456,
      "weighted_relation_loss": 0.001720433379523456
    },
    {
      "action_query_norm": 629.8199890136718,
      "action_relation_loss": 0.0014105798327364028,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.77268475210477,
      "flow_matching_loss": 0.023594943061470985,
      "grad_norm": 0.11780897527933121,
      "learning_rate": 3.810544455171044e-05,
      "loss": 0.0253,
      "relation_loss": 0.001744676660746336,
      "selected_pairs": 26901.2,
      "state_query_norm": 1126.9571166992187,
      "state_relation_loss": 0.0003340968367410824,
      "step": 17930,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001744676660746336,
      "weighted_relation_loss": 0.001744676660746336
    },
    {
      "action_query_norm": 628.3077453613281,
      "action_relation_loss": 0.00139129301533103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.78203928905519,
      "flow_matching_loss": 0.024071928672492506,
      "grad_norm": 0.08893746137619019,
      "learning_rate": 3.8051918398184114e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017202119808644057,
      "selected_pairs": 26666.4,
      "state_query_norm": 1131.4297729492187,
      "state_relation_loss": 0.0003289189771749079,
      "step": 17940,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0017202119808644057,
      "weighted_relation_loss": 0.0017202119808644057
    },
    {
      "action_query_norm": 622.1677368164062,
      "action_relation_loss": 0.0013594897813163699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.791393826005613,
      "flow_matching_loss": 0.023375419527292253,
      "grad_norm": 0.11940299719572067,
      "learning_rate": 3.799840676270356e-05,
      "loss": 0.0251,
      "relation_loss": 0.0016874741879291833,
      "selected_pairs": 26915.6,
      "state_query_norm": 1125.4969116210937,
      "state_relation_loss": 0.0003279844066128135,
      "step": 17950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0016874741879291833,
      "weighted_relation_loss": 0.0016874741879291833
    },
    {
      "action_query_norm": 626.3971130371094,
      "action_relation_loss": 0.0013704653596505524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.800748362956035,
      "flow_matching_loss": 0.02283230796456337,
      "grad_norm": 0.08960767835378647,
      "learning_rate": 3.794490971029048e-05,
      "loss": 0.0245,
      "relation_loss": 0.0017102773068472742,
      "selected_pairs": 26718.6,
      "state_query_norm": 1124.9034301757813,
      "state_relation_loss": 0.00033981196174863725,
      "step": 17960,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0017102773068472742,
      "weighted_relation_loss": 0.0017102773068472742
    },
    {
      "action_query_norm": 627.197802734375,
      "action_relation_loss": 0.0013438549707643689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.810102899906454,
      "flow_matching_loss": 0.02232309300452471,
      "grad_norm": 0.0853586345911026,
      "learning_rate": 3.789142730594881e-05,
      "loss": 0.024,
      "relation_loss": 0.0016856894479133188,
      "selected_pairs": 27102.6,
      "state_query_norm": 1119.5538818359375,
      "state_relation_loss": 0.00034183448005933317,
      "step": 17970,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0016856894479133188,
      "weighted_relation_loss": 0.0016856894479133188
    },
    {
      "action_query_norm": 628.2895141601563,
      "action_relation_loss": 0.0014044255134649576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.819457436856876,
      "flow_matching_loss": 0.024031673558056355,
      "grad_norm": 0.10388486087322235,
      "learning_rate": 3.783795961466471e-05,
      "loss": 0.0258,
      "relation_loss": 0.0017392150126397609,
      "selected_pairs": 27081.8,
      "state_query_norm": 1137.0371826171875,
      "state_relation_loss": 0.0003347895020851865,
      "step": 17980,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0017392150126397609,
      "weighted_relation_loss": 0.0017392150126397609
    },
    {
      "action_query_norm": 626.6415588378907,
      "action_relation_loss": 0.001353438338264823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.828811973807298,
      "flow_matching_loss": 0.022827334888279437,
      "grad_norm": 0.11886973679065704,
      "learning_rate": 3.778450670140651e-05,
      "loss": 0.0245,
      "relation_loss": 0.0016849359148181976,
      "selected_pairs": 27055.1,
      "state_query_norm": 1129.5626953125,
      "state_relation_loss": 0.00033149761147797106,
      "step": 17990,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0016849359148181976,
      "weighted_relation_loss": 0.0016849359148181976
    },
    {
      "action_query_norm": 629.55791015625,
      "action_relation_loss": 0.0013771473546512424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.838166510757716,
      "flow_matching_loss": 0.021688953414559363,
      "grad_norm": 0.08431291580200195,
      "learning_rate": 3.773106863112451e-05,
      "loss": 0.0234,
      "relation_loss": 0.0017079157405532897,
      "selected_pairs": 26365.2,
      "state_query_norm": 1137.7351318359374,
      "state_relation_loss": 0.00033076838590204714,
      "step": 18000,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0017079157405532897,
      "weighted_relation_loss": 0.0017079157405532897
    },
    {
      "action_query_norm": 626.9974304199219,
      "action_relation_loss": 0.0014118678052909672,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.84752104770814,
      "flow_matching_loss": 0.025579416193068027,
      "grad_norm": 0.14465197920799255,
      "learning_rate": 3.7677645468751e-05,
      "loss": 0.0273,
      "relation_loss": 0.0017564782872796058,
      "selected_pairs": 27034.3,
      "state_query_norm": 1113.3779541015624,
      "state_relation_loss": 0.00034461049363017083,
      "step": 18010,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0017564782872796058,
      "weighted_relation_loss": 0.0017564782872796058
    },
    {
      "action_query_norm": 628.35009765625,
      "action_relation_loss": 0.0013741981005296112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.85687558465856,
      "flow_matching_loss": 0.025377647392451762,
      "grad_norm": 0.11261099576950073,
      "learning_rate": 3.762423727920018e-05,
      "loss": 0.0271,
      "relation_loss": 0.0017021121922880412,
      "selected_pairs": 26987.7,
      "state_query_norm": 1116.9201782226562,
      "state_relation_loss": 0.0003279140772065148,
      "step": 18020,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0017021121922880412,
      "weighted_relation_loss": 0.0017021121922880412
    },
    {
      "action_query_norm": 629.7114929199219,
      "action_relation_loss": 0.0013993268948979676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.86623012160898,
      "flow_matching_loss": 0.022943519800901414,
      "grad_norm": 0.1281164586544037,
      "learning_rate": 3.7570844127367995e-05,
      "loss": 0.0247,
      "relation_loss": 0.001739911874756217,
      "selected_pairs": 26320.8,
      "state_query_norm": 1129.8620849609374,
      "state_relation_loss": 0.0003405849944101647,
      "step": 18030,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.001739911874756217,
      "weighted_relation_loss": 0.001739911874756217
    },
    {
      "action_query_norm": 629.7219543457031,
      "action_relation_loss": 0.0013863898580893873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.8755846585594,
      "flow_matching_loss": 0.02502450980246067,
      "grad_norm": 0.09443635493516922,
      "learning_rate": 3.751746607813221e-05,
      "loss": 0.0267,
      "relation_loss": 0.0017179449670948089,
      "selected_pairs": 26644.7,
      "state_query_norm": 1147.7786254882812,
      "state_relation_loss": 0.0003315551031846553,
      "step": 18040,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0017179449670948089,
      "weighted_relation_loss": 0.0017179449670948089
    },
    {
      "action_query_norm": 622.8798645019531,
      "action_relation_loss": 0.0013903164188377559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.884939195509823,
      "flow_matching_loss": 0.023250282928347587,
      "grad_norm": 0.0896725282073021,
      "learning_rate": 3.746410319635217e-05,
      "loss": 0.025,
      "relation_loss": 0.0017170735634863376,
      "selected_pairs": 26136.2,
      "state_query_norm": 1124.0457885742187,
      "state_relation_loss": 0.0003267571475589648,
      "step": 18050,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0017170735634863376,
      "weighted_relation_loss": 0.0017170735634863376
    },
    {
      "action_query_norm": 629.1203735351562,
      "action_relation_loss": 0.00137466115411371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.894293732460245,
      "flow_matching_loss": 0.023173591122031212,
      "grad_norm": 0.10533475130796432,
      "learning_rate": 3.7410755546868806e-05,
      "loss": 0.0249,
      "relation_loss": 0.0017051262082532048,
      "selected_pairs": 26662.5,
      "state_query_norm": 1129.0686889648437,
      "state_relation_loss": 0.0003304650424979627,
      "step": 18060,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0017051262082532048,
      "weighted_relation_loss": 0.0017051262082532048
    },
    {
      "action_query_norm": 630.3329345703125,
      "action_relation_loss": 0.0013543116161599754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.903648269410663,
      "flow_matching_loss": 0.023542282171547413,
      "grad_norm": 0.08470480144023895,
      "learning_rate": 3.735742319450454e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016966210445389152,
      "selected_pairs": 26920.2,
      "state_query_norm": 1130.7475952148438,
      "state_relation_loss": 0.0003423094254685566,
      "step": 18070,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0016966210445389152,
      "weighted_relation_loss": 0.0016966210445389152
    },
    {
      "action_query_norm": 631.2840576171875,
      "action_relation_loss": 0.0014174291165545583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.913002806361085,
      "flow_matching_loss": 0.024227123707532883,
      "grad_norm": 0.0745534598827362,
      "learning_rate": 3.7304106204063184e-05,
      "loss": 0.026,
      "relation_loss": 0.0017479460686445237,
      "selected_pairs": 26907.8,
      "state_query_norm": 1122.5872924804687,
      "state_relation_loss": 0.0003305169346276671,
      "step": 18080,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0017479460686445237,
      "weighted_relation_loss": 0.0017479460686445237
    },
    {
      "action_query_norm": 629.0041198730469,
      "action_relation_loss": 0.001384131726808846,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.922357343311507,
      "flow_matching_loss": 0.024325160309672356,
      "grad_norm": 0.08466724306344986,
      "learning_rate": 3.725080464032996e-05,
      "loss": 0.0261,
      "relation_loss": 0.0017291860887780786,
      "selected_pairs": 27315.1,
      "state_query_norm": 1128.3627197265625,
      "state_relation_loss": 0.0003450543561484665,
      "step": 18090,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0017291860887780786,
      "weighted_relation_loss": 0.0017291860887780786
    },
    {
      "action_query_norm": 628.1912658691406,
      "action_relation_loss": 0.001422418770380318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.931711880261926,
      "flow_matching_loss": 0.02220387738198042,
      "grad_norm": 0.10576967149972916,
      "learning_rate": 3.719751856807125e-05,
      "loss": 0.024,
      "relation_loss": 0.00177151532843709,
      "selected_pairs": 26604.6,
      "state_query_norm": 1115.1642944335938,
      "state_relation_loss": 0.0003490965609671548,
      "step": 18100,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.00177151532843709,
      "weighted_relation_loss": 0.00177151532843709
    },
    {
      "action_query_norm": 633.2191772460938,
      "action_relation_loss": 0.001370234217029065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.941066417212348,
      "flow_matching_loss": 0.02354244329035282,
      "grad_norm": 0.08462537825107574,
      "learning_rate": 3.71442480520347e-05,
      "loss": 0.0253,
      "relation_loss": 0.001711963547859341,
      "selected_pairs": 26166.6,
      "state_query_norm": 1123.8909912109375,
      "state_relation_loss": 0.00034172931918874383,
      "step": 18110,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.001711963547859341,
      "weighted_relation_loss": 0.001711963547859341
    },
    {
      "action_query_norm": 634.8204956054688,
      "action_relation_loss": 0.0013712095795199275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.95042095416277,
      "flow_matching_loss": 0.024760013073682786,
      "grad_norm": 0.09082183986902237,
      "learning_rate": 3.709099315694897e-05,
      "loss": 0.0265,
      "relation_loss": 0.001714189408812672,
      "selected_pairs": 26676.7,
      "state_query_norm": 1122.94130859375,
      "state_relation_loss": 0.00034297983220312747,
      "step": 18120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001714189408812672,
      "weighted_relation_loss": 0.001714189408812672
    },
    {
      "action_query_norm": 633.5480407714844,
      "action_relation_loss": 0.0013634152011945844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.95977549111319,
      "flow_matching_loss": 0.02502078302204609,
      "grad_norm": 0.08630600571632385,
      "learning_rate": 3.703775394752378e-05,
      "loss": 0.0267,
      "relation_loss": 0.0016961989109404384,
      "selected_pairs": 26386.2,
      "state_query_norm": 1140.8622680664062,
      "state_relation_loss": 0.0003327837068354711,
      "step": 18130,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0016961989109404384,
      "weighted_relation_loss": 0.0016961989109404384
    },
    {
      "action_query_norm": 628.6149353027344,
      "action_relation_loss": 0.0014172601047903299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.96913002806361,
      "flow_matching_loss": 0.023004568554461,
      "grad_norm": 0.0940346047282219,
      "learning_rate": 3.698453048844983e-05,
      "loss": 0.0248,
      "relation_loss": 0.001746272190939635,
      "selected_pairs": 26548.9,
      "state_query_norm": 1130.7495239257812,
      "state_relation_loss": 0.000329012080328539,
      "step": 18140,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001746272190939635,
      "weighted_relation_loss": 0.001746272190939635
    },
    {
      "action_query_norm": 637.9476745605468,
      "action_relation_loss": 0.0013934778631664812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.978484565014032,
      "flow_matching_loss": 0.025489661283791064,
      "grad_norm": 0.10230462998151779,
      "learning_rate": 3.693132284439861e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017349460860714317,
      "selected_pairs": 26621.1,
      "state_query_norm": 1132.3054931640625,
      "state_relation_loss": 0.0003414682229049504,
      "step": 18150,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0017349460860714317,
      "weighted_relation_loss": 0.0017349460860714317
    },
    {
      "action_query_norm": 633.7242431640625,
      "action_relation_loss": 0.0014291957137174903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.987839101964454,
      "flow_matching_loss": 0.023868854716420175,
      "grad_norm": 0.09032857418060303,
      "learning_rate": 3.6878131080022414e-05,
      "loss": 0.0256,
      "relation_loss": 0.0017678044619970023,
      "selected_pairs": 27038.6,
      "state_query_norm": 1122.301123046875,
      "state_relation_loss": 0.0003386087453691289,
      "step": 18160,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0017678044619970023,
      "weighted_relation_loss": 0.0017678044619970023
    },
    {
      "action_query_norm": 631.8928833007812,
      "action_relation_loss": 0.0014094663434661924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.997193638914872,
      "flow_matching_loss": 0.024409699812531473,
      "grad_norm": 0.11709552258253098,
      "learning_rate": 3.682495525995429e-05,
      "loss": 0.0262,
      "relation_loss": 0.001741266273893416,
      "selected_pairs": 27161.7,
      "state_query_norm": 1129.5684204101562,
      "state_relation_loss": 0.0003317999216960743,
      "step": 18170,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001741266273893416,
      "weighted_relation_loss": 0.001741266273893416
    },
    {
      "action_query_norm": 631.0314697265625,
      "action_relation_loss": 0.0013954723137430847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.006548175865294,
      "flow_matching_loss": 0.024416941590607165,
      "grad_norm": 0.09799652546644211,
      "learning_rate": 3.6771795448807846e-05,
      "loss": 0.0261,
      "relation_loss": 0.0017296161269769073,
      "selected_pairs": 24369.6,
      "state_query_norm": 1146.9383422851563,
      "state_relation_loss": 0.0003341438277857378,
      "step": 18180,
      "valid_action_samples": 214.5,
      "weighted_auxiliary_loss": 0.0017296161269769073,
      "weighted_relation_loss": 0.0017296161269769073
    },
    {
      "action_query_norm": 637.095556640625,
      "action_relation_loss": 0.0012863336363807321,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.015902712815716,
      "flow_matching_loss": 0.021588660776615143,
      "grad_norm": 0.11200515925884247,
      "learning_rate": 3.671865171117724e-05,
      "loss": 0.0232,
      "relation_loss": 0.0016258984338492154,
      "selected_pairs": 27020.0,
      "state_query_norm": 1121.0024780273438,
      "state_relation_loss": 0.0003395647945581004,
      "step": 18190,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0016258984338492154,
      "weighted_relation_loss": 0.0016258984338492154
    },
    {
      "action_query_norm": 631.0573669433594,
      "action_relation_loss": 0.001336846500635147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.02525724976614,
      "flow_matching_loss": 0.02272625584155321,
      "grad_norm": 0.09476152807474136,
      "learning_rate": 3.666552411163718e-05,
      "loss": 0.0244,
      "relation_loss": 0.001661875715944916,
      "selected_pairs": 26274.7,
      "state_query_norm": 1136.9153076171874,
      "state_relation_loss": 0.00032502921239938585,
      "step": 18200,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.001661875715944916,
      "weighted_relation_loss": 0.001661875715944916
    },
    {
      "action_query_norm": 641.75791015625,
      "action_relation_loss": 0.0013133782544173301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.034611786716557,
      "flow_matching_loss": 0.022679419070482255,
      "grad_norm": 0.11623981595039368,
      "learning_rate": 3.661241271474269e-05,
      "loss": 0.0243,
      "relation_loss": 0.001642469735816121,
      "selected_pairs": 27469.4,
      "state_query_norm": 1142.430322265625,
      "state_relation_loss": 0.00032909148430917414,
      "step": 18210,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.001642469735816121,
      "weighted_relation_loss": 0.001642469735816121
    },
    {
      "action_query_norm": 635.6366638183594,
      "action_relation_loss": 0.001343139773234725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.04396632366698,
      "flow_matching_loss": 0.022540644370019437,
      "grad_norm": 0.10114288330078125,
      "learning_rate": 3.6559317585029116e-05,
      "loss": 0.0242,
      "relation_loss": 0.0016818259609863162,
      "selected_pairs": 27208.8,
      "state_query_norm": 1132.5962036132812,
      "state_relation_loss": 0.0003386862052138895,
      "step": 18220,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0016818259609863162,
      "weighted_relation_loss": 0.0016818259609863162
    },
    {
      "action_query_norm": 630.2440002441406,
      "action_relation_loss": 0.0013172341976314783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.0533208606174,
      "flow_matching_loss": 0.02240510266274214,
      "grad_norm": 0.10483624786138535,
      "learning_rate": 3.650623878701204e-05,
      "loss": 0.0241,
      "relation_loss": 0.001652999664656818,
      "selected_pairs": 27229.6,
      "state_query_norm": 1123.7994506835937,
      "state_relation_loss": 0.0003357654553838074,
      "step": 18230,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.001652999664656818,
      "weighted_relation_loss": 0.001652999664656818
    },
    {
      "action_query_norm": 633.7101135253906,
      "action_relation_loss": 0.0012875744374468922,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.06267539756782,
      "flow_matching_loss": 0.023538621701300146,
      "grad_norm": 0.0866447314620018,
      "learning_rate": 3.645317638518721e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016191159142181278,
      "selected_pairs": 26947.0,
      "state_query_norm": 1141.1135986328125,
      "state_relation_loss": 0.0003315414767712355,
      "step": 18240,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016191159142181278,
      "weighted_relation_loss": 0.0016191159142181278
    },
    {
      "action_query_norm": 627.1490417480469,
      "action_relation_loss": 0.0013700692215934396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.07202993451824,
      "flow_matching_loss": 0.023948822543025015,
      "grad_norm": 0.11477138847112656,
      "learning_rate": 3.640013044403046e-05,
      "loss": 0.0257,
      "relation_loss": 0.001702029095031321,
      "selected_pairs": 26506.6,
      "state_query_norm": 1117.248193359375,
      "state_relation_loss": 0.00033195987052749845,
      "step": 18250,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.001702029095031321,
      "weighted_relation_loss": 0.001702029095031321
    },
    {
      "action_query_norm": 633.0737243652344,
      "action_relation_loss": 0.0012948762043379248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.081384471468663,
      "flow_matching_loss": 0.023698837496340275,
      "grad_norm": 0.09244029223918915,
      "learning_rate": 3.634710102799761e-05,
      "loss": 0.0253,
      "relation_loss": 0.0016358992434106767,
      "selected_pairs": 26735.2,
      "state_query_norm": 1129.0169921875,
      "state_relation_loss": 0.00034102303034160285,
      "step": 18260,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0016358992434106767,
      "weighted_relation_loss": 0.0016358992434106767
    },
    {
      "action_query_norm": 636.5569274902343,
      "action_relation_loss": 0.001312056230381131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.09073900841908,
      "flow_matching_loss": 0.02221496878191829,
      "grad_norm": 0.09049811214208603,
      "learning_rate": 3.6294088201524394e-05,
      "loss": 0.0238,
      "relation_loss": 0.0016286425525322556,
      "selected_pairs": 26966.8,
      "state_query_norm": 1133.8137084960938,
      "state_relation_loss": 0.00031658630759920923,
      "step": 18270,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016286425525322556,
      "weighted_relation_loss": 0.0016286425525322556
    },
    {
      "action_query_norm": 634.7226135253907,
      "action_relation_loss": 0.0012758349999785422,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.100093545369504,
      "flow_matching_loss": 0.023795466497540475,
      "grad_norm": 0.09717263281345367,
      "learning_rate": 3.62410920290264e-05,
      "loss": 0.0254,
      "relation_loss": 0.001602354075293988,
      "selected_pairs": 26620.9,
      "state_query_norm": 1136.2298706054687,
      "state_relation_loss": 0.00032651907531544566,
      "step": 18280,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001602354075293988,
      "weighted_relation_loss": 0.001602354075293988
    },
    {
      "action_query_norm": 632.512548828125,
      "action_relation_loss": 0.001354650210123509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.109448082319926,
      "flow_matching_loss": 0.023213955760002136,
      "grad_norm": 0.08802609145641327,
      "learning_rate": 3.6188112574898954e-05,
      "loss": 0.0249,
      "relation_loss": 0.0016898415167815983,
      "selected_pairs": 26571.9,
      "state_query_norm": 1136.0973754882812,
      "state_relation_loss": 0.00033519129792694005,
      "step": 18290,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0016898415167815983,
      "weighted_relation_loss": 0.0016898415167815983
    },
    {
      "action_query_norm": 627.6776489257812,
      "action_relation_loss": 0.0013468756107613445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.118802619270348,
      "flow_matching_loss": 0.024728119373321533,
      "grad_norm": 0.0829734206199646,
      "learning_rate": 3.6135149903517115e-05,
      "loss": 0.0264,
      "relation_loss": 0.0016824669670313596,
      "selected_pairs": 26864.4,
      "state_query_norm": 1143.028759765625,
      "state_relation_loss": 0.0003355913650011644,
      "step": 18300,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0016824669670313596,
      "weighted_relation_loss": 0.0016824669670313596
    },
    {
      "action_query_norm": 634.3055847167968,
      "action_relation_loss": 0.001291021122597158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.128157156220766,
      "flow_matching_loss": 0.02457327414304018,
      "grad_norm": 0.11676637083292007,
      "learning_rate": 3.608220407923552e-05,
      "loss": 0.0262,
      "relation_loss": 0.0016185962012968957,
      "selected_pairs": 26574.4,
      "state_query_norm": 1131.8240478515625,
      "state_relation_loss": 0.0003275750728789717,
      "step": 18310,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0016185962012968957,
      "weighted_relation_loss": 0.0016185962012968957
    },
    {
      "action_query_norm": 636.3154052734375,
      "action_relation_loss": 0.001286483451258391,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.137511693171188,
      "flow_matching_loss": 0.023264794051647185,
      "grad_norm": 0.11649197340011597,
      "learning_rate": 3.602927516638833e-05,
      "loss": 0.0249,
      "relation_loss": 0.001623480545822531,
      "selected_pairs": 27247.5,
      "state_query_norm": 1138.0782958984375,
      "state_relation_loss": 0.0003369970887433738,
      "step": 18320,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001623480545822531,
      "weighted_relation_loss": 0.001623480545822531
    },
    {
      "action_query_norm": 634.5320251464843,
      "action_relation_loss": 0.0013974865549243986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.14686623012161,
      "flow_matching_loss": 0.020613984577357768,
      "grad_norm": 0.09428569674491882,
      "learning_rate": 3.5976363229289165e-05,
      "loss": 0.0223,
      "relation_loss": 0.001728409226052463,
      "selected_pairs": 27034.1,
      "state_query_norm": 1118.4461791992187,
      "state_relation_loss": 0.00033092266821768137,
      "step": 18330,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001728409226052463,
      "weighted_relation_loss": 0.001728409226052463
    },
    {
      "action_query_norm": 630.5524291992188,
      "action_relation_loss": 0.001304079603869468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.15622076707203,
      "flow_matching_loss": 0.022263090312480926,
      "grad_norm": 0.0926445797085762,
      "learning_rate": 3.5923468332231e-05,
      "loss": 0.0239,
      "relation_loss": 0.001632386795245111,
      "selected_pairs": 26848.3,
      "state_query_norm": 1138.5767822265625,
      "state_relation_loss": 0.00032830718846525996,
      "step": 18340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001632386795245111,
      "weighted_relation_loss": 0.001632386795245111
    },
    {
      "action_query_norm": 631.95126953125,
      "action_relation_loss": 0.0013281321967951953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.16557530402245,
      "flow_matching_loss": 0.023056378401815892,
      "grad_norm": 0.08912496268749237,
      "learning_rate": 3.587059053948616e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016622738912701607,
      "selected_pairs": 27015.3,
      "state_query_norm": 1134.7048706054688,
      "state_relation_loss": 0.00033414171193726363,
      "step": 18350,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0016622738912701607,
      "weighted_relation_loss": 0.0016622738912701607
    },
    {
      "action_query_norm": 635.1708801269531,
      "action_relation_loss": 0.0013320565922185779,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.174929840972872,
      "flow_matching_loss": 0.024200870655477047,
      "grad_norm": 0.11619172990322113,
      "learning_rate": 3.5817729915306134e-05,
      "loss": 0.0259,
      "relation_loss": 0.0016677339444868267,
      "selected_pairs": 26570.7,
      "state_query_norm": 1137.1696166992188,
      "state_relation_loss": 0.00033567735517863183,
      "step": 18360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0016677339444868267,
      "weighted_relation_loss": 0.0016677339444868267
    },
    {
      "action_query_norm": 635.3183776855469,
      "action_relation_loss": 0.0014000382740050553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.184284377923294,
      "flow_matching_loss": 0.02135081570595503,
      "grad_norm": 0.11803487688302994,
      "learning_rate": 3.5764886523921566e-05,
      "loss": 0.0231,
      "relation_loss": 0.0017295989207923412,
      "selected_pairs": 26692.0,
      "state_query_norm": 1153.1992919921875,
      "state_relation_loss": 0.00032956064969766883,
      "step": 18370,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0017295989207923412,
      "weighted_relation_loss": 0.0017295989207923412
    },
    {
      "action_query_norm": 638.3829833984375,
      "action_relation_loss": 0.0013844102155417203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.193638914873713,
      "flow_matching_loss": 0.023153583705425262,
      "grad_norm": 0.10616245120763779,
      "learning_rate": 3.571206042954214e-05,
      "loss": 0.0249,
      "relation_loss": 0.001711646595504135,
      "selected_pairs": 27058.9,
      "state_query_norm": 1131.6380126953125,
      "state_relation_loss": 0.0003272363683208823,
      "step": 18380,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001711646595504135,
      "weighted_relation_loss": 0.001711646595504135
    },
    {
      "action_query_norm": 631.4995056152344,
      "action_relation_loss": 0.0013112102402374148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.202993451824135,
      "flow_matching_loss": 0.022176015190780163,
      "grad_norm": 0.10947487503290176,
      "learning_rate": 3.5659251696356566e-05,
      "loss": 0.0238,
      "relation_loss": 0.0016333753592334688,
      "selected_pairs": 26925.7,
      "state_query_norm": 1141.57587890625,
      "state_relation_loss": 0.0003221651189960539,
      "step": 18390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0016333753592334688,
      "weighted_relation_loss": 0.0016333753592334688
    },
    {
      "action_query_norm": 637.0333190917969,
      "action_relation_loss": 0.00126841178862378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.212347988774557,
      "flow_matching_loss": 0.022360284253954886,
      "grad_norm": 0.09296382963657379,
      "learning_rate": 3.56064603885324e-05,
      "loss": 0.024,
      "relation_loss": 0.0015950302593410015,
      "selected_pairs": 26805.2,
      "state_query_norm": 1134.4619018554688,
      "state_relation_loss": 0.00032661846198607237,
      "step": 18400,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015950302593410015,
      "weighted_relation_loss": 0.0015950302593410015
    },
    {
      "action_query_norm": 632.3865173339843,
      "action_relation_loss": 0.001317601278424263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.221702525724975,
      "flow_matching_loss": 0.022605041228234767,
      "grad_norm": 0.09543244540691376,
      "learning_rate": 3.5553686570216115e-05,
      "loss": 0.0243,
      "relation_loss": 0.0016555652488023042,
      "selected_pairs": 25846.2,
      "state_query_norm": 1126.8688354492188,
      "state_relation_loss": 0.0003379639907507226,
      "step": 18410,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.0016555652488023042,
      "weighted_relation_loss": 0.0016555652488023042
    },
    {
      "action_query_norm": 632.33564453125,
      "action_relation_loss": 0.0012881040456704795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.231057062675397,
      "flow_matching_loss": 0.02344239205121994,
      "grad_norm": 0.0880265086889267,
      "learning_rate": 3.550093030553284e-05,
      "loss": 0.0251,
      "relation_loss": 0.0016205194289796055,
      "selected_pairs": 26986.4,
      "state_query_norm": 1126.3634887695312,
      "state_relation_loss": 0.0003324153949506581,
      "step": 18420,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0016205194289796055,
      "weighted_relation_loss": 0.0016205194289796055
    },
    {
      "action_query_norm": 636.409228515625,
      "action_relation_loss": 0.0012705355533398689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.24041159962582,
      "flow_matching_loss": 0.022830940410494806,
      "grad_norm": 0.09687186777591705,
      "learning_rate": 3.544819165858642e-05,
      "loss": 0.0244,
      "relation_loss": 0.001594680710695684,
      "selected_pairs": 26486.2,
      "state_query_norm": 1145.0252319335937,
      "state_relation_loss": 0.0003241451573558152,
      "step": 18430,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001594680710695684,
      "weighted_relation_loss": 0.001594680710695684
    },
    {
      "action_query_norm": 640.87158203125,
      "action_relation_loss": 0.001292089221533388,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.24976613657624,
      "flow_matching_loss": 0.022921846807003023,
      "grad_norm": 0.0835244432091713,
      "learning_rate": 3.539547069345926e-05,
      "loss": 0.0245,
      "relation_loss": 0.0016170833958312869,
      "selected_pairs": 27104.0,
      "state_query_norm": 1140.46923828125,
      "state_relation_loss": 0.00032499417720828206,
      "step": 18440,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0016170833958312869,
      "weighted_relation_loss": 0.0016170833958312869
    },
    {
      "action_query_norm": 633.1661926269531,
      "action_relation_loss": 0.0013519428903236985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.25912067352666,
      "flow_matching_loss": 0.021620933525264265,
      "grad_norm": 0.09339065849781036,
      "learning_rate": 3.5342767474212344e-05,
      "loss": 0.0233,
      "relation_loss": 0.0016757740057073533,
      "selected_pairs": 27639.2,
      "state_query_norm": 1134.1082885742187,
      "state_relation_loss": 0.00032383112411480395,
      "step": 18450,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0016757740057073533,
      "weighted_relation_loss": 0.0016757740057073533
    },
    {
      "action_query_norm": 632.3404724121094,
      "action_relation_loss": 0.0013556292396970093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.26847521047708,
      "flow_matching_loss": 0.022517265751957893,
      "grad_norm": 0.10564825683832169,
      "learning_rate": 3.529008206488502e-05,
      "loss": 0.0242,
      "relation_loss": 0.001693439797963947,
      "selected_pairs": 27262.2,
      "state_query_norm": 1133.374609375,
      "state_relation_loss": 0.0003378105524461716,
      "step": 18460,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001693439797963947,
      "weighted_relation_loss": 0.001693439797963947
    },
    {
      "action_query_norm": 628.345361328125,
      "action_relation_loss": 0.0013441443326883018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.277829747427504,
      "flow_matching_loss": 0.02282196916639805,
      "grad_norm": 0.0908937081694603,
      "learning_rate": 3.5237414529495055e-05,
      "loss": 0.0245,
      "relation_loss": 0.001672541897278279,
      "selected_pairs": 26790.1,
      "state_query_norm": 1127.834619140625,
      "state_relation_loss": 0.00032839755876921115,
      "step": 18470,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001672541897278279,
      "weighted_relation_loss": 0.001672541897278279
    },
    {
      "action_query_norm": 632.2577209472656,
      "action_relation_loss": 0.0013993613538332283,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.287184284377922,
      "flow_matching_loss": 0.019972926843911408,
      "grad_norm": 0.09528064727783203,
      "learning_rate": 3.5184764932038454e-05,
      "loss": 0.0217,
      "relation_loss": 0.001728376536630094,
      "selected_pairs": 26943.2,
      "state_query_norm": 1142.0457763671875,
      "state_relation_loss": 0.00032901517988648266,
      "step": 18480,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001728376536630094,
      "weighted_relation_loss": 0.001728376536630094
    },
    {
      "action_query_norm": 631.1861877441406,
      "action_relation_loss": 0.0013080645352602006,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.296538821328344,
      "flow_matching_loss": 0.023609542287886143,
      "grad_norm": 0.08513154834508896,
      "learning_rate": 3.513213333648945e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016355378669686616,
      "selected_pairs": 26824.5,
      "state_query_norm": 1139.5649658203124,
      "state_relation_loss": 0.00032747333170846105,
      "step": 18490,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0016355378669686616,
      "weighted_relation_loss": 0.0016355378669686616
    },
    {
      "action_query_norm": 639.0111572265625,
      "action_relation_loss": 0.001338924956507981,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.305893358278766,
      "flow_matching_loss": 0.02139425426721573,
      "grad_norm": 0.09174446761608124,
      "learning_rate": 3.5079519806800374e-05,
      "loss": 0.0231,
      "relation_loss": 0.001670205732807517,
      "selected_pairs": 26961.5,
      "state_query_norm": 1145.2249633789063,
      "state_relation_loss": 0.00033128077338915316,
      "step": 18500,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001670205732807517,
      "weighted_relation_loss": 0.001670205732807517
    },
    {
      "action_query_norm": 631.860791015625,
      "action_relation_loss": 0.0012700349441729486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.315247895229184,
      "flow_matching_loss": 0.023597775772213937,
      "grad_norm": 0.1164904460310936,
      "learning_rate": 3.502692440690165e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016042578150518238,
      "selected_pairs": 26711.1,
      "state_query_norm": 1136.69765625,
      "state_relation_loss": 0.00033422286505810914,
      "step": 18510,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0016042578150518238,
      "weighted_relation_loss": 0.0016042578150518238
    },
    {
      "action_query_norm": 635.1047546386719,
      "action_relation_loss": 0.001333316881209612,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.324602432179606,
      "flow_matching_loss": 0.023726267367601396,
      "grad_norm": 0.10569051653146744,
      "learning_rate": 3.497434720070165e-05,
      "loss": 0.0254,
      "relation_loss": 0.0016662562964484095,
      "selected_pairs": 27124.8,
      "state_query_norm": 1130.3912963867188,
      "state_relation_loss": 0.0003329394123284146,
      "step": 18520,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0016662562964484095,
      "weighted_relation_loss": 0.0016662562964484095
    },
    {
      "action_query_norm": 632.1748352050781,
      "action_relation_loss": 0.0012880600290372968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.33395696913003,
      "flow_matching_loss": 0.02317981645464897,
      "grad_norm": 0.10915596038103104,
      "learning_rate": 3.4921788252086616e-05,
      "loss": 0.0248,
      "relation_loss": 0.0016180148697458208,
      "selected_pairs": 26673.5,
      "state_query_norm": 1130.023291015625,
      "state_relation_loss": 0.00032995485817082225,
      "step": 18530,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0016180148697458208,
      "weighted_relation_loss": 0.0016180148697458208
    },
    {
      "action_query_norm": 626.64736328125,
      "action_relation_loss": 0.0012767689535394312,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.34331150608045,
      "flow_matching_loss": 0.023692643083631993,
      "grad_norm": 0.10774267464876175,
      "learning_rate": 3.486924762492065e-05,
      "loss": 0.0253,
      "relation_loss": 0.0016024860786274075,
      "selected_pairs": 27406.1,
      "state_query_norm": 1133.208349609375,
      "state_relation_loss": 0.00032571711635682733,
      "step": 18540,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0016024860786274075,
      "weighted_relation_loss": 0.0016024860786274075
    },
    {
      "action_query_norm": 633.5020263671875,
      "action_relation_loss": 0.0012758648372255266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.35266604303087,
      "flow_matching_loss": 0.022021673247218133,
      "grad_norm": 0.10381375998258591,
      "learning_rate": 3.4816725383045534e-05,
      "loss": 0.0236,
      "relation_loss": 0.0015975313493981958,
      "selected_pairs": 27017.8,
      "state_query_norm": 1142.4462280273438,
      "state_relation_loss": 0.00032166652381420134,
      "step": 18550,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0015975313493981958,
      "weighted_relation_loss": 0.0015975313493981958
    },
    {
      "action_query_norm": 631.636181640625,
      "action_relation_loss": 0.001339979621116072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.36202057998129,
      "flow_matching_loss": 0.024137219414114952,
      "grad_norm": 0.09205857664346695,
      "learning_rate": 3.476422159028079e-05,
      "loss": 0.0258,
      "relation_loss": 0.0016665357165038587,
      "selected_pairs": 26832.8,
      "state_query_norm": 1129.9903442382813,
      "state_relation_loss": 0.0003265560924774036,
      "step": 18560,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0016665357165038587,
      "weighted_relation_loss": 0.0016665357165038587
    },
    {
      "action_query_norm": 635.847900390625,
      "action_relation_loss": 0.0012695526704192162,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.371375116931713,
      "flow_matching_loss": 0.022431345004588366,
      "grad_norm": 0.0999671146273613,
      "learning_rate": 3.471173631042345e-05,
      "loss": 0.024,
      "relation_loss": 0.0016090929973870515,
      "selected_pairs": 27248.1,
      "state_query_norm": 1129.3835205078126,
      "state_relation_loss": 0.0003395403386093676,
      "step": 18570,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0016090929973870515,
      "weighted_relation_loss": 0.0016090929973870515
    },
    {
      "action_query_norm": 631.6430297851563,
      "action_relation_loss": 0.0013322190730832516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.38072965388213,
      "flow_matching_loss": 0.022898970171809196,
      "grad_norm": 0.13818171620368958,
      "learning_rate": 3.465926960724808e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016777107142843307,
      "selected_pairs": 26944.1,
      "state_query_norm": 1134.7311767578126,
      "state_relation_loss": 0.0003454916412010789,
      "step": 18580,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016777107142843307,
      "weighted_relation_loss": 0.0016777107142843307
    },
    {
      "action_query_norm": 631.0523681640625,
      "action_relation_loss": 0.0013398867216892541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.390084190832553,
      "flow_matching_loss": 0.02173692490905523,
      "grad_norm": 0.08762598037719727,
      "learning_rate": 3.460682154450666e-05,
      "loss": 0.0234,
      "relation_loss": 0.0016608869424089789,
      "selected_pairs": 26937.2,
      "state_query_norm": 1135.9925903320313,
      "state_relation_loss": 0.0003210002352716401,
      "step": 18590,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016608869424089789,
      "weighted_relation_loss": 0.0016608869424089789
    },
    {
      "action_query_norm": 636.6424377441406,
      "action_relation_loss": 0.0012965946341864764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.399438727782975,
      "flow_matching_loss": 0.022458661533892154,
      "grad_norm": 0.11927514523267746,
      "learning_rate": 3.4554392185928564e-05,
      "loss": 0.0241,
      "relation_loss": 0.0016122621949762107,
      "selected_pairs": 26597.6,
      "state_query_norm": 1136.900537109375,
      "state_relation_loss": 0.0003156675637001172,
      "step": 18600,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0016122621949762107,
      "weighted_relation_loss": 0.0016122621949762107
    },
    {
      "action_query_norm": 629.498486328125,
      "action_relation_loss": 0.001308961387258023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.408793264733397,
      "flow_matching_loss": 0.02421797104179859,
      "grad_norm": 0.1086113303899765,
      "learning_rate": 3.450198159522037e-05,
      "loss": 0.0259,
      "relation_loss": 0.0016388855990953743,
      "selected_pairs": 26620.9,
      "state_query_norm": 1131.2633544921875,
      "state_relation_loss": 0.00032992422638926654,
      "step": 18610,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0016388855990953743,
      "weighted_relation_loss": 0.0016388855990953743
    },
    {
      "action_query_norm": 635.896142578125,
      "action_relation_loss": 0.001319095492362976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.418147801683816,
      "flow_matching_loss": 0.02160440254956484,
      "grad_norm": 0.09364011883735657,
      "learning_rate": 3.444958983606592e-05,
      "loss": 0.0233,
      "relation_loss": 0.0016480737132951618,
      "selected_pairs": 27264.3,
      "state_query_norm": 1126.6187622070313,
      "state_relation_loss": 0.00032897821511141957,
      "step": 18620,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0016480737132951618,
      "weighted_relation_loss": 0.0016480737132951618
    },
    {
      "action_query_norm": 632.3729736328125,
      "action_relation_loss": 0.0013603449566289782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.427502338634238,
      "flow_matching_loss": 0.023533522337675094,
      "grad_norm": 0.09010127186775208,
      "learning_rate": 3.439721697212612e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016705771326087416,
      "selected_pairs": 26484.2,
      "state_query_norm": 1139.6156005859375,
      "state_relation_loss": 0.0003102321963524446,
      "step": 18630,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0016705771326087416,
      "weighted_relation_loss": 0.0016705771326087416
    },
    {
      "action_query_norm": 633.9494201660157,
      "action_relation_loss": 0.0013281257124617696,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.43685687558466,
      "flow_matching_loss": 0.021350072696805,
      "grad_norm": 0.0889134481549263,
      "learning_rate": 3.4344863067038954e-05,
      "loss": 0.023,
      "relation_loss": 0.0016431293333880603,
      "selected_pairs": 27013.4,
      "state_query_norm": 1135.2753540039062,
      "state_relation_loss": 0.0003150036296574399,
      "step": 18640,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0016431293333880603,
      "weighted_relation_loss": 0.0016431293333880603
    },
    {
      "action_query_norm": 633.2046997070313,
      "action_relation_loss": 0.0013354307622648776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.446211412535078,
      "flow_matching_loss": 0.02299487292766571,
      "grad_norm": 0.11222070455551147,
      "learning_rate": 3.429252818441935e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016579614719375968,
      "selected_pairs": 27164.1,
      "state_query_norm": 1129.1796020507813,
      "state_relation_loss": 0.0003225307242246345,
      "step": 18650,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0016579614719375968,
      "weighted_relation_loss": 0.0016579614719375968
    },
    {
      "action_query_norm": 631.5591125488281,
      "action_relation_loss": 0.0012556401081383229,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.4555659494855,
      "flow_matching_loss": 0.022838452830910683,
      "grad_norm": 0.1315847933292389,
      "learning_rate": 3.4240212387859094e-05,
      "loss": 0.0244,
      "relation_loss": 0.0015901305479928852,
      "selected_pairs": 26848.0,
      "state_query_norm": 1134.1263916015625,
      "state_relation_loss": 0.00033449044567532835,
      "step": 18660,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0015901305479928852,
      "weighted_relation_loss": 0.0015901305479928852
    },
    {
      "action_query_norm": 642.7235046386719,
      "action_relation_loss": 0.0012661923537962139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.464920486435922,
      "flow_matching_loss": 0.023285092413425447,
      "grad_norm": 0.07628825306892395,
      "learning_rate": 3.418791574092686e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015833723125979304,
      "selected_pairs": 27369.4,
      "state_query_norm": 1123.53681640625,
      "state_relation_loss": 0.00031717995298095045,
      "step": 18670,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0015833723125979304,
      "weighted_relation_loss": 0.0015833723125979304
    },
    {
      "action_query_norm": 634.3192138671875,
      "action_relation_loss": 0.0012871520244516431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.474275023386344,
      "flow_matching_loss": 0.02374597955495119,
      "grad_norm": 0.10703379660844803,
      "learning_rate": 3.413563830716796e-05,
      "loss": 0.0254,
      "relation_loss": 0.0016188511508516967,
      "selected_pairs": 26993.8,
      "state_query_norm": 1134.653564453125,
      "state_relation_loss": 0.0003316991264000535,
      "step": 18680,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0016188511508516967,
      "weighted_relation_loss": 0.0016188511508516967
    },
    {
      "action_query_norm": 627.7886901855469,
      "action_relation_loss": 0.0013417746988125146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.483629560336762,
      "flow_matching_loss": 0.024133980087935925,
      "grad_norm": 0.11556889861822128,
      "learning_rate": 3.408338015010445e-05,
      "loss": 0.0258,
      "relation_loss": 0.0016683732392266393,
      "selected_pairs": 26410.1,
      "state_query_norm": 1122.6623168945312,
      "state_relation_loss": 0.00032659855205565693,
      "step": 18690,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0016683732392266393,
      "weighted_relation_loss": 0.0016683732392266393
    },
    {
      "action_query_norm": 637.3189880371094,
      "action_relation_loss": 0.001253144652582705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.492984097287184,
      "flow_matching_loss": 0.024534568935632706,
      "grad_norm": 0.10230479389429092,
      "learning_rate": 3.40311413332349e-05,
      "loss": 0.0261,
      "relation_loss": 0.0015743108815513551,
      "selected_pairs": 27367.3,
      "state_query_norm": 1133.3635620117188,
      "state_relation_loss": 0.00032116621441673486,
      "step": 18700,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0015743108815513551,
      "weighted_relation_loss": 0.0015743108815513551
    },
    {
      "action_query_norm": 626.8626892089844,
      "action_relation_loss": 0.0013149405713193119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.502338634237606,
      "flow_matching_loss": 0.022500409744679928,
      "grad_norm": 0.14765192568302155,
      "learning_rate": 3.3978921920034365e-05,
      "loss": 0.0241,
      "relation_loss": 0.0016375719453208149,
      "selected_pairs": 26653.3,
      "state_query_norm": 1129.5416870117188,
      "state_relation_loss": 0.00032263139437418433,
      "step": 18710,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0016375719453208149,
      "weighted_relation_loss": 0.0016375719453208149
    },
    {
      "action_query_norm": 629.4354858398438,
      "action_relation_loss": 0.0012751391273923217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.511693171188025,
      "flow_matching_loss": 0.020935579761862756,
      "grad_norm": 0.09226073324680328,
      "learning_rate": 3.392672197395441e-05,
      "loss": 0.0225,
      "relation_loss": 0.0015974294394254685,
      "selected_pairs": 26851.6,
      "state_query_norm": 1139.32392578125,
      "state_relation_loss": 0.0003222903062123805,
      "step": 18720,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0015974294394254685,
      "weighted_relation_loss": 0.0015974294394254685
    },
    {
      "action_query_norm": 627.7397766113281,
      "action_relation_loss": 0.0013494994025677442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.521047708138447,
      "flow_matching_loss": 0.022011217847466468,
      "grad_norm": 0.12737096846103668,
      "learning_rate": 3.387454155842287e-05,
      "loss": 0.0237,
      "relation_loss": 0.0016822784557007254,
      "selected_pairs": 26643.4,
      "state_query_norm": 1118.7740478515625,
      "state_relation_loss": 0.00033277904731221497,
      "step": 18730,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0016822784557007254,
      "weighted_relation_loss": 0.0016822784557007254
    },
    {
      "action_query_norm": 630.8393310546875,
      "action_relation_loss": 0.0012793794274330139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.53040224508887,
      "flow_matching_loss": 0.022130529023706914,
      "grad_norm": 0.10851020365953445,
      "learning_rate": 3.382238073684386e-05,
      "loss": 0.0237,
      "relation_loss": 0.0016080604982562363,
      "selected_pairs": 26969.5,
      "state_query_norm": 1125.804931640625,
      "state_relation_loss": 0.0003286810650024563,
      "step": 18740,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016080604982562363,
      "weighted_relation_loss": 0.0016080604982562363
    },
    {
      "action_query_norm": 639.4194030761719,
      "action_relation_loss": 0.0012771171517670155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.53975678203929,
      "flow_matching_loss": 0.02157265804708004,
      "grad_norm": 0.08677183091640472,
      "learning_rate": 3.377023957259771e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015990879386663438,
      "selected_pairs": 27088.0,
      "state_query_norm": 1132.8633422851562,
      "state_relation_loss": 0.0003219707927200943,
      "step": 18750,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0015990879386663438,
      "weighted_relation_loss": 0.0015990879386663438
    },
    {
      "action_query_norm": 636.3407043457031,
      "action_relation_loss": 0.001247428823262453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.54911131898971,
      "flow_matching_loss": 0.021326621249318122,
      "grad_norm": 0.12643346190452576,
      "learning_rate": 3.3718118129040835e-05,
      "loss": 0.0229,
      "relation_loss": 0.001571448496542871,
      "selected_pairs": 27180.2,
      "state_query_norm": 1125.0116333007813,
      "state_relation_loss": 0.0003240196645492688,
      "step": 18760,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001571448496542871,
      "weighted_relation_loss": 0.001571448496542871
    },
    {
      "action_query_norm": 637.7576110839843,
      "action_relation_loss": 0.001329900079872459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.55846585594013,
      "flow_matching_loss": 0.02128787972033024,
      "grad_norm": 0.09363893419504166,
      "learning_rate": 3.3666016469505724e-05,
      "loss": 0.0229,
      "relation_loss": 0.0016584814409725368,
      "selected_pairs": 26943.6,
      "state_query_norm": 1118.4994262695313,
      "state_relation_loss": 0.0003285813698312268,
      "step": 18770,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016584814409725368,
      "weighted_relation_loss": 0.0016584814409725368
    },
    {
      "action_query_norm": 634.2398742675781,
      "action_relation_loss": 0.0012573470827192067,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.567820392890553,
      "flow_matching_loss": 0.021412508748471736,
      "grad_norm": 0.10610703378915787,
      "learning_rate": 3.361393465730079e-05,
      "loss": 0.023,
      "relation_loss": 0.0015771191800013184,
      "selected_pairs": 26875.8,
      "state_query_norm": 1135.0732788085938,
      "state_relation_loss": 0.00031977209146134556,
      "step": 18780,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015771191800013184,
      "weighted_relation_loss": 0.0015771191800013184
    },
    {
      "action_query_norm": 635.7966613769531,
      "action_relation_loss": 0.00129773726221174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.57717492984097,
      "flow_matching_loss": 0.024164036847651005,
      "grad_norm": 0.09107500314712524,
      "learning_rate": 3.3561872755710366e-05,
      "loss": 0.0258,
      "relation_loss": 0.0016152896103449165,
      "selected_pairs": 26922.0,
      "state_query_norm": 1130.8198608398438,
      "state_relation_loss": 0.0003175523423124105,
      "step": 18790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0016152896103449165,
      "weighted_relation_loss": 0.0016152896103449165
    },
    {
      "action_query_norm": 631.7257873535157,
      "action_relation_loss": 0.0012468675267882645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.586529466791394,
      "flow_matching_loss": 0.021811109222471713,
      "grad_norm": 0.07542286813259125,
      "learning_rate": 3.350983082799456e-05,
      "loss": 0.0234,
      "relation_loss": 0.0015751380473375321,
      "selected_pairs": 26780.2,
      "state_query_norm": 1129.0197509765626,
      "state_relation_loss": 0.00032827050890773537,
      "step": 18800,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0015751380473375321,
      "weighted_relation_loss": 0.0015751380473375321
    },
    {
      "action_query_norm": 626.1490417480469,
      "action_relation_loss": 0.0013411758351139723,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.595884003741816,
      "flow_matching_loss": 0.02142925448715687,
      "grad_norm": 0.11433344334363937,
      "learning_rate": 3.34578089373892e-05,
      "loss": 0.0231,
      "relation_loss": 0.0016672874102368952,
      "selected_pairs": 26756.7,
      "state_query_norm": 1122.3639282226563,
      "state_relation_loss": 0.0003261115838540718,
      "step": 18810,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0016672874102368952,
      "weighted_relation_loss": 0.0016672874102368952
    },
    {
      "action_query_norm": 635.7419738769531,
      "action_relation_loss": 0.001240566186606884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.605238540692234,
      "flow_matching_loss": 0.021119138319045307,
      "grad_norm": 0.08598607778549194,
      "learning_rate": 3.340580714710581e-05,
      "loss": 0.0227,
      "relation_loss": 0.0015650281100533902,
      "selected_pairs": 26800.5,
      "state_query_norm": 1138.212451171875,
      "state_relation_loss": 0.0003244619234465063,
      "step": 18820,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015650281100533902,
      "weighted_relation_loss": 0.0015650281100533902
    },
    {
      "action_query_norm": 634.5749633789062,
      "action_relation_loss": 0.001235084712971002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.614593077642656,
      "flow_matching_loss": 0.022020890004932882,
      "grad_norm": 0.09625320881605148,
      "learning_rate": 3.3353825520331466e-05,
      "loss": 0.0236,
      "relation_loss": 0.0015618854784406721,
      "selected_pairs": 27627.5,
      "state_query_norm": 1120.008935546875,
      "state_relation_loss": 0.0003268007742008194,
      "step": 18830,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0015618854784406721,
      "weighted_relation_loss": 0.0015618854784406721
    },
    {
      "action_query_norm": 625.6326049804687,
      "action_relation_loss": 0.0012731908122077585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.623947614593078,
      "flow_matching_loss": 0.021355769596993923,
      "grad_norm": 0.08274749666452408,
      "learning_rate": 3.330186412022876e-05,
      "loss": 0.023,
      "relation_loss": 0.0015947425505146383,
      "selected_pairs": 26158.0,
      "state_query_norm": 1134.5521850585938,
      "state_relation_loss": 0.0003215517295757309,
      "step": 18840,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0015947425505146383,
      "weighted_relation_loss": 0.0015947425505146383
    },
    {
      "action_query_norm": 628.6058227539063,
      "action_relation_loss": 0.0012748914072290064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.6333021515435,
      "flow_matching_loss": 0.020160912070423365,
      "grad_norm": 0.11997514218091965,
      "learning_rate": 3.324992300993568e-05,
      "loss": 0.0218,
      "relation_loss": 0.0016018062247894703,
      "selected_pairs": 26622.7,
      "state_query_norm": 1123.9859619140625,
      "state_relation_loss": 0.0003269148146500811,
      "step": 18850,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0016018062247894703,
      "weighted_relation_loss": 0.0016018062247894703
    },
    {
      "action_query_norm": 627.2206176757812,
      "action_relation_loss": 0.0012863815063610673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.64265668849392,
      "flow_matching_loss": 0.021669875457882883,
      "grad_norm": 0.10928725451231003,
      "learning_rate": 3.319800225256556e-05,
      "loss": 0.0233,
      "relation_loss": 0.0015994861256331205,
      "selected_pairs": 26768.6,
      "state_query_norm": 1121.1978271484375,
      "state_relation_loss": 0.0003131046221824363,
      "step": 18860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0015994861256331205,
      "weighted_relation_loss": 0.0015994861256331205
    },
    {
      "action_query_norm": 628.4800354003906,
      "action_relation_loss": 0.0012255594483576715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.65201122544434,
      "flow_matching_loss": 0.02334211189299822,
      "grad_norm": 0.10845866054296494,
      "learning_rate": 3.314610191120702e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015398568706586958,
      "selected_pairs": 26481.6,
      "state_query_norm": 1135.1921630859374,
      "state_relation_loss": 0.00031429740774910895,
      "step": 18870,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0015398568706586958,
      "weighted_relation_loss": 0.0015398568706586958
    },
    {
      "action_query_norm": 629.8159301757812,
      "action_relation_loss": 0.0012675720849074424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.661365762394762,
      "flow_matching_loss": 0.021781396679580213,
      "grad_norm": 0.09760734438896179,
      "learning_rate": 3.30942220489239e-05,
      "loss": 0.0234,
      "relation_loss": 0.001596970623359084,
      "selected_pairs": 26968.2,
      "state_query_norm": 1124.2743774414062,
      "state_relation_loss": 0.00032939852389972656,
      "step": 18880,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001596970623359084,
      "weighted_relation_loss": 0.001596970623359084
    },
    {
      "action_query_norm": 630.3102966308594,
      "action_relation_loss": 0.0012784361955709756,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.67072029934518,
      "flow_matching_loss": 0.02313916403800249,
      "grad_norm": 0.08712024986743927,
      "learning_rate": 3.3042362728755086e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016100349719636142,
      "selected_pairs": 26341.2,
      "state_query_norm": 1107.9892700195312,
      "state_relation_loss": 0.00033159876766148956,
      "step": 18890,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0016100349719636142,
      "weighted_relation_loss": 0.0016100349719636142
    },
    {
      "action_query_norm": 631.5160522460938,
      "action_relation_loss": 0.0012991780531592667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.680074836295603,
      "flow_matching_loss": 0.02197108678519726,
      "grad_norm": 0.09794507175683975,
      "learning_rate": 3.299052401371456e-05,
      "loss": 0.0236,
      "relation_loss": 0.0016146003035828471,
      "selected_pairs": 27047.0,
      "state_query_norm": 1140.1400146484375,
      "state_relation_loss": 0.0003154222562443465,
      "step": 18900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0016146003035828471,
      "weighted_relation_loss": 0.0016146003035828471
    },
    {
      "action_query_norm": 631.3535339355469,
      "action_relation_loss": 0.001225748157594353,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.689429373246025,
      "flow_matching_loss": 0.02111545093357563,
      "grad_norm": 0.0949511006474495,
      "learning_rate": 3.293870596679125e-05,
      "loss": 0.0227,
      "relation_loss": 0.0015437654452398419,
      "selected_pairs": 27290.8,
      "state_query_norm": 1127.457861328125,
      "state_relation_loss": 0.0003180172847351059,
      "step": 18910,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0015437654452398419,
      "weighted_relation_loss": 0.0015437654452398419
    },
    {
      "action_query_norm": 633.0931274414063,
      "action_relation_loss": 0.0013037141994573175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.698783910196447,
      "flow_matching_loss": 0.022816881909966467,
      "grad_norm": 0.0878593847155571,
      "learning_rate": 3.288690865094895e-05,
      "loss": 0.0245,
      "relation_loss": 0.0016372588230296969,
      "selected_pairs": 27160.7,
      "state_query_norm": 1128.14697265625,
      "state_relation_loss": 0.00033354462357237935,
      "step": 18920,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0016372588230296969,
      "weighted_relation_loss": 0.0016372588230296969
    },
    {
      "action_query_norm": 626.2629455566406,
      "action_relation_loss": 0.0013481016736477613,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.708138447146865,
      "flow_matching_loss": 0.023745924420654773,
      "grad_norm": 0.09988231956958771,
      "learning_rate": 3.283513212912632e-05,
      "loss": 0.0254,
      "relation_loss": 0.0016805292572826148,
      "selected_pairs": 26593.2,
      "state_query_norm": 1123.3891845703124,
      "state_relation_loss": 0.0003324276069179177,
      "step": 18930,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0016805292572826148,
      "weighted_relation_loss": 0.0016805292572826148
    },
    {
      "action_query_norm": 627.5829833984375,
      "action_relation_loss": 0.0012389408424496652,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.717492984097287,
      "flow_matching_loss": 0.02258206680417061,
      "grad_norm": 0.08899005502462387,
      "learning_rate": 3.2783376464236684e-05,
      "loss": 0.0241,
      "relation_loss": 0.0015569266048260032,
      "selected_pairs": 26803.4,
      "state_query_norm": 1115.2439453125,
      "state_relation_loss": 0.00031798575364518913,
      "step": 18940,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015569266048260032,
      "weighted_relation_loss": 0.0015569266048260032
    },
    {
      "action_query_norm": 633.7586853027344,
      "action_relation_loss": 0.0013350676163099706,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.72684752104771,
      "flow_matching_loss": 0.02126810550689697,
      "grad_norm": 0.09324431419372559,
      "learning_rate": 3.273164171916806e-05,
      "loss": 0.0229,
      "relation_loss": 0.0016703418223187327,
      "selected_pairs": 26943.3,
      "state_query_norm": 1120.0039428710938,
      "state_relation_loss": 0.0003352742147399113,
      "step": 18950,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016703418223187327,
      "weighted_relation_loss": 0.0016703418223187327
    },
    {
      "action_query_norm": 629.9820068359375,
      "action_relation_loss": 0.0012868719757534563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.736202057998128,
      "flow_matching_loss": 0.02235442344099283,
      "grad_norm": 0.09862960129976273,
      "learning_rate": 3.267992795678306e-05,
      "loss": 0.024,
      "relation_loss": 0.001610689435619861,
      "selected_pairs": 26634.4,
      "state_query_norm": 1129.10673828125,
      "state_relation_loss": 0.0003238174511352554,
      "step": 18960,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001610689435619861,
      "weighted_relation_loss": 0.001610689435619861
    },
    {
      "action_query_norm": 626.4736206054688,
      "action_relation_loss": 0.0013140258728526534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.74555659494855,
      "flow_matching_loss": 0.02260650172829628,
      "grad_norm": 0.0876031219959259,
      "learning_rate": 3.2628235239918744e-05,
      "loss": 0.0242,
      "relation_loss": 0.0016354680294170976,
      "selected_pairs": 26881.1,
      "state_query_norm": 1123.9769897460938,
      "state_relation_loss": 0.0003214421623852104,
      "step": 18970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0016354680294170976,
      "weighted_relation_loss": 0.0016354680294170976
    },
    {
      "action_query_norm": 631.5345947265625,
      "action_relation_loss": 0.001290987920947373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.75491113189897,
      "flow_matching_loss": 0.022323341108858585,
      "grad_norm": 0.08255457133054733,
      "learning_rate": 3.2576563631386695e-05,
      "loss": 0.0239,
      "relation_loss": 0.00160240976838395,
      "selected_pairs": 26942.5,
      "state_query_norm": 1116.958203125,
      "state_relation_loss": 0.0003114218561677262,
      "step": 18980,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00160240976838395,
      "weighted_relation_loss": 0.00160240976838395
    },
    {
      "action_query_norm": 629.0051879882812,
      "action_relation_loss": 0.0012847303645685316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.764265668849394,
      "flow_matching_loss": 0.021107026003301145,
      "grad_norm": 0.09906817227602005,
      "learning_rate": 3.252491319397275e-05,
      "loss": 0.0227,
      "relation_loss": 0.001612312369979918,
      "selected_pairs": 26335.8,
      "state_query_norm": 1135.687939453125,
      "state_relation_loss": 0.00032758200250100346,
      "step": 18990,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.001612312369979918,
      "weighted_relation_loss": 0.001612312369979918
    },
    {
      "action_query_norm": 630.245654296875,
      "action_relation_loss": 0.001257682393770665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.773620205799812,
      "flow_matching_loss": 0.02218069490045309,
      "grad_norm": 0.10437574982643127,
      "learning_rate": 3.247328399043706e-05,
      "loss": 0.0238,
      "relation_loss": 0.0015829270239919424,
      "selected_pairs": 26663.1,
      "state_query_norm": 1117.48359375,
      "state_relation_loss": 0.0003252446302212775,
      "step": 19000,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0015829270239919424,
      "weighted_relation_loss": 0.0015829270239919424
    },
    {
      "action_query_norm": 630.1697448730469,
      "action_relation_loss": 0.0012963034911081194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.782974742750234,
      "flow_matching_loss": 0.022575890086591242,
      "grad_norm": 0.10862525552511215,
      "learning_rate": 3.242167608351399e-05,
      "loss": 0.0242,
      "relation_loss": 0.001604918541852385,
      "selected_pairs": 27358.8,
      "state_query_norm": 1138.9426147460938,
      "state_relation_loss": 0.0003086150623857975,
      "step": 19010,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001604918541852385,
      "weighted_relation_loss": 0.001604918541852385
    },
    {
      "action_query_norm": 629.3215454101562,
      "action_relation_loss": 0.0012804741854779422,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.792329279700656,
      "flow_matching_loss": 0.02301502525806427,
      "grad_norm": 0.09319277852773666,
      "learning_rate": 3.2370089535911986e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016011063707992435,
      "selected_pairs": 27173.5,
      "state_query_norm": 1127.2382080078125,
      "state_relation_loss": 0.00032063216203823683,
      "step": 19020,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0016011063707992435,
      "weighted_relation_loss": 0.0016011063707992435
    },
    {
      "action_query_norm": 624.9694946289062,
      "action_relation_loss": 0.001349494606256485,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.801683816651074,
      "flow_matching_loss": 0.022478694282472133,
      "grad_norm": 0.10544136166572571,
      "learning_rate": 3.23185244103136e-05,
      "loss": 0.0241,
      "relation_loss": 0.0016654658247716725,
      "selected_pairs": 26572.0,
      "state_query_norm": 1149.0042358398437,
      "state_relation_loss": 0.0003159712068736553,
      "step": 19030,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0016654658247716725,
      "weighted_relation_loss": 0.0016654658247716725
    },
    {
      "action_query_norm": 628.1525512695313,
      "action_relation_loss": 0.0012854931759648026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.811038353601496,
      "flow_matching_loss": 0.02310546487569809,
      "grad_norm": 0.0990852490067482,
      "learning_rate": 3.22669807693753e-05,
      "loss": 0.0247,
      "relation_loss": 0.001606919930782169,
      "selected_pairs": 26384.6,
      "state_query_norm": 1113.2612670898438,
      "state_relation_loss": 0.0003214267460862175,
      "step": 19040,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001606919930782169,
      "weighted_relation_loss": 0.001606919930782169
    },
    {
      "action_query_norm": 636.7289428710938,
      "action_relation_loss": 0.0012163324630819262,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.82039289055192,
      "flow_matching_loss": 0.02379950974136591,
      "grad_norm": 0.09034344553947449,
      "learning_rate": 3.2215458675727495e-05,
      "loss": 0.0253,
      "relation_loss": 0.0015384708531200886,
      "selected_pairs": 27579.6,
      "state_query_norm": 1143.2966552734374,
      "state_relation_loss": 0.0003221383900381625,
      "step": 19050,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0015384708531200886,
      "weighted_relation_loss": 0.0015384708531200886
    },
    {
      "action_query_norm": 631.6499450683593,
      "action_relation_loss": 0.0012903303722850978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.82974742750234,
      "flow_matching_loss": 0.023001446574926376,
      "grad_norm": 0.1024199053645134,
      "learning_rate": 3.216395819197438e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016193047631531954,
      "selected_pairs": 26638.7,
      "state_query_norm": 1110.3373901367188,
      "state_relation_loss": 0.0003289743821369484,
      "step": 19060,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0016193047631531954,
      "weighted_relation_loss": 0.0016193047631531954
    },
    {
      "action_query_norm": 627.3824890136718,
      "action_relation_loss": 0.0012862046016380191,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.83910196445276,
      "flow_matching_loss": 0.021950571238994597,
      "grad_norm": 0.10675457864999771,
      "learning_rate": 3.211247938069387e-05,
      "loss": 0.0236,
      "relation_loss": 0.0016132284537889064,
      "selected_pairs": 27025.6,
      "state_query_norm": 1132.006201171875,
      "state_relation_loss": 0.0003270238608820364,
      "step": 19070,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0016132284537889064,
      "weighted_relation_loss": 0.0016132284537889064
    },
    {
      "action_query_norm": 629.8177124023438,
      "action_relation_loss": 0.001294497400522232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.84845650140318,
      "flow_matching_loss": 0.02261295113712549,
      "grad_norm": 0.10345064103603363,
      "learning_rate": 3.206102230443759e-05,
      "loss": 0.0242,
      "relation_loss": 0.0016168055473826826,
      "selected_pairs": 26436.9,
      "state_query_norm": 1126.66005859375,
      "state_relation_loss": 0.00032230815559159965,
      "step": 19080,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0016168055473826826,
      "weighted_relation_loss": 0.0016168055473826826
    },
    {
      "action_query_norm": 625.6323120117188,
      "action_relation_loss": 0.0012786721461452543,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.857811038353603,
      "flow_matching_loss": 0.021775857359170914,
      "grad_norm": 0.10846404731273651,
      "learning_rate": 3.2009587025730764e-05,
      "loss": 0.0234,
      "relation_loss": 0.0016130446456372738,
      "selected_pairs": 26113.4,
      "state_query_norm": 1119.1625610351562,
      "state_relation_loss": 0.00033437249367125333,
      "step": 19090,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.0016130446456372738,
      "weighted_relation_loss": 0.0016130446456372738
    },
    {
      "action_query_norm": 632.0946655273438,
      "action_relation_loss": 0.001252622075844556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.86716557530402,
      "flow_matching_loss": 0.02261520568281412,
      "grad_norm": 0.1020943894982338,
      "learning_rate": 3.195817360707207e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015776607557199895,
      "selected_pairs": 27247.5,
      "state_query_norm": 1108.5778442382812,
      "state_relation_loss": 0.00032503866823390125,
      "step": 19100,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0015776607557199895,
      "weighted_relation_loss": 0.0015776607557199895
    },
    {
      "action_query_norm": 633.3958190917969,
      "action_relation_loss": 0.001237147394567728,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.876520112254443,
      "flow_matching_loss": 0.019977287761867045,
      "grad_norm": 0.07984153181314468,
      "learning_rate": 3.19067821109337e-05,
      "loss": 0.0215,
      "relation_loss": 0.0015597036690451205,
      "selected_pairs": 26526.9,
      "state_query_norm": 1123.3831176757812,
      "state_relation_loss": 0.00032255628029815854,
      "step": 19110,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0015597036690451205,
      "weighted_relation_loss": 0.0015597036690451205
    },
    {
      "action_query_norm": 634.1860595703125,
      "action_relation_loss": 0.0012567645870149135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.885874649204865,
      "flow_matching_loss": 0.021412076149135827,
      "grad_norm": 0.09205494821071625,
      "learning_rate": 3.185541259976114e-05,
      "loss": 0.023,
      "relation_loss": 0.001585880678612739,
      "selected_pairs": 26915.6,
      "state_query_norm": 1122.8915771484376,
      "state_relation_loss": 0.00032911608286667615,
      "step": 19120,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001585880678612739,
      "weighted_relation_loss": 0.001585880678612739
    },
    {
      "action_query_norm": 633.7286499023437,
      "action_relation_loss": 0.0012707033078186215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.895229186155284,
      "flow_matching_loss": 0.020315544679760933,
      "grad_norm": 0.11048512160778046,
      "learning_rate": 3.180406513597316e-05,
      "loss": 0.0219,
      "relation_loss": 0.0015817073988728225,
      "selected_pairs": 27181.6,
      "state_query_norm": 1131.7032470703125,
      "state_relation_loss": 0.00031100409978535026,
      "step": 19130,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0015817073988728225,
      "weighted_relation_loss": 0.0015817073988728225
    },
    {
      "action_query_norm": 634.0156921386719,
      "action_relation_loss": 0.0013027640292420984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.904583723105706,
      "flow_matching_loss": 0.023222183622419835,
      "grad_norm": 0.10967657715082169,
      "learning_rate": 3.1752739781961835e-05,
      "loss": 0.0248,
      "relation_loss": 0.0016098798834718764,
      "selected_pairs": 27081.5,
      "state_query_norm": 1129.1153930664063,
      "state_relation_loss": 0.00030711586587131026,
      "step": 19140,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0016098798834718764,
      "weighted_relation_loss": 0.0016098798834718764
    },
    {
      "action_query_norm": 630.0906799316406,
      "action_relation_loss": 0.0012388104107230902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.913938260056128,
      "flow_matching_loss": 0.023752215318381786,
      "grad_norm": 0.10029122978448868,
      "learning_rate": 3.170143660009228e-05,
      "loss": 0.0253,
      "relation_loss": 0.001553864707238972,
      "selected_pairs": 26319.1,
      "state_query_norm": 1130.9859985351563,
      "state_relation_loss": 0.00031505432561971245,
      "step": 19150,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.001553864707238972,
      "weighted_relation_loss": 0.001553864707238972
    },
    {
      "action_query_norm": 631.0542419433593,
      "action_relation_loss": 0.0012717071222141385,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.92329279700655,
      "flow_matching_loss": 0.021209209691733122,
      "grad_norm": 0.1077222228050232,
      "learning_rate": 3.16501556527027e-05,
      "loss": 0.0228,
      "relation_loss": 0.0015881043625995516,
      "selected_pairs": 26629.6,
      "state_query_norm": 1123.0621704101563,
      "state_relation_loss": 0.0003163972229231149,
      "step": 19160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015881043625995516,
      "weighted_relation_loss": 0.0015881043625995516
    },
    {
      "action_query_norm": 629.4582946777343,
      "action_relation_loss": 0.0012081602588295936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.932647333956968,
      "flow_matching_loss": 0.022356150299310686,
      "grad_norm": 0.11427503079175949,
      "learning_rate": 3.1598897002104265e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015263731009326875,
      "selected_pairs": 27526.3,
      "state_query_norm": 1133.6184326171874,
      "state_relation_loss": 0.00031821283046156167,
      "step": 19170,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0015263731009326875,
      "weighted_relation_loss": 0.0015263731009326875
    },
    {
      "action_query_norm": 635.3727600097657,
      "action_relation_loss": 0.0012487179599702358,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.94200187090739,
      "flow_matching_loss": 0.021334398537874222,
      "grad_norm": 0.11060461401939392,
      "learning_rate": 3.154766071058108e-05,
      "loss": 0.0229,
      "relation_loss": 0.0015716118738055228,
      "selected_pairs": 26483.2,
      "state_query_norm": 1124.5519897460938,
      "state_relation_loss": 0.0003228939138352871,
      "step": 19180,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0015716118738055228,
      "weighted_relation_loss": 0.0015716118738055228
    },
    {
      "action_query_norm": 627.9135131835938,
      "action_relation_loss": 0.0012760934187099338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.951356407857812,
      "flow_matching_loss": 0.02159272599965334,
      "grad_norm": 0.11072307825088501,
      "learning_rate": 3.149644684039008e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015949584660120309,
      "selected_pairs": 26896.1,
      "state_query_norm": 1123.6715087890625,
      "state_relation_loss": 0.00031886505894362926,
      "step": 19190,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0015949584660120309,
      "weighted_relation_loss": 0.0015949584660120309
    },
    {
      "action_query_norm": 628.0957702636719,
      "action_relation_loss": 0.0013062686193734407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.96071094480823,
      "flow_matching_loss": 0.02284709382802248,
      "grad_norm": 0.10796141624450684,
      "learning_rate": 3.144525545376095e-05,
      "loss": 0.0245,
      "relation_loss": 0.0016184990294277669,
      "selected_pairs": 26225.4,
      "state_query_norm": 1124.6452880859374,
      "state_relation_loss": 0.0003122304129647091,
      "step": 19200,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0016184990294277669,
      "weighted_relation_loss": 0.0016184990294277669
    },
    {
      "action_query_norm": 630.6394470214843,
      "action_relation_loss": 0.0012125958106480538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.970065481758652,
      "flow_matching_loss": 0.021570261940360068,
      "grad_norm": 0.09624869376420975,
      "learning_rate": 3.139408661289603e-05,
      "loss": 0.0231,
      "relation_loss": 0.0015355895040556789,
      "selected_pairs": 26388.2,
      "state_query_norm": 1136.8486083984376,
      "state_relation_loss": 0.00032299367303494363,
      "step": 19210,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0015355895040556789,
      "weighted_relation_loss": 0.0015355895040556789
    },
    {
      "action_query_norm": 638.0399658203125,
      "action_relation_loss": 0.0012642773916013538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.979420018709074,
      "flow_matching_loss": 0.021863310225307943,
      "grad_norm": 0.12080192565917969,
      "learning_rate": 3.134294037997032e-05,
      "loss": 0.0235,
      "relation_loss": 0.0015936765703372657,
      "selected_pairs": 26577.9,
      "state_query_norm": 1149.42822265625,
      "state_relation_loss": 0.00032939919037744403,
      "step": 19220,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015936765703372657,
      "weighted_relation_loss": 0.0015936765703372657
    },
    {
      "action_query_norm": 632.7705261230469,
      "action_relation_loss": 0.0012959463289007544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.988774555659496,
      "flow_matching_loss": 0.021146102249622344,
      "grad_norm": 0.0938555970788002,
      "learning_rate": 3.129181681713127e-05,
      "loss": 0.0228,
      "relation_loss": 0.0016181347891688347,
      "selected_pairs": 27057.2,
      "state_query_norm": 1140.9257568359376,
      "state_relation_loss": 0.0003221884719096124,
      "step": 19230,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0016181347891688347,
      "weighted_relation_loss": 0.0016181347891688347
    },
    {
      "action_query_norm": 629.1029357910156,
      "action_relation_loss": 0.0012764018611051142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.998129092609915,
      "flow_matching_loss": 0.022641271445900203,
      "grad_norm": 0.11414209753274918,
      "learning_rate": 3.1240715986498855e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015936306212097407,
      "selected_pairs": 26733.5,
      "state_query_norm": 1143.8015014648438,
      "state_relation_loss": 0.0003172287601046264,
      "step": 19240,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0015936306212097407,
      "weighted_relation_loss": 0.0015936306212097407
    },
    {
      "action_query_norm": 637.3771728515625,
      "action_relation_loss": 0.0012330409605056047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.007483629560337,
      "flow_matching_loss": 0.023304557800292967,
      "grad_norm": 0.10760628432035446,
      "learning_rate": 3.1189637950165394e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015601511695422232,
      "selected_pairs": 24955.9,
      "state_query_norm": 1122.070574951172,
      "state_relation_loss": 0.0003271102177677676,
      "step": 19250,
      "valid_action_samples": 217.0,
      "weighted_auxiliary_loss": 0.0015601511695422232,
      "weighted_relation_loss": 0.0015601511695422232
    },
    {
      "action_query_norm": 639.857373046875,
      "action_relation_loss": 0.0012209205073304475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.01683816651076,
      "flow_matching_loss": 0.022632231563329698,
      "grad_norm": 0.11122654378414154,
      "learning_rate": 3.1138582770195544e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015447585727088153,
      "selected_pairs": 27433.0,
      "state_query_norm": 1135.2204956054688,
      "state_relation_loss": 0.00032383807410951704,
      "step": 19260,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0015447585727088153,
      "weighted_relation_loss": 0.0015447585727088153
    },
    {
      "action_query_norm": 638.2702941894531,
      "action_relation_loss": 0.0012264399032574148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.026192703461177,
      "flow_matching_loss": 0.021779891103506088,
      "grad_norm": 0.07758080214262009,
      "learning_rate": 3.108755050862615e-05,
      "loss": 0.0233,
      "relation_loss": 0.001550578442402184,
      "selected_pairs": 26895.0,
      "state_query_norm": 1136.7725341796875,
      "state_relation_loss": 0.00032413854205515236,
      "step": 19270,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001550578442402184,
      "weighted_relation_loss": 0.001550578442402184
    },
    {
      "action_query_norm": 643.00849609375,
      "action_relation_loss": 0.0012046454357914627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.0355472404116,
      "flow_matching_loss": 0.02318968316540122,
      "grad_norm": 0.10645339637994766,
      "learning_rate": 3.1036541227466204e-05,
      "loss": 0.0247,
      "relation_loss": 0.001521088380832225,
      "selected_pairs": 27379.5,
      "state_query_norm": 1137.3222290039062,
      "state_relation_loss": 0.0003164429508615285,
      "step": 19280,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.001521088380832225,
      "weighted_relation_loss": 0.001521088380832225
    },
    {
      "action_query_norm": 639.990234375,
      "action_relation_loss": 0.0012667673639953137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.04490177736202,
      "flow_matching_loss": 0.02217223271727562,
      "grad_norm": 0.10685662925243378,
      "learning_rate": 3.0985554988696786e-05,
      "loss": 0.0238,
      "relation_loss": 0.0015849412535317241,
      "selected_pairs": 26622.2,
      "state_query_norm": 1149.664990234375,
      "state_relation_loss": 0.0003181738720741123,
      "step": 19290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015849412535317241,
      "weighted_relation_loss": 0.0015849412535317241
    },
    {
      "action_query_norm": 639.3247192382812,
      "action_relation_loss": 0.0012106975773349404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.054256314312443,
      "flow_matching_loss": 0.02281879410147667,
      "grad_norm": 0.11573152989149094,
      "learning_rate": 3.093459185427102e-05,
      "loss": 0.0244,
      "relation_loss": 0.0015330334194004537,
      "selected_pairs": 27150.4,
      "state_query_norm": 1136.5026123046875,
      "state_relation_loss": 0.00032233584788627924,
      "step": 19300,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0015330334194004537,
      "weighted_relation_loss": 0.0015330334194004537
    },
    {
      "action_query_norm": 634.4232238769531,
      "action_relation_loss": 0.0012408075504936277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.06361085126286,
      "flow_matching_loss": 0.02046498879790306,
      "grad_norm": 0.09027345478534698,
      "learning_rate": 3.0883651886113905e-05,
      "loss": 0.022,
      "relation_loss": 0.0015596068697050214,
      "selected_pairs": 26582.5,
      "state_query_norm": 1128.2210571289063,
      "state_relation_loss": 0.0003187993250321597,
      "step": 19310,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015596068697050214,
      "weighted_relation_loss": 0.0015596068697050214
    },
    {
      "action_query_norm": 634.6795776367187,
      "action_relation_loss": 0.0012500535231083632,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.072965388213284,
      "flow_matching_loss": 0.022992359660565852,
      "grad_norm": 0.10501904785633087,
      "learning_rate": 3.0832735146122294e-05,
      "loss": 0.0246,
      "relation_loss": 0.001565950200892985,
      "selected_pairs": 26733.4,
      "state_query_norm": 1130.6203491210938,
      "state_relation_loss": 0.0003158966719638556,
      "step": 19320,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001565950200892985,
      "weighted_relation_loss": 0.001565950200892985
    },
    {
      "action_query_norm": 642.31240234375,
      "action_relation_loss": 0.0011949015315622091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.082319925163706,
      "flow_matching_loss": 0.021926391683518887,
      "grad_norm": 0.11189498752355576,
      "learning_rate": 3.078184169616485e-05,
      "loss": 0.0234,
      "relation_loss": 0.0015210111509077252,
      "selected_pairs": 27105.8,
      "state_query_norm": 1142.1665161132812,
      "state_relation_loss": 0.00032610961061436684,
      "step": 19330,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0015210111509077252,
      "weighted_relation_loss": 0.0015210111509077252
    },
    {
      "action_query_norm": 634.4827514648438,
      "action_relation_loss": 0.0012023834744468332,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.091674462114124,
      "flow_matching_loss": 0.02209103312343359,
      "grad_norm": 0.0936136245727539,
      "learning_rate": 3.073097159808187e-05,
      "loss": 0.0236,
      "relation_loss": 0.0015245602582581342,
      "selected_pairs": 26442.9,
      "state_query_norm": 1130.3449584960938,
      "state_relation_loss": 0.00032217679254245015,
      "step": 19340,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0015245602582581342,
      "weighted_relation_loss": 0.0015245602582581342
    },
    {
      "action_query_norm": 632.8249389648438,
      "action_relation_loss": 0.0011996744200587274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.101028999064546,
      "flow_matching_loss": 0.020645433850586415,
      "grad_norm": 0.11843989044427872,
      "learning_rate": 3.068012491368537e-05,
      "loss": 0.0221,
      "relation_loss": 0.001503692928235978,
      "selected_pairs": 26190.4,
      "state_query_norm": 1133.3878662109375,
      "state_relation_loss": 0.0003040185023564845,
      "step": 19350,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.001503692928235978,
      "weighted_relation_loss": 0.001503692928235978
    },
    {
      "action_query_norm": 629.302294921875,
      "action_relation_loss": 0.0012495843460783363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.110383536014968,
      "flow_matching_loss": 0.022700146958231925,
      "grad_norm": 0.09833608567714691,
      "learning_rate": 3.062930170475885e-05,
      "loss": 0.0243,
      "relation_loss": 0.001572243613190949,
      "selected_pairs": 26709.4,
      "state_query_norm": 1115.9734008789062,
      "state_relation_loss": 0.00032265925838146357,
      "step": 19360,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001572243613190949,
      "weighted_relation_loss": 0.001572243613190949
    },
    {
      "action_query_norm": 639.5602844238281,
      "action_relation_loss": 0.0012400570209138095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.119738072965387,
      "flow_matching_loss": 0.021807304956018923,
      "grad_norm": 0.10687651485204697,
      "learning_rate": 3.057850203305729e-05,
      "loss": 0.0234,
      "relation_loss": 0.0015695374342612923,
      "selected_pairs": 27129.6,
      "state_query_norm": 1148.1597412109375,
      "state_relation_loss": 0.00032948041916824877,
      "step": 19370,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0015695374342612923,
      "weighted_relation_loss": 0.0015695374342612923
    },
    {
      "action_query_norm": 636.56005859375,
      "action_relation_loss": 0.0011933499947190285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.12909260991581,
      "flow_matching_loss": 0.021381343342363836,
      "grad_norm": 0.09337104856967926,
      "learning_rate": 3.052772596030708e-05,
      "loss": 0.0229,
      "relation_loss": 0.001514202740509063,
      "selected_pairs": 26898.7,
      "state_query_norm": 1132.1424438476563,
      "state_relation_loss": 0.0003208527632523328,
      "step": 19380,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001514202740509063,
      "weighted_relation_loss": 0.001514202740509063
    },
    {
      "action_query_norm": 636.480859375,
      "action_relation_loss": 0.0012422869098372758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.13844714686623,
      "flow_matching_loss": 0.021333753876388074,
      "grad_norm": 0.11078885197639465,
      "learning_rate": 3.0476973548205943e-05,
      "loss": 0.0229,
      "relation_loss": 0.0015463766059838236,
      "selected_pairs": 26962.1,
      "state_query_norm": 1141.9823974609376,
      "state_relation_loss": 0.00030408969905693084,
      "step": 19390,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015463766059838236,
      "weighted_relation_loss": 0.0015463766059838236
    },
    {
      "action_query_norm": 634.2836303710938,
      "action_relation_loss": 0.001200259942561388,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.147801683816652,
      "flow_matching_loss": 0.023371141962707044,
      "grad_norm": 0.10536983609199524,
      "learning_rate": 3.0426244858422847e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015236057690344751,
      "selected_pairs": 26778.1,
      "state_query_norm": 1124.8944091796875,
      "state_relation_loss": 0.0003233458090107888,
      "step": 19400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0015236057690344751,
      "weighted_relation_loss": 0.0015236057690344751
    },
    {
      "action_query_norm": 636.4584655761719,
      "action_relation_loss": 0.0012724095256999135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.15715622076707,
      "flow_matching_loss": 0.022335292398929597,
      "grad_norm": 0.09280847758054733,
      "learning_rate": 3.0375539952597943e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015928125940263271,
      "selected_pairs": 26666.9,
      "state_query_norm": 1111.0182739257812,
      "state_relation_loss": 0.0003204030770575628,
      "step": 19410,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0015928125940263271,
      "weighted_relation_loss": 0.0015928125940263271
    },
    {
      "action_query_norm": 634.2632263183593,
      "action_relation_loss": 0.0011910666711628438,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.166510757717493,
      "flow_matching_loss": 0.020030139200389385,
      "grad_norm": 0.07334985584020615,
      "learning_rate": 3.0324858892342468e-05,
      "loss": 0.0215,
      "relation_loss": 0.0014993321849033237,
      "selected_pairs": 26387.5,
      "state_query_norm": 1140.8829711914063,
      "state_relation_loss": 0.000308265516650863,
      "step": 19420,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0014993321849033237,
      "weighted_relation_loss": 0.0014993321849033237
    },
    {
      "action_query_norm": 628.3131774902344,
      "action_relation_loss": 0.0012189348693937064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.175865294667915,
      "flow_matching_loss": 0.02224799208343029,
      "grad_norm": 0.08280611783266068,
      "learning_rate": 3.0274201739238672e-05,
      "loss": 0.0238,
      "relation_loss": 0.0015310564544051886,
      "selected_pairs": 26388.4,
      "state_query_norm": 1115.76201171875,
      "state_relation_loss": 0.00031212156172841785,
      "step": 19430,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0015310564544051886,
      "weighted_relation_loss": 0.0015310564544051886
    },
    {
      "action_query_norm": 634.0628845214844,
      "action_relation_loss": 0.0011806805268861353,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.185219831618333,
      "flow_matching_loss": 0.021200532093644142,
      "grad_norm": 0.10160750895738602,
      "learning_rate": 3.0223568554839786e-05,
      "loss": 0.0227,
      "relation_loss": 0.001505326631013304,
      "selected_pairs": 27230.7,
      "state_query_norm": 1119.0501953125,
      "state_relation_loss": 0.00032464612158946695,
      "step": 19440,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.001505326631013304,
      "weighted_relation_loss": 0.001505326631013304
    },
    {
      "action_query_norm": 633.5197265625,
      "action_relation_loss": 0.0012419568840414286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.194574368568755,
      "flow_matching_loss": 0.021153673902153967,
      "grad_norm": 0.09903837740421295,
      "learning_rate": 3.0172959400669886e-05,
      "loss": 0.0227,
      "relation_loss": 0.001547567523084581,
      "selected_pairs": 26812.1,
      "state_query_norm": 1136.732568359375,
      "state_relation_loss": 0.00030561065359506755,
      "step": 19450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001547567523084581,
      "weighted_relation_loss": 0.001547567523084581
    },
    {
      "action_query_norm": 631.7026000976563,
      "action_relation_loss": 0.0012120779603719712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.203928905519177,
      "flow_matching_loss": 0.020620682276785374,
      "grad_norm": 0.11187232285737991,
      "learning_rate": 3.0122374338223902e-05,
      "loss": 0.0222,
      "relation_loss": 0.0015386623446829617,
      "selected_pairs": 26732.2,
      "state_query_norm": 1124.9798706054687,
      "state_relation_loss": 0.0003265843726694584,
      "step": 19460,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0015386623446829617,
      "weighted_relation_loss": 0.0015386623446829617
    },
    {
      "action_query_norm": 634.4633850097656,
      "action_relation_loss": 0.0011731345555745065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.2132834424696,
      "flow_matching_loss": 0.021643600426614285,
      "grad_norm": 0.0913887768983841,
      "learning_rate": 3.0071813428967427e-05,
      "loss": 0.0231,
      "relation_loss": 0.0014827837352640928,
      "selected_pairs": 26601.3,
      "state_query_norm": 1126.258544921875,
      "state_relation_loss": 0.0003096491825999692,
      "step": 19470,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0014827837352640928,
      "weighted_relation_loss": 0.0014827837352640928
    },
    {
      "action_query_norm": 639.2334106445312,
      "action_relation_loss": 0.0011632475303485989,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.222637979420018,
      "flow_matching_loss": 0.0216231694445014,
      "grad_norm": 0.09504889696836472,
      "learning_rate": 3.0021276734336746e-05,
      "loss": 0.0231,
      "relation_loss": 0.001474043307825923,
      "selected_pairs": 26733.7,
      "state_query_norm": 1130.4030883789062,
      "state_relation_loss": 0.000310795777477324,
      "step": 19480,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001474043307825923,
      "weighted_relation_loss": 0.001474043307825923
    },
    {
      "action_query_norm": 632.987451171875,
      "action_relation_loss": 0.001215347880497575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.23199251637044,
      "flow_matching_loss": 0.01966708041727543,
      "grad_norm": 0.133004292845726,
      "learning_rate": 2.997076431573871e-05,
      "loss": 0.0212,
      "relation_loss": 0.0015227941097691656,
      "selected_pairs": 26814.7,
      "state_query_norm": 1118.5815185546876,
      "state_relation_loss": 0.00030744622054044155,
      "step": 19490,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015227941097691656,
      "weighted_relation_loss": 0.0015227941097691656
    },
    {
      "action_query_norm": 637.3541625976562,
      "action_relation_loss": 0.001180208777077496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.24134705332086,
      "flow_matching_loss": 0.021699908562004565,
      "grad_norm": 0.10867103189229965,
      "learning_rate": 2.9920276234550636e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015025568194687367,
      "selected_pairs": 26666.1,
      "state_query_norm": 1121.8193969726562,
      "state_relation_loss": 0.0003223480540327728,
      "step": 19500,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0015025568194687367,
      "weighted_relation_loss": 0.0015025568194687367
    },
    {
      "action_query_norm": 634.4597473144531,
      "action_relation_loss": 0.0012068597483448685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.25070159027128,
      "flow_matching_loss": 0.023080096952617167,
      "grad_norm": 0.0978148877620697,
      "learning_rate": 2.986981255212035e-05,
      "loss": 0.0246,
      "relation_loss": 0.001513727323617786,
      "selected_pairs": 26322.2,
      "state_query_norm": 1139.2144775390625,
      "state_relation_loss": 0.00030686756654176863,
      "step": 19510,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.001513727323617786,
      "weighted_relation_loss": 0.001513727323617786
    },
    {
      "action_query_norm": 632.4679626464844,
      "action_relation_loss": 0.0012245957157574594,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.260056127221702,
      "flow_matching_loss": 0.023261993564665316,
      "grad_norm": 0.09097201377153397,
      "learning_rate": 2.9819373329765977e-05,
      "loss": 0.0248,
      "relation_loss": 0.001539010473061353,
      "selected_pairs": 26897.4,
      "state_query_norm": 1122.4871948242187,
      "state_relation_loss": 0.00031441478640772404,
      "step": 19520,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001539010473061353,
      "weighted_relation_loss": 0.001539010473061353
    },
    {
      "action_query_norm": 635.2858459472657,
      "action_relation_loss": 0.001155325979925692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.269410664172124,
      "flow_matching_loss": 0.019802266731858253,
      "grad_norm": 0.0993773341178894,
      "learning_rate": 2.9768958628775902e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014783709193579853,
      "selected_pairs": 26965.3,
      "state_query_norm": 1130.8083740234374,
      "state_relation_loss": 0.0003230449452530593,
      "step": 19530,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014783709193579853,
      "weighted_relation_loss": 0.0014783709193579853
    },
    {
      "action_query_norm": 632.2027954101562,
      "action_relation_loss": 0.001229756511747837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.278765201122546,
      "flow_matching_loss": 0.020201225765049458,
      "grad_norm": 0.07930261641740799,
      "learning_rate": 2.9718568510408762e-05,
      "loss": 0.0218,
      "relation_loss": 0.0015567241120152176,
      "selected_pairs": 26598.2,
      "state_query_norm": 1122.0209838867188,
      "state_relation_loss": 0.00032696761772967877,
      "step": 19540,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0015567241120152176,
      "weighted_relation_loss": 0.0015567241120152176
    },
    {
      "action_query_norm": 627.680859375,
      "action_relation_loss": 0.0012273647589609027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.288119738072965,
      "flow_matching_loss": 0.02334344945847988,
      "grad_norm": 0.10122979432344437,
      "learning_rate": 2.9668203035893272e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015365654835477471,
      "selected_pairs": 27171.5,
      "state_query_norm": 1134.606103515625,
      "state_relation_loss": 0.0003092007304076105,
      "step": 19550,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0015365654835477471,
      "weighted_relation_loss": 0.0015365654835477471
    },
    {
      "action_query_norm": 629.101025390625,
      "action_relation_loss": 0.0011746215051971376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.297474275023387,
      "flow_matching_loss": 0.022114007174968718,
      "grad_norm": 0.09422712028026581,
      "learning_rate": 2.9617862266428287e-05,
      "loss": 0.0236,
      "relation_loss": 0.0014918530127033592,
      "selected_pairs": 26642.5,
      "state_query_norm": 1127.3372192382812,
      "state_relation_loss": 0.0003172315307892859,
      "step": 19560,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0014918530127033592,
      "weighted_relation_loss": 0.0014918530127033592
    },
    {
      "action_query_norm": 633.0647033691406,
      "action_relation_loss": 0.001243452972266823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.30682881197381,
      "flow_matching_loss": 0.02219969742000103,
      "grad_norm": 0.09882036596536636,
      "learning_rate": 2.9567546263182556e-05,
      "loss": 0.0238,
      "relation_loss": 0.001562151254620403,
      "selected_pairs": 27784.2,
      "state_query_norm": 1125.8523315429688,
      "state_relation_loss": 0.00031869829108472915,
      "step": 19570,
      "valid_action_samples": 236.2,
      "weighted_auxiliary_loss": 0.001562151254620403,
      "weighted_relation_loss": 0.001562151254620403
    },
    {
      "action_query_norm": 633.018017578125,
      "action_relation_loss": 0.0012188700959086419,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.316183348924227,
      "flow_matching_loss": 0.023222401179373263,
      "grad_norm": 0.11857938766479492,
      "learning_rate": 2.951725508729476e-05,
      "loss": 0.0248,
      "relation_loss": 0.0015305489068850875,
      "selected_pairs": 26463.2,
      "state_query_norm": 1110.8998046875,
      "state_relation_loss": 0.0003116788109764457,
      "step": 19580,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0015305489068850875,
      "weighted_relation_loss": 0.0015305489068850875
    },
    {
      "action_query_norm": 629.4285339355469,
      "action_relation_loss": 0.001276136247906834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.32553788587465,
      "flow_matching_loss": 0.02210985906422138,
      "grad_norm": 0.09812229871749878,
      "learning_rate": 2.946698879987344e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015901520266197622,
      "selected_pairs": 26761.1,
      "state_query_norm": 1135.72412109375,
      "state_relation_loss": 0.00031401577580254526,
      "step": 19590,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0015901520266197622,
      "weighted_relation_loss": 0.0015901520266197622
    },
    {
      "action_query_norm": 643.4744873046875,
      "action_relation_loss": 0.0012276462628506124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.33489242282507,
      "flow_matching_loss": 0.02217540070414543,
      "grad_norm": 0.10231472551822662,
      "learning_rate": 2.9416747461996853e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015535410842858256,
      "selected_pairs": 26625.5,
      "state_query_norm": 1140.7831420898438,
      "state_relation_loss": 0.0003258948214352131,
      "step": 19600,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015535410842858256,
      "weighted_relation_loss": 0.0015535410842858256
    },
    {
      "action_query_norm": 632.9771667480469,
      "action_relation_loss": 0.001251848496031016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.34424695977549,
      "flow_matching_loss": 0.020659351721405983,
      "grad_norm": 0.1125338077545166,
      "learning_rate": 2.9366531134712973e-05,
      "loss": 0.0222,
      "relation_loss": 0.0015719852526672184,
      "selected_pairs": 26521.6,
      "state_query_norm": 1142.4040771484374,
      "state_relation_loss": 0.0003201367537258193,
      "step": 19610,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0015719852526672184,
      "weighted_relation_loss": 0.0015719852526672184
    },
    {
      "action_query_norm": 635.9981994628906,
      "action_relation_loss": 0.001298286102246493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.35360149672591,
      "flow_matching_loss": 0.020213166251778603,
      "grad_norm": 0.10077274590730667,
      "learning_rate": 2.9316339879039367e-05,
      "loss": 0.0218,
      "relation_loss": 0.0016088150907307864,
      "selected_pairs": 26971.1,
      "state_query_norm": 1128.7410400390625,
      "state_relation_loss": 0.00031052898557391016,
      "step": 19620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016088150907307864,
      "weighted_relation_loss": 0.0016088150907307864
    },
    {
      "action_query_norm": 628.6877563476562,
      "action_relation_loss": 0.0012558770831674337,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.362956033676333,
      "flow_matching_loss": 0.02086911555379629,
      "grad_norm": 0.10100391507148743,
      "learning_rate": 2.9266173755963167e-05,
      "loss": 0.0224,
      "relation_loss": 0.0015674544963985681,
      "selected_pairs": 27152.3,
      "state_query_norm": 1128.700634765625,
      "state_relation_loss": 0.0003115774306934327,
      "step": 19630,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0015674544963985681,
      "weighted_relation_loss": 0.0015674544963985681
    },
    {
      "action_query_norm": 627.7573425292969,
      "action_relation_loss": 0.0011955924797803163,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.372310570626755,
      "flow_matching_loss": 0.0204897397197783,
      "grad_norm": 0.09181862324476242,
      "learning_rate": 2.9216032826440926e-05,
      "loss": 0.022,
      "relation_loss": 0.0015085338731296361,
      "selected_pairs": 26963.2,
      "state_query_norm": 1136.5634765625,
      "state_relation_loss": 0.0003129413875285536,
      "step": 19640,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015085338731296361,
      "weighted_relation_loss": 0.0015085338731296361
    },
    {
      "action_query_norm": 639.0361267089844,
      "action_relation_loss": 0.0011745434952899815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.381665107577174,
      "flow_matching_loss": 0.020557422190904617,
      "grad_norm": 0.13008077442646027,
      "learning_rate": 2.9165917151398592e-05,
      "loss": 0.022,
      "relation_loss": 0.001490843656938523,
      "selected_pairs": 26805.9,
      "state_query_norm": 1130.71005859375,
      "state_relation_loss": 0.0003163001878419891,
      "step": 19650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001490843656938523,
      "weighted_relation_loss": 0.001490843656938523
    },
    {
      "action_query_norm": 639.5439208984375,
      "action_relation_loss": 0.001149257260840386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.391019644527596,
      "flow_matching_loss": 0.020790945179760455,
      "grad_norm": 0.08292422443628311,
      "learning_rate": 2.9115826791731428e-05,
      "loss": 0.0223,
      "relation_loss": 0.0014679871732369066,
      "selected_pairs": 27093.6,
      "state_query_norm": 1138.0520141601562,
      "state_relation_loss": 0.00031872989784460516,
      "step": 19660,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0014679871732369066,
      "weighted_relation_loss": 0.0014679871732369066
    },
    {
      "action_query_norm": 635.0820373535156,
      "action_relation_loss": 0.0012588799116201698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.400374181478018,
      "flow_matching_loss": 0.020881429314613342,
      "grad_norm": 0.10472220927476883,
      "learning_rate": 2.906576180830398e-05,
      "loss": 0.0225,
      "relation_loss": 0.0015808458789251744,
      "selected_pairs": 26831.8,
      "state_query_norm": 1146.7131591796874,
      "state_relation_loss": 0.0003219659731257707,
      "step": 19670,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0015808458789251744,
      "weighted_relation_loss": 0.0015808458789251744
    },
    {
      "action_query_norm": 632.9177429199219,
      "action_relation_loss": 0.0012066533090546727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.409728718428436,
      "flow_matching_loss": 0.019721673149615527,
      "grad_norm": 0.1022087037563324,
      "learning_rate": 2.9015722261949917e-05,
      "loss": 0.0212,
      "relation_loss": 0.0015260784653946756,
      "selected_pairs": 26712.5,
      "state_query_norm": 1112.0433837890625,
      "state_relation_loss": 0.0003194251738023013,
      "step": 19680,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0015260784653946756,
      "weighted_relation_loss": 0.0015260784653946756
    },
    {
      "action_query_norm": 631.3795104980469,
      "action_relation_loss": 0.0012060185079462826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.419083255378858,
      "flow_matching_loss": 0.02222275510430336,
      "grad_norm": 0.08704183995723724,
      "learning_rate": 2.8965708213471986e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015153975458815695,
      "selected_pairs": 26984.7,
      "state_query_norm": 1120.3848754882813,
      "state_relation_loss": 0.0003093790350249037,
      "step": 19690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0015153975458815695,
      "weighted_relation_loss": 0.0015153975458815695
    },
    {
      "action_query_norm": 634.0754211425781,
      "action_relation_loss": 0.0012015319196507334,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.42843779232928,
      "flow_matching_loss": 0.021170837432146074,
      "grad_norm": 0.10379839688539505,
      "learning_rate": 2.8915719723641975e-05,
      "loss": 0.0227,
      "relation_loss": 0.0015185988042503596,
      "selected_pairs": 26802.2,
      "state_query_norm": 1120.59189453125,
      "state_relation_loss": 0.00031706687877886,
      "step": 19700,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015185988042503596,
      "weighted_relation_loss": 0.0015185988042503596
    },
    {
      "action_query_norm": 634.9016662597656,
      "action_relation_loss": 0.0012330358964391052,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.437792329279702,
      "flow_matching_loss": 0.021260248403996228,
      "grad_norm": 0.1020195260643959,
      "learning_rate": 2.8865756853200604e-05,
      "loss": 0.0228,
      "relation_loss": 0.0015329045359976589,
      "selected_pairs": 26972.9,
      "state_query_norm": 1144.4761840820313,
      "state_relation_loss": 0.0002998686337377876,
      "step": 19710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015329045359976589,
      "weighted_relation_loss": 0.0015329045359976589
    },
    {
      "action_query_norm": 632.7293029785156,
      "action_relation_loss": 0.0012082353001460432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.44714686623012,
      "flow_matching_loss": 0.02025108812376857,
      "grad_norm": 0.11554710566997528,
      "learning_rate": 2.8815819662857503e-05,
      "loss": 0.0218,
      "relation_loss": 0.0015188616118393839,
      "selected_pairs": 26343.1,
      "state_query_norm": 1143.4966064453124,
      "state_relation_loss": 0.0003106263233348727,
      "step": 19720,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0015188616118393839,
      "weighted_relation_loss": 0.0015188616118393839
    },
    {
      "action_query_norm": 631.5139404296875,
      "action_relation_loss": 0.0012535112095065416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.456501403180543,
      "flow_matching_loss": 0.022823341004550458,
      "grad_norm": 0.08786405622959137,
      "learning_rate": 2.876590821329105e-05,
      "loss": 0.0244,
      "relation_loss": 0.0015631902730092405,
      "selected_pairs": 27195.8,
      "state_query_norm": 1145.0889892578125,
      "state_relation_loss": 0.00030967906932346525,
      "step": 19730,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0015631902730092405,
      "weighted_relation_loss": 0.0015631902730092405
    },
    {
      "action_query_norm": 639.7924743652344,
      "action_relation_loss": 0.0012030486715957523,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.465855940130965,
      "flow_matching_loss": 0.02181006297469139,
      "grad_norm": 0.11313784867525101,
      "learning_rate": 2.871602256514836e-05,
      "loss": 0.0233,
      "relation_loss": 0.0015155172208324076,
      "selected_pairs": 26966.2,
      "state_query_norm": 1131.9734130859374,
      "state_relation_loss": 0.0003124685579678044,
      "step": 19740,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015155172208324076,
      "weighted_relation_loss": 0.0015155172208324076
    },
    {
      "action_query_norm": 633.7749816894532,
      "action_relation_loss": 0.0012179246521554887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.475210477081383,
      "flow_matching_loss": 0.020011895708739756,
      "grad_norm": 0.08500000834465027,
      "learning_rate": 2.8666162779045203e-05,
      "loss": 0.0215,
      "relation_loss": 0.0015264852438122035,
      "selected_pairs": 27036.5,
      "state_query_norm": 1134.5685180664063,
      "state_relation_loss": 0.0003085605887463316,
      "step": 19750,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0015264852438122035,
      "weighted_relation_loss": 0.0015264852438122035
    },
    {
      "action_query_norm": 627.94892578125,
      "action_relation_loss": 0.001196602056734264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.484565014031805,
      "flow_matching_loss": 0.02125694938004017,
      "grad_norm": 0.09157174825668335,
      "learning_rate": 2.8616328915565904e-05,
      "loss": 0.0228,
      "relation_loss": 0.0014959124499000608,
      "selected_pairs": 27267.6,
      "state_query_norm": 1124.9412475585937,
      "state_relation_loss": 0.00029931039025541393,
      "step": 19760,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0014959124499000608,
      "weighted_relation_loss": 0.0014959124499000608
    },
    {
      "action_query_norm": 637.9133911132812,
      "action_relation_loss": 0.0011738430242985488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.493919550982227,
      "flow_matching_loss": 0.0226490942761302,
      "grad_norm": 0.09629174321889877,
      "learning_rate": 2.856652103526334e-05,
      "loss": 0.0241,
      "relation_loss": 0.0014810628141276538,
      "selected_pairs": 26648.7,
      "state_query_norm": 1129.5236206054688,
      "state_relation_loss": 0.0003072197752771899,
      "step": 19770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0014810628141276538,
      "weighted_relation_loss": 0.0014810628141276538
    },
    {
      "action_query_norm": 633.4025390625,
      "action_relation_loss": 0.0011303985142149032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.50327408793265,
      "flow_matching_loss": 0.020598868653178216,
      "grad_norm": 0.09526895731687546,
      "learning_rate": 2.8516739198658755e-05,
      "loss": 0.022,
      "relation_loss": 0.0014413867844268679,
      "selected_pairs": 27010.2,
      "state_query_norm": 1128.0334594726562,
      "state_relation_loss": 0.0003109882673015818,
      "step": 19780,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014413867844268679,
      "weighted_relation_loss": 0.0014413867844268679
    },
    {
      "action_query_norm": 642.8267517089844,
      "action_relation_loss": 0.0012493145070038736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.512628624883067,
      "flow_matching_loss": 0.020290403813123702,
      "grad_norm": 0.08467172086238861,
      "learning_rate": 2.846698346624177e-05,
      "loss": 0.0219,
      "relation_loss": 0.0015683627920225264,
      "selected_pairs": 27197.8,
      "state_query_norm": 1122.7929565429688,
      "state_relation_loss": 0.0003190482791978866,
      "step": 19790,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0015683627920225264,
      "weighted_relation_loss": 0.0015683627920225264
    },
    {
      "action_query_norm": 630.5041137695313,
      "action_relation_loss": 0.0012093650875613093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.52198316183349,
      "flow_matching_loss": 0.01949886754155159,
      "grad_norm": 0.1028248742222786,
      "learning_rate": 2.841725389847032e-05,
      "loss": 0.021,
      "relation_loss": 0.0015260460902936756,
      "selected_pairs": 27264.6,
      "state_query_norm": 1117.3065307617187,
      "state_relation_loss": 0.0003166809969116002,
      "step": 19800,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0015260460902936756,
      "weighted_relation_loss": 0.0015260460902936756
    },
    {
      "action_query_norm": 638.0294799804688,
      "action_relation_loss": 0.0011361546348780394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.53133769878391,
      "flow_matching_loss": 0.020677806064486503,
      "grad_norm": 0.06614850461483002,
      "learning_rate": 2.8367550555770506e-05,
      "loss": 0.0221,
      "relation_loss": 0.0014502011938020587,
      "selected_pairs": 26622.3,
      "state_query_norm": 1139.1294677734375,
      "state_relation_loss": 0.0003140465501928702,
      "step": 19810,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0014502011938020587,
      "weighted_relation_loss": 0.0014502011938020587
    },
    {
      "action_query_norm": 641.5960754394531,
      "action_relation_loss": 0.0011180808884091674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.54069223573433,
      "flow_matching_loss": 0.020747210830450058,
      "grad_norm": 0.08567024022340775,
      "learning_rate": 2.831787349853655e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014349176664836705,
      "selected_pairs": 27105.9,
      "state_query_norm": 1128.2625122070312,
      "state_relation_loss": 0.00031683676061220465,
      "step": 19820,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0014349176664836705,
      "weighted_relation_loss": 0.0014349176664836705
    },
    {
      "action_query_norm": 625.7484130859375,
      "action_relation_loss": 0.0011947737890295683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.55004677268475,
      "flow_matching_loss": 0.020558080449700356,
      "grad_norm": 0.08737114816904068,
      "learning_rate": 2.8268222787130806e-05,
      "loss": 0.0221,
      "relation_loss": 0.001518147101160139,
      "selected_pairs": 27079.4,
      "state_query_norm": 1121.7944091796876,
      "state_relation_loss": 0.0003233733208617195,
      "step": 19830,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001518147101160139,
      "weighted_relation_loss": 0.001518147101160139
    },
    {
      "action_query_norm": 635.0270202636718,
      "action_relation_loss": 0.0011997816618531942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.559401309635174,
      "flow_matching_loss": 0.022222376056015493,
      "grad_norm": 0.10075189918279648,
      "learning_rate": 2.821859848188355e-05,
      "loss": 0.0237,
      "relation_loss": 0.001510370383039117,
      "selected_pairs": 26827.1,
      "state_query_norm": 1141.3253295898437,
      "state_relation_loss": 0.0003105887124547735,
      "step": 19840,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001510370383039117,
      "weighted_relation_loss": 0.001510370383039117
    },
    {
      "action_query_norm": 631.9247497558594,
      "action_relation_loss": 0.0012073151534423231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.568755846585596,
      "flow_matching_loss": 0.0215188167989254,
      "grad_norm": 0.10878598690032959,
      "learning_rate": 2.816900064309299e-05,
      "loss": 0.023,
      "relation_loss": 0.001505588658619672,
      "selected_pairs": 26459.3,
      "state_query_norm": 1124.9032104492187,
      "state_relation_loss": 0.00029827351099811497,
      "step": 19850,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001505588658619672,
      "weighted_relation_loss": 0.001505588658619672
    },
    {
      "action_query_norm": 635.0316040039063,
      "action_relation_loss": 0.001226972439326346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.578110383536014,
      "flow_matching_loss": 0.023392818868160248,
      "grad_norm": 0.08945124596357346,
      "learning_rate": 2.811942933102517e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015398286865092813,
      "selected_pairs": 26871.9,
      "state_query_norm": 1122.5412475585938,
      "state_relation_loss": 0.0003128562471829355,
      "step": 19860,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015398286865092813,
      "weighted_relation_loss": 0.0015398286865092813
    },
    {
      "action_query_norm": 631.9350708007812,
      "action_relation_loss": 0.0011700758477672935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.587464920486436,
      "flow_matching_loss": 0.020199263095855714,
      "grad_norm": 0.09901782870292664,
      "learning_rate": 2.806988460591391e-05,
      "loss": 0.0217,
      "relation_loss": 0.0014855825575068593,
      "selected_pairs": 26916.6,
      "state_query_norm": 1136.7815185546874,
      "state_relation_loss": 0.0003155066951876506,
      "step": 19870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0014855825575068593,
      "weighted_relation_loss": 0.0014855825575068593
    },
    {
      "action_query_norm": 641.5282592773438,
      "action_relation_loss": 0.0011929864645935594,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.596819457436858,
      "flow_matching_loss": 0.02071949541568756,
      "grad_norm": 0.08814449608325958,
      "learning_rate": 2.802036652796074e-05,
      "loss": 0.0222,
      "relation_loss": 0.001507240778300911,
      "selected_pairs": 27410.0,
      "state_query_norm": 1123.0183227539062,
      "state_relation_loss": 0.00031425429915543646,
      "step": 19880,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001507240778300911,
      "weighted_relation_loss": 0.001507240778300911
    },
    {
      "action_query_norm": 632.8529541015625,
      "action_relation_loss": 0.001267061405815184,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.606173994387277,
      "flow_matching_loss": 0.02048409152776003,
      "grad_norm": 0.10651124268770218,
      "learning_rate": 2.797087515733478e-05,
      "loss": 0.0221,
      "relation_loss": 0.0015770675148814917,
      "selected_pairs": 26227.0,
      "state_query_norm": 1134.2290649414062,
      "state_relation_loss": 0.00031000610324554144,
      "step": 19890,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0015770675148814917,
      "weighted_relation_loss": 0.0015770675148814917
    },
    {
      "action_query_norm": 636.7598999023437,
      "action_relation_loss": 0.0011937582050450147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.6155285313377,
      "flow_matching_loss": 0.0214330380782485,
      "grad_norm": 0.08901827037334442,
      "learning_rate": 2.7921410554172723e-05,
      "loss": 0.0229,
      "relation_loss": 0.001489180966746062,
      "selected_pairs": 27180.8,
      "state_query_norm": 1138.4529541015625,
      "state_relation_loss": 0.00029542275005951525,
      "step": 19900,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001489180966746062,
      "weighted_relation_loss": 0.001489180966746062
    },
    {
      "action_query_norm": 633.3110229492188,
      "action_relation_loss": 0.0011797687970101834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.62488306828812,
      "flow_matching_loss": 0.022345317434519528,
      "grad_norm": 0.11161239445209503,
      "learning_rate": 2.787197277857871e-05,
      "loss": 0.0238,
      "relation_loss": 0.0014908361830748619,
      "selected_pairs": 26676.2,
      "state_query_norm": 1138.3958740234375,
      "state_relation_loss": 0.00031106736860238017,
      "step": 19910,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0014908361830748619,
      "weighted_relation_loss": 0.0014908361830748619
    },
    {
      "action_query_norm": 630.8449523925781,
      "action_relation_loss": 0.0012107212212868034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.63423760523854,
      "flow_matching_loss": 0.021517649851739408,
      "grad_norm": 0.09016955643892288,
      "learning_rate": 2.782256189062429e-05,
      "loss": 0.023,
      "relation_loss": 0.0015173735446296632,
      "selected_pairs": 26990.8,
      "state_query_norm": 1138.4715698242187,
      "state_relation_loss": 0.0003066523437155411,
      "step": 19920,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0015173735446296632,
      "weighted_relation_loss": 0.0015173735446296632
    },
    {
      "action_query_norm": 631.6648376464843,
      "action_relation_loss": 0.0012019184185191989,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.64359214218896,
      "flow_matching_loss": 0.020477302744984625,
      "grad_norm": 0.09741230309009552,
      "learning_rate": 2.777317795034839e-05,
      "loss": 0.022,
      "relation_loss": 0.0015180552378296851,
      "selected_pairs": 27018.5,
      "state_query_norm": 1133.821875,
      "state_relation_loss": 0.0003161368164001033,
      "step": 19930,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0015180552378296851,
      "weighted_relation_loss": 0.0015180552378296851
    },
    {
      "action_query_norm": 636.77626953125,
      "action_relation_loss": 0.0011627911357209087,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.652946679139383,
      "flow_matching_loss": 0.02066122367978096,
      "grad_norm": 0.10898862779140472,
      "learning_rate": 2.7723821017757112e-05,
      "loss": 0.0221,
      "relation_loss": 0.0014787625172175468,
      "selected_pairs": 27455.1,
      "state_query_norm": 1134.1598754882812,
      "state_relation_loss": 0.00031597136694472283,
      "step": 19940,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0014787625172175468,
      "weighted_relation_loss": 0.0014787625172175468
    },
    {
      "action_query_norm": 642.4442077636719,
      "action_relation_loss": 0.0011262356769293548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.662301216089805,
      "flow_matching_loss": 0.02155925501137972,
      "grad_norm": 0.11588132381439209,
      "learning_rate": 2.7674491152823822e-05,
      "loss": 0.023,
      "relation_loss": 0.0014348519500344992,
      "selected_pairs": 26758.1,
      "state_query_norm": 1127.8272705078125,
      "state_relation_loss": 0.0003086162818362936,
      "step": 19950,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0014348519500344992,
      "weighted_relation_loss": 0.0014348519500344992
    },
    {
      "action_query_norm": 635.6782287597656,
      "action_relation_loss": 0.0011805011425167322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.671655753040223,
      "flow_matching_loss": 0.020409927517175675,
      "grad_norm": 0.12383143603801727,
      "learning_rate": 2.7625188415488944e-05,
      "loss": 0.0219,
      "relation_loss": 0.0014870663871988654,
      "selected_pairs": 27031.6,
      "state_query_norm": 1128.106298828125,
      "state_relation_loss": 0.0003065652388613671,
      "step": 19960,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0014870663871988654,
      "weighted_relation_loss": 0.0014870663871988654
    },
    {
      "action_query_norm": 640.4021545410156,
      "action_relation_loss": 0.001202896755421534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.681010289990645,
      "flow_matching_loss": 0.02121133226901293,
      "grad_norm": 0.08691024035215378,
      "learning_rate": 2.7575912865659924e-05,
      "loss": 0.0227,
      "relation_loss": 0.0015089705120772123,
      "selected_pairs": 26490.1,
      "state_query_norm": 1139.855517578125,
      "state_relation_loss": 0.00030607375665567813,
      "step": 19970,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0015089705120772123,
      "weighted_relation_loss": 0.0015089705120772123
    },
    {
      "action_query_norm": 640.1147399902344,
      "action_relation_loss": 0.0012170791625976563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.690364826941067,
      "flow_matching_loss": 0.022130353003740312,
      "grad_norm": 0.09692387282848358,
      "learning_rate": 2.7526664563211245e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015236729173921048,
      "selected_pairs": 26576.1,
      "state_query_norm": 1122.4046142578125,
      "state_relation_loss": 0.00030659375188406556,
      "step": 19980,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015236729173921048,
      "weighted_relation_loss": 0.0015236729173921048
    },
    {
      "action_query_norm": 629.0135498046875,
      "action_relation_loss": 0.0012074603349901737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.699719363891486,
      "flow_matching_loss": 0.021846463344991206,
      "grad_norm": 0.08692856132984161,
      "learning_rate": 2.7477443567984224e-05,
      "loss": 0.0234,
      "relation_loss": 0.0015178341302089394,
      "selected_pairs": 26978.5,
      "state_query_norm": 1131.0747802734375,
      "state_relation_loss": 0.0003103737806668505,
      "step": 19990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015178341302089394,
      "weighted_relation_loss": 0.0015178341302089394
    },
    {
      "action_query_norm": 633.3221923828125,
      "action_relation_loss": 0.0011907969717867672,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.709073900841908,
      "flow_matching_loss": 0.020510172471404076,
      "grad_norm": 0.08628813177347183,
      "learning_rate": 2.7428249939787e-05,
      "loss": 0.022,
      "relation_loss": 0.0015099531272426247,
      "selected_pairs": 27381.1,
      "state_query_norm": 1117.0392211914063,
      "state_relation_loss": 0.0003191561350831762,
      "step": 20000,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0015099531272426247,
      "weighted_relation_loss": 0.0015099531272426247
    },
    {
      "action_query_norm": 631.7862915039062,
      "action_relation_loss": 0.0012017335509881377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.71842843779233,
      "flow_matching_loss": 0.021350820921361445,
      "grad_norm": 0.11465238779783249,
      "learning_rate": 2.7379083738394484e-05,
      "loss": 0.0229,
      "relation_loss": 0.0015062411548569798,
      "selected_pairs": 27512.7,
      "state_query_norm": 1119.8834228515625,
      "state_relation_loss": 0.00030450758640654385,
      "step": 20010,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0015062411548569798,
      "weighted_relation_loss": 0.0015062411548569798
    },
    {
      "action_query_norm": 623.1393127441406,
      "action_relation_loss": 0.001173031225334853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.72778297474275,
      "flow_matching_loss": 0.021774460095912218,
      "grad_norm": 0.1148027777671814,
      "learning_rate": 2.732994502354823e-05,
      "loss": 0.0233,
      "relation_loss": 0.001485804351978004,
      "selected_pairs": 26662.0,
      "state_query_norm": 1130.92177734375,
      "state_relation_loss": 0.0003127731411950663,
      "step": 20020,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001485804351978004,
      "weighted_relation_loss": 0.001485804351978004
    },
    {
      "action_query_norm": 628.372802734375,
      "action_relation_loss": 0.001172443211544305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.73713751169317,
      "flow_matching_loss": 0.021254471503198147,
      "grad_norm": 0.089840367436409,
      "learning_rate": 2.72808338549564e-05,
      "loss": 0.0227,
      "relation_loss": 0.0014904482290148736,
      "selected_pairs": 26573.9,
      "state_query_norm": 1120.8422607421876,
      "state_relation_loss": 0.0003180050320224836,
      "step": 20030,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0014904482290148736,
      "weighted_relation_loss": 0.0014904482290148736
    },
    {
      "action_query_norm": 629.2567260742187,
      "action_relation_loss": 0.0012359129264950751,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.746492048643592,
      "flow_matching_loss": 0.020372072607278822,
      "grad_norm": 0.11242366582155228,
      "learning_rate": 2.723175029229374e-05,
      "loss": 0.0219,
      "relation_loss": 0.0015438289963640272,
      "selected_pairs": 26844.8,
      "state_query_norm": 1129.53974609375,
      "state_relation_loss": 0.0003079160553170368,
      "step": 20040,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0015438289963640272,
      "weighted_relation_loss": 0.0015438289963640272
    },
    {
      "action_query_norm": 630.5152038574219,
      "action_relation_loss": 0.0012456097174435854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.755846585594014,
      "flow_matching_loss": 0.020227960124611855,
      "grad_norm": 0.1032010093331337,
      "learning_rate": 2.718269439520138e-05,
      "loss": 0.0218,
      "relation_loss": 0.0015538239502348005,
      "selected_pairs": 26881.7,
      "state_query_norm": 1140.9618774414062,
      "state_relation_loss": 0.0003082142095081508,
      "step": 20050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015538239502348005,
      "weighted_relation_loss": 0.0015538239502348005
    },
    {
      "action_query_norm": 632.7030395507812,
      "action_relation_loss": 0.001148641644977033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.765201122544433,
      "flow_matching_loss": 0.020455983094871046,
      "grad_norm": 0.08414849638938904,
      "learning_rate": 2.7133666223286858e-05,
      "loss": 0.0219,
      "relation_loss": 0.001456179830711335,
      "selected_pairs": 26004.6,
      "state_query_norm": 1126.7532592773437,
      "state_relation_loss": 0.0003075381857343018,
      "step": 20060,
      "valid_action_samples": 228.5,
      "weighted_auxiliary_loss": 0.001456179830711335,
      "weighted_relation_loss": 0.001456179830711335
    },
    {
      "action_query_norm": 628.3820434570313,
      "action_relation_loss": 0.0012013538274914027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.774555659494855,
      "flow_matching_loss": 0.01992686791345477,
      "grad_norm": 0.1129656508564949,
      "learning_rate": 2.7084665836124006e-05,
      "loss": 0.0214,
      "relation_loss": 0.0015040886355563998,
      "selected_pairs": 27195.3,
      "state_query_norm": 1127.215478515625,
      "state_relation_loss": 0.0003027348167961463,
      "step": 20070,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0015040886355563998,
      "weighted_relation_loss": 0.0015040886355563998
    },
    {
      "action_query_norm": 634.6457824707031,
      "action_relation_loss": 0.0011268702917732298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.783910196445277,
      "flow_matching_loss": 0.020692129246890546,
      "grad_norm": 0.1194310188293457,
      "learning_rate": 2.703569329325296e-05,
      "loss": 0.0221,
      "relation_loss": 0.0014443284366279841,
      "selected_pairs": 27321.4,
      "state_query_norm": 1116.3929809570313,
      "state_relation_loss": 0.00031745814194437114,
      "step": 20080,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0014443284366279841,
      "weighted_relation_loss": 0.0014443284366279841
    },
    {
      "action_query_norm": 637.84599609375,
      "action_relation_loss": 0.0011858975049108267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.7932647333957,
      "flow_matching_loss": 0.019599780440330505,
      "grad_norm": 0.10763972997665405,
      "learning_rate": 2.698674865417994e-05,
      "loss": 0.0211,
      "relation_loss": 0.001495234400499612,
      "selected_pairs": 27007.5,
      "state_query_norm": 1115.7889404296875,
      "state_relation_loss": 0.00030933688394725324,
      "step": 20090,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001495234400499612,
      "weighted_relation_loss": 0.001495234400499612
    },
    {
      "action_query_norm": 635.859912109375,
      "action_relation_loss": 0.0011459561646915972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.802619270346117,
      "flow_matching_loss": 0.021247290633618833,
      "grad_norm": 0.11292105913162231,
      "learning_rate": 2.6937831978377325e-05,
      "loss": 0.0227,
      "relation_loss": 0.0014580016140826046,
      "selected_pairs": 26550.1,
      "state_query_norm": 1124.6157836914062,
      "state_relation_loss": 0.0003120454581221566,
      "step": 20100,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0014580016140826046,
      "weighted_relation_loss": 0.0014580016140826046
    },
    {
      "action_query_norm": 625.06220703125,
      "action_relation_loss": 0.0012278985348530113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.81197380729654,
      "flow_matching_loss": 0.021314029954373835,
      "grad_norm": 0.09633142501115799,
      "learning_rate": 2.6888943325283485e-05,
      "loss": 0.0228,
      "relation_loss": 0.0015353384660556912,
      "selected_pairs": 26858.8,
      "state_query_norm": 1121.8252807617187,
      "state_relation_loss": 0.000307439939933829,
      "step": 20110,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0015353384660556912,
      "weighted_relation_loss": 0.0015353384660556912
    },
    {
      "action_query_norm": 626.4639282226562,
      "action_relation_loss": 0.0011732840328477323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.82132834424696,
      "flow_matching_loss": 0.022116219624876976,
      "grad_norm": 0.09212563931941986,
      "learning_rate": 2.6840082754302733e-05,
      "loss": 0.0236,
      "relation_loss": 0.0014706305461004376,
      "selected_pairs": 26878.1,
      "state_query_norm": 1128.437548828125,
      "state_relation_loss": 0.0002973465045215562,
      "step": 20120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0014706305461004376,
      "weighted_relation_loss": 0.0014706305461004376
    },
    {
      "action_query_norm": 628.3197021484375,
      "action_relation_loss": 0.0011970512219704687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.83068288119738,
      "flow_matching_loss": 0.020805630087852477,
      "grad_norm": 0.08947515487670898,
      "learning_rate": 2.679125032480525e-05,
      "loss": 0.0223,
      "relation_loss": 0.0014960602624341846,
      "selected_pairs": 26710.1,
      "state_query_norm": 1120.66044921875,
      "state_relation_loss": 0.00029900903173256664,
      "step": 20130,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0014960602624341846,
      "weighted_relation_loss": 0.0014960602624341846
    },
    {
      "action_query_norm": 632.9255493164062,
      "action_relation_loss": 0.0011066945735365152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.8400374181478,
      "flow_matching_loss": 0.021030708216130733,
      "grad_norm": 0.08278564363718033,
      "learning_rate": 2.674244609612708e-05,
      "loss": 0.0225,
      "relation_loss": 0.001419767364859581,
      "selected_pairs": 27152.8,
      "state_query_norm": 1121.525048828125,
      "state_relation_loss": 0.00031307278259191663,
      "step": 20140,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001419767364859581,
      "weighted_relation_loss": 0.001419767364859581
    },
    {
      "action_query_norm": 634.4744201660156,
      "action_relation_loss": 0.0010939819214399904,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.849391955098223,
      "flow_matching_loss": 0.01904244301840663,
      "grad_norm": 0.09396405518054962,
      "learning_rate": 2.6693670127569958e-05,
      "loss": 0.0204,
      "relation_loss": 0.0014018792076967656,
      "selected_pairs": 27010.9,
      "state_query_norm": 1112.8455688476563,
      "state_relation_loss": 0.00030789727752562613,
      "step": 20150,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014018792076967656,
      "weighted_relation_loss": 0.0014018792076967656
    },
    {
      "action_query_norm": 626.7887512207031,
      "action_relation_loss": 0.0012372665107250214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.858746492048645,
      "flow_matching_loss": 0.02008363837376237,
      "grad_norm": 0.09896522015333176,
      "learning_rate": 2.6644922478401268e-05,
      "loss": 0.0216,
      "relation_loss": 0.0015364689053967596,
      "selected_pairs": 26413.8,
      "state_query_norm": 1128.9101928710938,
      "state_relation_loss": 0.00029920238303020595,
      "step": 20160,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0015364689053967596,
      "weighted_relation_loss": 0.0015364689053967596
    },
    {
      "action_query_norm": 635.0778564453125,
      "action_relation_loss": 0.0011778931715525688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.868101028999064,
      "flow_matching_loss": 0.021697035618126394,
      "grad_norm": 0.08663499355316162,
      "learning_rate": 2.6596203207854004e-05,
      "loss": 0.0232,
      "relation_loss": 0.0014771584654226899,
      "selected_pairs": 26964.3,
      "state_query_norm": 1130.1324096679687,
      "state_relation_loss": 0.000299265279318206,
      "step": 20170,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014771584654226899,
      "weighted_relation_loss": 0.0014771584654226899
    },
    {
      "action_query_norm": 630.39501953125,
      "action_relation_loss": 0.0011426398006733507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.877455565949486,
      "flow_matching_loss": 0.0202456284314394,
      "grad_norm": 0.07242237031459808,
      "learning_rate": 2.654751237512666e-05,
      "loss": 0.0217,
      "relation_loss": 0.0014503040933050216,
      "selected_pairs": 26754.7,
      "state_query_norm": 1128.3037475585938,
      "state_relation_loss": 0.0003076642897212878,
      "step": 20180,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0014503040933050216,
      "weighted_relation_loss": 0.0014503040933050216
    },
    {
      "action_query_norm": 630.9533081054688,
      "action_relation_loss": 0.0011648021987639367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.886810102899908,
      "flow_matching_loss": 0.02096455208957195,
      "grad_norm": 0.10556569695472717,
      "learning_rate": 2.649885003938323e-05,
      "loss": 0.0224,
      "relation_loss": 0.001462692639324814,
      "selected_pairs": 27019.8,
      "state_query_norm": 1129.4057739257812,
      "state_relation_loss": 0.0002978904318297282,
      "step": 20190,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001462692639324814,
      "weighted_relation_loss": 0.001462692639324814
    },
    {
      "action_query_norm": 636.850244140625,
      "action_relation_loss": 0.0011387873208150267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.896164639850326,
      "flow_matching_loss": 0.019828513078391553,
      "grad_norm": 0.08578937500715256,
      "learning_rate": 2.6450216259753e-05,
      "loss": 0.0213,
      "relation_loss": 0.001441277761477977,
      "selected_pairs": 26439.1,
      "state_query_norm": 1140.8985107421875,
      "state_relation_loss": 0.00030249045812524855,
      "step": 20200,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001441277761477977,
      "weighted_relation_loss": 0.001441277761477977
    },
    {
      "action_query_norm": 631.8511779785156,
      "action_relation_loss": 0.0011513848323374987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.90551917680075,
      "flow_matching_loss": 0.018984506465494634,
      "grad_norm": 0.08646184951066971,
      "learning_rate": 2.640161109533063e-05,
      "loss": 0.0204,
      "relation_loss": 0.001454680401366204,
      "selected_pairs": 27197.3,
      "state_query_norm": 1136.7161254882812,
      "state_relation_loss": 0.00030329556320793925,
      "step": 20210,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001454680401366204,
      "weighted_relation_loss": 0.001454680401366204
    },
    {
      "action_query_norm": 629.3013305664062,
      "action_relation_loss": 0.0011885497835464776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.91487371375117,
      "flow_matching_loss": 0.02039231304079294,
      "grad_norm": 0.08375190198421478,
      "learning_rate": 2.6353034605175936e-05,
      "loss": 0.0219,
      "relation_loss": 0.001479534280952066,
      "selected_pairs": 26914.6,
      "state_query_norm": 1130.1964721679688,
      "state_relation_loss": 0.00029098450613673776,
      "step": 20220,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001479534280952066,
      "weighted_relation_loss": 0.001479534280952066
    },
    {
      "action_query_norm": 627.0099670410157,
      "action_relation_loss": 0.001168780610896647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.92422825070159,
      "flow_matching_loss": 0.023023297637701036,
      "grad_norm": 0.10391745716333389,
      "learning_rate": 2.630448684831398e-05,
      "loss": 0.0245,
      "relation_loss": 0.0014658585190773011,
      "selected_pairs": 26709.7,
      "state_query_norm": 1132.4970458984376,
      "state_relation_loss": 0.0002970778994495049,
      "step": 20230,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0014658585190773011,
      "weighted_relation_loss": 0.0014658585190773011
    },
    {
      "action_query_norm": 621.1625854492188,
      "action_relation_loss": 0.001146995380986482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.93358278765201,
      "flow_matching_loss": 0.02162288185209036,
      "grad_norm": 0.11953119933605194,
      "learning_rate": 2.625596788373482e-05,
      "loss": 0.0231,
      "relation_loss": 0.0014410121133551002,
      "selected_pairs": 27075.3,
      "state_query_norm": 1135.87880859375,
      "state_relation_loss": 0.00029401675274129955,
      "step": 20240,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0014410121133551002,
      "weighted_relation_loss": 0.0014410121133551002
    },
    {
      "action_query_norm": 633.793017578125,
      "action_relation_loss": 0.0011566839995793999,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.942937324602433,
      "flow_matching_loss": 0.022696915082633495,
      "grad_norm": 0.08772508800029755,
      "learning_rate": 2.620747777039363e-05,
      "loss": 0.0242,
      "relation_loss": 0.001459657307714224,
      "selected_pairs": 26990.8,
      "state_query_norm": 1131.3764526367188,
      "state_relation_loss": 0.0003029733168659732,
      "step": 20250,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001459657307714224,
      "weighted_relation_loss": 0.001459657307714224
    },
    {
      "action_query_norm": 623.2318603515625,
      "action_relation_loss": 0.0012085892376489937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.952291861552855,
      "flow_matching_loss": 0.01963895345106721,
      "grad_norm": 0.11246191710233688,
      "learning_rate": 2.6159016567210424e-05,
      "loss": 0.0212,
      "relation_loss": 0.0015146365854889154,
      "selected_pairs": 26409.9,
      "state_query_norm": 1128.1805419921875,
      "state_relation_loss": 0.0003060473594814539,
      "step": 20260,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0015146365854889154,
      "weighted_relation_loss": 0.0015146365854889154
    },
    {
      "action_query_norm": 627.7005798339844,
      "action_relation_loss": 0.0011668180115520955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.961646398503273,
      "flow_matching_loss": 0.01907559996470809,
      "grad_norm": 0.08225645869970322,
      "learning_rate": 2.6110584333070153e-05,
      "loss": 0.0206,
      "relation_loss": 0.001478942937683314,
      "selected_pairs": 26270.4,
      "state_query_norm": 1115.3717651367188,
      "state_relation_loss": 0.0003121249086689204,
      "step": 20270,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.001478942937683314,
      "weighted_relation_loss": 0.001478942937683314
    },
    {
      "action_query_norm": 632.0154296875,
      "action_relation_loss": 0.0011806428316049277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.971000935453695,
      "flow_matching_loss": 0.021090050227940083,
      "grad_norm": 0.08607790619134903,
      "learning_rate": 2.6062181126822537e-05,
      "loss": 0.0226,
      "relation_loss": 0.0014959530555643142,
      "selected_pairs": 26590.9,
      "state_query_norm": 1129.52158203125,
      "state_relation_loss": 0.00031531021813862025,
      "step": 20280,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0014959530555643142,
      "weighted_relation_loss": 0.0014959530555643142
    },
    {
      "action_query_norm": 629.7949096679688,
      "action_relation_loss": 0.0011646664352156222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.980355472404117,
      "flow_matching_loss": 0.021835544146597385,
      "grad_norm": 0.08952189236879349,
      "learning_rate": 2.601380700728203e-05,
      "loss": 0.0233,
      "relation_loss": 0.0014759056735783815,
      "selected_pairs": 26617.1,
      "state_query_norm": 1127.2935424804687,
      "state_relation_loss": 0.0003112392412731424,
      "step": 20290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0014759056735783815,
      "weighted_relation_loss": 0.0014759056735783815
    },
    {
      "action_query_norm": 629.7391906738281,
      "action_relation_loss": 0.0011880101286806167,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.989710009354535,
      "flow_matching_loss": 0.020673260278999805,
      "grad_norm": 0.1037406399846077,
      "learning_rate": 2.596546203322777e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014978347928263247,
      "selected_pairs": 27290.2,
      "state_query_norm": 1147.7135375976563,
      "state_relation_loss": 0.00030982465832494197,
      "step": 20300,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0014978347928263247,
      "weighted_relation_loss": 0.0014978347928263247
    },
    {
      "action_query_norm": 632.9541320800781,
      "action_relation_loss": 0.0011204335489310325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.999064546304957,
      "flow_matching_loss": 0.019433601386845113,
      "grad_norm": 0.1351628601551056,
      "learning_rate": 2.5917146263403457e-05,
      "loss": 0.0209,
      "relation_loss": 0.0014239145326428115,
      "selected_pairs": 26641.6,
      "state_query_norm": 1138.5669799804687,
      "state_relation_loss": 0.00030348097498063,
      "step": 20310,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0014239145326428115,
      "weighted_relation_loss": 0.0014239145326428115
    },
    {
      "action_query_norm": 634.447314453125,
      "action_relation_loss": 0.0011945053585805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.00841908325538,
      "flow_matching_loss": 0.019484975002706052,
      "grad_norm": 0.08598130941390991,
      "learning_rate": 2.5868859756517296e-05,
      "loss": 0.021,
      "relation_loss": 0.001500095147639513,
      "selected_pairs": 24783.1,
      "state_query_norm": 1138.615625,
      "state_relation_loss": 0.0003055897919693962,
      "step": 20320,
      "valid_action_samples": 216.0,
      "weighted_auxiliary_loss": 0.001500095147639513,
      "weighted_relation_loss": 0.001500095147639513
    },
    {
      "action_query_norm": 629.72900390625,
      "action_relation_loss": 0.0011602471582591533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.0177736202058,
      "flow_matching_loss": 0.021326168440282345,
      "grad_norm": 0.10107726603746414,
      "learning_rate": 2.5820602571241947e-05,
      "loss": 0.0228,
      "relation_loss": 0.0014631168800406159,
      "selected_pairs": 26899.7,
      "state_query_norm": 1127.55927734375,
      "state_relation_loss": 0.00030286973051261155,
      "step": 20330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0014631168800406159,
      "weighted_relation_loss": 0.0014631168800406159
    },
    {
      "action_query_norm": 633.0774597167969,
      "action_relation_loss": 0.0011889011890161782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.02712815715622,
      "flow_matching_loss": 0.01921827718615532,
      "grad_norm": 0.09613008797168732,
      "learning_rate": 2.577237476621442e-05,
      "loss": 0.0207,
      "relation_loss": 0.0014978502644225956,
      "selected_pairs": 27153.6,
      "state_query_norm": 1134.3138305664063,
      "state_relation_loss": 0.0003089490521233529,
      "step": 20340,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0014978502644225956,
      "weighted_relation_loss": 0.0014978502644225956
    },
    {
      "action_query_norm": 639.6824462890625,
      "action_relation_loss": 0.0011102436343207954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.036482694106642,
      "flow_matching_loss": 0.020997287519276143,
      "grad_norm": 0.09484737366437912,
      "learning_rate": 2.572417640003609e-05,
      "loss": 0.0224,
      "relation_loss": 0.0014070941018871964,
      "selected_pairs": 27034.5,
      "state_query_norm": 1132.4129028320312,
      "state_relation_loss": 0.0002968504879390821,
      "step": 20350,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0014070941018871964,
      "weighted_relation_loss": 0.0014070941018871964
    },
    {
      "action_query_norm": 626.2400573730469,
      "action_relation_loss": 0.0011230197502300142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.045837231057064,
      "flow_matching_loss": 0.018436082080006598,
      "grad_norm": 0.09897594153881073,
      "learning_rate": 2.5676007531272473e-05,
      "loss": 0.0199,
      "relation_loss": 0.001437341992277652,
      "selected_pairs": 26386.6,
      "state_query_norm": 1131.008642578125,
      "state_relation_loss": 0.00031432223913725467,
      "step": 20360,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001437341992277652,
      "weighted_relation_loss": 0.001437341992277652
    },
    {
      "action_query_norm": 629.727197265625,
      "action_relation_loss": 0.0011437840468715877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.055191768007482,
      "flow_matching_loss": 0.02102069277316332,
      "grad_norm": 0.09383122622966766,
      "learning_rate": 2.5627868218453332e-05,
      "loss": 0.0225,
      "relation_loss": 0.0014538250863552094,
      "selected_pairs": 26512.9,
      "state_query_norm": 1130.953173828125,
      "state_relation_loss": 0.00031004104239400475,
      "step": 20370,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0014538250863552094,
      "weighted_relation_loss": 0.0014538250863552094
    },
    {
      "action_query_norm": 637.0789855957031,
      "action_relation_loss": 0.0010881068999879062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.064546304957904,
      "flow_matching_loss": 0.02014915943145752,
      "grad_norm": 0.07743103057146072,
      "learning_rate": 2.5579758520072443e-05,
      "loss": 0.0215,
      "relation_loss": 0.0013893755152821542,
      "selected_pairs": 26735.2,
      "state_query_norm": 1130.9482666015624,
      "state_relation_loss": 0.00030126862111501396,
      "step": 20380,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013893755152821542,
      "weighted_relation_loss": 0.0013893755152821542
    },
    {
      "action_query_norm": 637.2355529785157,
      "action_relation_loss": 0.0011183510650880635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.073900841908326,
      "flow_matching_loss": 0.01991206929087639,
      "grad_norm": 0.11217934638261795,
      "learning_rate": 2.5531678494587615e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014286456047557295,
      "selected_pairs": 26735.0,
      "state_query_norm": 1116.6116088867188,
      "state_relation_loss": 0.0003102945251157507,
      "step": 20390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0014286456047557295,
      "weighted_relation_loss": 0.0014286456047557295
    },
    {
      "action_query_norm": 639.8490539550781,
      "action_relation_loss": 0.0011548500624485314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.08325537885875,
      "flow_matching_loss": 0.01995983850210905,
      "grad_norm": 0.09463959187269211,
      "learning_rate": 2.5483628200420644e-05,
      "loss": 0.0214,
      "relation_loss": 0.0014627703931182623,
      "selected_pairs": 26723.0,
      "state_query_norm": 1123.4678588867187,
      "state_relation_loss": 0.00030792032484896483,
      "step": 20400,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0014627703931182623,
      "weighted_relation_loss": 0.0014627703931182623
    },
    {
      "action_query_norm": 642.8377868652344,
      "action_relation_loss": 0.0010868637298699468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.092609915809167,
      "flow_matching_loss": 0.019599221274256708,
      "grad_norm": 0.08122692257165909,
      "learning_rate": 2.5435607695957154e-05,
      "loss": 0.021,
      "relation_loss": 0.0013916378375142813,
      "selected_pairs": 26894.4,
      "state_query_norm": 1147.2587646484376,
      "state_relation_loss": 0.00030477410182356835,
      "step": 20410,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013916378375142813,
      "weighted_relation_loss": 0.0013916378375142813
    },
    {
      "action_query_norm": 642.8234069824218,
      "action_relation_loss": 0.0011230960837565363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.10196445275959,
      "flow_matching_loss": 0.019269271474331617,
      "grad_norm": 0.11700990051031113,
      "learning_rate": 2.5387617039546586e-05,
      "loss": 0.0207,
      "relation_loss": 0.001429444458335638,
      "selected_pairs": 27322.6,
      "state_query_norm": 1123.9043090820312,
      "state_relation_loss": 0.00030634837166871877,
      "step": 20420,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.001429444458335638,
      "weighted_relation_loss": 0.001429444458335638
    },
    {
      "action_query_norm": 638.9612609863282,
      "action_relation_loss": 0.0011113764194305986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.11131898971001,
      "flow_matching_loss": 0.019179305247962474,
      "grad_norm": 0.08732371032238007,
      "learning_rate": 2.5339656289502106e-05,
      "loss": 0.0206,
      "relation_loss": 0.0014211197500117122,
      "selected_pairs": 26778.8,
      "state_query_norm": 1124.0817626953126,
      "state_relation_loss": 0.0003097433393122628,
      "step": 20430,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0014211197500117122,
      "weighted_relation_loss": 0.0014211197500117122
    },
    {
      "action_query_norm": 637.6563781738281,
      "action_relation_loss": 0.0011021943180821836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.12067352666043,
      "flow_matching_loss": 0.019906548969447612,
      "grad_norm": 0.11643086373806,
      "learning_rate": 2.529172550410056e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014142578933387996,
      "selected_pairs": 26789.0,
      "state_query_norm": 1124.860400390625,
      "state_relation_loss": 0.00031206359271891416,
      "step": 20440,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0014142578933387996,
      "weighted_relation_loss": 0.0014142578933387996
    },
    {
      "action_query_norm": 633.8500183105468,
      "action_relation_loss": 0.0011594711570069194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.13002806361085,
      "flow_matching_loss": 0.019937192648649217,
      "grad_norm": 0.09525395929813385,
      "learning_rate": 2.524382474158234e-05,
      "loss": 0.0214,
      "relation_loss": 0.0014712388743646442,
      "selected_pairs": 27079.8,
      "state_query_norm": 1127.7063842773437,
      "state_relation_loss": 0.00031176772026810797,
      "step": 20450,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0014712388743646442,
      "weighted_relation_loss": 0.0014712388743646442
    },
    {
      "action_query_norm": 635.1483276367187,
      "action_relation_loss": 0.0010998585144989192,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.139382600561273,
      "flow_matching_loss": 0.02056251633912325,
      "grad_norm": 0.08484090119600296,
      "learning_rate": 2.5195954060151433e-05,
      "loss": 0.022,
      "relation_loss": 0.0014063389971852303,
      "selected_pairs": 26898.7,
      "state_query_norm": 1133.9428344726562,
      "state_relation_loss": 0.00030648050305899234,
      "step": 20460,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0014063389971852303,
      "weighted_relation_loss": 0.0014063389971852303
    },
    {
      "action_query_norm": 636.1989440917969,
      "action_relation_loss": 0.0011472865124233067,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.14873713751169,
      "flow_matching_loss": 0.020513503439724447,
      "grad_norm": 0.09577398002147675,
      "learning_rate": 2.5148113517975212e-05,
      "loss": 0.022,
      "relation_loss": 0.0014466263703070581,
      "selected_pairs": 26411.9,
      "state_query_norm": 1146.5678466796876,
      "state_relation_loss": 0.0002993398782564327,
      "step": 20470,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0014466263703070581,
      "weighted_relation_loss": 0.0014466263703070581
    },
    {
      "action_query_norm": 631.0699401855469,
      "action_relation_loss": 0.001121220795903355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.158091674462113,
      "flow_matching_loss": 0.021417533047497272,
      "grad_norm": 0.09263801574707031,
      "learning_rate": 2.5100303173184447e-05,
      "loss": 0.0228,
      "relation_loss": 0.001415213046129793,
      "selected_pairs": 27056.8,
      "state_query_norm": 1129.3739379882813,
      "state_relation_loss": 0.00029399226768873633,
      "step": 20480,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001415213046129793,
      "weighted_relation_loss": 0.001415213046129793
    },
    {
      "action_query_norm": 630.6753479003906,
      "action_relation_loss": 0.0010451822075992824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.167446211412535,
      "flow_matching_loss": 0.019241663068532942,
      "grad_norm": 0.10823887586593628,
      "learning_rate": 2.5052523083873196e-05,
      "loss": 0.0206,
      "relation_loss": 0.0013497990788891912,
      "selected_pairs": 26247.3,
      "state_query_norm": 1117.547265625,
      "state_relation_loss": 0.0003046168858418241,
      "step": 20490,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0013497990788891912,
      "weighted_relation_loss": 0.0013497990788891912
    },
    {
      "action_query_norm": 633.89326171875,
      "action_relation_loss": 0.0011093725333921611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.176800748362957,
      "flow_matching_loss": 0.019408766366541385,
      "grad_norm": 0.10218432545661926,
      "learning_rate": 2.500477330809881e-05,
      "loss": 0.0208,
      "relation_loss": 0.001413094368763268,
      "selected_pairs": 26597.2,
      "state_query_norm": 1130.8819580078125,
      "state_relation_loss": 0.0003037218353711069,
      "step": 20500,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001413094368763268,
      "weighted_relation_loss": 0.001413094368763268
    },
    {
      "action_query_norm": 635.6740417480469,
      "action_relation_loss": 0.0011148139950819314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.186155285313376,
      "flow_matching_loss": 0.02082091588526964,
      "grad_norm": 0.10516571998596191,
      "learning_rate": 2.4957053903881733e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014120587147772312,
      "selected_pairs": 26641.8,
      "state_query_norm": 1132.3069580078125,
      "state_relation_loss": 0.0002972447080537677,
      "step": 20510,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0014120587147772312,
      "weighted_relation_loss": 0.0014120587147772312
    },
    {
      "action_query_norm": 639.6145568847656,
      "action_relation_loss": 0.0011324208928272129,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.195509822263798,
      "flow_matching_loss": 0.021840436011552812,
      "grad_norm": 0.10940311849117279,
      "learning_rate": 2.4909364929205576e-05,
      "loss": 0.0233,
      "relation_loss": 0.0014227789943106472,
      "selected_pairs": 26461.8,
      "state_query_norm": 1124.85419921875,
      "state_relation_loss": 0.0002903580985730514,
      "step": 20520,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0014227789943106472,
      "weighted_relation_loss": 0.0014227789943106472
    },
    {
      "action_query_norm": 636.3102294921875,
      "action_relation_loss": 0.0010760632867459208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.20486435921422,
      "flow_matching_loss": 0.020051414240151645,
      "grad_norm": 0.10383930057287216,
      "learning_rate": 2.4861706442016923e-05,
      "loss": 0.0214,
      "relation_loss": 0.0013787190779112279,
      "selected_pairs": 26049.9,
      "state_query_norm": 1144.001611328125,
      "state_relation_loss": 0.0003026557998964563,
      "step": 20530,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0013787190779112279,
      "weighted_relation_loss": 0.0013787190779112279
    },
    {
      "action_query_norm": 629.3487609863281,
      "action_relation_loss": 0.0011267199530266224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.21421889616464,
      "flow_matching_loss": 0.01990182725712657,
      "grad_norm": 0.08755488693714142,
      "learning_rate": 2.481407850022533e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014315802487544715,
      "selected_pairs": 26840.1,
      "state_query_norm": 1120.0175170898438,
      "state_relation_loss": 0.00030486028699669985,
      "step": 20540,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0014315802487544715,
      "weighted_relation_loss": 0.0014315802487544715
    },
    {
      "action_query_norm": 630.9333312988281,
      "action_relation_loss": 0.0011012220289558171,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.22357343311506,
      "flow_matching_loss": 0.021132951974868773,
      "grad_norm": 0.09495718032121658,
      "learning_rate": 2.4766481161703213e-05,
      "loss": 0.0225,
      "relation_loss": 0.001401445479132235,
      "selected_pairs": 26528.2,
      "state_query_norm": 1126.2839477539062,
      "state_relation_loss": 0.00030022344726603477,
      "step": 20550,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001401445479132235,
      "weighted_relation_loss": 0.001401445479132235
    },
    {
      "action_query_norm": 632.5077941894531,
      "action_relation_loss": 0.0011186257761437446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.232927970065482,
      "flow_matching_loss": 0.01896065454930067,
      "grad_norm": 0.10791240632534027,
      "learning_rate": 2.4718914484285876e-05,
      "loss": 0.0204,
      "relation_loss": 0.0014089757460169494,
      "selected_pairs": 26800.8,
      "state_query_norm": 1123.0541870117188,
      "state_relation_loss": 0.0002903499553212896,
      "step": 20560,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014089757460169494,
      "weighted_relation_loss": 0.0014089757460169494
    },
    {
      "action_query_norm": 631.9328979492187,
      "action_relation_loss": 0.0010523259406909347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.242282507015904,
      "flow_matching_loss": 0.01877409527078271,
      "grad_norm": 0.09777192026376724,
      "learning_rate": 2.467137852577129e-05,
      "loss": 0.0201,
      "relation_loss": 0.001353205752093345,
      "selected_pairs": 26780.2,
      "state_query_norm": 1134.8498046875,
      "state_relation_loss": 0.0003008798201335594,
      "step": 20570,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001353205752093345,
      "weighted_relation_loss": 0.001353205752093345
    },
    {
      "action_query_norm": 631.3471252441407,
      "action_relation_loss": 0.0011470329947769642,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.251637043966323,
      "flow_matching_loss": 0.020767507329583167,
      "grad_norm": 0.09813981503248215,
      "learning_rate": 2.4623873343920123e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014593491214327514,
      "selected_pairs": 26623.3,
      "state_query_norm": 1129.669970703125,
      "state_relation_loss": 0.00031231611501425505,
      "step": 20580,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0014593491214327514,
      "weighted_relation_loss": 0.0014593491214327514
    },
    {
      "action_query_norm": 635.0584899902344,
      "action_relation_loss": 0.001073427707888186,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.260991580916745,
      "flow_matching_loss": 0.019888621848076583,
      "grad_norm": 0.07526665180921555,
      "learning_rate": 2.4576398996455656e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013767460477538407,
      "selected_pairs": 27343.1,
      "state_query_norm": 1120.2528564453125,
      "state_relation_loss": 0.00030331834859680383,
      "step": 20590,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0013767460477538407,
      "weighted_relation_loss": 0.0013767460477538407
    },
    {
      "action_query_norm": 634.1120910644531,
      "action_relation_loss": 0.0010919656720943748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.270346117867167,
      "flow_matching_loss": 0.020908704865723847,
      "grad_norm": 0.09818597882986069,
      "learning_rate": 2.452895554106368e-05,
      "loss": 0.0223,
      "relation_loss": 0.0013955421396531164,
      "selected_pairs": 27548.9,
      "state_query_norm": 1124.6818115234375,
      "state_relation_loss": 0.00030357645300682634,
      "step": 20600,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0013955421396531164,
      "weighted_relation_loss": 0.0013955421396531164
    },
    {
      "action_query_norm": 628.1722778320312,
      "action_relation_loss": 0.0011287207482382656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.279700654817585,
      "flow_matching_loss": 0.02104802690446377,
      "grad_norm": 0.08691925555467606,
      "learning_rate": 2.4481543035392506e-05,
      "loss": 0.0225,
      "relation_loss": 0.001424575666896999,
      "selected_pairs": 27197.5,
      "state_query_norm": 1132.800341796875,
      "state_relation_loss": 0.0002958549244794995,
      "step": 20610,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001424575666896999,
      "weighted_relation_loss": 0.001424575666896999
    },
    {
      "action_query_norm": 625.810791015625,
      "action_relation_loss": 0.0011766727780923246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.289055191768007,
      "flow_matching_loss": 0.02011649599298835,
      "grad_norm": 0.08084197342395782,
      "learning_rate": 2.4434161537052774e-05,
      "loss": 0.0216,
      "relation_loss": 0.001476845913566649,
      "selected_pairs": 26575.2,
      "state_query_norm": 1133.5704223632813,
      "state_relation_loss": 0.000300173120922409,
      "step": 20620,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.001476845913566649,
      "weighted_relation_loss": 0.001476845913566649
    },
    {
      "action_query_norm": 633.1138977050781,
      "action_relation_loss": 0.0011146250180900097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.29840972871843,
      "flow_matching_loss": 0.019785390608012675,
      "grad_norm": 0.10362110286951065,
      "learning_rate": 2.4386811103617474e-05,
      "loss": 0.0212,
      "relation_loss": 0.0014298905502073468,
      "selected_pairs": 26802.2,
      "state_query_norm": 1122.8212280273438,
      "state_relation_loss": 0.00031526553502772006,
      "step": 20630,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014298905502073468,
      "weighted_relation_loss": 0.0014298905502073468
    },
    {
      "action_query_norm": 631.6695617675781,
      "action_relation_loss": 0.001123633305542171,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.30776426566885,
      "flow_matching_loss": 0.019891894049942495,
      "grad_norm": 0.11014976352453232,
      "learning_rate": 2.4339491792621833e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014259531046263874,
      "selected_pairs": 27085.8,
      "state_query_norm": 1132.8937255859375,
      "state_relation_loss": 0.00030231979908421633,
      "step": 20640,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0014259531046263874,
      "weighted_relation_loss": 0.0014259531046263874
    },
    {
      "action_query_norm": 633.7378112792969,
      "action_relation_loss": 0.0011894695926457644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.31711880261927,
      "flow_matching_loss": 0.01931616896763444,
      "grad_norm": 0.0945403054356575,
      "learning_rate": 2.4292203661563312e-05,
      "loss": 0.0208,
      "relation_loss": 0.0014836872578598559,
      "selected_pairs": 26920.3,
      "state_query_norm": 1126.3302124023437,
      "state_relation_loss": 0.0002942176477517933,
      "step": 20650,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0014836872578598559,
      "weighted_relation_loss": 0.0014836872578598559
    },
    {
      "action_query_norm": 633.2370666503906,
      "action_relation_loss": 0.0011249180184677242,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.32647333956969,
      "flow_matching_loss": 0.01902198800817132,
      "grad_norm": 0.0947394147515297,
      "learning_rate": 2.424494676790141e-05,
      "loss": 0.0205,
      "relation_loss": 0.0014379117055796087,
      "selected_pairs": 27289.7,
      "state_query_norm": 1108.8348754882813,
      "state_relation_loss": 0.0003129936842015013,
      "step": 20660,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0014379117055796087,
      "weighted_relation_loss": 0.0014379117055796087
    },
    {
      "action_query_norm": 631.5983703613281,
      "action_relation_loss": 0.0011169628822244703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.335827876520113,
      "flow_matching_loss": 0.02119328621774912,
      "grad_norm": 0.10440226644277573,
      "learning_rate": 2.4197721169057745e-05,
      "loss": 0.0226,
      "relation_loss": 0.0014179685269482434,
      "selected_pairs": 26524.0,
      "state_query_norm": 1127.1416625976562,
      "state_relation_loss": 0.0003010056592756882,
      "step": 20670,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0014179685269482434,
      "weighted_relation_loss": 0.0014179685269482434
    },
    {
      "action_query_norm": 633.0513366699219,
      "action_relation_loss": 0.0010718467645347118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.345182413470532,
      "flow_matching_loss": 0.019355231523513795,
      "grad_norm": 0.10699158906936646,
      "learning_rate": 2.4150526922415857e-05,
      "loss": 0.0207,
      "relation_loss": 0.001372446969617158,
      "selected_pairs": 27110.4,
      "state_query_norm": 1139.0548095703125,
      "state_relation_loss": 0.0003006001963512972,
      "step": 20680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001372446969617158,
      "weighted_relation_loss": 0.001372446969617158
    },
    {
      "action_query_norm": 622.6311767578125,
      "action_relation_loss": 0.0011608276166953147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.354536950420954,
      "flow_matching_loss": 0.021101210080087186,
      "grad_norm": 0.1023956686258316,
      "learning_rate": 2.41033640853212e-05,
      "loss": 0.0226,
      "relation_loss": 0.0014649667195044459,
      "selected_pairs": 26951.1,
      "state_query_norm": 1125.7480346679688,
      "state_relation_loss": 0.00030413911445066334,
      "step": 20690,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0014649667195044459,
      "weighted_relation_loss": 0.0014649667195044459
    },
    {
      "action_query_norm": 626.8418518066406,
      "action_relation_loss": 0.001088463730411604,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.363891487371376,
      "flow_matching_loss": 0.01849294202402234,
      "grad_norm": 0.09031783044338226,
      "learning_rate": 2.405623271508108e-05,
      "loss": 0.0199,
      "relation_loss": 0.0013995512505061924,
      "selected_pairs": 26876.6,
      "state_query_norm": 1119.6135498046874,
      "state_relation_loss": 0.00031108752300497146,
      "step": 20700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0013995512505061924,
      "weighted_relation_loss": 0.0013995512505061924
    },
    {
      "action_query_norm": 627.7042419433594,
      "action_relation_loss": 0.0012040099361911416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.373246024321794,
      "flow_matching_loss": 0.018754491582512856,
      "grad_norm": 0.09294956177473068,
      "learning_rate": 2.4009132868964522e-05,
      "loss": 0.0203,
      "relation_loss": 0.0015065488987602293,
      "selected_pairs": 26755.6,
      "state_query_norm": 1141.107080078125,
      "state_relation_loss": 0.00030253896838985386,
      "step": 20710,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0015065488987602293,
      "weighted_relation_loss": 0.0015065488987602293
    },
    {
      "action_query_norm": 625.14541015625,
      "action_relation_loss": 0.0010288301564287394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.382600561272216,
      "flow_matching_loss": 0.02047725096344948,
      "grad_norm": 0.07831887155771255,
      "learning_rate": 2.3962064604202327e-05,
      "loss": 0.0218,
      "relation_loss": 0.0013341687736101449,
      "selected_pairs": 26416.6,
      "state_query_norm": 1097.6932006835937,
      "state_relation_loss": 0.00030533863173332065,
      "step": 20720,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0013341687736101449,
      "weighted_relation_loss": 0.0013341687736101449
    },
    {
      "action_query_norm": 634.0751770019531,
      "action_relation_loss": 0.001112742378609255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.39195509822264,
      "flow_matching_loss": 0.01882035322487354,
      "grad_norm": 0.10346104204654694,
      "learning_rate": 2.3915027977986858e-05,
      "loss": 0.0202,
      "relation_loss": 0.0014088337658904493,
      "selected_pairs": 26696.2,
      "state_query_norm": 1141.8386962890625,
      "state_relation_loss": 0.00029609140765387567,
      "step": 20730,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0014088337658904493,
      "weighted_relation_loss": 0.0014088337658904493
    },
    {
      "action_query_norm": 627.7819519042969,
      "action_relation_loss": 0.0011043691425584256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.40130963517306,
      "flow_matching_loss": 0.021795864589512347,
      "grad_norm": 0.09760236740112305,
      "learning_rate": 2.386802304747205e-05,
      "loss": 0.0232,
      "relation_loss": 0.0014027353958226741,
      "selected_pairs": 27178.0,
      "state_query_norm": 1123.8866577148438,
      "state_relation_loss": 0.0002983662445330992,
      "step": 20740,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0014027353958226741,
      "weighted_relation_loss": 0.0014027353958226741
    },
    {
      "action_query_norm": 623.3025268554687,
      "action_relation_loss": 0.001092894357861951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.41066417212348,
      "flow_matching_loss": 0.020515633188188075,
      "grad_norm": 0.08633215725421906,
      "learning_rate": 2.382104986977332e-05,
      "loss": 0.0219,
      "relation_loss": 0.0013962417258881032,
      "selected_pairs": 26940.8,
      "state_query_norm": 1117.94052734375,
      "state_relation_loss": 0.00030334736802615223,
      "step": 20750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013962417258881032,
      "weighted_relation_loss": 0.0013962417258881032
    },
    {
      "action_query_norm": 634.8764282226563,
      "action_relation_loss": 0.0011174043465871364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.4200187090739,
      "flow_matching_loss": 0.019757350254803897,
      "grad_norm": 0.09422764927148819,
      "learning_rate": 2.377410850196749e-05,
      "loss": 0.0212,
      "relation_loss": 0.001406487263739109,
      "selected_pairs": 27525.3,
      "state_query_norm": 1132.0449462890624,
      "state_relation_loss": 0.0002890829317038879,
      "step": 20760,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.001406487263739109,
      "weighted_relation_loss": 0.001406487263739109
    },
    {
      "action_query_norm": 633.8330261230469,
      "action_relation_loss": 0.0010834691813215614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.429373246024323,
      "flow_matching_loss": 0.020917982701212166,
      "grad_norm": 0.10123953223228455,
      "learning_rate": 2.3727199001092797e-05,
      "loss": 0.0223,
      "relation_loss": 0.001386729395017028,
      "selected_pairs": 26920.2,
      "state_query_norm": 1126.7504150390625,
      "state_relation_loss": 0.0003032602049643174,
      "step": 20770,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001386729395017028,
      "weighted_relation_loss": 0.001386729395017028
    },
    {
      "action_query_norm": 630.9371887207031,
      "action_relation_loss": 0.0010951055446639657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.43872778297474,
      "flow_matching_loss": 0.01884123105555773,
      "grad_norm": 0.09151928871870041,
      "learning_rate": 2.3680321424148677e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013935548719018697,
      "selected_pairs": 27217.7,
      "state_query_norm": 1134.6859008789063,
      "state_relation_loss": 0.0002984493476105854,
      "step": 20780,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0013935548719018697,
      "weighted_relation_loss": 0.0013935548719018697
    },
    {
      "action_query_norm": 629.102880859375,
      "action_relation_loss": 0.0011657138355076313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.448082319925163,
      "flow_matching_loss": 0.02073945999145508,
      "grad_norm": 0.0832754448056221,
      "learning_rate": 2.3633475828095793e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014686507522128523,
      "selected_pairs": 27456.6,
      "state_query_norm": 1122.3593383789062,
      "state_relation_loss": 0.0003029369254363701,
      "step": 20790,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0014686507522128523,
      "weighted_relation_loss": 0.0014686507522128523
    },
    {
      "action_query_norm": 627.8016967773438,
      "action_relation_loss": 0.0011679654009640217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.457436856875585,
      "flow_matching_loss": 0.020895326882600783,
      "grad_norm": 0.11265381425619125,
      "learning_rate": 2.358666226985599e-05,
      "loss": 0.0224,
      "relation_loss": 0.0014640568639151751,
      "selected_pairs": 26966.4,
      "state_query_norm": 1106.1820068359375,
      "state_relation_loss": 0.0002960914687719196,
      "step": 20800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014640568639151751,
      "weighted_relation_loss": 0.0014640568639151751
    },
    {
      "action_query_norm": 630.3448425292969,
      "action_relation_loss": 0.0011249186005443334,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.466791393826007,
      "flow_matching_loss": 0.019727262668311596,
      "grad_norm": 0.08595053851604462,
      "learning_rate": 2.353988080631213e-05,
      "loss": 0.0211,
      "relation_loss": 0.0014157053898088633,
      "selected_pairs": 26176.7,
      "state_query_norm": 1141.9026611328125,
      "state_relation_loss": 0.0002907867718022317,
      "step": 20810,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0014157053898088633,
      "weighted_relation_loss": 0.0014157053898088633
    },
    {
      "action_query_norm": 631.1092529296875,
      "action_relation_loss": 0.0010887758224271237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.476145930776426,
      "flow_matching_loss": 0.018900782242417334,
      "grad_norm": 0.07865728437900543,
      "learning_rate": 2.3493131494308142e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013851063675247132,
      "selected_pairs": 27035.6,
      "state_query_norm": 1137.2958740234376,
      "state_relation_loss": 0.00029633054800797253,
      "step": 20820,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0013851063675247132,
      "weighted_relation_loss": 0.0013851063675247132
    },
    {
      "action_query_norm": 634.7131469726562,
      "action_relation_loss": 0.0010953176533803344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.485500467726848,
      "flow_matching_loss": 0.01954537909477949,
      "grad_norm": 0.07190040498971939,
      "learning_rate": 2.344641439064881e-05,
      "loss": 0.0209,
      "relation_loss": 0.0013934434158727528,
      "selected_pairs": 26828.2,
      "state_query_norm": 1136.5460205078125,
      "state_relation_loss": 0.0002981257770443335,
      "step": 20830,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013934434158727528,
      "weighted_relation_loss": 0.0013934434158727528
    },
    {
      "action_query_norm": 630.3574768066406,
      "action_relation_loss": 0.0011246787966229022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.49485500467727,
      "flow_matching_loss": 0.01930782413110137,
      "grad_norm": 0.11243503540754318,
      "learning_rate": 2.339972955209984e-05,
      "loss": 0.0207,
      "relation_loss": 0.0014331895392388106,
      "selected_pairs": 26813.1,
      "state_query_norm": 1120.7393432617187,
      "state_relation_loss": 0.0003085107513470575,
      "step": 20840,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014331895392388106,
      "weighted_relation_loss": 0.0014331895392388106
    },
    {
      "action_query_norm": 630.7146911621094,
      "action_relation_loss": 0.0011069900006987154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.504209541627688,
      "flow_matching_loss": 0.020971996244043112,
      "grad_norm": 0.08384054154157639,
      "learning_rate": 2.335307703538771e-05,
      "loss": 0.0224,
      "relation_loss": 0.001412045385222882,
      "selected_pairs": 27317.5,
      "state_query_norm": 1140.324072265625,
      "state_relation_loss": 0.0003050553932553157,
      "step": 20850,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.001412045385222882,
      "weighted_relation_loss": 0.001412045385222882
    },
    {
      "action_query_norm": 640.4305358886719,
      "action_relation_loss": 0.0010949028190225362,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.51356407857811,
      "flow_matching_loss": 0.02032248694449663,
      "grad_norm": 0.10557889938354492,
      "learning_rate": 2.330645689719962e-05,
      "loss": 0.0217,
      "relation_loss": 0.0013860794249922037,
      "selected_pairs": 26987.8,
      "state_query_norm": 1128.1982788085938,
      "state_relation_loss": 0.0002911765914177522,
      "step": 20860,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013860794249922037,
      "weighted_relation_loss": 0.0013860794249922037
    },
    {
      "action_query_norm": 634.1779907226562,
      "action_relation_loss": 0.0010797137278132142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.522918615528532,
      "flow_matching_loss": 0.019542219303548335,
      "grad_norm": 0.10681922733783722,
      "learning_rate": 2.3259869194183414e-05,
      "loss": 0.0209,
      "relation_loss": 0.0013845963985659182,
      "selected_pairs": 27150.1,
      "state_query_norm": 1130.733251953125,
      "state_relation_loss": 0.00030488265329040585,
      "step": 20870,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013845963985659182,
      "weighted_relation_loss": 0.0013845963985659182
    },
    {
      "action_query_norm": 637.6811096191407,
      "action_relation_loss": 0.0011312323971651495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.532273152478954,
      "flow_matching_loss": 0.019293308816850187,
      "grad_norm": 0.08665243536233902,
      "learning_rate": 2.321331398294759e-05,
      "loss": 0.0207,
      "relation_loss": 0.0014322496950626372,
      "selected_pairs": 26482.7,
      "state_query_norm": 1134.90625,
      "state_relation_loss": 0.000301017306628637,
      "step": 20880,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0014322496950626372,
      "weighted_relation_loss": 0.0014322496950626372
    },
    {
      "action_query_norm": 633.0061584472656,
      "action_relation_loss": 0.0011488718912005425,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.541627689429372,
      "flow_matching_loss": 0.018960953690111637,
      "grad_norm": 0.09974946081638336,
      "learning_rate": 2.3166791320061094e-05,
      "loss": 0.0204,
      "relation_loss": 0.001442838041111827,
      "selected_pairs": 27058.6,
      "state_query_norm": 1134.84375,
      "state_relation_loss": 0.0002939661440905184,
      "step": 20890,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001442838041111827,
      "weighted_relation_loss": 0.001442838041111827
    },
    {
      "action_query_norm": 638.6691589355469,
      "action_relation_loss": 0.0010757420561276377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.550982226379794,
      "flow_matching_loss": 0.01918940767645836,
      "grad_norm": 0.09700435400009155,
      "learning_rate": 2.312030126205335e-05,
      "loss": 0.0206,
      "relation_loss": 0.0013731240876950323,
      "selected_pairs": 27041.8,
      "state_query_norm": 1124.4420776367188,
      "state_relation_loss": 0.0002973820344777778,
      "step": 20900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0013731240876950323,
      "weighted_relation_loss": 0.0013731240876950323
    },
    {
      "action_query_norm": 624.3268920898438,
      "action_relation_loss": 0.001130802428815514,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.560336763330216,
      "flow_matching_loss": 0.018614143785089254,
      "grad_norm": 0.09178650379180908,
      "learning_rate": 2.3073843865414162e-05,
      "loss": 0.02,
      "relation_loss": 0.0014173764153383671,
      "selected_pairs": 26666.1,
      "state_query_norm": 1127.5736083984375,
      "state_relation_loss": 0.00028657398652285335,
      "step": 20910,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0014173764153383671,
      "weighted_relation_loss": 0.0014173764153383671
    },
    {
      "action_query_norm": 630.9415222167969,
      "action_relation_loss": 0.0010873377788811923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.569691300280635,
      "flow_matching_loss": 0.020379555970430376,
      "grad_norm": 0.08667808771133423,
      "learning_rate": 2.302741918659363e-05,
      "loss": 0.0218,
      "relation_loss": 0.0013845224166288972,
      "selected_pairs": 26618.1,
      "state_query_norm": 1126.2569702148437,
      "state_relation_loss": 0.00029718464356847106,
      "step": 20920,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0013845224166288972,
      "weighted_relation_loss": 0.0013845224166288972
    },
    {
      "action_query_norm": 624.1600463867187,
      "action_relation_loss": 0.0011211684555746614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.579045837231057,
      "flow_matching_loss": 0.01849925983697176,
      "grad_norm": 0.09095831960439682,
      "learning_rate": 2.2981027282002154e-05,
      "loss": 0.0199,
      "relation_loss": 0.0014182121376506984,
      "selected_pairs": 26803.9,
      "state_query_norm": 1136.9963012695312,
      "state_relation_loss": 0.0002970436820760369,
      "step": 20930,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014182121376506984,
      "weighted_relation_loss": 0.0014182121376506984
    },
    {
      "action_query_norm": 622.6601806640625,
      "action_relation_loss": 0.0010818342270795256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.58840037418148,
      "flow_matching_loss": 0.019610763154923917,
      "grad_norm": 0.08733896166086197,
      "learning_rate": 2.2934668208010236e-05,
      "loss": 0.021,
      "relation_loss": 0.0013732068007811904,
      "selected_pairs": 26559.1,
      "state_query_norm": 1126.0460205078125,
      "state_relation_loss": 0.0002913725591497496,
      "step": 20940,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013732068007811904,
      "weighted_relation_loss": 0.0013732068007811904
    },
    {
      "action_query_norm": 621.7988464355469,
      "action_relation_loss": 0.0011244151275604964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.5977549111319,
      "flow_matching_loss": 0.0194806651212275,
      "grad_norm": 0.10963037610054016,
      "learning_rate": 2.2888342020948555e-05,
      "loss": 0.0209,
      "relation_loss": 0.001421042380388826,
      "selected_pairs": 26922.7,
      "state_query_norm": 1119.9527954101563,
      "state_relation_loss": 0.0002966272470075637,
      "step": 20950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001421042380388826,
      "weighted_relation_loss": 0.001421042380388826
    },
    {
      "action_query_norm": 626.6846313476562,
      "action_relation_loss": 0.0011123644071631134,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.60710944808232,
      "flow_matching_loss": 0.019223944656550883,
      "grad_norm": 0.14806786179542542,
      "learning_rate": 2.2842048777107783e-05,
      "loss": 0.0206,
      "relation_loss": 0.0014030390419065952,
      "selected_pairs": 27180.3,
      "state_query_norm": 1125.6594360351562,
      "state_relation_loss": 0.0002906746260123327,
      "step": 20960,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0014030390419065952,
      "weighted_relation_loss": 0.0014030390419065952
    },
    {
      "action_query_norm": 628.8847534179688,
      "action_relation_loss": 0.0010486793122254311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.61646398503274,
      "flow_matching_loss": 0.01839955672621727,
      "grad_norm": 0.07897933572530746,
      "learning_rate": 2.2795788532738553e-05,
      "loss": 0.0197,
      "relation_loss": 0.001338736992329359,
      "selected_pairs": 27013.6,
      "state_query_norm": 1130.3038208007813,
      "state_relation_loss": 0.0002900576742831618,
      "step": 20970,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001338736992329359,
      "weighted_relation_loss": 0.001338736992329359
    },
    {
      "action_query_norm": 624.7996765136719,
      "action_relation_loss": 0.0011560327489860355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.625818521983163,
      "flow_matching_loss": 0.020282577071338893,
      "grad_norm": 0.09781701117753983,
      "learning_rate": 2.2749561344051468e-05,
      "loss": 0.0217,
      "relation_loss": 0.001447719638235867,
      "selected_pairs": 26831.8,
      "state_query_norm": 1132.56904296875,
      "state_relation_loss": 0.00029168690380174664,
      "step": 20980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001447719638235867,
      "weighted_relation_loss": 0.001447719638235867
    },
    {
      "action_query_norm": 629.5380310058594,
      "action_relation_loss": 0.00107255702605471,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.63517305893358,
      "flow_matching_loss": 0.020553207769989968,
      "grad_norm": 0.09640155732631683,
      "learning_rate": 2.2703367267216896e-05,
      "loss": 0.0219,
      "relation_loss": 0.0013737863861024379,
      "selected_pairs": 26964.2,
      "state_query_norm": 1116.1708129882813,
      "state_relation_loss": 0.0003012293600477278,
      "step": 20990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0013737863861024379,
      "weighted_relation_loss": 0.0013737863861024379
    },
    {
      "action_query_norm": 627.2695922851562,
      "action_relation_loss": 0.0011853598291054368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.644527595884004,
      "flow_matching_loss": 0.018302234634757043,
      "grad_norm": 0.11355791240930557,
      "learning_rate": 2.2657206358364997e-05,
      "loss": 0.0198,
      "relation_loss": 0.0014857662841677665,
      "selected_pairs": 27180.4,
      "state_query_norm": 1111.423681640625,
      "state_relation_loss": 0.00030040643468964845,
      "step": 21000,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0014857662841677665,
      "weighted_relation_loss": 0.0014857662841677665
    },
    {
      "action_query_norm": 625.5024475097656,
      "action_relation_loss": 0.0011071902699768543,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.653882132834426,
      "flow_matching_loss": 0.020081241428852082,
      "grad_norm": 0.09941791743040085,
      "learning_rate": 2.261107867358563e-05,
      "loss": 0.0215,
      "relation_loss": 0.0014067829470150173,
      "selected_pairs": 26969.9,
      "state_query_norm": 1106.816748046875,
      "state_relation_loss": 0.0002995926741277799,
      "step": 21010,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014067829470150173,
      "weighted_relation_loss": 0.0014067829470150173
    },
    {
      "action_query_norm": 628.9065124511719,
      "action_relation_loss": 0.001065153896342963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.663236669784844,
      "flow_matching_loss": 0.018856918811798094,
      "grad_norm": 0.112624891102314,
      "learning_rate": 2.2564984268928263e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013614768395200372,
      "selected_pairs": 26945.5,
      "state_query_norm": 1145.8774169921876,
      "state_relation_loss": 0.0002963229373563081,
      "step": 21020,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013614768395200372,
      "weighted_relation_loss": 0.0013614768395200372
    },
    {
      "action_query_norm": 623.9837097167969,
      "action_relation_loss": 0.0010978212929330766,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.672591206735266,
      "flow_matching_loss": 0.020963545236736537,
      "grad_norm": 0.10514701157808304,
      "learning_rate": 2.251892320040198e-05,
      "loss": 0.0224,
      "relation_loss": 0.0013989133411087095,
      "selected_pairs": 26432.8,
      "state_query_norm": 1132.893359375,
      "state_relation_loss": 0.00030109203653410076,
      "step": 21030,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0013989133411087095,
      "weighted_relation_loss": 0.0013989133411087095
    },
    {
      "action_query_norm": 627.9160766601562,
      "action_relation_loss": 0.001096680318005383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.681945743685688,
      "flow_matching_loss": 0.01882625399157405,
      "grad_norm": 0.0948949083685875,
      "learning_rate": 2.2472895523975312e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013821577187627554,
      "selected_pairs": 27149.0,
      "state_query_norm": 1135.4154296875,
      "state_relation_loss": 0.00028547740366775544,
      "step": 21040,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013821577187627554,
      "weighted_relation_loss": 0.0013821577187627554
    },
    {
      "action_query_norm": 632.7490234375,
      "action_relation_loss": 0.0011345836566761136,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.69130028063611,
      "flow_matching_loss": 0.01979395356029272,
      "grad_norm": 0.07775857299566269,
      "learning_rate": 2.2426901295576212e-05,
      "loss": 0.0212,
      "relation_loss": 0.0014237643568776547,
      "selected_pairs": 26943.2,
      "state_query_norm": 1130.0584228515625,
      "state_relation_loss": 0.0002891806943807751,
      "step": 21050,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0014237643568776547,
      "weighted_relation_loss": 0.0014237643568776547
    },
    {
      "action_query_norm": 625.7182006835938,
      "action_relation_loss": 0.00116993059637025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.70065481758653,
      "flow_matching_loss": 0.020142051950097085,
      "grad_norm": 0.10752597451210022,
      "learning_rate": 2.2380940571092013e-05,
      "loss": 0.0216,
      "relation_loss": 0.001473419729154557,
      "selected_pairs": 27365.3,
      "state_query_norm": 1125.7287353515626,
      "state_relation_loss": 0.000303489132784307,
      "step": 21060,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001473419729154557,
      "weighted_relation_loss": 0.001473419729154557
    },
    {
      "action_query_norm": 630.3252502441406,
      "action_relation_loss": 0.0010971469164360315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.71000935453695,
      "flow_matching_loss": 0.020766185782849788,
      "grad_norm": 0.14391356706619263,
      "learning_rate": 2.2335013406369302e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013951061409898102,
      "selected_pairs": 27344.0,
      "state_query_norm": 1139.7857666015625,
      "state_relation_loss": 0.00029795924492646007,
      "step": 21070,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0013951061409898102,
      "weighted_relation_loss": 0.0013951061409898102
    },
    {
      "action_query_norm": 628.3642456054688,
      "action_relation_loss": 0.0011138973175548017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.719363891487372,
      "flow_matching_loss": 0.019200656842440367,
      "grad_norm": 0.10298698395490646,
      "learning_rate": 2.228911985721397e-05,
      "loss": 0.0206,
      "relation_loss": 0.0014114318299107253,
      "selected_pairs": 26414.5,
      "state_query_norm": 1138.3292724609375,
      "state_relation_loss": 0.0002975345094455406,
      "step": 21080,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0014114318299107253,
      "weighted_relation_loss": 0.0014114318299107253
    },
    {
      "action_query_norm": 629.0505920410156,
      "action_relation_loss": 0.0011171403923071921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.72871842843779,
      "flow_matching_loss": 0.019030411262065172,
      "grad_norm": 0.107845738530159,
      "learning_rate": 2.224325997939095e-05,
      "loss": 0.0204,
      "relation_loss": 0.0014132791198790073,
      "selected_pairs": 27083.4,
      "state_query_norm": 1125.9227416992187,
      "state_relation_loss": 0.0002961387188406661,
      "step": 21090,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0014132791198790073,
      "weighted_relation_loss": 0.0014132791198790073
    },
    {
      "action_query_norm": 625.6367004394531,
      "action_relation_loss": 0.0011172042228281497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.738072965388213,
      "flow_matching_loss": 0.02036837227642536,
      "grad_norm": 0.09033071249723434,
      "learning_rate": 2.219743382862437e-05,
      "loss": 0.0218,
      "relation_loss": 0.0014095313497819006,
      "selected_pairs": 26762.6,
      "state_query_norm": 1134.0782470703125,
      "state_relation_loss": 0.0002923271502368152,
      "step": 21100,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0014095313497819006,
      "weighted_relation_loss": 0.0014095313497819006
    },
    {
      "action_query_norm": 629.1279724121093,
      "action_relation_loss": 0.0011575815151445568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.747427502338635,
      "flow_matching_loss": 0.019326486717909576,
      "grad_norm": 0.09916922450065613,
      "learning_rate": 2.2151641460597294e-05,
      "loss": 0.0208,
      "relation_loss": 0.0014504445251077413,
      "selected_pairs": 26363.2,
      "state_query_norm": 1140.0593139648438,
      "state_relation_loss": 0.0002928630186943337,
      "step": 21110,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0014504445251077413,
      "weighted_relation_loss": 0.0014504445251077413
    },
    {
      "action_query_norm": 628.2603698730469,
      "action_relation_loss": 0.0011055016657337546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.756782039289057,
      "flow_matching_loss": 0.016239474434405564,
      "grad_norm": 0.08185867220163345,
      "learning_rate": 2.210588293095177e-05,
      "loss": 0.0176,
      "relation_loss": 0.0014079829445108772,
      "selected_pairs": 27502.1,
      "state_query_norm": 1114.8037475585938,
      "state_relation_loss": 0.0003024812787771225,
      "step": 21120,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0014079829445108772,
      "weighted_relation_loss": 0.0014079829445108772
    },
    {
      "action_query_norm": 625.9635437011718,
      "action_relation_loss": 0.0011057895550038665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.766136576239475,
      "flow_matching_loss": 0.020080178417265414,
      "grad_norm": 0.10007982701063156,
      "learning_rate": 2.2060158295288714e-05,
      "loss": 0.0215,
      "relation_loss": 0.0014133400982245803,
      "selected_pairs": 27009.7,
      "state_query_norm": 1118.5439697265624,
      "state_relation_loss": 0.00030755056359339507,
      "step": 21130,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014133400982245803,
      "weighted_relation_loss": 0.0014133400982245803
    },
    {
      "action_query_norm": 639.8644897460938,
      "action_relation_loss": 0.0010945430200081318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.775491113189897,
      "flow_matching_loss": 0.0196987914852798,
      "grad_norm": 0.0940791592001915,
      "learning_rate": 2.2014467609167905e-05,
      "loss": 0.0211,
      "relation_loss": 0.001387086056638509,
      "selected_pairs": 27247.0,
      "state_query_norm": 1118.8739868164062,
      "state_relation_loss": 0.0002925430308096111,
      "step": 21140,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001387086056638509,
      "weighted_relation_loss": 0.001387086056638509
    },
    {
      "action_query_norm": 624.6520080566406,
      "action_relation_loss": 0.0011253954144194722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.78484565014032,
      "flow_matching_loss": 0.018856918811798094,
      "grad_norm": 0.10076834261417389,
      "learning_rate": 2.1968810928107806e-05,
      "loss": 0.0203,
      "relation_loss": 0.001421244186349213,
      "selected_pairs": 26671.1,
      "state_query_norm": 1127.46103515625,
      "state_relation_loss": 0.0002958487602882087,
      "step": 21150,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001421244186349213,
      "weighted_relation_loss": 0.001421244186349213
    },
    {
      "action_query_norm": 627.1333068847656,
      "action_relation_loss": 0.0011084111290983855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.794200187090738,
      "flow_matching_loss": 0.01896183770149946,
      "grad_norm": 0.08988817036151886,
      "learning_rate": 2.1923188307585606e-05,
      "loss": 0.0204,
      "relation_loss": 0.00140814637998119,
      "selected_pairs": 26278.6,
      "state_query_norm": 1138.76552734375,
      "state_relation_loss": 0.0002997352363308892,
      "step": 21160,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.00140814637998119,
      "weighted_relation_loss": 0.00140814637998119
    },
    {
      "action_query_norm": 636.5389465332031,
      "action_relation_loss": 0.0010350328346248716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.80355472404116,
      "flow_matching_loss": 0.019252875354140997,
      "grad_norm": 0.10604177415370941,
      "learning_rate": 2.187759980303708e-05,
      "loss": 0.0206,
      "relation_loss": 0.0013332689763046802,
      "selected_pairs": 27202.0,
      "state_query_norm": 1135.0290283203126,
      "state_relation_loss": 0.00029823614459019155,
      "step": 21170,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0013332689763046802,
      "weighted_relation_loss": 0.0013332689763046802
    },
    {
      "action_query_norm": 630.0462585449219,
      "action_relation_loss": 0.0011585084605030716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.81290926099158,
      "flow_matching_loss": 0.021031046472489833,
      "grad_norm": 0.12410499155521393,
      "learning_rate": 2.1832045469856545e-05,
      "loss": 0.0225,
      "relation_loss": 0.0014502352802082895,
      "selected_pairs": 26600.1,
      "state_query_norm": 1118.00712890625,
      "state_relation_loss": 0.00029172682261560114,
      "step": 21180,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0014502352802082895,
      "weighted_relation_loss": 0.0014502352802082895
    },
    {
      "action_query_norm": 629.985498046875,
      "action_relation_loss": 0.0010440855345223098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.822263797942004,
      "flow_matching_loss": 0.018926900625228883,
      "grad_norm": 0.09269928932189941,
      "learning_rate": 2.1786525363396844e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013375566806644202,
      "selected_pairs": 26992.0,
      "state_query_norm": 1120.5410034179688,
      "state_relation_loss": 0.00029347115487325935,
      "step": 21190,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013375566806644202,
      "weighted_relation_loss": 0.0013375566806644202
    },
    {
      "action_query_norm": 633.1510131835937,
      "action_relation_loss": 0.0010856521257665008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.831618334892422,
      "flow_matching_loss": 0.019748937524855136,
      "grad_norm": 0.09439534693956375,
      "learning_rate": 2.1741039538969183e-05,
      "loss": 0.0211,
      "relation_loss": 0.0013876848854124546,
      "selected_pairs": 26465.1,
      "state_query_norm": 1126.463427734375,
      "state_relation_loss": 0.00030203277128748596,
      "step": 21200,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0013876848854124546,
      "weighted_relation_loss": 0.0013876848854124546
    },
    {
      "action_query_norm": 626.918017578125,
      "action_relation_loss": 0.001134794019162655,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.840972871842844,
      "flow_matching_loss": 0.019676341954618693,
      "grad_norm": 0.1107037141919136,
      "learning_rate": 2.169558805184313e-05,
      "loss": 0.0211,
      "relation_loss": 0.0014386420953087509,
      "selected_pairs": 26685.3,
      "state_query_norm": 1115.0736206054687,
      "state_relation_loss": 0.00030384807323571297,
      "step": 21210,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0014386420953087509,
      "weighted_relation_loss": 0.0014386420953087509
    },
    {
      "action_query_norm": 631.2827514648437,
      "action_relation_loss": 0.0010867609351407736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.850327408793266,
      "flow_matching_loss": 0.02040236406028271,
      "grad_norm": 0.08537586033344269,
      "learning_rate": 2.165017095724651e-05,
      "loss": 0.0218,
      "relation_loss": 0.0013818528619594871,
      "selected_pairs": 26649.2,
      "state_query_norm": 1128.6030883789062,
      "state_relation_loss": 0.0002950919297290966,
      "step": 21220,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0013818528619594871,
      "weighted_relation_loss": 0.0013818528619594871
    },
    {
      "action_query_norm": 627.9688415527344,
      "action_relation_loss": 0.0010968712565954774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.859681945743684,
      "flow_matching_loss": 0.018892418965697287,
      "grad_norm": 0.09278294444084167,
      "learning_rate": 2.1604788310365405e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013987810001708568,
      "selected_pairs": 26599.7,
      "state_query_norm": 1127.3723999023437,
      "state_relation_loss": 0.00030190974939614534,
      "step": 21230,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013987810001708568,
      "weighted_relation_loss": 0.0013987810001708568
    },
    {
      "action_query_norm": 624.3710021972656,
      "action_relation_loss": 0.0010978201229590923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.869036482694106,
      "flow_matching_loss": 0.018812431767582894,
      "grad_norm": 0.0871850848197937,
      "learning_rate": 2.155944016634398e-05,
      "loss": 0.0202,
      "relation_loss": 0.001396041561383754,
      "selected_pairs": 26420.0,
      "state_query_norm": 1124.8944580078125,
      "state_relation_loss": 0.0002982214151415974,
      "step": 21240,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001396041561383754,
      "weighted_relation_loss": 0.001396041561383754
    },
    {
      "action_query_norm": 631.77568359375,
      "action_relation_loss": 0.0010712748335208744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.87839101964453,
      "flow_matching_loss": 0.017047160491347312,
      "grad_norm": 0.09032978117465973,
      "learning_rate": 2.1514126580284538e-05,
      "loss": 0.0184,
      "relation_loss": 0.001365585089661181,
      "selected_pairs": 26823.3,
      "state_query_norm": 1114.8207641601562,
      "state_relation_loss": 0.00029431027069222183,
      "step": 21250,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001365585089661181,
      "weighted_relation_loss": 0.001365585089661181
    },
    {
      "action_query_norm": 632.3759826660156,
      "action_relation_loss": 0.0010613489081151783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.88774555659495,
      "flow_matching_loss": 0.017746938671916724,
      "grad_norm": 0.08815903961658478,
      "learning_rate": 2.1468847607247344e-05,
      "loss": 0.0191,
      "relation_loss": 0.0013605832238681615,
      "selected_pairs": 27406.0,
      "state_query_norm": 1115.6101684570312,
      "state_relation_loss": 0.00029923429538030176,
      "step": 21260,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0013605832238681615,
      "weighted_relation_loss": 0.0013605832238681615
    },
    {
      "action_query_norm": 636.68876953125,
      "action_relation_loss": 0.0010744911734946073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.89710009354537,
      "flow_matching_loss": 0.019210171513259412,
      "grad_norm": 0.09042360633611679,
      "learning_rate": 2.1423603302250623e-05,
      "loss": 0.0206,
      "relation_loss": 0.0013748245546594262,
      "selected_pairs": 27173.8,
      "state_query_norm": 1119.0761840820312,
      "state_relation_loss": 0.0003003333695232868,
      "step": 21270,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0013748245546594262,
      "weighted_relation_loss": 0.0013748245546594262
    },
    {
      "action_query_norm": 636.0943115234375,
      "action_relation_loss": 0.0011007313732989132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.90645463049579,
      "flow_matching_loss": 0.020068733580410482,
      "grad_norm": 0.09635687619447708,
      "learning_rate": 2.1378393720270468e-05,
      "loss": 0.0215,
      "relation_loss": 0.0013855515280738473,
      "selected_pairs": 26893.0,
      "state_query_norm": 1128.907861328125,
      "state_relation_loss": 0.00028482015477493405,
      "step": 21280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013855515280738473,
      "weighted_relation_loss": 0.0013855515280738473
    },
    {
      "action_query_norm": 628.5717956542969,
      "action_relation_loss": 0.0010704076790716498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.915809167446213,
      "flow_matching_loss": 0.019360346999019385,
      "grad_norm": 0.0782998576760292,
      "learning_rate": 2.133321891624076e-05,
      "loss": 0.0207,
      "relation_loss": 0.001361660740803927,
      "selected_pairs": 26942.4,
      "state_query_norm": 1125.0966552734376,
      "state_relation_loss": 0.0002912530617322773,
      "step": 21290,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001361660740803927,
      "weighted_relation_loss": 0.001361660740803927
    },
    {
      "action_query_norm": 631.2539306640625,
      "action_relation_loss": 0.0010626483126543462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.92516370439663,
      "flow_matching_loss": 0.019433369860053062,
      "grad_norm": 0.08651555329561234,
      "learning_rate": 2.1288078945053193e-05,
      "loss": 0.0208,
      "relation_loss": 0.0013712465879507363,
      "selected_pairs": 26506.7,
      "state_query_norm": 1116.685693359375,
      "state_relation_loss": 0.00030859828693792225,
      "step": 21300,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0013712465879507363,
      "weighted_relation_loss": 0.0013712465879507363
    },
    {
      "action_query_norm": 623.7267761230469,
      "action_relation_loss": 0.0011033137678168714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.934518241347053,
      "flow_matching_loss": 0.017477078549563884,
      "grad_norm": 0.08351249992847443,
      "learning_rate": 2.1242973861557064e-05,
      "loss": 0.0189,
      "relation_loss": 0.0013976765680126847,
      "selected_pairs": 26523.8,
      "state_query_norm": 1123.2732055664062,
      "state_relation_loss": 0.0002943628089269623,
      "step": 21310,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0013976765680126847,
      "weighted_relation_loss": 0.0013976765680126847
    },
    {
      "action_query_norm": 632.6150939941406,
      "action_relation_loss": 0.001084260205971077,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.943872778297475,
      "flow_matching_loss": 0.018917683605104686,
      "grad_norm": 0.10496022552251816,
      "learning_rate": 2.1197903720559304e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013739019515924156,
      "selected_pairs": 27152.9,
      "state_query_norm": 1117.77451171875,
      "state_relation_loss": 0.00028964176308363674,
      "step": 21320,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013739019515924156,
      "weighted_relation_loss": 0.0013739019515924156
    },
    {
      "action_query_norm": 626.4581726074218,
      "action_relation_loss": 0.0010813010507263244,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.953227315247894,
      "flow_matching_loss": 0.018629999458789827,
      "grad_norm": 0.10699880123138428,
      "learning_rate": 2.115286857682438e-05,
      "loss": 0.02,
      "relation_loss": 0.00137139402795583,
      "selected_pairs": 26255.2,
      "state_query_norm": 1133.1288818359376,
      "state_relation_loss": 0.00029009299469180406,
      "step": 21330,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.00137139402795583,
      "weighted_relation_loss": 0.00137139402795583
    },
    {
      "action_query_norm": 629.7041625976562,
      "action_relation_loss": 0.0010313831444364041,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.962581852198316,
      "flow_matching_loss": 0.02077956385910511,
      "grad_norm": 0.09712783992290497,
      "learning_rate": 2.1107868485074228e-05,
      "loss": 0.0221,
      "relation_loss": 0.001329599181190133,
      "selected_pairs": 26736.3,
      "state_query_norm": 1115.4476440429687,
      "state_relation_loss": 0.00029821604257449506,
      "step": 21340,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001329599181190133,
      "weighted_relation_loss": 0.001329599181190133
    },
    {
      "action_query_norm": 629.4667907714844,
      "action_relation_loss": 0.0010638096544425934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.971936389148738,
      "flow_matching_loss": 0.021865399740636348,
      "grad_norm": 0.10290898382663727,
      "learning_rate": 2.1062903499988234e-05,
      "loss": 0.0232,
      "relation_loss": 0.001358863792847842,
      "selected_pairs": 27263.0,
      "state_query_norm": 1113.7686401367187,
      "state_relation_loss": 0.00029505414131563156,
      "step": 21350,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.001358863792847842,
      "weighted_relation_loss": 0.001358863792847842
    },
    {
      "action_query_norm": 631.4590637207032,
      "action_relation_loss": 0.001032474497333169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.98129092609916,
      "flow_matching_loss": 0.018225488346070053,
      "grad_norm": 0.0881592407822609,
      "learning_rate": 2.101797367620308e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013424212695099414,
      "selected_pairs": 26132.5,
      "state_query_norm": 1126.0958618164063,
      "state_relation_loss": 0.00030994679545983674,
      "step": 21360,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0013424212695099414,
      "weighted_relation_loss": 0.0013424212695099414
    },
    {
      "action_query_norm": 631.8772033691406,
      "action_relation_loss": 0.0011139150359667838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.990645463049578,
      "flow_matching_loss": 0.018852922786027192,
      "grad_norm": 0.09277836978435516,
      "learning_rate": 2.0973079068312713e-05,
      "loss": 0.0203,
      "relation_loss": 0.0014079186716116966,
      "selected_pairs": 26946.2,
      "state_query_norm": 1134.898291015625,
      "state_relation_loss": 0.00029400362982414665,
      "step": 21370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0014079186716116966,
      "weighted_relation_loss": 0.0014079186716116966
    },
    {
      "action_query_norm": 635.0219116210938,
      "action_relation_loss": 0.001074137561954558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.0,
      "flow_matching_loss": 0.019802646711468697,
      "grad_norm": 0.16573193669319153,
      "learning_rate": 2.0928219730868358e-05,
      "loss": 0.0212,
      "relation_loss": 0.0013695344096049666,
      "selected_pairs": 24176.4,
      "state_query_norm": 1122.6954711914063,
      "state_relation_loss": 0.00029539684765040876,
      "step": 21380,
      "valid_action_samples": 213.9,
      "weighted_auxiliary_loss": 0.0013695344096049666,
      "weighted_relation_loss": 0.0013695344096049666
    },
    {
      "action_query_norm": 621.7656860351562,
      "action_relation_loss": 0.001102370984153822,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.009354536950422,
      "flow_matching_loss": 0.019826338440179826,
      "grad_norm": 0.10969994217157364,
      "learning_rate": 2.0883395718378302e-05,
      "loss": 0.0212,
      "relation_loss": 0.0013973541441373526,
      "selected_pairs": 26738.5,
      "state_query_norm": 1138.8284912109375,
      "state_relation_loss": 0.00029498314252123236,
      "step": 21390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013973541441373526,
      "weighted_relation_loss": 0.0013973541441373526
    },
    {
      "action_query_norm": 630.1669921875,
      "action_relation_loss": 0.0010562972747720778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.01870907390084,
      "flow_matching_loss": 0.017971530091017485,
      "grad_norm": 0.1147761270403862,
      "learning_rate": 2.083860708530798e-05,
      "loss": 0.0193,
      "relation_loss": 0.001349124824628234,
      "selected_pairs": 27358.8,
      "state_query_norm": 1132.7486694335937,
      "state_relation_loss": 0.00029282755276653915,
      "step": 21400,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001349124824628234,
      "weighted_relation_loss": 0.001349124824628234
    },
    {
      "action_query_norm": 630.6947143554687,
      "action_relation_loss": 0.0010345292219426482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.028063610851262,
      "flow_matching_loss": 0.02039053849875927,
      "grad_norm": 0.09225686639547348,
      "learning_rate": 2.0793853886079796e-05,
      "loss": 0.0217,
      "relation_loss": 0.0013223684974946082,
      "selected_pairs": 27408.7,
      "state_query_norm": 1132.4095947265625,
      "state_relation_loss": 0.0002878392580896616,
      "step": 21410,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0013223684974946082,
      "weighted_relation_loss": 0.0013223684974946082
    },
    {
      "action_query_norm": 633.143798828125,
      "action_relation_loss": 0.0010546350968070328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.037418147801684,
      "flow_matching_loss": 0.018193181324750184,
      "grad_norm": 0.09559226036071777,
      "learning_rate": 2.074913617507309e-05,
      "loss": 0.0195,
      "relation_loss": 0.0013436106033623218,
      "selected_pairs": 26354.4,
      "state_query_norm": 1127.4049194335937,
      "state_relation_loss": 0.00028897551528643816,
      "step": 21420,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0013436106033623218,
      "weighted_relation_loss": 0.0013436106033623218
    },
    {
      "action_query_norm": 631.748828125,
      "action_relation_loss": 0.0010573382140137255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.046772684752106,
      "flow_matching_loss": 0.018341523129492998,
      "grad_norm": 0.08070962876081467,
      "learning_rate": 2.070445400662411e-05,
      "loss": 0.0197,
      "relation_loss": 0.001342712133191526,
      "selected_pairs": 27312.9,
      "state_query_norm": 1128.935791015625,
      "state_relation_loss": 0.0002853739220881835,
      "step": 21430,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.001342712133191526,
      "weighted_relation_loss": 0.001342712133191526
    },
    {
      "action_query_norm": 633.34990234375,
      "action_relation_loss": 0.0010436365497298538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.056127221702525,
      "flow_matching_loss": 0.017651340272277594,
      "grad_norm": 0.10678461194038391,
      "learning_rate": 2.0659807435025907e-05,
      "loss": 0.019,
      "relation_loss": 0.001327723916620016,
      "selected_pairs": 27452.7,
      "state_query_norm": 1134.689453125,
      "state_relation_loss": 0.0002840873581590131,
      "step": 21440,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.001327723916620016,
      "weighted_relation_loss": 0.001327723916620016
    },
    {
      "action_query_norm": 636.7813354492188,
      "action_relation_loss": 0.0010256815527100115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.065481758652947,
      "flow_matching_loss": 0.018461955711245537,
      "grad_norm": 0.09543163329362869,
      "learning_rate": 2.061519651452825e-05,
      "loss": 0.0198,
      "relation_loss": 0.0013169888406991959,
      "selected_pairs": 26577.0,
      "state_query_norm": 1113.30537109375,
      "state_relation_loss": 0.00029130727925803513,
      "step": 21450,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0013169888406991959,
      "weighted_relation_loss": 0.0013169888406991959
    },
    {
      "action_query_norm": 632.6620544433594,
      "action_relation_loss": 0.001048360380809754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.07483629560337,
      "flow_matching_loss": 0.0180291342549026,
      "grad_norm": 0.0786086842417717,
      "learning_rate": 2.0570621299337656e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013408039812929928,
      "selected_pairs": 27388.4,
      "state_query_norm": 1134.0300537109374,
      "state_relation_loss": 0.0002924435917520896,
      "step": 21460,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0013408039812929928,
      "weighted_relation_loss": 0.0013408039812929928
    },
    {
      "action_query_norm": 631.86748046875,
      "action_relation_loss": 0.0010505693790037185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.084190832553787,
      "flow_matching_loss": 0.019938535615801813,
      "grad_norm": 0.09234333038330078,
      "learning_rate": 2.0526081843617183e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013379544718191028,
      "selected_pairs": 26558.4,
      "state_query_norm": 1126.0410278320312,
      "state_relation_loss": 0.0002873850869946182,
      "step": 21470,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013379544718191028,
      "weighted_relation_loss": 0.0013379544718191028
    },
    {
      "action_query_norm": 625.1919006347656,
      "action_relation_loss": 0.001063767890445888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.09354536950421,
      "flow_matching_loss": 0.019512847997248173,
      "grad_norm": 0.09101781994104385,
      "learning_rate": 2.0481578201486485e-05,
      "loss": 0.0209,
      "relation_loss": 0.0013552833464927972,
      "selected_pairs": 26984.8,
      "state_query_norm": 1125.28896484375,
      "state_relation_loss": 0.0002915154647780582,
      "step": 21480,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013552833464927972,
      "weighted_relation_loss": 0.0013552833464927972
    },
    {
      "action_query_norm": 620.94111328125,
      "action_relation_loss": 0.0010552016959991307,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.10289990645463,
      "flow_matching_loss": 0.017248804308474065,
      "grad_norm": 0.094306081533432,
      "learning_rate": 2.0437110427021677e-05,
      "loss": 0.0186,
      "relation_loss": 0.001341498689725995,
      "selected_pairs": 26360.7,
      "state_query_norm": 1138.308544921875,
      "state_relation_loss": 0.0002862969704438001,
      "step": 21490,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001341498689725995,
      "weighted_relation_loss": 0.001341498689725995
    },
    {
      "action_query_norm": 625.1877746582031,
      "action_relation_loss": 0.001103641331428662,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.112254443405053,
      "flow_matching_loss": 0.01840657154098153,
      "grad_norm": 0.10990298539400101,
      "learning_rate": 2.0392678574255282e-05,
      "loss": 0.0198,
      "relation_loss": 0.0013955525588244199,
      "selected_pairs": 26643.8,
      "state_query_norm": 1117.6187866210937,
      "state_relation_loss": 0.0002919112099334598,
      "step": 21500,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0013955525588244199,
      "weighted_relation_loss": 0.0013955525588244199
    },
    {
      "action_query_norm": 629.2258911132812,
      "action_relation_loss": 0.0010359301406424493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.12160898035547,
      "flow_matching_loss": 0.019311301968991758,
      "grad_norm": 0.08375857025384903,
      "learning_rate": 2.0348282697176217e-05,
      "loss": 0.0207,
      "relation_loss": 0.0013442661031149328,
      "selected_pairs": 27246.9,
      "state_query_norm": 1132.57939453125,
      "state_relation_loss": 0.0003083359712036327,
      "step": 21510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0013442661031149328,
      "weighted_relation_loss": 0.0013442661031149328
    },
    {
      "action_query_norm": 622.8981872558594,
      "action_relation_loss": 0.0011335120536386967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.130963517305894,
      "flow_matching_loss": 0.020051291212439537,
      "grad_norm": 0.09300903975963593,
      "learning_rate": 2.030392284972964e-05,
      "loss": 0.0215,
      "relation_loss": 0.0014394106925465166,
      "selected_pairs": 27201.9,
      "state_query_norm": 1131.3224731445312,
      "state_relation_loss": 0.00030589863017667086,
      "step": 21520,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0014394106925465166,
      "weighted_relation_loss": 0.0014394106925465166
    },
    {
      "action_query_norm": 628.9852172851563,
      "action_relation_loss": 0.0010398703918326647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.140318054256316,
      "flow_matching_loss": 0.018309885822236537,
      "grad_norm": 0.08752834796905518,
      "learning_rate": 2.0259599085816973e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013334357994608582,
      "selected_pairs": 26943.3,
      "state_query_norm": 1135.902587890625,
      "state_relation_loss": 0.0002935654105385765,
      "step": 21530,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013334357994608582,
      "weighted_relation_loss": 0.0013334357994608582
    },
    {
      "action_query_norm": 628.2124389648437,
      "action_relation_loss": 0.0011169873061589897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.149672591206734,
      "flow_matching_loss": 0.018818697705864907,
      "grad_norm": 0.09516709297895432,
      "learning_rate": 2.0215311459295755e-05,
      "loss": 0.0202,
      "relation_loss": 0.0014020859263837338,
      "selected_pairs": 26604.4,
      "state_query_norm": 1128.5986328125,
      "state_relation_loss": 0.00028509860567282885,
      "step": 21540,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0014020859263837338,
      "weighted_relation_loss": 0.0014020859263837338
    },
    {
      "action_query_norm": 628.31435546875,
      "action_relation_loss": 0.0010512262757401913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.159027128157156,
      "flow_matching_loss": 0.020287859812378885,
      "grad_norm": 0.08476497232913971,
      "learning_rate": 2.0171060023979606e-05,
      "loss": 0.0216,
      "relation_loss": 0.0013378996751271187,
      "selected_pairs": 26413.9,
      "state_query_norm": 1140.2119506835938,
      "state_relation_loss": 0.00028667339647654445,
      "step": 21550,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0013378996751271187,
      "weighted_relation_loss": 0.0013378996751271187
    },
    {
      "action_query_norm": 630.7193115234375,
      "action_relation_loss": 0.001021838973974809,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.168381665107578,
      "flow_matching_loss": 0.018761265091598034,
      "grad_norm": 0.09611838310956955,
      "learning_rate": 2.012684483363823e-05,
      "loss": 0.0201,
      "relation_loss": 0.0013129330356605351,
      "selected_pairs": 26552.2,
      "state_query_norm": 1116.1055053710938,
      "state_relation_loss": 0.0002910940645961091,
      "step": 21560,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013129330356605351,
      "weighted_relation_loss": 0.0013129330356605351
    },
    {
      "action_query_norm": 630.7652587890625,
      "action_relation_loss": 0.0010349173564463854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.177736202057996,
      "flow_matching_loss": 0.018085385393351316,
      "grad_norm": 0.07605188339948654,
      "learning_rate": 2.0082665941997237e-05,
      "loss": 0.0194,
      "relation_loss": 0.001318598329089582,
      "selected_pairs": 26717.2,
      "state_query_norm": 1128.748486328125,
      "state_relation_loss": 0.00028368096973281355,
      "step": 21570,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001318598329089582,
      "weighted_relation_loss": 0.001318598329089582
    },
    {
      "action_query_norm": 630.7107238769531,
      "action_relation_loss": 0.0010484342579729854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.18709073900842,
      "flow_matching_loss": 0.018628083076328038,
      "grad_norm": 0.10199601203203201,
      "learning_rate": 2.0038523402738146e-05,
      "loss": 0.02,
      "relation_loss": 0.0013366369646973908,
      "selected_pairs": 27012.1,
      "state_query_norm": 1135.5143188476563,
      "state_relation_loss": 0.0002882026979932562,
      "step": 21580,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013366369646973908,
      "weighted_relation_loss": 0.0013366369646973908
    },
    {
      "action_query_norm": 633.7453552246094,
      "action_relation_loss": 0.001023705559782684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.19644527595884,
      "flow_matching_loss": 0.018729111459106208,
      "grad_norm": 0.08826499432325363,
      "learning_rate": 1.99944172694983e-05,
      "loss": 0.02,
      "relation_loss": 0.0013057580217719078,
      "selected_pairs": 26811.3,
      "state_query_norm": 1126.2889404296875,
      "state_relation_loss": 0.00028205245616845785,
      "step": 21590,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0013057580217719078,
      "weighted_relation_loss": 0.0013057580217719078
    },
    {
      "action_query_norm": 631.3627197265625,
      "action_relation_loss": 0.0010800905525684356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.205799812909262,
      "flow_matching_loss": 0.019389746338129045,
      "grad_norm": 0.09556613862514496,
      "learning_rate": 1.99503475958708e-05,
      "loss": 0.0208,
      "relation_loss": 0.001378564105834812,
      "selected_pairs": 27094.4,
      "state_query_norm": 1120.5408813476563,
      "state_relation_loss": 0.0002984735503559932,
      "step": 21600,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001378564105834812,
      "weighted_relation_loss": 0.001378564105834812
    },
    {
      "action_query_norm": 632.4695861816406,
      "action_relation_loss": 0.0010582178947515786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.21515434985968,
      "flow_matching_loss": 0.01804886618629098,
      "grad_norm": 0.08701008558273315,
      "learning_rate": 1.9906314435404484e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013536776998080314,
      "selected_pairs": 26645.8,
      "state_query_norm": 1129.415966796875,
      "state_relation_loss": 0.00029545979341492055,
      "step": 21610,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0013536776998080314,
      "weighted_relation_loss": 0.0013536776998080314
    },
    {
      "action_query_norm": 628.4592102050781,
      "action_relation_loss": 0.0009965450735762715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.224508886810103,
      "flow_matching_loss": 0.01807105392217636,
      "grad_norm": 0.06618009507656097,
      "learning_rate": 1.986231784160378e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012888427241705357,
      "selected_pairs": 27220.9,
      "state_query_norm": 1114.5498657226562,
      "state_relation_loss": 0.00029229764477349817,
      "step": 21620,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0012888427241705357,
      "weighted_relation_loss": 0.0012888427241705357
    },
    {
      "action_query_norm": 632.6413879394531,
      "action_relation_loss": 0.0010372599703259767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.233863423760525,
      "flow_matching_loss": 0.018689964804798363,
      "grad_norm": 0.09767691045999527,
      "learning_rate": 1.9818357867928693e-05,
      "loss": 0.02,
      "relation_loss": 0.0013301608385518193,
      "selected_pairs": 26689.7,
      "state_query_norm": 1119.9135864257812,
      "state_relation_loss": 0.0002929008740466088,
      "step": 21630,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0013301608385518193,
      "weighted_relation_loss": 0.0013301608385518193
    },
    {
      "action_query_norm": 630.2491271972656,
      "action_relation_loss": 0.0010396109311841429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.243217960710943,
      "flow_matching_loss": 0.019673576578497887,
      "grad_norm": 0.0837792381644249,
      "learning_rate": 1.977443456779474e-05,
      "loss": 0.021,
      "relation_loss": 0.0013470203615725041,
      "selected_pairs": 26761.5,
      "state_query_norm": 1113.5243530273438,
      "state_relation_loss": 0.0003074094420298934,
      "step": 21640,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0013470203615725041,
      "weighted_relation_loss": 0.0013470203615725041
    },
    {
      "action_query_norm": 631.4033630371093,
      "action_relation_loss": 0.0010606581228785217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.252572497661365,
      "flow_matching_loss": 0.019952877145260572,
      "grad_norm": 0.09416051208972931,
      "learning_rate": 1.9730547994572858e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013522704946808517,
      "selected_pairs": 26594.9,
      "state_query_norm": 1142.5325561523437,
      "state_relation_loss": 0.0002916123834438622,
      "step": 21650,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013522704946808517,
      "weighted_relation_loss": 0.0013522704946808517
    },
    {
      "action_query_norm": 629.6759033203125,
      "action_relation_loss": 0.0010672687320038676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.261927034611787,
      "flow_matching_loss": 0.019657864607870577,
      "grad_norm": 0.1177932471036911,
      "learning_rate": 1.9686698201589393e-05,
      "loss": 0.021,
      "relation_loss": 0.0013639746350236237,
      "selected_pairs": 26300.7,
      "state_query_norm": 1139.2131958007812,
      "state_relation_loss": 0.0002967059175716713,
      "step": 21660,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0013639746350236237,
      "weighted_relation_loss": 0.0013639746350236237
    },
    {
      "action_query_norm": 630.8770263671875,
      "action_relation_loss": 0.0011396475718356669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.27128157156221,
      "flow_matching_loss": 0.01994167873635888,
      "grad_norm": 0.10828340798616409,
      "learning_rate": 1.9642885242125964e-05,
      "loss": 0.0214,
      "relation_loss": 0.0014395335456356406,
      "selected_pairs": 27103.9,
      "state_query_norm": 1132.7396484375,
      "state_relation_loss": 0.0002998859650688246,
      "step": 21670,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0014395335456356406,
      "weighted_relation_loss": 0.0014395335456356406
    },
    {
      "action_query_norm": 626.8566040039062,
      "action_relation_loss": 0.0010549204307608306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.280636108512628,
      "flow_matching_loss": 0.018249980732798578,
      "grad_norm": 0.1074146181344986,
      "learning_rate": 1.9599109169419466e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013458982459269464,
      "selected_pairs": 26213.2,
      "state_query_norm": 1120.3395385742188,
      "state_relation_loss": 0.0002909778122557327,
      "step": 21680,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0013458982459269464,
      "weighted_relation_loss": 0.0013458982459269464
    },
    {
      "action_query_norm": 636.4671264648438,
      "action_relation_loss": 0.0010051663732156158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.28999064546305,
      "flow_matching_loss": 0.018223705887794494,
      "grad_norm": 0.08730243146419525,
      "learning_rate": 1.9555370036661948e-05,
      "loss": 0.0195,
      "relation_loss": 0.0012933201505802572,
      "selected_pairs": 26351.1,
      "state_query_norm": 1112.0041870117188,
      "state_relation_loss": 0.0002881537599023432,
      "step": 21690,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0012933201505802572,
      "weighted_relation_loss": 0.0012933201505802572
    },
    {
      "action_query_norm": 637.3707702636718,
      "action_relation_loss": 0.0010186453757341952,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.29934518241347,
      "flow_matching_loss": 0.017402686923742295,
      "grad_norm": 0.11503661423921585,
      "learning_rate": 1.9511667897000574e-05,
      "loss": 0.0187,
      "relation_loss": 0.001301996491383761,
      "selected_pairs": 27743.2,
      "state_query_norm": 1121.1519897460937,
      "state_relation_loss": 0.0002833511243807152,
      "step": 21700,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.001301996491383761,
      "weighted_relation_loss": 0.001301996491383761
    },
    {
      "action_query_norm": 631.0951782226563,
      "action_relation_loss": 0.0010297736269421875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.30869971936389,
      "flow_matching_loss": 0.018259679805487393,
      "grad_norm": 0.08855868130922318,
      "learning_rate": 1.946800280353755e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013168904813937842,
      "selected_pairs": 26919.3,
      "state_query_norm": 1116.1156372070313,
      "state_relation_loss": 0.0002871168457204476,
      "step": 21710,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0013168904813937842,
      "weighted_relation_loss": 0.0013168904813937842
    },
    {
      "action_query_norm": 628.278515625,
      "action_relation_loss": 0.0010051999823190271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.318054256314312,
      "flow_matching_loss": 0.01788651179522276,
      "grad_norm": 0.09371621906757355,
      "learning_rate": 1.9424374809330114e-05,
      "loss": 0.0192,
      "relation_loss": 0.0012980859959498049,
      "selected_pairs": 26574.7,
      "state_query_norm": 1128.9271240234375,
      "state_relation_loss": 0.00029288601654116065,
      "step": 21720,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012980859959498049,
      "weighted_relation_loss": 0.0012980859959498049
    },
    {
      "action_query_norm": 636.0644348144531,
      "action_relation_loss": 0.001007754122838378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.327408793264734,
      "flow_matching_loss": 0.018850764632225035,
      "grad_norm": 0.10338090360164642,
      "learning_rate": 1.9380783967390382e-05,
      "loss": 0.0201,
      "relation_loss": 0.001298591319937259,
      "selected_pairs": 26828.6,
      "state_query_norm": 1117.0335327148437,
      "state_relation_loss": 0.00029083719418849797,
      "step": 21730,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001298591319937259,
      "weighted_relation_loss": 0.001298591319937259
    },
    {
      "action_query_norm": 630.7465881347656,
      "action_relation_loss": 0.0010542723233811558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.336763330215156,
      "flow_matching_loss": 0.01824661223217845,
      "grad_norm": 0.0911029800772667,
      "learning_rate": 1.933723033068533e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013488224474713206,
      "selected_pairs": 26989.8,
      "state_query_norm": 1131.8748779296875,
      "state_relation_loss": 0.00029455013282131404,
      "step": 21740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013488224474713206,
      "weighted_relation_loss": 0.0013488224474713206
    },
    {
      "action_query_norm": 627.3468811035157,
      "action_relation_loss": 0.0010994158161338418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.346117867165574,
      "flow_matching_loss": 0.018602099269628525,
      "grad_norm": 0.10837605595588684,
      "learning_rate": 1.929371395213674e-05,
      "loss": 0.02,
      "relation_loss": 0.0013949770480394363,
      "selected_pairs": 26476.8,
      "state_query_norm": 1127.7122436523437,
      "state_relation_loss": 0.0002955612493678927,
      "step": 21750,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0013949770480394363,
      "weighted_relation_loss": 0.0013949770480394363
    },
    {
      "action_query_norm": 627.6191345214844,
      "action_relation_loss": 0.001048993709264323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.355472404115996,
      "flow_matching_loss": 0.017643715627491473,
      "grad_norm": 0.08141323924064636,
      "learning_rate": 1.9250234884621092e-05,
      "loss": 0.019,
      "relation_loss": 0.001337140693794936,
      "selected_pairs": 26893.6,
      "state_query_norm": 1118.2681518554687,
      "state_relation_loss": 0.0002881469816202298,
      "step": 21760,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001337140693794936,
      "weighted_relation_loss": 0.001337140693794936
    },
    {
      "action_query_norm": 631.89521484375,
      "action_relation_loss": 0.0010118865931872279,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.36482694106642,
      "flow_matching_loss": 0.018563267309218645,
      "grad_norm": 0.07954834401607513,
      "learning_rate": 1.920679318096959e-05,
      "loss": 0.0199,
      "relation_loss": 0.0013075589435175062,
      "selected_pairs": 26781.5,
      "state_query_norm": 1112.4674194335937,
      "state_relation_loss": 0.0002956723445095122,
      "step": 21770,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0013075589435175062,
      "weighted_relation_loss": 0.0013075589435175062
    },
    {
      "action_query_norm": 627.4003051757812,
      "action_relation_loss": 0.0010314745595678687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.374181478016837,
      "flow_matching_loss": 0.018503420893102883,
      "grad_norm": 0.09556560218334198,
      "learning_rate": 1.9163388893967982e-05,
      "loss": 0.0198,
      "relation_loss": 0.0013239243533462286,
      "selected_pairs": 26751.3,
      "state_query_norm": 1125.614306640625,
      "state_relation_loss": 0.00029244977631606164,
      "step": 21780,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013239243533462286,
      "weighted_relation_loss": 0.0013239243533462286
    },
    {
      "action_query_norm": 637.3916931152344,
      "action_relation_loss": 0.0009846024331636727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.38353601496726,
      "flow_matching_loss": 0.01890127221122384,
      "grad_norm": 0.09080760180950165,
      "learning_rate": 1.9120022076356575e-05,
      "loss": 0.0202,
      "relation_loss": 0.0012756121926940978,
      "selected_pairs": 26578.3,
      "state_query_norm": 1118.146630859375,
      "state_relation_loss": 0.00029100977117195724,
      "step": 21790,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012756121926940978,
      "weighted_relation_loss": 0.0012756121926940978
    },
    {
      "action_query_norm": 630.1657653808594,
      "action_relation_loss": 0.0010139239253476261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.39289055191768,
      "flow_matching_loss": 0.01885300436988473,
      "grad_norm": 0.07967393845319748,
      "learning_rate": 1.9076692780830114e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013066609273664654,
      "selected_pairs": 26872.9,
      "state_query_norm": 1134.9087036132812,
      "state_relation_loss": 0.00029273699910845605,
      "step": 21800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0013066609273664654,
      "weighted_relation_loss": 0.0013066609273664654
    },
    {
      "action_query_norm": 635.8519287109375,
      "action_relation_loss": 0.0010381005646195262,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.4022450888681,
      "flow_matching_loss": 0.017363074887543916,
      "grad_norm": 0.0864311158657074,
      "learning_rate": 1.903340106003782e-05,
      "loss": 0.0187,
      "relation_loss": 0.0013260391075164079,
      "selected_pairs": 27103.6,
      "state_query_norm": 1122.263818359375,
      "state_relation_loss": 0.00028793853998649866,
      "step": 21810,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0013260391075164079,
      "weighted_relation_loss": 0.0013260391075164079
    },
    {
      "action_query_norm": 625.4762451171875,
      "action_relation_loss": 0.0010325415059924126,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.41159962581852,
      "flow_matching_loss": 0.019460069201886653,
      "grad_norm": 0.08076313883066177,
      "learning_rate": 1.8990146966583183e-05,
      "loss": 0.0208,
      "relation_loss": 0.00133040864020586,
      "selected_pairs": 26508.3,
      "state_query_norm": 1127.3894287109374,
      "state_relation_loss": 0.0002978671225719154,
      "step": 21820,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.00133040864020586,
      "weighted_relation_loss": 0.00133040864020586
    },
    {
      "action_query_norm": 626.9765686035156,
      "action_relation_loss": 0.0010179997712839394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.420954162768943,
      "flow_matching_loss": 0.020057907328009605,
      "grad_norm": 0.11590699851512909,
      "learning_rate": 1.8946930553024035e-05,
      "loss": 0.0214,
      "relation_loss": 0.0013181287213228643,
      "selected_pairs": 26831.3,
      "state_query_norm": 1121.433984375,
      "state_relation_loss": 0.00030012895877007395,
      "step": 21830,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013181287213228643,
      "weighted_relation_loss": 0.0013181287213228643
    },
    {
      "action_query_norm": 629.0532165527344,
      "action_relation_loss": 0.001056515466189012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.430308699719365,
      "flow_matching_loss": 0.019734665378928186,
      "grad_norm": 0.10237783193588257,
      "learning_rate": 1.8903751871872378e-05,
      "loss": 0.0211,
      "relation_loss": 0.0013474643346853554,
      "selected_pairs": 27083.6,
      "state_query_norm": 1130.90068359375,
      "state_relation_loss": 0.00029094885976519433,
      "step": 21840,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0013474643346853554,
      "weighted_relation_loss": 0.0013474643346853554
    },
    {
      "action_query_norm": 628.0304992675781,
      "action_relation_loss": 0.0010320288245566189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.439663236669784,
      "flow_matching_loss": 0.019295779801905156,
      "grad_norm": 0.11156118661165237,
      "learning_rate": 1.8860610975594382e-05,
      "loss": 0.0206,
      "relation_loss": 0.0013193691382184624,
      "selected_pairs": 27196.7,
      "state_query_norm": 1111.2694946289062,
      "state_relation_loss": 0.00028734031366184354,
      "step": 21850,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0013193691382184624,
      "weighted_relation_loss": 0.0013193691382184624
    },
    {
      "action_query_norm": 632.9369018554687,
      "action_relation_loss": 0.0010224675526842475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.449017773620206,
      "flow_matching_loss": 0.01810413058847189,
      "grad_norm": 0.08520418405532837,
      "learning_rate": 1.8817507916610307e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013128122896887361,
      "selected_pairs": 26897.6,
      "state_query_norm": 1121.3078125,
      "state_relation_loss": 0.0002903447370044887,
      "step": 21860,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013128122896887361,
      "weighted_relation_loss": 0.0013128122896887361
    },
    {
      "action_query_norm": 625.3419128417969,
      "action_relation_loss": 0.0010770313732791693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.458372310570628,
      "flow_matching_loss": 0.019480668380856513,
      "grad_norm": 0.07576938718557358,
      "learning_rate": 1.8774442747294407e-05,
      "loss": 0.0208,
      "relation_loss": 0.0013605240383185446,
      "selected_pairs": 26890.6,
      "state_query_norm": 1129.1971923828125,
      "state_relation_loss": 0.00028349265339784325,
      "step": 21870,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013605240383185446,
      "weighted_relation_loss": 0.0013605240383185446
    },
    {
      "action_query_norm": 634.6500793457031,
      "action_relation_loss": 0.001055858173640445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.467726847521046,
      "flow_matching_loss": 0.020037451945245265,
      "grad_norm": 0.08982940018177032,
      "learning_rate": 1.8731415519974965e-05,
      "loss": 0.0214,
      "relation_loss": 0.0013304106541909277,
      "selected_pairs": 27057.9,
      "state_query_norm": 1122.0419921875,
      "state_relation_loss": 0.00027455246017780153,
      "step": 21880,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0013304106541909277,
      "weighted_relation_loss": 0.0013304106541909277
    },
    {
      "action_query_norm": 629.5547668457032,
      "action_relation_loss": 0.001004559569992125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.477081384471468,
      "flow_matching_loss": 0.017764515615999697,
      "grad_norm": 0.08791456371545792,
      "learning_rate": 1.86884262869341e-05,
      "loss": 0.0191,
      "relation_loss": 0.0012922368128784,
      "selected_pairs": 26800.1,
      "state_query_norm": 1122.3549560546876,
      "state_relation_loss": 0.0002876772428862751,
      "step": 21890,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012922368128784,
      "weighted_relation_loss": 0.0012922368128784
    },
    {
      "action_query_norm": 637.5503234863281,
      "action_relation_loss": 0.001063506817445159,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.48643592142189,
      "flow_matching_loss": 0.019070072565227747,
      "grad_norm": 0.09544450789690018,
      "learning_rate": 1.8645475100407788e-05,
      "loss": 0.0204,
      "relation_loss": 0.0013581449864432215,
      "selected_pairs": 27361.3,
      "state_query_norm": 1109.035595703125,
      "state_relation_loss": 0.0002946381748188287,
      "step": 21900,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0013581449864432215,
      "weighted_relation_loss": 0.0013581449864432215
    },
    {
      "action_query_norm": 636.7919311523438,
      "action_relation_loss": 0.0010098312806803734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.495790458372312,
      "flow_matching_loss": 0.017901837546378373,
      "grad_norm": 0.10563573986291885,
      "learning_rate": 1.8602562012585768e-05,
      "loss": 0.0192,
      "relation_loss": 0.001300163718406111,
      "selected_pairs": 26368.4,
      "state_query_norm": 1135.933056640625,
      "state_relation_loss": 0.00029033244645688684,
      "step": 21910,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001300163718406111,
      "weighted_relation_loss": 0.001300163718406111
    },
    {
      "action_query_norm": 624.5778381347657,
      "action_relation_loss": 0.001092304178746417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.50514499532273,
      "flow_matching_loss": 0.01992968413978815,
      "grad_norm": 0.10565078258514404,
      "learning_rate": 1.8559687075611466e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013860323466360568,
      "selected_pairs": 27059.1,
      "state_query_norm": 1123.8381225585938,
      "state_relation_loss": 0.00029372816788963974,
      "step": 21920,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0013860323466360568,
      "weighted_relation_loss": 0.0013860323466360568
    },
    {
      "action_query_norm": 630.153173828125,
      "action_relation_loss": 0.001065124513115734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.514499532273152,
      "flow_matching_loss": 0.01969719585031271,
      "grad_norm": 0.08490443974733353,
      "learning_rate": 1.8516850341582017e-05,
      "loss": 0.0211,
      "relation_loss": 0.001356931950431317,
      "selected_pairs": 26295.7,
      "state_query_norm": 1120.5145263671875,
      "state_relation_loss": 0.000291807422763668,
      "step": 21930,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.001356931950431317,
      "weighted_relation_loss": 0.001356931950431317
    },
    {
      "action_query_norm": 639.0093566894532,
      "action_relation_loss": 0.00100944135338068,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.523854069223574,
      "flow_matching_loss": 0.01752747418358922,
      "grad_norm": 0.08837505429983139,
      "learning_rate": 1.847405186254807e-05,
      "loss": 0.0188,
      "relation_loss": 0.0013047407381236553,
      "selected_pairs": 27030.2,
      "state_query_norm": 1120.5859619140624,
      "state_relation_loss": 0.00029529938765335826,
      "step": 21940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0013047407381236553,
      "weighted_relation_loss": 0.0013047407381236553
    },
    {
      "action_query_norm": 636.8811340332031,
      "action_relation_loss": 0.0010591964703053236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.533208606173993,
      "flow_matching_loss": 0.017602087929844856,
      "grad_norm": 0.08368344604969025,
      "learning_rate": 1.843129169051379e-05,
      "loss": 0.0189,
      "relation_loss": 0.0013460071408189834,
      "selected_pairs": 27551.2,
      "state_query_norm": 1118.4004150390624,
      "state_relation_loss": 0.00028681066760327667,
      "step": 21950,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0013460071408189834,
      "weighted_relation_loss": 0.0013460071408189834
    },
    {
      "action_query_norm": 629.6830871582031,
      "action_relation_loss": 0.001051820459542796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.542563143124415,
      "flow_matching_loss": 0.016542603448033332,
      "grad_norm": 0.0887065902352333,
      "learning_rate": 1.8388569877436863e-05,
      "loss": 0.0179,
      "relation_loss": 0.0013443213421851397,
      "selected_pairs": 27107.6,
      "state_query_norm": 1110.2070190429688,
      "state_relation_loss": 0.0002925008622696623,
      "step": 21960,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0013443213421851397,
      "weighted_relation_loss": 0.0013443213421851397
    },
    {
      "action_query_norm": 630.318310546875,
      "action_relation_loss": 0.0010166452964767813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.551917680074837,
      "flow_matching_loss": 0.018742647487670183,
      "grad_norm": 0.09744662046432495,
      "learning_rate": 1.834588647522828e-05,
      "loss": 0.02,
      "relation_loss": 0.001305816532112658,
      "selected_pairs": 26734.5,
      "state_query_norm": 1129.4041625976563,
      "state_relation_loss": 0.00028917121526319536,
      "step": 21970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001305816532112658,
      "weighted_relation_loss": 0.001305816532112658
    },
    {
      "action_query_norm": 628.3983520507812,
      "action_relation_loss": 0.001061824633507058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.56127221702526,
      "flow_matching_loss": 0.017888545989990234,
      "grad_norm": 0.07584354281425476,
      "learning_rate": 1.8303241535752435e-05,
      "loss": 0.0192,
      "relation_loss": 0.0013422330026514827,
      "selected_pairs": 26545.4,
      "state_query_norm": 1134.6615478515625,
      "state_relation_loss": 0.00028040836041327567,
      "step": 21980,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013422330026514827,
      "weighted_relation_loss": 0.0013422330026514827
    },
    {
      "action_query_norm": 631.2869567871094,
      "action_relation_loss": 0.0010281674680300058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.570626753975677,
      "flow_matching_loss": 0.017922523897141218,
      "grad_norm": 0.09309180825948715,
      "learning_rate": 1.8260635110826934e-05,
      "loss": 0.0192,
      "relation_loss": 0.001321362005546689,
      "selected_pairs": 26805.9,
      "state_query_norm": 1129.3594604492187,
      "state_relation_loss": 0.00029319454042706636,
      "step": 21990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001321362005546689,
      "weighted_relation_loss": 0.001321362005546689
    },
    {
      "action_query_norm": 630.5916931152344,
      "action_relation_loss": 0.0009983494121115655,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.5799812909261,
      "flow_matching_loss": 0.018042154517024754,
      "grad_norm": 0.10614234209060669,
      "learning_rate": 1.8218067252222597e-05,
      "loss": 0.0193,
      "relation_loss": 0.001288419310003519,
      "selected_pairs": 27059.3,
      "state_query_norm": 1112.6069091796876,
      "state_relation_loss": 0.0002900698833400384,
      "step": 22000,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001288419310003519,
      "weighted_relation_loss": 0.001288419310003519
    },
    {
      "action_query_norm": 626.326513671875,
      "action_relation_loss": 0.001020535477437079,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.58933582787652,
      "flow_matching_loss": 0.01764317797496915,
      "grad_norm": 0.08252272009849548,
      "learning_rate": 1.817553801166339e-05,
      "loss": 0.019,
      "relation_loss": 0.0013173420913517476,
      "selected_pairs": 26858.9,
      "state_query_norm": 1099.520849609375,
      "state_relation_loss": 0.0002968066197354347,
      "step": 22010,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0013173420913517476,
      "weighted_relation_loss": 0.0013173420913517476
    },
    {
      "action_query_norm": 629.1784729003906,
      "action_relation_loss": 0.0010269825812429189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.59869036482694,
      "flow_matching_loss": 0.01818981934338808,
      "grad_norm": 0.09868372231721878,
      "learning_rate": 1.8133047440826333e-05,
      "loss": 0.0195,
      "relation_loss": 0.0013155801803804933,
      "selected_pairs": 26898.8,
      "state_query_norm": 1128.8597412109375,
      "state_relation_loss": 0.00028859760204795747,
      "step": 22020,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013155801803804933,
      "weighted_relation_loss": 0.0013155801803804933
    },
    {
      "action_query_norm": 629.9401245117188,
      "action_relation_loss": 0.0010192763991653919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.60804490177736,
      "flow_matching_loss": 0.018364062532782555,
      "grad_norm": 0.11671803891658783,
      "learning_rate": 1.809059559134151e-05,
      "loss": 0.0197,
      "relation_loss": 0.0013022308354265987,
      "selected_pairs": 27251.2,
      "state_query_norm": 1126.3865356445312,
      "state_relation_loss": 0.000282954444992356,
      "step": 22030,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0013022308354265987,
      "weighted_relation_loss": 0.0013022308354265987
    },
    {
      "action_query_norm": 627.9128051757813,
      "action_relation_loss": 0.0010387927119154482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.617399438727784,
      "flow_matching_loss": 0.019926314987242223,
      "grad_norm": 0.1109720841050148,
      "learning_rate": 1.80481825147919e-05,
      "loss": 0.0213,
      "relation_loss": 0.001334646821487695,
      "selected_pairs": 26663.5,
      "state_query_norm": 1129.5643310546875,
      "state_relation_loss": 0.00029585408919956534,
      "step": 22040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001334646821487695,
      "weighted_relation_loss": 0.001334646821487695
    },
    {
      "action_query_norm": 633.0750671386719,
      "action_relation_loss": 0.0010437516146339476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.626753975678206,
      "flow_matching_loss": 0.01979524679481983,
      "grad_norm": 0.08348878473043442,
      "learning_rate": 1.8005808262713396e-05,
      "loss": 0.0211,
      "relation_loss": 0.0013287185924127699,
      "selected_pairs": 26439.1,
      "state_query_norm": 1134.9838134765625,
      "state_relation_loss": 0.0002849669981515035,
      "step": 22050,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0013287185924127699,
      "weighted_relation_loss": 0.0013287185924127699
    },
    {
      "action_query_norm": 628.7299194335938,
      "action_relation_loss": 0.0010632729332428425,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.636108512628624,
      "flow_matching_loss": 0.017780604399740697,
      "grad_norm": 0.08311794698238373,
      "learning_rate": 1.7963472886594714e-05,
      "loss": 0.0191,
      "relation_loss": 0.0013586084009148181,
      "selected_pairs": 26734.3,
      "state_query_norm": 1118.4851440429688,
      "state_relation_loss": 0.00029533546476159247,
      "step": 22060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013586084009148181,
      "weighted_relation_loss": 0.0013586084009148181
    },
    {
      "action_query_norm": 626.1337890625,
      "action_relation_loss": 0.0010301803471520544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.645463049579046,
      "flow_matching_loss": 0.01793008269742131,
      "grad_norm": 0.09930205345153809,
      "learning_rate": 1.7921176437877302e-05,
      "loss": 0.0192,
      "relation_loss": 0.0013109572813846172,
      "selected_pairs": 27153.4,
      "state_query_norm": 1125.4527709960937,
      "state_relation_loss": 0.00028077695169486104,
      "step": 22070,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013109572813846172,
      "weighted_relation_loss": 0.0013109572813846172
    },
    {
      "action_query_norm": 629.26123046875,
      "action_relation_loss": 0.0010899626708123834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.654817586529468,
      "flow_matching_loss": 0.01863693669438362,
      "grad_norm": 0.09415395557880402,
      "learning_rate": 1.7878918967955366e-05,
      "loss": 0.02,
      "relation_loss": 0.0013904690509662031,
      "selected_pairs": 27012.4,
      "state_query_norm": 1131.3645874023437,
      "state_relation_loss": 0.00030050640052650126,
      "step": 22080,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013904690509662031,
      "weighted_relation_loss": 0.0013904690509662031
    },
    {
      "action_query_norm": 628.5194274902344,
      "action_relation_loss": 0.001038995647104457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.664172123479887,
      "flow_matching_loss": 0.01806389568373561,
      "grad_norm": 0.10847495496273041,
      "learning_rate": 1.7836700528175693e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013199356966651977,
      "selected_pairs": 27285.6,
      "state_query_norm": 1121.3038940429688,
      "state_relation_loss": 0.00028094006702303884,
      "step": 22090,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0013199356966651977,
      "weighted_relation_loss": 0.0013199356966651977
    },
    {
      "action_query_norm": 632.4941284179688,
      "action_relation_loss": 0.001048301171977073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.67352666043031,
      "flow_matching_loss": 0.019018523022532464,
      "grad_norm": 0.08597090095281601,
      "learning_rate": 1.779452116983769e-05,
      "loss": 0.0204,
      "relation_loss": 0.0013431059895083308,
      "selected_pairs": 26159.4,
      "state_query_norm": 1113.0302734375,
      "state_relation_loss": 0.00029480480297934265,
      "step": 22100,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0013431059895083308,
      "weighted_relation_loss": 0.0013431059895083308
    },
    {
      "action_query_norm": 627.58828125,
      "action_relation_loss": 0.0010460161371156572,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.68288119738073,
      "flow_matching_loss": 0.018366299010813235,
      "grad_norm": 0.10572686046361923,
      "learning_rate": 1.775238094419325e-05,
      "loss": 0.0197,
      "relation_loss": 0.0013373123831115663,
      "selected_pairs": 27571.0,
      "state_query_norm": 1119.1521606445312,
      "state_relation_loss": 0.00029129622562322765,
      "step": 22110,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0013373123831115663,
      "weighted_relation_loss": 0.0013373123831115663
    },
    {
      "action_query_norm": 628.8320739746093,
      "action_relation_loss": 0.0010158744582440705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.69223573433115,
      "flow_matching_loss": 0.018001923430711032,
      "grad_norm": 0.09627516567707062,
      "learning_rate": 1.7710279902446708e-05,
      "loss": 0.0193,
      "relation_loss": 0.0013105814810842275,
      "selected_pairs": 26951.5,
      "state_query_norm": 1128.2417846679687,
      "state_relation_loss": 0.000294707011198625,
      "step": 22120,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013105814810842275,
      "weighted_relation_loss": 0.0013105814810842275
    },
    {
      "action_query_norm": 622.3798461914063,
      "action_relation_loss": 0.0010121673112735152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.70159027128157,
      "flow_matching_loss": 0.01877726363018155,
      "grad_norm": 0.09314122051000595,
      "learning_rate": 1.7668218095754795e-05,
      "loss": 0.0201,
      "relation_loss": 0.0012976123252883554,
      "selected_pairs": 26456.8,
      "state_query_norm": 1117.941162109375,
      "state_relation_loss": 0.00028544501692522316,
      "step": 22130,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012976123252883554,
      "weighted_relation_loss": 0.0012976123252883554
    },
    {
      "action_query_norm": 632.8471008300781,
      "action_relation_loss": 0.001027479919139296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.710944808231993,
      "flow_matching_loss": 0.018097475450485945,
      "grad_norm": 0.09920438379049301,
      "learning_rate": 1.7626195575226594e-05,
      "loss": 0.0194,
      "relation_loss": 0.001323754503391683,
      "selected_pairs": 26898.1,
      "state_query_norm": 1112.4870239257812,
      "state_relation_loss": 0.0002962745929835364,
      "step": 22140,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001323754503391683,
      "weighted_relation_loss": 0.001323754503391683
    },
    {
      "action_query_norm": 628.0772277832032,
      "action_relation_loss": 0.0010665785637684166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.720299345182415,
      "flow_matching_loss": 0.018882748670876026,
      "grad_norm": 0.09953907877206802,
      "learning_rate": 1.758421239192343e-05,
      "loss": 0.0202,
      "relation_loss": 0.001358653837814927,
      "selected_pairs": 26824.9,
      "state_query_norm": 1125.7500610351562,
      "state_relation_loss": 0.00029207527404651046,
      "step": 22150,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001358653837814927,
      "weighted_relation_loss": 0.001358653837814927
    },
    {
      "action_query_norm": 629.8704956054687,
      "action_relation_loss": 0.001043705886695534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.729653882132833,
      "flow_matching_loss": 0.019786464236676692,
      "grad_norm": 0.09309758991003036,
      "learning_rate": 1.7542268596858814e-05,
      "loss": 0.0211,
      "relation_loss": 0.0013300086138769984,
      "selected_pairs": 27184.2,
      "state_query_norm": 1138.0122802734375,
      "state_relation_loss": 0.0002863027126295492,
      "step": 22160,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0013300086138769984,
      "weighted_relation_loss": 0.0013300086138769984
    },
    {
      "action_query_norm": 630.0294799804688,
      "action_relation_loss": 0.0010226341488305478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.739008419083255,
      "flow_matching_loss": 0.01889063660055399,
      "grad_norm": 0.08519380539655685,
      "learning_rate": 1.750036424099841e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013171346741728484,
      "selected_pairs": 26924.0,
      "state_query_norm": 1119.0673706054688,
      "state_relation_loss": 0.00029450051952153446,
      "step": 22170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0013171346741728484,
      "weighted_relation_loss": 0.0013171346741728484
    },
    {
      "action_query_norm": 624.7974060058593,
      "action_relation_loss": 0.000993578159250319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.748362956033677,
      "flow_matching_loss": 0.01623186357319355,
      "grad_norm": 0.08913127332925797,
      "learning_rate": 1.7458499375259956e-05,
      "loss": 0.0175,
      "relation_loss": 0.0012738943914882839,
      "selected_pairs": 26410.8,
      "state_query_norm": 1117.0375,
      "state_relation_loss": 0.0002803162467898801,
      "step": 22180,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0012738943914882839,
      "weighted_relation_loss": 0.0012738943914882839
    },
    {
      "action_query_norm": 632.4586669921875,
      "action_relation_loss": 0.000998197152512148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.757717492984096,
      "flow_matching_loss": 0.017806862946599722,
      "grad_norm": 0.10184511542320251,
      "learning_rate": 1.7416674050513245e-05,
      "loss": 0.0191,
      "relation_loss": 0.0012914332444779574,
      "selected_pairs": 26833.5,
      "state_query_norm": 1128.6539306640625,
      "state_relation_loss": 0.00029323608905542644,
      "step": 22190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012914332444779574,
      "weighted_relation_loss": 0.0012914332444779574
    },
    {
      "action_query_norm": 621.98681640625,
      "action_relation_loss": 0.0010280144750140607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.767072029934518,
      "flow_matching_loss": 0.0188869409263134,
      "grad_norm": 0.08120355010032654,
      "learning_rate": 1.737488831757997e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013295412994921208,
      "selected_pairs": 27476.9,
      "state_query_norm": 1118.6114990234375,
      "state_relation_loss": 0.00030152682447806003,
      "step": 22200,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0013295412994921208,
      "weighted_relation_loss": 0.0013295412994921208
    },
    {
      "action_query_norm": 627.8803344726563,
      "action_relation_loss": 0.0009887362888548524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.77642656688494,
      "flow_matching_loss": 0.017848583590239286,
      "grad_norm": 0.07194679230451584,
      "learning_rate": 1.733314222723373e-05,
      "loss": 0.0191,
      "relation_loss": 0.0012796776834875345,
      "selected_pairs": 27080.9,
      "state_query_norm": 1121.5868530273438,
      "state_relation_loss": 0.0002909413742600009,
      "step": 22210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012796776834875345,
      "weighted_relation_loss": 0.0012796776834875345
    },
    {
      "action_query_norm": 630.3043823242188,
      "action_relation_loss": 0.0010619528067763895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.78578110383536,
      "flow_matching_loss": 0.017785706929862498,
      "grad_norm": 0.08873964101076126,
      "learning_rate": 1.7291435830199954e-05,
      "loss": 0.0191,
      "relation_loss": 0.0013465756084769965,
      "selected_pairs": 26988.9,
      "state_query_norm": 1126.63837890625,
      "state_relation_loss": 0.00028462278132792564,
      "step": 22220,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013465756084769965,
      "weighted_relation_loss": 0.0013465756084769965
    },
    {
      "action_query_norm": 638.1064636230469,
      "action_relation_loss": 0.000982249272055924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.79513564078578,
      "flow_matching_loss": 0.019634125102311373,
      "grad_norm": 0.09251616150140762,
      "learning_rate": 1.7249769177155878e-05,
      "loss": 0.0209,
      "relation_loss": 0.0012727390276268125,
      "selected_pairs": 27081.1,
      "state_query_norm": 1128.7034912109375,
      "state_relation_loss": 0.0002904897759435698,
      "step": 22230,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012727390276268125,
      "weighted_relation_loss": 0.0012727390276268125
    },
    {
      "action_query_norm": 635.32890625,
      "action_relation_loss": 0.0009788996889255942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.804490177736202,
      "flow_matching_loss": 0.020206867158412932,
      "grad_norm": 0.09574772417545319,
      "learning_rate": 1.720814231873038e-05,
      "loss": 0.0215,
      "relation_loss": 0.001268826227169484,
      "selected_pairs": 26829.7,
      "state_query_norm": 1108.8384765625,
      "state_relation_loss": 0.00028992654115427284,
      "step": 22240,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001268826227169484,
      "weighted_relation_loss": 0.001268826227169484
    },
    {
      "action_query_norm": 634.4502746582032,
      "action_relation_loss": 0.0010252986394334584,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.813844714686624,
      "flow_matching_loss": 0.018888287153095006,
      "grad_norm": 0.10161510109901428,
      "learning_rate": 1.716655530550405e-05,
      "loss": 0.0202,
      "relation_loss": 0.0013148449012078344,
      "selected_pairs": 27316.2,
      "state_query_norm": 1124.36513671875,
      "state_relation_loss": 0.00028954627050552516,
      "step": 22250,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0013148449012078344,
      "weighted_relation_loss": 0.0013148449012078344
    },
    {
      "action_query_norm": 632.3427429199219,
      "action_relation_loss": 0.0010582453687675297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.823199251637043,
      "flow_matching_loss": 0.019683385640382765,
      "grad_norm": 0.08925843983888626,
      "learning_rate": 1.7125008188009017e-05,
      "loss": 0.021,
      "relation_loss": 0.00134327607229352,
      "selected_pairs": 27339.7,
      "state_query_norm": 1117.95302734375,
      "state_relation_loss": 0.00028503070352599026,
      "step": 22260,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.00134327607229352,
      "weighted_relation_loss": 0.00134327607229352
    },
    {
      "action_query_norm": 629.188916015625,
      "action_relation_loss": 0.0009894988092128188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.832553788587465,
      "flow_matching_loss": 0.01759641310200095,
      "grad_norm": 0.10234920680522919,
      "learning_rate": 1.708350101672894e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012685080175288022,
      "selected_pairs": 26755.1,
      "state_query_norm": 1115.7827270507812,
      "state_relation_loss": 0.0002790092054056004,
      "step": 22270,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012685080175288022,
      "weighted_relation_loss": 0.0012685080175288022
    },
    {
      "action_query_norm": 629.3777038574219,
      "action_relation_loss": 0.0009911977627780288,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.841908325537887,
      "flow_matching_loss": 0.019209834281355143,
      "grad_norm": 0.10528970509767532,
      "learning_rate": 1.7042033842098957e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012756991083733738,
      "selected_pairs": 26828.5,
      "state_query_norm": 1118.3112426757812,
      "state_relation_loss": 0.00028450133395381273,
      "step": 22280,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012756991083733738,
      "weighted_relation_loss": 0.0012756991083733738
    },
    {
      "action_query_norm": 627.6649780273438,
      "action_relation_loss": 0.0010080473555717617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.85126286248831,
      "flow_matching_loss": 0.021171654388308526,
      "grad_norm": 0.08391083031892776,
      "learning_rate": 1.7000606714505567e-05,
      "loss": 0.0225,
      "relation_loss": 0.0012989673297852278,
      "selected_pairs": 26630.9,
      "state_query_norm": 1119.4688110351562,
      "state_relation_loss": 0.00029091996839269996,
      "step": 22290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012989673297852278,
      "weighted_relation_loss": 0.0012989673297852278
    },
    {
      "action_query_norm": 635.4939392089843,
      "action_relation_loss": 0.0010202634730376303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.860617399438727,
      "flow_matching_loss": 0.01897835936397314,
      "grad_norm": 0.08985300362110138,
      "learning_rate": 1.695921968428668e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013052211608737708,
      "selected_pairs": 27014.8,
      "state_query_norm": 1123.5453002929687,
      "state_relation_loss": 0.0002849576820153743,
      "step": 22300,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013052211608737708,
      "weighted_relation_loss": 0.0013052211608737708
    },
    {
      "action_query_norm": 632.5916564941406,
      "action_relation_loss": 0.0010525931604206562,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.86997193638915,
      "flow_matching_loss": 0.01723001981154084,
      "grad_norm": 0.10099584609270096,
      "learning_rate": 1.6917872801731416e-05,
      "loss": 0.0186,
      "relation_loss": 0.001332375593483448,
      "selected_pairs": 26419.6,
      "state_query_norm": 1116.140869140625,
      "state_relation_loss": 0.0002797824505250901,
      "step": 22310,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001332375593483448,
      "weighted_relation_loss": 0.001332375593483448
    },
    {
      "action_query_norm": 627.0656188964844,
      "action_relation_loss": 0.0010365445632487535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.87932647333957,
      "flow_matching_loss": 0.018262311443686487,
      "grad_norm": 0.07891087979078293,
      "learning_rate": 1.687656611708014e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013390376465395094,
      "selected_pairs": 26506.9,
      "state_query_norm": 1112.7713134765625,
      "state_relation_loss": 0.00030249307746998966,
      "step": 22320,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0013390376465395094,
      "weighted_relation_loss": 0.0013390376465395094
    },
    {
      "action_query_norm": 628.5621643066406,
      "action_relation_loss": 0.0009745322051458061,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.88868101028999,
      "flow_matching_loss": 0.01946717398241162,
      "grad_norm": 0.08031956106424332,
      "learning_rate": 1.6835299680524368e-05,
      "loss": 0.0207,
      "relation_loss": 0.0012721385108307004,
      "selected_pairs": 26667.1,
      "state_query_norm": 1119.3083740234374,
      "state_relation_loss": 0.0002976063115056604,
      "step": 22330,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012721385108307004,
      "weighted_relation_loss": 0.0012721385108307004
    },
    {
      "action_query_norm": 626.4580444335937,
      "action_relation_loss": 0.0010003623145166784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.89803554724041,
      "flow_matching_loss": 0.01709290165454149,
      "grad_norm": 0.08447465300559998,
      "learning_rate": 1.679407354220669e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012975226622074843,
      "selected_pairs": 26601.8,
      "state_query_norm": 1117.429833984375,
      "state_relation_loss": 0.00029716035933233795,
      "step": 22340,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0012975226622074843,
      "weighted_relation_loss": 0.0012975226622074843
    },
    {
      "action_query_norm": 633.3434326171875,
      "action_relation_loss": 0.0009960353083442897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.907390084190833,
      "flow_matching_loss": 0.01586221354082227,
      "grad_norm": 0.09805990755558014,
      "learning_rate": 1.6752887752220793e-05,
      "loss": 0.0171,
      "relation_loss": 0.0012809803592972457,
      "selected_pairs": 26535.4,
      "state_query_norm": 1110.399658203125,
      "state_relation_loss": 0.0002849450393114239,
      "step": 22350,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0012809803592972457,
      "weighted_relation_loss": 0.0012809803592972457
    },
    {
      "action_query_norm": 626.9453369140625,
      "action_relation_loss": 0.0010284631280228496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.916744621141255,
      "flow_matching_loss": 0.016687069088220596,
      "grad_norm": 0.10357022285461426,
      "learning_rate": 1.6711742360611277e-05,
      "loss": 0.018,
      "relation_loss": 0.0013090878375805915,
      "selected_pairs": 26734.7,
      "state_query_norm": 1111.8517211914063,
      "state_relation_loss": 0.0002806246979162097,
      "step": 22360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013090878375805915,
      "weighted_relation_loss": 0.0013090878375805915
    },
    {
      "action_query_norm": 633.0004821777344,
      "action_relation_loss": 0.0010239953291602433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.926099158091674,
      "flow_matching_loss": 0.01824226463213563,
      "grad_norm": 0.09733494371175766,
      "learning_rate": 1.6670637417373652e-05,
      "loss": 0.0196,
      "relation_loss": 0.001309990359004587,
      "selected_pairs": 26741.2,
      "state_query_norm": 1113.2634643554688,
      "state_relation_loss": 0.00028599502693396064,
      "step": 22370,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001309990359004587,
      "weighted_relation_loss": 0.001309990359004587
    },
    {
      "action_query_norm": 631.7234436035156,
      "action_relation_loss": 0.0010233718145173042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.935453695042096,
      "flow_matching_loss": 0.019339110422879457,
      "grad_norm": 0.12511010468006134,
      "learning_rate": 1.6629572972454333e-05,
      "loss": 0.0207,
      "relation_loss": 0.0013134822132997216,
      "selected_pairs": 27665.1,
      "state_query_norm": 1119.1583862304688,
      "state_relation_loss": 0.0002901104075135663,
      "step": 22380,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0013134822132997216,
      "weighted_relation_loss": 0.0013134822132997216
    },
    {
      "action_query_norm": 621.9290893554687,
      "action_relation_loss": 0.000983027828624472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.944808231992518,
      "flow_matching_loss": 0.016926049254834652,
      "grad_norm": 0.10373812168836594,
      "learning_rate": 1.6588549075750464e-05,
      "loss": 0.0182,
      "relation_loss": 0.0012647068360820413,
      "selected_pairs": 26462.8,
      "state_query_norm": 1123.6442749023438,
      "state_relation_loss": 0.0002816790132783353,
      "step": 22390,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012647068360820413,
      "weighted_relation_loss": 0.0012647068360820413
    },
    {
      "action_query_norm": 626.4488891601562,
      "action_relation_loss": 0.0010110928909853102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.954162768942936,
      "flow_matching_loss": 0.017682718671858312,
      "grad_norm": 0.09521610289812088,
      "learning_rate": 1.6547565777109976e-05,
      "loss": 0.019,
      "relation_loss": 0.001305647613480687,
      "selected_pairs": 26876.1,
      "state_query_norm": 1107.965185546875,
      "state_relation_loss": 0.00029455472249537704,
      "step": 22400,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001305647613480687,
      "weighted_relation_loss": 0.001305647613480687
    },
    {
      "action_query_norm": 629.2025390625,
      "action_relation_loss": 0.001051681803073734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.963517305893358,
      "flow_matching_loss": 0.018102767132222652,
      "grad_norm": 0.09369374066591263,
      "learning_rate": 1.650662312633143e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013398557784967124,
      "selected_pairs": 26590.7,
      "state_query_norm": 1130.0285522460938,
      "state_relation_loss": 0.0002881739608710632,
      "step": 22410,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013398557784967124,
      "weighted_relation_loss": 0.0013398557784967124
    },
    {
      "action_query_norm": 624.68525390625,
      "action_relation_loss": 0.0010273238469380886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.97287184284378,
      "flow_matching_loss": 0.018291497696191074,
      "grad_norm": 0.1012118011713028,
      "learning_rate": 1.6465721173164002e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013259039493277668,
      "selected_pairs": 26179.4,
      "state_query_norm": 1117.6335693359374,
      "state_relation_loss": 0.0002985800907481462,
      "step": 22420,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0013259039493277668,
      "weighted_relation_loss": 0.0013259039493277668
    },
    {
      "action_query_norm": 620.1818969726562,
      "action_relation_loss": 0.0010601011104881763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.9822263797942,
      "flow_matching_loss": 0.018375710863620043,
      "grad_norm": 0.08716459572315216,
      "learning_rate": 1.6424859967307428e-05,
      "loss": 0.0197,
      "relation_loss": 0.001349369005765766,
      "selected_pairs": 26741.2,
      "state_query_norm": 1098.6384765625,
      "state_relation_loss": 0.00028926788945682347,
      "step": 22430,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001349369005765766,
      "weighted_relation_loss": 0.001349369005765766
    },
    {
      "action_query_norm": 631.6368469238281,
      "action_relation_loss": 0.0009621993289329111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.99158091674462,
      "flow_matching_loss": 0.016485729813575746,
      "grad_norm": 0.12090153992176056,
      "learning_rate": 1.6384039558411903e-05,
      "loss": 0.0177,
      "relation_loss": 0.0012384850531816482,
      "selected_pairs": 27182.4,
      "state_query_norm": 1118.725927734375,
      "state_relation_loss": 0.00027628570969682185,
      "step": 22440,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0012384850531816482,
      "weighted_relation_loss": 0.0012384850531816482
    },
    {
      "action_query_norm": 624.8426025390625,
      "action_relation_loss": 0.0009804937290027738,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.000935453695043,
      "flow_matching_loss": 0.017651805374771356,
      "grad_norm": 0.11771222203969955,
      "learning_rate": 1.634325999607811e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012642799643799662,
      "selected_pairs": 24271.0,
      "state_query_norm": 1111.8760864257813,
      "state_relation_loss": 0.0002837862237356603,
      "step": 22450,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.0012642799643799662,
      "weighted_relation_loss": 0.0012642799643799662
    },
    {
      "action_query_norm": 626.8955627441406,
      "action_relation_loss": 0.000980685168178752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.010289990645465,
      "flow_matching_loss": 0.018051253352314233,
      "grad_norm": 0.09622563421726227,
      "learning_rate": 1.6302521329857045e-05,
      "loss": 0.0193,
      "relation_loss": 0.0012694406672380865,
      "selected_pairs": 27089.3,
      "state_query_norm": 1126.0963500976563,
      "state_relation_loss": 0.00028875552234239874,
      "step": 22460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012694406672380865,
      "weighted_relation_loss": 0.0012694406672380865
    },
    {
      "action_query_norm": 633.6439086914063,
      "action_relation_loss": 0.0009410387021489442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.019644527595883,
      "flow_matching_loss": 0.01896800398826599,
      "grad_norm": 0.09285632520914078,
      "learning_rate": 1.6261823609250027e-05,
      "loss": 0.0202,
      "relation_loss": 0.001223608886357397,
      "selected_pairs": 27010.6,
      "state_query_norm": 1105.8539428710938,
      "state_relation_loss": 0.00028257016092538836,
      "step": 22470,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001223608886357397,
      "weighted_relation_loss": 0.001223608886357397
    },
    {
      "action_query_norm": 632.4100769042968,
      "action_relation_loss": 0.0010063856898341328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.028999064546305,
      "flow_matching_loss": 0.016999637894332408,
      "grad_norm": 0.0836559385061264,
      "learning_rate": 1.6221166883708627e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012854803120717407,
      "selected_pairs": 27222.1,
      "state_query_norm": 1134.87099609375,
      "state_relation_loss": 0.00027909463678952304,
      "step": 22480,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0012854803120717407,
      "weighted_relation_loss": 0.0012854803120717407
    },
    {
      "action_query_norm": 625.5078186035156,
      "action_relation_loss": 0.0009575508767738939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.038353601496727,
      "flow_matching_loss": 0.018349161371588707,
      "grad_norm": 0.09243756532669067,
      "learning_rate": 1.6180551202634603e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012378611485473812,
      "selected_pairs": 26906.2,
      "state_query_norm": 1129.2337646484375,
      "state_relation_loss": 0.00028031027759425344,
      "step": 22490,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0012378611485473812,
      "weighted_relation_loss": 0.0012378611485473812
    },
    {
      "action_query_norm": 626.4893005371093,
      "action_relation_loss": 0.0010261145594995468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.047708138447145,
      "flow_matching_loss": 0.016477574687451126,
      "grad_norm": 0.0750335231423378,
      "learning_rate": 1.613997661537981e-05,
      "loss": 0.0178,
      "relation_loss": 0.0013071209308691324,
      "selected_pairs": 26996.3,
      "state_query_norm": 1118.3755493164062,
      "state_relation_loss": 0.0002810063597280532,
      "step": 22500,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013071209308691324,
      "weighted_relation_loss": 0.0013071209308691324
    },
    {
      "action_query_norm": 624.1188232421875,
      "action_relation_loss": 0.0010137911653146149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.057062675397567,
      "flow_matching_loss": 0.017370104696601628,
      "grad_norm": 0.07916776835918427,
      "learning_rate": 1.609944317124624e-05,
      "loss": 0.0187,
      "relation_loss": 0.001290725206490606,
      "selected_pairs": 27057.3,
      "state_query_norm": 1112.144873046875,
      "state_relation_loss": 0.00027693402662407605,
      "step": 22510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001290725206490606,
      "weighted_relation_loss": 0.001290725206490606
    },
    {
      "action_query_norm": 634.9937438964844,
      "action_relation_loss": 0.0009598744218237698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.06641721234799,
      "flow_matching_loss": 0.01718118544667959,
      "grad_norm": 0.10587392747402191,
      "learning_rate": 1.605895091948582e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012508679646998644,
      "selected_pairs": 26638.9,
      "state_query_norm": 1119.0861938476562,
      "state_relation_loss": 0.00029099354869686065,
      "step": 22520,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0012508679646998644,
      "weighted_relation_loss": 0.0012508679646998644
    },
    {
      "action_query_norm": 627.07080078125,
      "action_relation_loss": 0.0010179112723562866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.07577174929841,
      "flow_matching_loss": 0.017112852539867163,
      "grad_norm": 0.08224185556173325,
      "learning_rate": 1.6018499909300477e-05,
      "loss": 0.0184,
      "relation_loss": 0.001304020534735173,
      "selected_pairs": 27110.1,
      "state_query_norm": 1129.8492431640625,
      "state_relation_loss": 0.0002861092449165881,
      "step": 22530,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001304020534735173,
      "weighted_relation_loss": 0.001304020534735173
    },
    {
      "action_query_norm": 628.0502136230468,
      "action_relation_loss": 0.0009985463810153305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.08512628624883,
      "flow_matching_loss": 0.016582836210727692,
      "grad_norm": 0.08781689405441284,
      "learning_rate": 1.597809018984199e-05,
      "loss": 0.0179,
      "relation_loss": 0.0012842279742471875,
      "selected_pairs": 27153.8,
      "state_query_norm": 1117.4783813476563,
      "state_relation_loss": 0.00028568156994879246,
      "step": 22540,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0012842279742471875,
      "weighted_relation_loss": 0.0012842279742471875
    },
    {
      "action_query_norm": 622.7855529785156,
      "action_relation_loss": 0.0009780657885130495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.09448082319925,
      "flow_matching_loss": 0.017959956265985965,
      "grad_norm": 0.08165633678436279,
      "learning_rate": 1.593772181021196e-05,
      "loss": 0.0192,
      "relation_loss": 0.001254026358947158,
      "selected_pairs": 27152.7,
      "state_query_norm": 1111.5291748046875,
      "state_relation_loss": 0.0002759605646133423,
      "step": 22550,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001254026358947158,
      "weighted_relation_loss": 0.001254026358947158
    },
    {
      "action_query_norm": 630.4549560546875,
      "action_relation_loss": 0.0009953943081200123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.103835360149674,
      "flow_matching_loss": 0.017667207680642606,
      "grad_norm": 0.10958808660507202,
      "learning_rate": 1.5897394819461813e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012820518109947442,
      "selected_pairs": 27009.7,
      "state_query_norm": 1119.1810668945313,
      "state_relation_loss": 0.00028665750287473204,
      "step": 22560,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012820518109947442,
      "weighted_relation_loss": 0.0012820518109947442
    },
    {
      "action_query_norm": 635.5561584472656,
      "action_relation_loss": 0.0009847377717960627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.113189897100092,
      "flow_matching_loss": 0.017063549254089594,
      "grad_norm": 0.08531832695007324,
      "learning_rate": 1.5857109266592625e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012815633323043584,
      "selected_pairs": 27103.4,
      "state_query_norm": 1124.0931762695313,
      "state_relation_loss": 0.00029682555759791285,
      "step": 22570,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012815633323043584,
      "weighted_relation_loss": 0.0012815633323043584
    },
    {
      "action_query_norm": 627.5181335449219,
      "action_relation_loss": 0.0010455240204464643,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.122544434050514,
      "flow_matching_loss": 0.019334872253239155,
      "grad_norm": 0.08768408745527267,
      "learning_rate": 1.581686520055514e-05,
      "loss": 0.0207,
      "relation_loss": 0.0013310859212651849,
      "selected_pairs": 26226.8,
      "state_query_norm": 1125.2701171875,
      "state_relation_loss": 0.00028556189499795437,
      "step": 22580,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0013310859212651849,
      "weighted_relation_loss": 0.0013310859212651849
    },
    {
      "action_query_norm": 631.053271484375,
      "action_relation_loss": 0.0009990423975978046,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.131898971000936,
      "flow_matching_loss": 0.018713372852653266,
      "grad_norm": 0.08730889856815338,
      "learning_rate": 1.5776662670249704e-05,
      "loss": 0.02,
      "relation_loss": 0.0012827594648115337,
      "selected_pairs": 26575.8,
      "state_query_norm": 1125.8422973632812,
      "state_relation_loss": 0.000283717067213729,
      "step": 22590,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012827594648115337,
      "weighted_relation_loss": 0.0012827594648115337
    },
    {
      "action_query_norm": 634.7957763671875,
      "action_relation_loss": 0.0009921438118908553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.141253507951358,
      "flow_matching_loss": 0.018125802092254163,
      "grad_norm": 0.10181265324354172,
      "learning_rate": 1.573650172452615e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012752614798955618,
      "selected_pairs": 26731.9,
      "state_query_norm": 1115.9413452148438,
      "state_relation_loss": 0.00028311766945989805,
      "step": 22600,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012752614798955618,
      "weighted_relation_loss": 0.0012752614798955618
    },
    {
      "action_query_norm": 625.9205078125,
      "action_relation_loss": 0.000988507131114602,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.150608044901777,
      "flow_matching_loss": 0.01694240905344486,
      "grad_norm": 0.09189992398023605,
      "learning_rate": 1.5696382412183852e-05,
      "loss": 0.0182,
      "relation_loss": 0.0012724359869025648,
      "selected_pairs": 27125.9,
      "state_query_norm": 1108.8268310546875,
      "state_relation_loss": 0.00028392884705681355,
      "step": 22610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0012724359869025648,
      "weighted_relation_loss": 0.0012724359869025648
    },
    {
      "action_query_norm": 625.3957214355469,
      "action_relation_loss": 0.0009727227210532874,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.1599625818522,
      "flow_matching_loss": 0.017955870274454354,
      "grad_norm": 0.08151203393936157,
      "learning_rate": 1.565630478197155e-05,
      "loss": 0.0192,
      "relation_loss": 0.0012535374611616136,
      "selected_pairs": 26965.0,
      "state_query_norm": 1112.2791137695312,
      "state_relation_loss": 0.00028081474592909215,
      "step": 22620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0012535374611616136,
      "weighted_relation_loss": 0.0012535374611616136
    },
    {
      "action_query_norm": 623.6198669433594,
      "action_relation_loss": 0.0009766724368091672,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.16931711880262,
      "flow_matching_loss": 0.018327099736779928,
      "grad_norm": 0.09579827636480331,
      "learning_rate": 1.5616268882587332e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012513991445302964,
      "selected_pairs": 27688.0,
      "state_query_norm": 1112.8072265625,
      "state_relation_loss": 0.0002747266960795969,
      "step": 22630,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.0012513991445302964,
      "weighted_relation_loss": 0.0012513991445302964
    },
    {
      "action_query_norm": 621.7735961914062,
      "action_relation_loss": 0.000978595862397924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.17867165575304,
      "flow_matching_loss": 0.01775082666426897,
      "grad_norm": 0.10427568107843399,
      "learning_rate": 1.5576274762678594e-05,
      "loss": 0.019,
      "relation_loss": 0.0012673672288656234,
      "selected_pairs": 26874.7,
      "state_query_norm": 1126.0668212890625,
      "state_relation_loss": 0.00028877136355731636,
      "step": 22640,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012673672288656234,
      "weighted_relation_loss": 0.0012673672288656234
    },
    {
      "action_query_norm": 624.2470275878907,
      "action_relation_loss": 0.000986288662534207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.18802619270346,
      "flow_matching_loss": 0.01655553113669157,
      "grad_norm": 0.07892598956823349,
      "learning_rate": 1.5536322470841952e-05,
      "loss": 0.0178,
      "relation_loss": 0.0012731325929053127,
      "selected_pairs": 26579.3,
      "state_query_norm": 1109.2760864257812,
      "state_relation_loss": 0.0002868439187295735,
      "step": 22650,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012731325929053127,
      "weighted_relation_loss": 0.0012731325929053127
    },
    {
      "action_query_norm": 624.4050598144531,
      "action_relation_loss": 0.0010330770106520503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.197380729653883,
      "flow_matching_loss": 0.018747149500995874,
      "grad_norm": 0.07405727356672287,
      "learning_rate": 1.549641205562324e-05,
      "loss": 0.0201,
      "relation_loss": 0.0013129798811860383,
      "selected_pairs": 26673.1,
      "state_query_norm": 1122.2555053710937,
      "state_relation_loss": 0.0002799028909066692,
      "step": 22660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0013129798811860383,
      "weighted_relation_loss": 0.0013129798811860383
    },
    {
      "action_query_norm": 626.0954772949219,
      "action_relation_loss": 0.0009638042189180851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.2067352666043,
      "flow_matching_loss": 0.018158752005547285,
      "grad_norm": 0.08251000195741653,
      "learning_rate": 1.5456543565517334e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012513653258793057,
      "selected_pairs": 26879.4,
      "state_query_norm": 1123.515576171875,
      "state_relation_loss": 0.00028756112151313574,
      "step": 22670,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012513653258793057,
      "weighted_relation_loss": 0.0012513653258793057
    },
    {
      "action_query_norm": 627.4053833007813,
      "action_relation_loss": 0.00100249569513835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.216089803554723,
      "flow_matching_loss": 0.017018663790076972,
      "grad_norm": 0.10114271193742752,
      "learning_rate": 1.541671704896825e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012885960983112454,
      "selected_pairs": 26164.2,
      "state_query_norm": 1132.0751831054688,
      "state_relation_loss": 0.0002861004031728953,
      "step": 22680,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0012885960983112454,
      "weighted_relation_loss": 0.0012885960983112454
    },
    {
      "action_query_norm": 625.7338500976563,
      "action_relation_loss": 0.0010077823884785176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.225444340505145,
      "flow_matching_loss": 0.018200772535055875,
      "grad_norm": 0.09004431962966919,
      "learning_rate": 1.5376932554368938e-05,
      "loss": 0.0195,
      "relation_loss": 0.0012881866074167191,
      "selected_pairs": 27015.0,
      "state_query_norm": 1114.3955444335938,
      "state_relation_loss": 0.00028040421020705255,
      "step": 22690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012881866074167191,
      "weighted_relation_loss": 0.0012881866074167191
    },
    {
      "action_query_norm": 626.1321960449219,
      "action_relation_loss": 0.000995603419141844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.234798877455567,
      "flow_matching_loss": 0.0169501269236207,
      "grad_norm": 0.07952887564897537,
      "learning_rate": 1.5337190130061318e-05,
      "loss": 0.0182,
      "relation_loss": 0.0012806128012016415,
      "selected_pairs": 27407.3,
      "state_query_norm": 1123.4730224609375,
      "state_relation_loss": 0.0002850093791494146,
      "step": 22700,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0012806128012016415,
      "weighted_relation_loss": 0.0012806128012016415
    },
    {
      "action_query_norm": 633.5643127441406,
      "action_relation_loss": 0.0009832670562900602,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.244153414405986,
      "flow_matching_loss": 0.017128103598952293,
      "grad_norm": 0.08823844790458679,
      "learning_rate": 1.5297489824336165e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012686852365732192,
      "selected_pairs": 26504.0,
      "state_query_norm": 1118.9779663085938,
      "state_relation_loss": 0.00028541818319354205,
      "step": 22710,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0012686852365732192,
      "weighted_relation_loss": 0.0012686852365732192
    },
    {
      "action_query_norm": 627.2800537109375,
      "action_relation_loss": 0.0009971149265766144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.253507951356408,
      "flow_matching_loss": 0.01582465274259448,
      "grad_norm": 0.09406666457653046,
      "learning_rate": 1.5257831685433121e-05,
      "loss": 0.0171,
      "relation_loss": 0.0012722406070679426,
      "selected_pairs": 27012.1,
      "state_query_norm": 1134.5052856445313,
      "state_relation_loss": 0.00027512568922247735,
      "step": 22720,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012722406070679426,
      "weighted_relation_loss": 0.0012722406070679426
    },
    {
      "action_query_norm": 623.3131042480469,
      "action_relation_loss": 0.0009733796352520585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.26286248830683,
      "flow_matching_loss": 0.017096607945859432,
      "grad_norm": 0.10664727538824081,
      "learning_rate": 1.521821576154055e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012602528207935394,
      "selected_pairs": 26580.1,
      "state_query_norm": 1120.4128784179688,
      "state_relation_loss": 0.0002868731797207147,
      "step": 22730,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012602528207935394,
      "weighted_relation_loss": 0.0012602528207935394
    },
    {
      "action_query_norm": 630.8521362304688,
      "action_relation_loss": 0.0009777618746738882,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.27221702525725,
      "flow_matching_loss": 0.017208827938884497,
      "grad_norm": 0.10532413423061371,
      "learning_rate": 1.5178642100795542e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012669796356931328,
      "selected_pairs": 26806.9,
      "state_query_norm": 1121.1092895507813,
      "state_relation_loss": 0.0002892177493777126,
      "step": 22740,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012669796356931328,
      "weighted_relation_loss": 0.0012669796356931328
    },
    {
      "action_query_norm": 626.3361206054688,
      "action_relation_loss": 0.000971312349429354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.28157156220767,
      "flow_matching_loss": 0.01733210813254118,
      "grad_norm": 0.0922953262925148,
      "learning_rate": 1.5139110751283819e-05,
      "loss": 0.0186,
      "relation_loss": 0.0012563373660668732,
      "selected_pairs": 26852.8,
      "state_query_norm": 1124.7685424804688,
      "state_relation_loss": 0.00028502501372713597,
      "step": 22750,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0012563373660668732,
      "weighted_relation_loss": 0.0012563373660668732
    },
    {
      "action_query_norm": 627.8460388183594,
      "action_relation_loss": 0.0010133396019227803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.290926099158092,
      "flow_matching_loss": 0.020249905437231062,
      "grad_norm": 0.12353036552667618,
      "learning_rate": 1.5099621761039683e-05,
      "loss": 0.0215,
      "relation_loss": 0.00129346395842731,
      "selected_pairs": 27105.1,
      "state_query_norm": 1121.4032958984376,
      "state_relation_loss": 0.00028012436523567886,
      "step": 22760,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.00129346395842731,
      "weighted_relation_loss": 0.00129346395842731
    },
    {
      "action_query_norm": 628.9602172851562,
      "action_relation_loss": 0.000966664933366701,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.300280636108514,
      "flow_matching_loss": 0.01813504584133625,
      "grad_norm": 0.0924667939543724,
      "learning_rate": 1.5060175178046016e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012425611261278392,
      "selected_pairs": 27452.8,
      "state_query_norm": 1128.7642700195313,
      "state_relation_loss": 0.00027589619858190415,
      "step": 22770,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0012425611261278392,
      "weighted_relation_loss": 0.0012425611261278392
    },
    {
      "action_query_norm": 634.7840515136719,
      "action_relation_loss": 0.0009597810334526002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.309635173058933,
      "flow_matching_loss": 0.020032110530883072,
      "grad_norm": 0.11690088361501694,
      "learning_rate": 1.5020771050234117e-05,
      "loss": 0.0213,
      "relation_loss": 0.001260401983745396,
      "selected_pairs": 26911.3,
      "state_query_norm": 1113.3707763671875,
      "state_relation_loss": 0.00030062095902394503,
      "step": 22780,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001260401983745396,
      "weighted_relation_loss": 0.001260401983745396
    },
    {
      "action_query_norm": 626.1227600097657,
      "action_relation_loss": 0.0009734048333484679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.318989710009355,
      "flow_matching_loss": 0.018512012157589196,
      "grad_norm": 0.0882638618350029,
      "learning_rate": 1.4981409425483716e-05,
      "loss": 0.0198,
      "relation_loss": 0.001261045434512198,
      "selected_pairs": 26138.1,
      "state_query_norm": 1117.342724609375,
      "state_relation_loss": 0.0002876406040741131,
      "step": 22790,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.001261045434512198,
      "weighted_relation_loss": 0.001261045434512198
    },
    {
      "action_query_norm": 627.7195251464843,
      "action_relation_loss": 0.0009776921244338155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.328344246959777,
      "flow_matching_loss": 0.015291226375848056,
      "grad_norm": 0.08414724469184875,
      "learning_rate": 1.4942090351622883e-05,
      "loss": 0.0166,
      "relation_loss": 0.0012597943423315883,
      "selected_pairs": 26670.9,
      "state_query_norm": 1116.6851440429687,
      "state_relation_loss": 0.00028210221498738977,
      "step": 22800,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012597943423315883,
      "weighted_relation_loss": 0.0012597943423315883
    },
    {
      "action_query_norm": 625.7418151855469,
      "action_relation_loss": 0.000981689797481522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.337698783910195,
      "flow_matching_loss": 0.016723806876689194,
      "grad_norm": 0.08218243718147278,
      "learning_rate": 1.490281387642804e-05,
      "loss": 0.018,
      "relation_loss": 0.0012596353888511657,
      "selected_pairs": 26455.4,
      "state_query_norm": 1118.5792724609375,
      "state_relation_loss": 0.00027794560592155906,
      "step": 22810,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012596353888511657,
      "weighted_relation_loss": 0.0012596353888511657
    },
    {
      "action_query_norm": 623.3606079101562,
      "action_relation_loss": 0.0009700494294520468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.347053320860617,
      "flow_matching_loss": 0.01703812899067998,
      "grad_norm": 0.09860147535800934,
      "learning_rate": 1.4863580047623765e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012555513298138977,
      "selected_pairs": 26869.5,
      "state_query_norm": 1118.1543701171875,
      "state_relation_loss": 0.0002855018945410848,
      "step": 22820,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012555513298138977,
      "weighted_relation_loss": 0.0012555513298138977
    },
    {
      "action_query_norm": 627.4415588378906,
      "action_relation_loss": 0.0009631054825149476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.35640785781104,
      "flow_matching_loss": 0.016185141075402497,
      "grad_norm": 0.07636240869760513,
      "learning_rate": 1.4824388912882897e-05,
      "loss": 0.0174,
      "relation_loss": 0.0012360947206616403,
      "selected_pairs": 26595.4,
      "state_query_norm": 1120.1267578125,
      "state_relation_loss": 0.00027298923523630945,
      "step": 22830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0012360947206616403,
      "weighted_relation_loss": 0.0012360947206616403
    },
    {
      "action_query_norm": 623.3981323242188,
      "action_relation_loss": 0.0009801742387935519,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.36576239476146,
      "flow_matching_loss": 0.017616539262235165,
      "grad_norm": 0.09161963313817978,
      "learning_rate": 1.4785240519826344e-05,
      "loss": 0.0189,
      "relation_loss": 0.001270446926355362,
      "selected_pairs": 27242.7,
      "state_query_norm": 1119.642236328125,
      "state_relation_loss": 0.0002902726759202778,
      "step": 22840,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001270446926355362,
      "weighted_relation_loss": 0.001270446926355362
    },
    {
      "action_query_norm": 627.1255798339844,
      "action_relation_loss": 0.0009796395141165703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.37511693171188,
      "flow_matching_loss": 0.01541021754965186,
      "grad_norm": 0.07624445855617523,
      "learning_rate": 1.4746134916023097e-05,
      "loss": 0.0167,
      "relation_loss": 0.001263424230273813,
      "selected_pairs": 26755.4,
      "state_query_norm": 1119.0541870117188,
      "state_relation_loss": 0.00028378472197800877,
      "step": 22850,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001263424230273813,
      "weighted_relation_loss": 0.001263424230273813
    },
    {
      "action_query_norm": 620.4394287109375,
      "action_relation_loss": 0.0009681217023171484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.3844714686623,
      "flow_matching_loss": 0.01833252776414156,
      "grad_norm": 0.09836961328983307,
      "learning_rate": 1.4707072148990141e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012498955591581761,
      "selected_pairs": 26438.2,
      "state_query_norm": 1119.8887084960938,
      "state_relation_loss": 0.0002817738655721769,
      "step": 22860,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0012498955591581761,
      "weighted_relation_loss": 0.0012498955591581761
    },
    {
      "action_query_norm": 625.2719665527344,
      "action_relation_loss": 0.0010156051313970238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.393826005612723,
      "flow_matching_loss": 0.017472954653203487,
      "grad_norm": 0.10078362375497818,
      "learning_rate": 1.4668052266192423e-05,
      "loss": 0.0188,
      "relation_loss": 0.0012906801654025911,
      "selected_pairs": 26616.0,
      "state_query_norm": 1113.7738525390625,
      "state_relation_loss": 0.00027507504855748264,
      "step": 22870,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012906801654025911,
      "weighted_relation_loss": 0.0012906801654025911
    },
    {
      "action_query_norm": 627.2295104980469,
      "action_relation_loss": 0.0009714061510749161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.403180542563142,
      "flow_matching_loss": 0.01815312886610627,
      "grad_norm": 0.10041575133800507,
      "learning_rate": 1.4629075315042795e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012564408592879773,
      "selected_pairs": 27481.5,
      "state_query_norm": 1109.4860961914062,
      "state_relation_loss": 0.00028503470530267806,
      "step": 22880,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0012564408592879773,
      "weighted_relation_loss": 0.0012564408592879773
    },
    {
      "action_query_norm": 626.545751953125,
      "action_relation_loss": 0.0010170042398385704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.412535079513564,
      "flow_matching_loss": 0.018074668385088442,
      "grad_norm": 0.10604207217693329,
      "learning_rate": 1.4590141342901925e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012999512138776482,
      "selected_pairs": 27038.0,
      "state_query_norm": 1110.7086791992188,
      "state_relation_loss": 0.0002829469653079286,
      "step": 22890,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0012999512138776482,
      "weighted_relation_loss": 0.0012999512138776482
    },
    {
      "action_query_norm": 637.1918823242188,
      "action_relation_loss": 0.0009160256711766123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.421889616463986,
      "flow_matching_loss": 0.01853794474154711,
      "grad_norm": 0.09047907590866089,
      "learning_rate": 1.4551250397078253e-05,
      "loss": 0.0197,
      "relation_loss": 0.00119668401312083,
      "selected_pairs": 26945.3,
      "state_query_norm": 1119.7680419921876,
      "state_relation_loss": 0.0002806583477649838,
      "step": 22900,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00119668401312083,
      "weighted_relation_loss": 0.00119668401312083
    },
    {
      "action_query_norm": 636.2775024414062,
      "action_relation_loss": 0.0009635204856749624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.431244153414404,
      "flow_matching_loss": 0.017249488830566408,
      "grad_norm": 0.09867790341377258,
      "learning_rate": 1.4512402524827945e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012446617241948844,
      "selected_pairs": 26899.5,
      "state_query_norm": 1139.1675170898438,
      "state_relation_loss": 0.0002811412501614541,
      "step": 22910,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0012446617241948844,
      "weighted_relation_loss": 0.0012446617241948844
    },
    {
      "action_query_norm": 631.2905029296875,
      "action_relation_loss": 0.0009786365961190313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.440598690364826,
      "flow_matching_loss": 0.016477262135595083,
      "grad_norm": 0.08558560907840729,
      "learning_rate": 1.447359777335482e-05,
      "loss": 0.0177,
      "relation_loss": 0.001254958240315318,
      "selected_pairs": 27034.0,
      "state_query_norm": 1123.8642211914062,
      "state_relation_loss": 0.00027632163546513765,
      "step": 22920,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001254958240315318,
      "weighted_relation_loss": 0.001254958240315318
    },
    {
      "action_query_norm": 627.1995422363282,
      "action_relation_loss": 0.0009393551095854491,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.44995322731525,
      "flow_matching_loss": 0.017487337812781334,
      "grad_norm": 0.07625293731689453,
      "learning_rate": 1.4434836189810335e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012129252892918885,
      "selected_pairs": 26735.0,
      "state_query_norm": 1124.7795532226562,
      "state_relation_loss": 0.0002735701680649072,
      "step": 22930,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012129252892918885,
      "weighted_relation_loss": 0.0012129252892918885
    },
    {
      "action_query_norm": 627.0952880859375,
      "action_relation_loss": 0.0009817382728215308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.45930776426567,
      "flow_matching_loss": 0.017778001725673676,
      "grad_norm": 0.08870445191860199,
      "learning_rate": 1.4396117821293453e-05,
      "loss": 0.019,
      "relation_loss": 0.0012598592904396356,
      "selected_pairs": 26759.8,
      "state_query_norm": 1112.6227294921875,
      "state_relation_loss": 0.00027812100888695567,
      "step": 22940,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012598592904396356,
      "weighted_relation_loss": 0.0012598592904396356
    },
    {
      "action_query_norm": 627.2834045410157,
      "action_relation_loss": 0.0009891783294733614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.46866230121609,
      "flow_matching_loss": 0.018656582105904818,
      "grad_norm": 0.08620529621839523,
      "learning_rate": 1.4357442714850633e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012734490795992316,
      "selected_pairs": 26393.2,
      "state_query_norm": 1112.6288940429688,
      "state_relation_loss": 0.0002842707501258701,
      "step": 22950,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0012734490795992316,
      "weighted_relation_loss": 0.0012734490795992316
    },
    {
      "action_query_norm": 627.5911193847656,
      "action_relation_loss": 0.0009739419794641436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.47801683816651,
      "flow_matching_loss": 0.01685608262196183,
      "grad_norm": 0.08317047357559204,
      "learning_rate": 1.43188109174758e-05,
      "loss": 0.0181,
      "relation_loss": 0.0012481410522013903,
      "selected_pairs": 27010.7,
      "state_query_norm": 1126.6605834960938,
      "state_relation_loss": 0.0002741990756476298,
      "step": 22960,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012481410522013903,
      "weighted_relation_loss": 0.0012481410522013903
    },
    {
      "action_query_norm": 625.9920043945312,
      "action_relation_loss": 0.0009972648229449987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.487371375116933,
      "flow_matching_loss": 0.018709318526089192,
      "grad_norm": 0.08398683369159698,
      "learning_rate": 1.4280222476110206e-05,
      "loss": 0.02,
      "relation_loss": 0.0012831459869630634,
      "selected_pairs": 26364.4,
      "state_query_norm": 1123.648193359375,
      "state_relation_loss": 0.00028588115237653254,
      "step": 22970,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0012831459869630634,
      "weighted_relation_loss": 0.0012831459869630634
    },
    {
      "action_query_norm": 625.7668701171875,
      "action_relation_loss": 0.0009735303057823331,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.49672591206735,
      "flow_matching_loss": 0.017029360588639974,
      "grad_norm": 0.07550147920846939,
      "learning_rate": 1.4241677437642469e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012620184570550919,
      "selected_pairs": 26737.7,
      "state_query_norm": 1127.8489990234375,
      "state_relation_loss": 0.0002884881570935249,
      "step": 22980,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012620184570550919,
      "weighted_relation_loss": 0.0012620184570550919
    },
    {
      "action_query_norm": 625.7178405761719,
      "action_relation_loss": 0.0010007399192545562,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.506080449017773,
      "flow_matching_loss": 0.016879377700388432,
      "grad_norm": 0.10861845314502716,
      "learning_rate": 1.420317584890844e-05,
      "loss": 0.0182,
      "relation_loss": 0.0012927504838444292,
      "selected_pairs": 26805.0,
      "state_query_norm": 1110.6199462890625,
      "state_relation_loss": 0.00029201056458987295,
      "step": 22990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012927504838444292,
      "weighted_relation_loss": 0.0012927504838444292
    },
    {
      "action_query_norm": 624.1625122070312,
      "action_relation_loss": 0.0009743485250510275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.515434985968195,
      "flow_matching_loss": 0.018362618423998354,
      "grad_norm": 0.10552380234003067,
      "learning_rate": 1.4164717756691176e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012568348785862327,
      "selected_pairs": 27086.5,
      "state_query_norm": 1125.4450317382812,
      "state_relation_loss": 0.0002824863535352051,
      "step": 23000,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012568348785862327,
      "weighted_relation_loss": 0.0012568348785862327
    },
    {
      "action_query_norm": 631.2580322265625,
      "action_relation_loss": 0.0009644343634136021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.524789522918617,
      "flow_matching_loss": 0.018707345891743898,
      "grad_norm": 0.0983782410621643,
      "learning_rate": 1.4126303207720882e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012418269063346088,
      "selected_pairs": 27081.2,
      "state_query_norm": 1125.31201171875,
      "state_relation_loss": 0.0002773925574729219,
      "step": 23010,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012418269063346088,
      "weighted_relation_loss": 0.0012418269063346088
    },
    {
      "action_query_norm": 629.2315795898437,
      "action_relation_loss": 0.0009498370985966176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.534144059869035,
      "flow_matching_loss": 0.018557163327932356,
      "grad_norm": 0.08809313923120499,
      "learning_rate": 1.408793224867484e-05,
      "loss": 0.0198,
      "relation_loss": 0.001230813469737768,
      "selected_pairs": 26716.1,
      "state_query_norm": 1127.5504028320313,
      "state_relation_loss": 0.0002809763594996184,
      "step": 23020,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001230813469737768,
      "weighted_relation_loss": 0.001230813469737768
    },
    {
      "action_query_norm": 627.5556945800781,
      "action_relation_loss": 0.000978592608589679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.543498596819457,
      "flow_matching_loss": 0.01867638239637017,
      "grad_norm": 0.09747227281332016,
      "learning_rate": 1.4049604926177423e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012552703265100718,
      "selected_pairs": 26599.0,
      "state_query_norm": 1110.5808715820312,
      "state_relation_loss": 0.000276677732472308,
      "step": 23030,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0012552703265100718,
      "weighted_relation_loss": 0.0012552703265100718
    },
    {
      "action_query_norm": 623.6310180664062,
      "action_relation_loss": 0.000996486924123019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.55285313376988,
      "flow_matching_loss": 0.018110196106135845,
      "grad_norm": 0.08184969425201416,
      "learning_rate": 1.4011321286799916e-05,
      "loss": 0.0194,
      "relation_loss": 0.001274452789220959,
      "selected_pairs": 27154.9,
      "state_query_norm": 1119.227197265625,
      "state_relation_loss": 0.0002779658534564078,
      "step": 23040,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001274452789220959,
      "weighted_relation_loss": 0.001274452789220959
    },
    {
      "action_query_norm": 624.4365661621093,
      "action_relation_loss": 0.0009461412148084492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.562207670720298,
      "flow_matching_loss": 0.016061183251440524,
      "grad_norm": 0.0853332057595253,
      "learning_rate": 1.3973081377060565e-05,
      "loss": 0.0173,
      "relation_loss": 0.001221541129052639,
      "selected_pairs": 26348.0,
      "state_query_norm": 1133.6128051757812,
      "state_relation_loss": 0.0002753999156993814,
      "step": 23050,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.001221541129052639,
      "weighted_relation_loss": 0.001221541129052639
    },
    {
      "action_query_norm": 625.3047119140625,
      "action_relation_loss": 0.0009911486529745161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.57156220767072,
      "flow_matching_loss": 0.019085075985640288,
      "grad_norm": 0.11414434760808945,
      "learning_rate": 1.393488524342445e-05,
      "loss": 0.0204,
      "relation_loss": 0.0012724712607450784,
      "selected_pairs": 27112.2,
      "state_query_norm": 1120.4299438476562,
      "state_relation_loss": 0.0002813226135913283,
      "step": 23060,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012724712607450784,
      "weighted_relation_loss": 0.0012724712607450784
    },
    {
      "action_query_norm": 622.3708435058594,
      "action_relation_loss": 0.0009259915794245898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.580916744621142,
      "flow_matching_loss": 0.01955934762954712,
      "grad_norm": 0.10343869030475616,
      "learning_rate": 1.3896732932303486e-05,
      "loss": 0.0208,
      "relation_loss": 0.00121104649733752,
      "selected_pairs": 26231.7,
      "state_query_norm": 1119.4799926757812,
      "state_relation_loss": 0.000285054900450632,
      "step": 23070,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.00121104649733752,
      "weighted_relation_loss": 0.00121104649733752
    },
    {
      "action_query_norm": 626.1392822265625,
      "action_relation_loss": 0.0009645782003644854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.590271281571564,
      "flow_matching_loss": 0.01750868959352374,
      "grad_norm": 0.09935671091079712,
      "learning_rate": 1.3858624490056304e-05,
      "loss": 0.0188,
      "relation_loss": 0.001241798885166645,
      "selected_pairs": 27064.2,
      "state_query_norm": 1111.1725341796875,
      "state_relation_loss": 0.00027722068480215965,
      "step": 23080,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001241798885166645,
      "weighted_relation_loss": 0.001241798885166645
    },
    {
      "action_query_norm": 629.6446899414062,
      "action_relation_loss": 0.0009816947858780622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.599625818521982,
      "flow_matching_loss": 0.01741375299170613,
      "grad_norm": 0.1009039506316185,
      "learning_rate": 1.3820559962988299e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012674633879214525,
      "selected_pairs": 26916.9,
      "state_query_norm": 1121.701123046875,
      "state_relation_loss": 0.00028576859913300725,
      "step": 23090,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0012674633879214525,
      "weighted_relation_loss": 0.0012674633879214525
    },
    {
      "action_query_norm": 621.7869750976563,
      "action_relation_loss": 0.000988171575590968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.608980355472404,
      "flow_matching_loss": 0.01715037366375327,
      "grad_norm": 0.10419796407222748,
      "learning_rate": 1.3782539397351418e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012671802775003015,
      "selected_pairs": 26618.8,
      "state_query_norm": 1118.0056640625,
      "state_relation_loss": 0.00027900871355086566,
      "step": 23100,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012671802775003015,
      "weighted_relation_loss": 0.0012671802775003015
    },
    {
      "action_query_norm": 629.4390930175781,
      "action_relation_loss": 0.0009706890326924622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.618334892422826,
      "flow_matching_loss": 0.0184787230566144,
      "grad_norm": 0.10738188773393631,
      "learning_rate": 1.3744562839344266e-05,
      "loss": 0.0197,
      "relation_loss": 0.0012500855140388013,
      "selected_pairs": 27317.3,
      "state_query_norm": 1121.5462890625,
      "state_relation_loss": 0.00027939649007748815,
      "step": 23110,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0012500855140388013,
      "weighted_relation_loss": 0.0012500855140388013
    },
    {
      "action_query_norm": 626.8723693847656,
      "action_relation_loss": 0.0009882383397780358,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.627689429373245,
      "flow_matching_loss": 0.01779205072671175,
      "grad_norm": 0.08107045292854309,
      "learning_rate": 1.370663033511193e-05,
      "loss": 0.0191,
      "relation_loss": 0.0012586130993440746,
      "selected_pairs": 26872.3,
      "state_query_norm": 1125.4921997070312,
      "state_relation_loss": 0.00027037477266276253,
      "step": 23120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012586130993440746,
      "weighted_relation_loss": 0.0012586130993440746
    },
    {
      "action_query_norm": 627.8673034667969,
      "action_relation_loss": 0.0009516878810245544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.637043966323667,
      "flow_matching_loss": 0.017629912495613097,
      "grad_norm": 0.09162060171365738,
      "learning_rate": 1.3668741930745964e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012361486325971783,
      "selected_pairs": 26998.6,
      "state_query_norm": 1116.5205810546875,
      "state_relation_loss": 0.00028446073847590017,
      "step": 23130,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0012361486325971783,
      "weighted_relation_loss": 0.0012361486325971783
    },
    {
      "action_query_norm": 628.0322021484375,
      "action_relation_loss": 0.0009551834198646247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.64639850327409,
      "flow_matching_loss": 0.017257984168827532,
      "grad_norm": 0.0920632854104042,
      "learning_rate": 1.3630897672284382e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012392158270813525,
      "selected_pairs": 26917.9,
      "state_query_norm": 1114.5536987304688,
      "state_relation_loss": 0.00028403242176864294,
      "step": 23140,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0012392158270813525,
      "weighted_relation_loss": 0.0012392158270813525
    },
    {
      "action_query_norm": 623.9176025390625,
      "action_relation_loss": 0.0009638674790039658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.65575304022451,
      "flow_matching_loss": 0.01644821995869279,
      "grad_norm": 0.07943113893270493,
      "learning_rate": 1.3593097605711508e-05,
      "loss": 0.0177,
      "relation_loss": 0.0012497397372499108,
      "selected_pairs": 26144.3,
      "state_query_norm": 1107.3259033203126,
      "state_relation_loss": 0.00028587227570824324,
      "step": 23150,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0012497397372499108,
      "weighted_relation_loss": 0.0012497397372499108
    },
    {
      "action_query_norm": 623.92412109375,
      "action_relation_loss": 0.0010798984672874211,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.66510757717493,
      "flow_matching_loss": 0.018455391842871906,
      "grad_norm": 0.09176526963710785,
      "learning_rate": 1.3555341776957992e-05,
      "loss": 0.0198,
      "relation_loss": 0.0013517303508706392,
      "selected_pairs": 26964.3,
      "state_query_norm": 1115.461962890625,
      "state_relation_loss": 0.00027183186903130264,
      "step": 23160,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0013517303508706392,
      "weighted_relation_loss": 0.0013517303508706392
    },
    {
      "action_query_norm": 626.4652709960938,
      "action_relation_loss": 0.0010115793265867978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.67446211412535,
      "flow_matching_loss": 0.01767182582989335,
      "grad_norm": 0.08526375144720078,
      "learning_rate": 1.3517630231900724e-05,
      "loss": 0.019,
      "relation_loss": 0.0012895352556370199,
      "selected_pairs": 26755.4,
      "state_query_norm": 1122.3950561523438,
      "state_relation_loss": 0.00027795590867754074,
      "step": 23170,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012895352556370199,
      "weighted_relation_loss": 0.0012895352556370199
    },
    {
      "action_query_norm": 631.6050720214844,
      "action_relation_loss": 0.000929813424590975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.683816651075773,
      "flow_matching_loss": 0.01637113941833377,
      "grad_norm": 0.08419084548950195,
      "learning_rate": 1.3479963016362767e-05,
      "loss": 0.0176,
      "relation_loss": 0.0012100015650503338,
      "selected_pairs": 26801.5,
      "state_query_norm": 1113.764892578125,
      "state_relation_loss": 0.00028018814045935867,
      "step": 23180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012100015650503338,
      "weighted_relation_loss": 0.0012100015650503338
    },
    {
      "action_query_norm": 626.558056640625,
      "action_relation_loss": 0.0009671292791608721,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.69317118802619,
      "flow_matching_loss": 0.017659994401037694,
      "grad_norm": 0.08216964453458786,
      "learning_rate": 1.3442340176113377e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012403280008584261,
      "selected_pairs": 27126.0,
      "state_query_norm": 1115.525830078125,
      "state_relation_loss": 0.00027319873042870314,
      "step": 23190,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0012403280008584261,
      "weighted_relation_loss": 0.0012403280008584261
    },
    {
      "action_query_norm": 626.1032775878906,
      "action_relation_loss": 0.000984304177109152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.702525724976613,
      "flow_matching_loss": 0.018467839062213897,
      "grad_norm": 0.07686944305896759,
      "learning_rate": 1.340476175686784e-05,
      "loss": 0.0197,
      "relation_loss": 0.0012654330115765334,
      "selected_pairs": 27014.2,
      "state_query_norm": 1122.8225463867188,
      "state_relation_loss": 0.0002811288315569982,
      "step": 23200,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012654330115765334,
      "weighted_relation_loss": 0.0012654330115765334
    },
    {
      "action_query_norm": 626.8890625,
      "action_relation_loss": 0.0009695469634607434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.711880261927035,
      "flow_matching_loss": 0.018166438955813647,
      "grad_norm": 0.10931262373924255,
      "learning_rate": 1.3367227804287469e-05,
      "loss": 0.0194,
      "relation_loss": 0.001254964026156813,
      "selected_pairs": 26758.7,
      "state_query_norm": 1125.1619140625,
      "state_relation_loss": 0.0002854170685168356,
      "step": 23210,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001254964026156813,
      "weighted_relation_loss": 0.001254964026156813
    },
    {
      "action_query_norm": 638.8223388671875,
      "action_relation_loss": 0.0009443245071452111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.721234798877454,
      "flow_matching_loss": 0.017783632781356574,
      "grad_norm": 0.08232899010181427,
      "learning_rate": 1.332973836397956e-05,
      "loss": 0.019,
      "relation_loss": 0.001218801224604249,
      "selected_pairs": 27199.2,
      "state_query_norm": 1129.7066040039062,
      "state_relation_loss": 0.00027447671454865484,
      "step": 23220,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001218801224604249,
      "weighted_relation_loss": 0.001218801224604249
    },
    {
      "action_query_norm": 628.0599853515625,
      "action_relation_loss": 0.0009888346889056266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.730589335827876,
      "flow_matching_loss": 0.01853393781930208,
      "grad_norm": 0.08344972133636475,
      "learning_rate": 1.329229348149731e-05,
      "loss": 0.0198,
      "relation_loss": 0.0012683924287557602,
      "selected_pairs": 27246.6,
      "state_query_norm": 1127.3739013671875,
      "state_relation_loss": 0.0002795577544020489,
      "step": 23230,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0012683924287557602,
      "weighted_relation_loss": 0.0012683924287557602
    },
    {
      "action_query_norm": 623.5550903320312,
      "action_relation_loss": 0.0010007150238379836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.739943872778298,
      "flow_matching_loss": 0.017061627376824616,
      "grad_norm": 0.10062181949615479,
      "learning_rate": 1.3254893202339796e-05,
      "loss": 0.0183,
      "relation_loss": 0.0012773869442753495,
      "selected_pairs": 27058.4,
      "state_query_norm": 1108.8151245117188,
      "state_relation_loss": 0.0002766719204373658,
      "step": 23240,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0012773869442753495,
      "weighted_relation_loss": 0.0012773869442753495
    },
    {
      "action_query_norm": 622.1927124023438,
      "action_relation_loss": 0.0009522895037662238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.74929840972872,
      "flow_matching_loss": 0.019461223110556602,
      "grad_norm": 0.07797827571630478,
      "learning_rate": 1.3217537571951872e-05,
      "loss": 0.0207,
      "relation_loss": 0.0012356771156191825,
      "selected_pairs": 26618.5,
      "state_query_norm": 1123.6616577148438,
      "state_relation_loss": 0.00028338762349449097,
      "step": 23250,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012356771156191825,
      "weighted_relation_loss": 0.0012356771156191825
    },
    {
      "action_query_norm": 631.1835388183594,
      "action_relation_loss": 0.0009672726853750646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.75865294667914,
      "flow_matching_loss": 0.01722159618511796,
      "grad_norm": 0.07983960956335068,
      "learning_rate": 1.3180226635724169e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012497763498686254,
      "selected_pairs": 26738.3,
      "state_query_norm": 1105.542529296875,
      "state_relation_loss": 0.00028250367031432686,
      "step": 23260,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012497763498686254,
      "weighted_relation_loss": 0.0012497763498686254
    },
    {
      "action_query_norm": 630.6263488769531,
      "action_relation_loss": 0.0009527058165986091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.76800748362956,
      "flow_matching_loss": 0.017534707952290774,
      "grad_norm": 0.11051914095878601,
      "learning_rate": 1.314296043899298e-05,
      "loss": 0.0188,
      "relation_loss": 0.0012429984053596854,
      "selected_pairs": 27341.7,
      "state_query_norm": 1116.1575073242188,
      "state_relation_loss": 0.0002902925829403102,
      "step": 23270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0012429984053596854,
      "weighted_relation_loss": 0.0012429984053596854
    },
    {
      "action_query_norm": 629.295556640625,
      "action_relation_loss": 0.0009459027962293476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.777362020579982,
      "flow_matching_loss": 0.017440866213291883,
      "grad_norm": 0.10145628452301025,
      "learning_rate": 1.3105739027040248e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012336233048699796,
      "selected_pairs": 26989.5,
      "state_query_norm": 1112.5318481445313,
      "state_relation_loss": 0.0002877205028198659,
      "step": 23280,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0012336233048699796,
      "weighted_relation_loss": 0.0012336233048699796
    },
    {
      "action_query_norm": 626.5411987304688,
      "action_relation_loss": 0.0009512414108030498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.7867165575304,
      "flow_matching_loss": 0.017470771074295045,
      "grad_norm": 0.0809263214468956,
      "learning_rate": 1.3068562445093491e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012221875018440186,
      "selected_pairs": 26484.2,
      "state_query_norm": 1121.1893920898438,
      "state_relation_loss": 0.00027094609831692653,
      "step": 23290,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0012221875018440186,
      "weighted_relation_loss": 0.0012221875018440186
    },
    {
      "action_query_norm": 634.8027893066406,
      "action_relation_loss": 0.0009321027551777661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.796071094480823,
      "flow_matching_loss": 0.01917941477149725,
      "grad_norm": 0.09495111554861069,
      "learning_rate": 1.3031430738325778e-05,
      "loss": 0.0204,
      "relation_loss": 0.0012103632092475892,
      "selected_pairs": 26737.0,
      "state_query_norm": 1116.679541015625,
      "state_relation_loss": 0.00027826046571135523,
      "step": 23300,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012103632092475892,
      "weighted_relation_loss": 0.0012103632092475892
    },
    {
      "action_query_norm": 625.6629028320312,
      "action_relation_loss": 0.000971942552132532,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.805425631431245,
      "flow_matching_loss": 0.017653853446245194,
      "grad_norm": 0.07603238523006439,
      "learning_rate": 1.299434395185563e-05,
      "loss": 0.0189,
      "relation_loss": 0.001251285325270146,
      "selected_pairs": 26374.4,
      "state_query_norm": 1111.1448486328125,
      "state_relation_loss": 0.00027934276004089044,
      "step": 23310,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001251285325270146,
      "weighted_relation_loss": 0.001251285325270146
    },
    {
      "action_query_norm": 622.9552001953125,
      "action_relation_loss": 0.0009812720934860408,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.814780168381667,
      "flow_matching_loss": 0.016549505945295095,
      "grad_norm": 0.08834788203239441,
      "learning_rate": 1.295730213074699e-05,
      "loss": 0.0178,
      "relation_loss": 0.0012584966025315225,
      "selected_pairs": 26757.3,
      "state_query_norm": 1114.79755859375,
      "state_relation_loss": 0.0002772245119558647,
      "step": 23320,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012584966025315225,
      "weighted_relation_loss": 0.0012584966025315225
    },
    {
      "action_query_norm": 631.9506225585938,
      "action_relation_loss": 0.0009834883909206837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.824134705332085,
      "flow_matching_loss": 0.01718679266050458,
      "grad_norm": 0.1065806970000267,
      "learning_rate": 1.2920305320009152e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012694910168647767,
      "selected_pairs": 26567.8,
      "state_query_norm": 1124.9195190429687,
      "state_relation_loss": 0.0002860026259440929,
      "step": 23330,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012694910168647767,
      "weighted_relation_loss": 0.0012694910168647767
    },
    {
      "action_query_norm": 625.9373168945312,
      "action_relation_loss": 0.0009449849370867014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.833489242282507,
      "flow_matching_loss": 0.017094111628830434,
      "grad_norm": 0.084725521504879,
      "learning_rate": 1.2883353564596729e-05,
      "loss": 0.0183,
      "relation_loss": 0.001223569887224585,
      "selected_pairs": 26917.6,
      "state_query_norm": 1113.3479248046874,
      "state_relation_loss": 0.00027858492685481904,
      "step": 23340,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001223569887224585,
      "weighted_relation_loss": 0.001223569887224585
    },
    {
      "action_query_norm": 624.3432678222656,
      "action_relation_loss": 0.0009625822014641017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.84284377923293,
      "flow_matching_loss": 0.018766346946358682,
      "grad_norm": 0.08964525908231735,
      "learning_rate": 1.2846446909409604e-05,
      "loss": 0.02,
      "relation_loss": 0.001245737634599209,
      "selected_pairs": 26803.6,
      "state_query_norm": 1103.067919921875,
      "state_relation_loss": 0.0002831554273143411,
      "step": 23350,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001245737634599209,
      "weighted_relation_loss": 0.001245737634599209
    },
    {
      "action_query_norm": 628.612255859375,
      "action_relation_loss": 0.000984056870220229,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.852198316183348,
      "flow_matching_loss": 0.017105264589190483,
      "grad_norm": 0.10681109130382538,
      "learning_rate": 1.280958539929284e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012570169172249734,
      "selected_pairs": 26615.1,
      "state_query_norm": 1123.5880859375,
      "state_relation_loss": 0.00027296004700474443,
      "step": 23360,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012570169172249734,
      "weighted_relation_loss": 0.0012570169172249734
    },
    {
      "action_query_norm": 630.7172607421875,
      "action_relation_loss": 0.0009651984844822436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.86155285313377,
      "flow_matching_loss": 0.016542728431522845,
      "grad_norm": 0.08149230480194092,
      "learning_rate": 1.277276907903664e-05,
      "loss": 0.0178,
      "relation_loss": 0.0012491272762417793,
      "selected_pairs": 26872.7,
      "state_query_norm": 1124.0857055664062,
      "state_relation_loss": 0.0002839287742972374,
      "step": 23370,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012491272762417793,
      "weighted_relation_loss": 0.0012491272762417793
    },
    {
      "action_query_norm": 622.8249633789062,
      "action_relation_loss": 0.0009727395605295897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.87090739008419,
      "flow_matching_loss": 0.018925883807241917,
      "grad_norm": 0.0774708166718483,
      "learning_rate": 1.2735997993376298e-05,
      "loss": 0.0202,
      "relation_loss": 0.0012553705484606326,
      "selected_pairs": 26530.6,
      "state_query_norm": 1114.780029296875,
      "state_relation_loss": 0.00028263099666219204,
      "step": 23380,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0012553705484606326,
      "weighted_relation_loss": 0.0012553705484606326
    },
    {
      "action_query_norm": 628.9680969238282,
      "action_relation_loss": 0.0009245373366866261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.880261927034613,
      "flow_matching_loss": 0.01663265023380518,
      "grad_norm": 0.08766774088144302,
      "learning_rate": 1.2699272186992166e-05,
      "loss": 0.0178,
      "relation_loss": 0.0012051263824105264,
      "selected_pairs": 26803.8,
      "state_query_norm": 1117.252490234375,
      "state_relation_loss": 0.0002805890719173476,
      "step": 23390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012051263824105264,
      "weighted_relation_loss": 0.0012051263824105264
    },
    {
      "action_query_norm": 625.0606140136719,
      "action_relation_loss": 0.0009852303774096072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.889616463985032,
      "flow_matching_loss": 0.016093716118484737,
      "grad_norm": 0.09778140485286713,
      "learning_rate": 1.2662591704509547e-05,
      "loss": 0.0174,
      "relation_loss": 0.0012608292163349688,
      "selected_pairs": 26523.3,
      "state_query_norm": 1115.80771484375,
      "state_relation_loss": 0.00027559883310459556,
      "step": 23400,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0012608292163349688,
      "weighted_relation_loss": 0.0012608292163349688
    },
    {
      "action_query_norm": 624.815380859375,
      "action_relation_loss": 0.0009365932084619999,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.898971000935454,
      "flow_matching_loss": 0.019326117262244223,
      "grad_norm": 0.0916890949010849,
      "learning_rate": 1.2625956590498711e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012199393822811544,
      "selected_pairs": 26500.1,
      "state_query_norm": 1123.6084228515624,
      "state_relation_loss": 0.00028334616508800535,
      "step": 23410,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0012199393822811544,
      "weighted_relation_loss": 0.0012199393822811544
    },
    {
      "action_query_norm": 625.1905578613281,
      "action_relation_loss": 0.0009545425593387336,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.908325537885876,
      "flow_matching_loss": 0.016460321843624115,
      "grad_norm": 0.08836133778095245,
      "learning_rate": 1.2589366889474757e-05,
      "loss": 0.0177,
      "relation_loss": 0.001232280652038753,
      "selected_pairs": 26940.0,
      "state_query_norm": 1114.4384155273438,
      "state_relation_loss": 0.0002777380665065721,
      "step": 23420,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001232280652038753,
      "weighted_relation_loss": 0.001232280652038753
    },
    {
      "action_query_norm": 631.7764953613281,
      "action_relation_loss": 0.0009501933411229402,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.917680074836294,
      "flow_matching_loss": 0.0187612546607852,
      "grad_norm": 0.11719593405723572,
      "learning_rate": 1.2552822645897621e-05,
      "loss": 0.02,
      "relation_loss": 0.001228228339459747,
      "selected_pairs": 27174.7,
      "state_query_norm": 1102.3809326171875,
      "state_relation_loss": 0.0002780350303510204,
      "step": 23430,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001228228339459747,
      "weighted_relation_loss": 0.001228228339459747
    },
    {
      "action_query_norm": 627.3449035644531,
      "action_relation_loss": 0.0009394609660375863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.927034611786716,
      "flow_matching_loss": 0.017706461902707814,
      "grad_norm": 0.0994090586900711,
      "learning_rate": 1.2516323904172e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012239925912581385,
      "selected_pairs": 27317.0,
      "state_query_norm": 1104.9032348632813,
      "state_relation_loss": 0.000284531619399786,
      "step": 23440,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0012239925912581385,
      "weighted_relation_loss": 0.0012239925912581385
    },
    {
      "action_query_norm": 625.4352478027344,
      "action_relation_loss": 0.0009286121698096394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.93638914873714,
      "flow_matching_loss": 0.017895919177681206,
      "grad_norm": 0.10337372124195099,
      "learning_rate": 1.2479870708647324e-05,
      "loss": 0.0191,
      "relation_loss": 0.001216859626583755,
      "selected_pairs": 27244.7,
      "state_query_norm": 1106.9362426757812,
      "state_relation_loss": 0.0002882474887883291,
      "step": 23450,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001216859626583755,
      "weighted_relation_loss": 0.001216859626583755
    },
    {
      "action_query_norm": 619.7585021972657,
      "action_relation_loss": 0.0010264291369821877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.94574368568756,
      "flow_matching_loss": 0.01805981285870075,
      "grad_norm": 0.11683736741542816,
      "learning_rate": 1.2443463103617658e-05,
      "loss": 0.0194,
      "relation_loss": 0.0013018358731642365,
      "selected_pairs": 27040.2,
      "state_query_norm": 1114.8954711914062,
      "state_relation_loss": 0.00027540671289898454,
      "step": 23460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0013018358731642365,
      "weighted_relation_loss": 0.0013018358731642365
    },
    {
      "action_query_norm": 624.0626403808594,
      "action_relation_loss": 0.0009486736438702792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.95509822263798,
      "flow_matching_loss": 0.015708750858902932,
      "grad_norm": 0.09149272739887238,
      "learning_rate": 1.240710113332167e-05,
      "loss": 0.0169,
      "relation_loss": 0.0012295893277041614,
      "selected_pairs": 26668.8,
      "state_query_norm": 1132.5325439453125,
      "state_relation_loss": 0.00028091567219235003,
      "step": 23470,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012295893277041614,
      "weighted_relation_loss": 0.0012295893277041614
    },
    {
      "action_query_norm": 627.4773071289062,
      "action_relation_loss": 0.0009825881395954638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.9644527595884,
      "flow_matching_loss": 0.01868219589814544,
      "grad_norm": 0.10043013095855713,
      "learning_rate": 1.2370784841942584e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012611777521669865,
      "selected_pairs": 27175.5,
      "state_query_norm": 1109.6004150390625,
      "state_relation_loss": 0.00027858961548190566,
      "step": 23480,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0012611777521669865,
      "weighted_relation_loss": 0.0012611777521669865
    },
    {
      "action_query_norm": 625.5152160644532,
      "action_relation_loss": 0.0009636934730224312,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.973807296538823,
      "flow_matching_loss": 0.016146337054669856,
      "grad_norm": 0.1020468920469284,
      "learning_rate": 1.2334514273608117e-05,
      "loss": 0.0174,
      "relation_loss": 0.0012463182443752884,
      "selected_pairs": 26445.2,
      "state_query_norm": 1113.57021484375,
      "state_relation_loss": 0.0002826247800840065,
      "step": 23490,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0012463182443752884,
      "weighted_relation_loss": 0.0012463182443752884
    },
    {
      "action_query_norm": 621.8896301269531,
      "action_relation_loss": 0.0010239737806841732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.98316183348924,
      "flow_matching_loss": 0.017582580167800187,
      "grad_norm": 0.10264801234006882,
      "learning_rate": 1.2298289472390417e-05,
      "loss": 0.0189,
      "relation_loss": 0.0013049131957814097,
      "selected_pairs": 26881.2,
      "state_query_norm": 1113.39853515625,
      "state_relation_loss": 0.00028093941218685357,
      "step": 23500,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0013049131957814097,
      "weighted_relation_loss": 0.0013049131957814097
    },
    {
      "action_query_norm": 625.7705932617188,
      "action_relation_loss": 0.0009690276230685413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.992516370439663,
      "flow_matching_loss": 0.017700661066919567,
      "grad_norm": 0.088166743516922,
      "learning_rate": 1.226211048230606e-05,
      "loss": 0.019,
      "relation_loss": 0.001252253947313875,
      "selected_pairs": 27470.3,
      "state_query_norm": 1110.3853881835937,
      "state_relation_loss": 0.0002832263242453337,
      "step": 23510,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.001252253947313875,
      "weighted_relation_loss": 0.001252253947313875
    },
    {
      "action_query_norm": 623.123876953125,
      "action_relation_loss": 0.0009855499491095542,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.001870907390085,
      "flow_matching_loss": 0.01564423507079482,
      "grad_norm": 0.08787639439105988,
      "learning_rate": 1.2225977347315921e-05,
      "loss": 0.0169,
      "relation_loss": 0.0012660556472837924,
      "selected_pairs": 24058.7,
      "state_query_norm": 1116.255322265625,
      "state_relation_loss": 0.0002805056981742382,
      "step": 23520,
      "valid_action_samples": 213.3,
      "weighted_auxiliary_loss": 0.0012660556472837924,
      "weighted_relation_loss": 0.0012660556472837924
    },
    {
      "action_query_norm": 619.8474304199219,
      "action_relation_loss": 0.0009560797072481364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.011225444340504,
      "flow_matching_loss": 0.017562652099877597,
      "grad_norm": 0.08353235572576523,
      "learning_rate": 1.2189890111325148e-05,
      "loss": 0.0188,
      "relation_loss": 0.0012396027916111052,
      "selected_pairs": 25891.5,
      "state_query_norm": 1101.2215209960937,
      "state_relation_loss": 0.00028352307854220273,
      "step": 23530,
      "valid_action_samples": 228.0,
      "weighted_auxiliary_loss": 0.0012396027916111052,
      "weighted_relation_loss": 0.0012396027916111052
    },
    {
      "action_query_norm": 620.2858642578125,
      "action_relation_loss": 0.0008918243926018477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.020579981290926,
      "flow_matching_loss": 0.017828752659261225,
      "grad_norm": 0.08754824846982956,
      "learning_rate": 1.2153848818183161e-05,
      "loss": 0.019,
      "relation_loss": 0.001176721160300076,
      "selected_pairs": 26090.6,
      "state_query_norm": 1114.9545043945313,
      "state_relation_loss": 0.0002848967706086114,
      "step": 23540,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.001176721160300076,
      "weighted_relation_loss": 0.001176721160300076
    },
    {
      "action_query_norm": 629.2173278808593,
      "action_relation_loss": 0.0009073125838767737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.029934518241348,
      "flow_matching_loss": 0.01675945147871971,
      "grad_norm": 0.1115172877907753,
      "learning_rate": 1.2117853511683507e-05,
      "loss": 0.0179,
      "relation_loss": 0.0011863235966302454,
      "selected_pairs": 26874.1,
      "state_query_norm": 1112.783544921875,
      "state_relation_loss": 0.0002790109952911735,
      "step": 23550,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011863235966302454,
      "weighted_relation_loss": 0.0011863235966302454
    },
    {
      "action_query_norm": 624.4469970703125,
      "action_relation_loss": 0.0009716941218357533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.03928905519177,
      "flow_matching_loss": 0.016005935985594986,
      "grad_norm": 0.08522064983844757,
      "learning_rate": 1.2081904235563906e-05,
      "loss": 0.0173,
      "relation_loss": 0.0012540758703835308,
      "selected_pairs": 27342.8,
      "state_query_norm": 1100.1092407226563,
      "state_relation_loss": 0.0002823817543685436,
      "step": 23560,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0012540758703835308,
      "weighted_relation_loss": 0.0012540758703835308
    },
    {
      "action_query_norm": 623.4437744140625,
      "action_relation_loss": 0.0009293457900639623,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.048643592142188,
      "flow_matching_loss": 0.01677443804219365,
      "grad_norm": 0.08609475940465927,
      "learning_rate": 1.20460010335061e-05,
      "loss": 0.018,
      "relation_loss": 0.001205716235563159,
      "selected_pairs": 26782.3,
      "state_query_norm": 1117.8319946289062,
      "state_relation_loss": 0.0002763704484095797,
      "step": 23570,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001205716235563159,
      "weighted_relation_loss": 0.001205716235563159
    },
    {
      "action_query_norm": 628.3936157226562,
      "action_relation_loss": 0.0009313595248386264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.05799812909261,
      "flow_matching_loss": 0.016793577279895545,
      "grad_norm": 0.10219251364469528,
      "learning_rate": 1.2010143949135865e-05,
      "loss": 0.018,
      "relation_loss": 0.0012040854897350072,
      "selected_pairs": 26564.5,
      "state_query_norm": 1122.265478515625,
      "state_relation_loss": 0.00027272596926195545,
      "step": 23580,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0012040854897350072,
      "weighted_relation_loss": 0.0012040854897350072
    },
    {
      "action_query_norm": 633.0917419433594,
      "action_relation_loss": 0.0009147073025815189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.067352666043032,
      "flow_matching_loss": 0.01713209496811032,
      "grad_norm": 0.08738763630390167,
      "learning_rate": 1.1974333026022938e-05,
      "loss": 0.0183,
      "relation_loss": 0.0011893971357494592,
      "selected_pairs": 27461.7,
      "state_query_norm": 1118.4524658203125,
      "state_relation_loss": 0.00027468982443679124,
      "step": 23590,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0011893971357494592,
      "weighted_relation_loss": 0.0011893971357494592
    },
    {
      "action_query_norm": 626.7625732421875,
      "action_relation_loss": 0.0009395594068337231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.07670720299345,
      "flow_matching_loss": 0.018414507154375315,
      "grad_norm": 0.10373042523860931,
      "learning_rate": 1.1938568307680964e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012241603923030197,
      "selected_pairs": 27181.0,
      "state_query_norm": 1109.693505859375,
      "state_relation_loss": 0.0002846009883796796,
      "step": 23600,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0012241603923030197,
      "weighted_relation_loss": 0.0012241603923030197
    },
    {
      "action_query_norm": 626.3872436523437,
      "action_relation_loss": 0.000947267800802365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.086061739943872,
      "flow_matching_loss": 0.016440204996615648,
      "grad_norm": 0.09122225642204285,
      "learning_rate": 1.1902849837567464e-05,
      "loss": 0.0177,
      "relation_loss": 0.0012318898690864444,
      "selected_pairs": 26660.5,
      "state_query_norm": 1120.0644775390624,
      "state_relation_loss": 0.00028462206537369637,
      "step": 23610,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012318898690864444,
      "weighted_relation_loss": 0.0012318898690864444
    },
    {
      "action_query_norm": 625.6518493652344,
      "action_relation_loss": 0.0009778613515663893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.095416276894294,
      "flow_matching_loss": 0.017655718326568603,
      "grad_norm": 0.08618925511837006,
      "learning_rate": 1.1867177659083739e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012588520301505923,
      "selected_pairs": 26571.1,
      "state_query_norm": 1122.8099487304687,
      "state_relation_loss": 0.00028099067567382007,
      "step": 23620,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012588520301505923,
      "weighted_relation_loss": 0.0012588520301505923
    },
    {
      "action_query_norm": 628.0611206054688,
      "action_relation_loss": 0.0009536042052786797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.104770813844716,
      "flow_matching_loss": 0.01744376765564084,
      "grad_norm": 0.08162370324134827,
      "learning_rate": 1.1831551815574848e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012274716282263398,
      "selected_pairs": 27736.0,
      "state_query_norm": 1105.355859375,
      "state_relation_loss": 0.00027386741276131944,
      "step": 23630,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.0012274716282263398,
      "weighted_relation_loss": 0.0012274716282263398
    },
    {
      "action_query_norm": 629.2442260742188,
      "action_relation_loss": 0.000938455294817686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.114125350795135,
      "flow_matching_loss": 0.01816376531496644,
      "grad_norm": 0.0830083042383194,
      "learning_rate": 1.1795972350329554e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012123638181947172,
      "selected_pairs": 26579.5,
      "state_query_norm": 1122.1380859375,
      "state_relation_loss": 0.00027390855102567003,
      "step": 23640,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012123638181947172,
      "weighted_relation_loss": 0.0012123638181947172
    },
    {
      "action_query_norm": 620.0355407714844,
      "action_relation_loss": 0.000925789587199688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.123479887745557,
      "flow_matching_loss": 0.01746931215748191,
      "grad_norm": 0.07801233232021332,
      "learning_rate": 1.176043930658024e-05,
      "loss": 0.0187,
      "relation_loss": 0.0012059195665642618,
      "selected_pairs": 26965.6,
      "state_query_norm": 1104.9548706054688,
      "state_relation_loss": 0.0002801299997372553,
      "step": 23650,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0012059195665642618,
      "weighted_relation_loss": 0.0012059195665642618
    },
    {
      "action_query_norm": 624.2014770507812,
      "action_relation_loss": 0.0009033254580572247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.13283442469598,
      "flow_matching_loss": 0.016874491516500712,
      "grad_norm": 0.08254937082529068,
      "learning_rate": 1.1724952727502942e-05,
      "loss": 0.0181,
      "relation_loss": 0.0011842441977933048,
      "selected_pairs": 26669.5,
      "state_query_norm": 1111.8479248046874,
      "state_relation_loss": 0.00028091873682569715,
      "step": 23660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011842441977933048,
      "weighted_relation_loss": 0.0011842441977933048
    },
    {
      "action_query_norm": 628.9782836914062,
      "action_relation_loss": 0.0009562815248500556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.142188961646397,
      "flow_matching_loss": 0.017552572209388016,
      "grad_norm": 0.09897949546575546,
      "learning_rate": 1.1689512656217178e-05,
      "loss": 0.0188,
      "relation_loss": 0.0012252576067112385,
      "selected_pairs": 26948.6,
      "state_query_norm": 1093.3077880859375,
      "state_relation_loss": 0.0002689760905923322,
      "step": 23670,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0012252576067112385,
      "weighted_relation_loss": 0.0012252576067112385
    },
    {
      "action_query_norm": 623.8184692382813,
      "action_relation_loss": 0.000923717173282057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.15154349859682,
      "flow_matching_loss": 0.016990593075752257,
      "grad_norm": 0.09363167732954025,
      "learning_rate": 1.1654119135785962e-05,
      "loss": 0.0182,
      "relation_loss": 0.001203026610892266,
      "selected_pairs": 26555.9,
      "state_query_norm": 1120.278076171875,
      "state_relation_loss": 0.00027930942887905986,
      "step": 23680,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001203026610892266,
      "weighted_relation_loss": 0.001203026610892266
    },
    {
      "action_query_norm": 624.878955078125,
      "action_relation_loss": 0.000897024042205885,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.16089803554724,
      "flow_matching_loss": 0.017620996851474047,
      "grad_norm": 0.08746818453073502,
      "learning_rate": 1.1618772209215794e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011688099475577474,
      "selected_pairs": 27019.2,
      "state_query_norm": 1113.5434448242188,
      "state_relation_loss": 0.0002717859169933945,
      "step": 23690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011688099475577474,
      "weighted_relation_loss": 0.0011688099475577474
    },
    {
      "action_query_norm": 627.9261413574219,
      "action_relation_loss": 0.0009160964866168797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.170252572497663,
      "flow_matching_loss": 0.016237127222120763,
      "grad_norm": 0.10169682651758194,
      "learning_rate": 1.1583471919456506e-05,
      "loss": 0.0174,
      "relation_loss": 0.0011874359799548984,
      "selected_pairs": 27126.0,
      "state_query_norm": 1119.5586303710938,
      "state_relation_loss": 0.0002713394846068695,
      "step": 23700,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011874359799548984,
      "weighted_relation_loss": 0.0011874359799548984
    },
    {
      "action_query_norm": 627.1811584472656,
      "action_relation_loss": 0.0009185750328470022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.17960710944808,
      "flow_matching_loss": 0.016757633071392775,
      "grad_norm": 0.095832958817482,
      "learning_rate": 1.1548218309401266e-05,
      "loss": 0.018,
      "relation_loss": 0.0011976309935562313,
      "selected_pairs": 26476.7,
      "state_query_norm": 1128.4324096679688,
      "state_relation_loss": 0.00027905596944037824,
      "step": 23710,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011976309935562313,
      "weighted_relation_loss": 0.0011976309935562313
    },
    {
      "action_query_norm": 625.4381469726562,
      "action_relation_loss": 0.0009338795556686818,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.188961646398504,
      "flow_matching_loss": 0.01725984998047352,
      "grad_norm": 0.08816289156675339,
      "learning_rate": 1.1513011421886555e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012142684194259346,
      "selected_pairs": 26556.4,
      "state_query_norm": 1107.5007080078126,
      "state_relation_loss": 0.0002803888579364866,
      "step": 23720,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0012142684194259346,
      "weighted_relation_loss": 0.0012142684194259346
    },
    {
      "action_query_norm": 628.5321960449219,
      "action_relation_loss": 0.0009329324529971927,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.198316183348926,
      "flow_matching_loss": 0.015154281351715326,
      "grad_norm": 0.09203328937292099,
      "learning_rate": 1.1477851299692056e-05,
      "loss": 0.0164,
      "relation_loss": 0.0012067708186805249,
      "selected_pairs": 27320.2,
      "state_query_norm": 1127.457763671875,
      "state_relation_loss": 0.0002738383802352473,
      "step": 23730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0012067708186805249,
      "weighted_relation_loss": 0.0012067708186805249
    },
    {
      "action_query_norm": 627.7163391113281,
      "action_relation_loss": 0.0009716188651509583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.207670720299344,
      "flow_matching_loss": 0.016296511236578225,
      "grad_norm": 0.0803537666797638,
      "learning_rate": 1.144273798554063e-05,
      "loss": 0.0175,
      "relation_loss": 0.001251354836858809,
      "selected_pairs": 27289.5,
      "state_query_norm": 1127.1765991210937,
      "state_relation_loss": 0.00027973599499091505,
      "step": 23740,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.001251354836858809,
      "weighted_relation_loss": 0.001251354836858809
    },
    {
      "action_query_norm": 629.2200439453125,
      "action_relation_loss": 0.0009100130875594914,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.217025257249766,
      "flow_matching_loss": 0.016909300163388253,
      "grad_norm": 0.08314582705497742,
      "learning_rate": 1.1407671522098262e-05,
      "loss": 0.0181,
      "relation_loss": 0.0011833502561785282,
      "selected_pairs": 26846.0,
      "state_query_norm": 1134.303857421875,
      "state_relation_loss": 0.0002733371627982706,
      "step": 23750,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011833502561785282,
      "weighted_relation_loss": 0.0011833502561785282
    },
    {
      "action_query_norm": 622.48486328125,
      "action_relation_loss": 0.000913145358208567,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.226379794200188,
      "flow_matching_loss": 0.016367793083190918,
      "grad_norm": 0.0752091184258461,
      "learning_rate": 1.1372651951974e-05,
      "loss": 0.0176,
      "relation_loss": 0.0011912949266843498,
      "selected_pairs": 27152.3,
      "state_query_norm": 1127.0181274414062,
      "state_relation_loss": 0.0002781495801173151,
      "step": 23760,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011912949266843498,
      "weighted_relation_loss": 0.0011912949266843498
    },
    {
      "action_query_norm": 625.6349182128906,
      "action_relation_loss": 0.0009451972029637546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.235734331150606,
      "flow_matching_loss": 0.016428015287965538,
      "grad_norm": 0.10053635388612747,
      "learning_rate": 1.1337679317719952e-05,
      "loss": 0.0176,
      "relation_loss": 0.001216098153963685,
      "selected_pairs": 26909.5,
      "state_query_norm": 1110.5975830078125,
      "state_relation_loss": 0.0002709009335376322,
      "step": 23770,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001216098153963685,
      "weighted_relation_loss": 0.001216098153963685
    },
    {
      "action_query_norm": 626.1532775878907,
      "action_relation_loss": 0.0009112529281992465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.24508886810103,
      "flow_matching_loss": 0.015608241502195596,
      "grad_norm": 0.08479201048612595,
      "learning_rate": 1.1302753661831151e-05,
      "loss": 0.0168,
      "relation_loss": 0.0011780755943618716,
      "selected_pairs": 26299.2,
      "state_query_norm": 1123.9433349609376,
      "state_relation_loss": 0.0002668226879904978,
      "step": 23780,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0011780755943618716,
      "weighted_relation_loss": 0.0011780755943618716
    },
    {
      "action_query_norm": 621.1146789550781,
      "action_relation_loss": 0.0009248409594874829,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.25444340505145,
      "flow_matching_loss": 0.01642724433913827,
      "grad_norm": 0.07601167261600494,
      "learning_rate": 1.1267875026745561e-05,
      "loss": 0.0176,
      "relation_loss": 0.001201976288575679,
      "selected_pairs": 26805.6,
      "state_query_norm": 1126.5409545898438,
      "state_relation_loss": 0.0002771353436401114,
      "step": 23790,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001201976288575679,
      "weighted_relation_loss": 0.001201976288575679
    },
    {
      "action_query_norm": 636.9323852539062,
      "action_relation_loss": 0.0009063062607310712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.263797942001872,
      "flow_matching_loss": 0.016557082999497652,
      "grad_norm": 0.08515956997871399,
      "learning_rate": 1.1233043454844016e-05,
      "loss": 0.0177,
      "relation_loss": 0.0011825829627923668,
      "selected_pairs": 26963.9,
      "state_query_norm": 1120.6592407226562,
      "state_relation_loss": 0.00027627667877823115,
      "step": 23800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011825829627923668,
      "weighted_relation_loss": 0.0011825829627923668
    },
    {
      "action_query_norm": 631.4666870117187,
      "action_relation_loss": 0.0009495744714513421,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.27315247895229,
      "flow_matching_loss": 0.018053250201046468,
      "grad_norm": 0.07397450506687164,
      "learning_rate": 1.1198258988450144e-05,
      "loss": 0.0193,
      "relation_loss": 0.0012204710859805346,
      "selected_pairs": 27015.4,
      "state_query_norm": 1113.5121948242188,
      "state_relation_loss": 0.00027089660143246873,
      "step": 23810,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012204710859805346,
      "weighted_relation_loss": 0.0012204710859805346
    },
    {
      "action_query_norm": 628.6008605957031,
      "action_relation_loss": 0.000950473453849554,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.282507015902713,
      "flow_matching_loss": 0.017126324586570263,
      "grad_norm": 0.08721360564231873,
      "learning_rate": 1.1163521669830368e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012282276060432197,
      "selected_pairs": 27405.1,
      "state_query_norm": 1105.4154296875,
      "state_relation_loss": 0.0002777541521936655,
      "step": 23820,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0012282276060432197,
      "weighted_relation_loss": 0.0012282276060432197
    },
    {
      "action_query_norm": 619.4263122558593,
      "action_relation_loss": 0.0009456423460505903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.291861552853135,
      "flow_matching_loss": 0.01814935337752104,
      "grad_norm": 0.0983329489827156,
      "learning_rate": 1.1128831541193818e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012160466401837766,
      "selected_pairs": 26051.9,
      "state_query_norm": 1114.3656860351562,
      "state_relation_loss": 0.0002704043232370168,
      "step": 23830,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0012160466401837766,
      "weighted_relation_loss": 0.0012160466401837766
    },
    {
      "action_query_norm": 623.7941223144531,
      "action_relation_loss": 0.0009549289883580059,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.301216089803553,
      "flow_matching_loss": 0.017025099601596594,
      "grad_norm": 0.10503718256950378,
      "learning_rate": 1.1094188644692255e-05,
      "loss": 0.0182,
      "relation_loss": 0.0012235736474394798,
      "selected_pairs": 27009.0,
      "state_query_norm": 1118.4712524414062,
      "state_relation_loss": 0.0002686446750885807,
      "step": 23840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012235736474394798,
      "weighted_relation_loss": 0.0012235736474394798
    },
    {
      "action_query_norm": 622.7093872070312,
      "action_relation_loss": 0.0009407683974131942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.310570626753975,
      "flow_matching_loss": 0.01726138051599264,
      "grad_norm": 0.08570794761180878,
      "learning_rate": 1.1059593022420067e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012229433748871088,
      "selected_pairs": 26184.5,
      "state_query_norm": 1091.8586791992188,
      "state_relation_loss": 0.0002821749832946807,
      "step": 23850,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0012229433748871088,
      "weighted_relation_loss": 0.0012229433748871088
    },
    {
      "action_query_norm": 625.3185119628906,
      "action_relation_loss": 0.0009172960591968149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.319925163704397,
      "flow_matching_loss": 0.016916987579315902,
      "grad_norm": 0.09613995254039764,
      "learning_rate": 1.1025044716414184e-05,
      "loss": 0.0181,
      "relation_loss": 0.0011867015971802175,
      "selected_pairs": 26757.7,
      "state_query_norm": 1126.2817260742188,
      "state_relation_loss": 0.00026940551470033824,
      "step": 23860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011867015971802175,
      "weighted_relation_loss": 0.0011867015971802175
    },
    {
      "action_query_norm": 624.1173767089844,
      "action_relation_loss": 0.0008921839180402458,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.32927970065482,
      "flow_matching_loss": 0.015792643837630748,
      "grad_norm": 0.08031278103590012,
      "learning_rate": 1.0990543768654082e-05,
      "loss": 0.017,
      "relation_loss": 0.0011673434055410326,
      "selected_pairs": 26804.4,
      "state_query_norm": 1105.1963012695312,
      "state_relation_loss": 0.0002751594904111698,
      "step": 23870,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011673434055410326,
      "weighted_relation_loss": 0.0011673434055410326
    },
    {
      "action_query_norm": 617.4167663574219,
      "action_relation_loss": 0.0009621594159398228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.338634237605238,
      "flow_matching_loss": 0.016348804347217083,
      "grad_norm": 0.10646867752075195,
      "learning_rate": 1.0956090221061655e-05,
      "loss": 0.0176,
      "relation_loss": 0.001239713141694665,
      "selected_pairs": 26689.7,
      "state_query_norm": 1107.0222045898438,
      "state_relation_loss": 0.000277553717023693,
      "step": 23880,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.001239713141694665,
      "weighted_relation_loss": 0.001239713141694665
    },
    {
      "action_query_norm": 624.5875549316406,
      "action_relation_loss": 0.0009215721220243722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.34798877455566,
      "flow_matching_loss": 0.017226759064942598,
      "grad_norm": 0.10221229493618011,
      "learning_rate": 1.0921684115501208e-05,
      "loss": 0.0184,
      "relation_loss": 0.0011967116734012961,
      "selected_pairs": 26859.9,
      "state_query_norm": 1143.716552734375,
      "state_relation_loss": 0.00027513955719769,
      "step": 23890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011967116734012961,
      "weighted_relation_loss": 0.0011967116734012961
    },
    {
      "action_query_norm": 622.382080078125,
      "action_relation_loss": 0.0009469216864090412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.35734331150608,
      "flow_matching_loss": 0.017322841007262467,
      "grad_norm": 0.09580028057098389,
      "learning_rate": 1.0887325493779405e-05,
      "loss": 0.0186,
      "relation_loss": 0.0012299841968342663,
      "selected_pairs": 26838.5,
      "state_query_norm": 1094.5892333984375,
      "state_relation_loss": 0.00028306250751484183,
      "step": 23900,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012299841968342663,
      "weighted_relation_loss": 0.0012299841968342663
    },
    {
      "action_query_norm": 623.2830871582031,
      "action_relation_loss": 0.0009232209529727697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.3666978484565,
      "flow_matching_loss": 0.015108714252710343,
      "grad_norm": 0.08531693369150162,
      "learning_rate": 1.0853014397645207e-05,
      "loss": 0.0163,
      "relation_loss": 0.0012080879881978036,
      "selected_pairs": 26695.7,
      "state_query_norm": 1126.1055786132813,
      "state_relation_loss": 0.0002848670381354168,
      "step": 23910,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0012080879881978036,
      "weighted_relation_loss": 0.0012080879881978036
    },
    {
      "action_query_norm": 622.7206359863281,
      "action_relation_loss": 0.0009667467908002436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.376052385406922,
      "flow_matching_loss": 0.016537322476506234,
      "grad_norm": 0.08996012061834335,
      "learning_rate": 1.0818750868789829e-05,
      "loss": 0.0178,
      "relation_loss": 0.001241770468186587,
      "selected_pairs": 26433.6,
      "state_query_norm": 1123.0723876953125,
      "state_relation_loss": 0.0002750236861174926,
      "step": 23920,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001241770468186587,
      "weighted_relation_loss": 0.001241770468186587
    },
    {
      "action_query_norm": 627.8175903320313,
      "action_relation_loss": 0.0009054087393451482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.385406922357344,
      "flow_matching_loss": 0.017359329480677844,
      "grad_norm": 0.10113988071680069,
      "learning_rate": 1.0784534948846703e-05,
      "loss": 0.0185,
      "relation_loss": 0.0011794850928708912,
      "selected_pairs": 27245.9,
      "state_query_norm": 1097.396875,
      "state_relation_loss": 0.0002740763477049768,
      "step": 23930,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011794850928708912,
      "weighted_relation_loss": 0.0011794850928708912
    },
    {
      "action_query_norm": 623.7198913574218,
      "action_relation_loss": 0.0008999524929095059,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.394761459307766,
      "flow_matching_loss": 0.01771057499572635,
      "grad_norm": 0.09300343692302704,
      "learning_rate": 1.0750366679391393e-05,
      "loss": 0.0189,
      "relation_loss": 0.001177609758451581,
      "selected_pairs": 26440.8,
      "state_query_norm": 1126.8218017578124,
      "state_relation_loss": 0.00027765725972130897,
      "step": 23940,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001177609758451581,
      "weighted_relation_loss": 0.001177609758451581
    },
    {
      "action_query_norm": 625.3261596679688,
      "action_relation_loss": 0.0008897511521354318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.404115996258184,
      "flow_matching_loss": 0.017136078141629697,
      "grad_norm": 0.09712083637714386,
      "learning_rate": 1.0716246101941557e-05,
      "loss": 0.0183,
      "relation_loss": 0.0011630253517068922,
      "selected_pairs": 27157.4,
      "state_query_norm": 1115.5921020507812,
      "state_relation_loss": 0.00027327417337801306,
      "step": 23950,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011630253517068922,
      "weighted_relation_loss": 0.0011630253517068922
    },
    {
      "action_query_norm": 621.3221374511719,
      "action_relation_loss": 0.0009662496100645513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.413470533208606,
      "flow_matching_loss": 0.015829595644026993,
      "grad_norm": 0.10249636322259903,
      "learning_rate": 1.0682173257956934e-05,
      "loss": 0.0171,
      "relation_loss": 0.0012446863926015793,
      "selected_pairs": 26709.9,
      "state_query_norm": 1132.6158447265625,
      "state_relation_loss": 0.00027843677380587903,
      "step": 23960,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0012446863926015793,
      "weighted_relation_loss": 0.0012446863926015793
    },
    {
      "action_query_norm": 624.0689392089844,
      "action_relation_loss": 0.0009059640171471984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.42282507015903,
      "flow_matching_loss": 0.014843128342181443,
      "grad_norm": 0.08420716226100922,
      "learning_rate": 1.064814818883922e-05,
      "loss": 0.016,
      "relation_loss": 0.0011892294394783675,
      "selected_pairs": 26502.5,
      "state_query_norm": 1116.9788940429687,
      "state_relation_loss": 0.00028326542815193534,
      "step": 23970,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011892294394783675,
      "weighted_relation_loss": 0.0011892294394783675
    },
    {
      "action_query_norm": 621.9559814453125,
      "action_relation_loss": 0.0009424721531104297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.432179607109447,
      "flow_matching_loss": 0.017404037341475487,
      "grad_norm": 0.08935491740703583,
      "learning_rate": 1.06141709359321e-05,
      "loss": 0.0186,
      "relation_loss": 0.0012144806561991573,
      "selected_pairs": 26835.2,
      "state_query_norm": 1115.2066284179687,
      "state_relation_loss": 0.0002720084914471954,
      "step": 23980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012144806561991573,
      "weighted_relation_loss": 0.0012144806561991573
    },
    {
      "action_query_norm": 617.5783996582031,
      "action_relation_loss": 0.0009294993127696216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.44153414405987,
      "flow_matching_loss": 0.015928572043776512,
      "grad_norm": 0.08401769399642944,
      "learning_rate": 1.058024154052114e-05,
      "loss": 0.0171,
      "relation_loss": 0.0012012785766273737,
      "selected_pairs": 26475.9,
      "state_query_norm": 1108.1447875976562,
      "state_relation_loss": 0.00027177926094736906,
      "step": 23990,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0012012785766273737,
      "weighted_relation_loss": 0.0012012785766273737
    },
    {
      "action_query_norm": 624.6928466796875,
      "action_relation_loss": 0.0008990272704977542,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.45088868101029,
      "flow_matching_loss": 0.015883292257785796,
      "grad_norm": 0.09725919365882874,
      "learning_rate": 1.0546360043833747e-05,
      "loss": 0.0171,
      "relation_loss": 0.0011749073048122228,
      "selected_pairs": 27126.5,
      "state_query_norm": 1114.5388061523438,
      "state_relation_loss": 0.0002758800284937024,
      "step": 24000,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011749073048122228,
      "weighted_relation_loss": 0.0011749073048122228
    },
    {
      "action_query_norm": 619.1935546875,
      "action_relation_loss": 0.000953997578471899,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.46024321796071,
      "flow_matching_loss": 0.017131969798356295,
      "grad_norm": 0.08824776113033295,
      "learning_rate": 1.0512526487039138e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012361495872028172,
      "selected_pairs": 27335.3,
      "state_query_norm": 1106.5672607421875,
      "state_relation_loss": 0.0002821520232828334,
      "step": 24010,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0012361495872028172,
      "weighted_relation_loss": 0.0012361495872028172
    },
    {
      "action_query_norm": 625.0677429199219,
      "action_relation_loss": 0.0009079712501261384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.46959775491113,
      "flow_matching_loss": 0.01510337544605136,
      "grad_norm": 0.0853482037782669,
      "learning_rate": 1.0478740911248258e-05,
      "loss": 0.0163,
      "relation_loss": 0.0011828524176962674,
      "selected_pairs": 27018.1,
      "state_query_norm": 1116.7538330078125,
      "state_relation_loss": 0.00027488116757012904,
      "step": 24020,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011828524176962674,
      "weighted_relation_loss": 0.0011828524176962674
    },
    {
      "action_query_norm": 625.3356811523438,
      "action_relation_loss": 0.0008936845057178289,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.478952291861553,
      "flow_matching_loss": 0.015818602032959463,
      "grad_norm": 0.07842511683702469,
      "learning_rate": 1.0445003357513793e-05,
      "loss": 0.017,
      "relation_loss": 0.0011646422441117465,
      "selected_pairs": 27104.7,
      "state_query_norm": 1107.19228515625,
      "state_relation_loss": 0.0002709577209316194,
      "step": 24030,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011646422441117465,
      "weighted_relation_loss": 0.0011646422441117465
    },
    {
      "action_query_norm": 623.3574462890625,
      "action_relation_loss": 0.0009153197286650538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.488306828811975,
      "flow_matching_loss": 0.015788966789841652,
      "grad_norm": 0.10987845063209534,
      "learning_rate": 1.0411313866830042e-05,
      "loss": 0.017,
      "relation_loss": 0.0011837521917186678,
      "selected_pairs": 27182.1,
      "state_query_norm": 1109.176171875,
      "state_relation_loss": 0.00026843246887438,
      "step": 24040,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0011837521917186678,
      "weighted_relation_loss": 0.0011837521917186678
    },
    {
      "action_query_norm": 624.73740234375,
      "action_relation_loss": 0.0009510871139355003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.497661365762394,
      "flow_matching_loss": 0.014473484922200441,
      "grad_norm": 0.08046193420886993,
      "learning_rate": 1.0377672480132915e-05,
      "loss": 0.0157,
      "relation_loss": 0.0012234525056555867,
      "selected_pairs": 27081.9,
      "state_query_norm": 1121.6555053710938,
      "state_relation_loss": 0.00027236539463046936,
      "step": 24050,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012234525056555867,
      "weighted_relation_loss": 0.0012234525056555867
    },
    {
      "action_query_norm": 626.1436157226562,
      "action_relation_loss": 0.0009637404466047883,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.507015902712816,
      "flow_matching_loss": 0.01575333746150136,
      "grad_norm": 0.11608105897903442,
      "learning_rate": 1.0344079238299864e-05,
      "loss": 0.017,
      "relation_loss": 0.0012346910312771796,
      "selected_pairs": 27527.0,
      "state_query_norm": 1127.058349609375,
      "state_relation_loss": 0.0002709505759412423,
      "step": 24060,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0012346910312771796,
      "weighted_relation_loss": 0.0012346910312771796
    },
    {
      "action_query_norm": 628.1817016601562,
      "action_relation_loss": 0.00089403708698228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.516370439663238,
      "flow_matching_loss": 0.016403579618781805,
      "grad_norm": 0.0873137041926384,
      "learning_rate": 1.0310534182149833e-05,
      "loss": 0.0176,
      "relation_loss": 0.0011793251847848295,
      "selected_pairs": 26483.9,
      "state_query_norm": 1121.541845703125,
      "state_relation_loss": 0.00028528811235446484,
      "step": 24070,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011793251847848295,
      "weighted_relation_loss": 0.0011793251847848295
    },
    {
      "action_query_norm": 620.4293212890625,
      "action_relation_loss": 0.0009418693429324776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.525724976613656,
      "flow_matching_loss": 0.016768023371696472,
      "grad_norm": 0.10699266195297241,
      "learning_rate": 1.0277037352443258e-05,
      "loss": 0.018,
      "relation_loss": 0.0012155497912317514,
      "selected_pairs": 26625.2,
      "state_query_norm": 1125.7248901367188,
      "state_relation_loss": 0.00027368044829927384,
      "step": 24080,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012155497912317514,
      "weighted_relation_loss": 0.0012155497912317514
    },
    {
      "action_query_norm": 623.8970031738281,
      "action_relation_loss": 0.000936268683290109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.535079513564078,
      "flow_matching_loss": 0.01685273926705122,
      "grad_norm": 0.08334676176309586,
      "learning_rate": 1.024358878988193e-05,
      "loss": 0.0181,
      "relation_loss": 0.0012217646930366755,
      "selected_pairs": 27366.9,
      "state_query_norm": 1122.2923583984375,
      "state_relation_loss": 0.0002854960155673325,
      "step": 24090,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0012217646930366755,
      "weighted_relation_loss": 0.0012217646930366755
    },
    {
      "action_query_norm": 624.608642578125,
      "action_relation_loss": 0.0009265411994419992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.5444340505145,
      "flow_matching_loss": 0.016033924743533136,
      "grad_norm": 0.09553567320108414,
      "learning_rate": 1.0210188535108995e-05,
      "loss": 0.0172,
      "relation_loss": 0.0012043557595461606,
      "selected_pairs": 26328.0,
      "state_query_norm": 1114.7335327148437,
      "state_relation_loss": 0.0002778145601041615,
      "step": 24100,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0012043557595461606,
      "weighted_relation_loss": 0.0012043557595461606
    },
    {
      "action_query_norm": 630.4080139160156,
      "action_relation_loss": 0.0008932181110139937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.553788587464922,
      "flow_matching_loss": 0.015456871688365936,
      "grad_norm": 0.09014163911342621,
      "learning_rate": 1.0176836628708935e-05,
      "loss": 0.0166,
      "relation_loss": 0.0011586814071051776,
      "selected_pairs": 27130.6,
      "state_query_norm": 1114.654345703125,
      "state_relation_loss": 0.00026546328736003487,
      "step": 24110,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011586814071051776,
      "weighted_relation_loss": 0.0011586814071051776
    },
    {
      "action_query_norm": 624.0698059082031,
      "action_relation_loss": 0.000913130835397169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.56314312441534,
      "flow_matching_loss": 0.016741930786520243,
      "grad_norm": 0.0996289998292923,
      "learning_rate": 1.0143533111207454e-05,
      "loss": 0.0179,
      "relation_loss": 0.0011828270158730446,
      "selected_pairs": 26878.9,
      "state_query_norm": 1119.2070556640624,
      "state_relation_loss": 0.00026969616301357745,
      "step": 24120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011828270158730446,
      "weighted_relation_loss": 0.0011828270158730446
    },
    {
      "action_query_norm": 622.2196655273438,
      "action_relation_loss": 0.0009222474880516529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.572497661365762,
      "flow_matching_loss": 0.01727447658777237,
      "grad_norm": 0.08814887702465057,
      "learning_rate": 1.0110278023071446e-05,
      "loss": 0.0185,
      "relation_loss": 0.001199721812736243,
      "selected_pairs": 26956.0,
      "state_query_norm": 1130.9921264648438,
      "state_relation_loss": 0.00027747432759497317,
      "step": 24130,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001199721812736243,
      "weighted_relation_loss": 0.001199721812736243
    },
    {
      "action_query_norm": 616.9968811035156,
      "action_relation_loss": 0.0009342006524093449,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.581852198316184,
      "flow_matching_loss": 0.01586693190038204,
      "grad_norm": 0.07716716825962067,
      "learning_rate": 1.007707140470901e-05,
      "loss": 0.0171,
      "relation_loss": 0.001210959500167519,
      "selected_pairs": 27023.3,
      "state_query_norm": 1106.3737182617188,
      "state_relation_loss": 0.00027675883466145025,
      "step": 24140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001210959500167519,
      "weighted_relation_loss": 0.001210959500167519
    },
    {
      "action_query_norm": 619.7622985839844,
      "action_relation_loss": 0.0008688150614034384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.591206735266603,
      "flow_matching_loss": 0.015246872510761022,
      "grad_norm": 0.08197775483131409,
      "learning_rate": 1.0043913296469298e-05,
      "loss": 0.0164,
      "relation_loss": 0.0011462002294138074,
      "selected_pairs": 26734.7,
      "state_query_norm": 1098.6701171875,
      "state_relation_loss": 0.00027738517674151807,
      "step": 24150,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011462002294138074,
      "weighted_relation_loss": 0.0011462002294138074
    },
    {
      "action_query_norm": 623.5232360839843,
      "action_relation_loss": 0.0009310889698099345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.600561272217025,
      "flow_matching_loss": 0.015968756563961507,
      "grad_norm": 0.08296110481023788,
      "learning_rate": 1.001080373864255e-05,
      "loss": 0.0172,
      "relation_loss": 0.001203179534059018,
      "selected_pairs": 26760.6,
      "state_query_norm": 1113.3149047851562,
      "state_relation_loss": 0.0002720905700698495,
      "step": 24160,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001203179534059018,
      "weighted_relation_loss": 0.001203179534059018
    },
    {
      "action_query_norm": 622.1101623535156,
      "action_relation_loss": 0.0009660072450060397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.609915809167447,
      "flow_matching_loss": 0.01657915338873863,
      "grad_norm": 0.07310114055871964,
      "learning_rate": 9.977742771459991e-06,
      "loss": 0.0178,
      "relation_loss": 0.0012480676406994462,
      "selected_pairs": 26699.1,
      "state_query_norm": 1113.6751220703125,
      "state_relation_loss": 0.00028206038114149125,
      "step": 24170,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0012480676406994462,
      "weighted_relation_loss": 0.0012480676406994462
    },
    {
      "action_query_norm": 621.8888671875,
      "action_relation_loss": 0.0009054901311174035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.61927034611787,
      "flow_matching_loss": 0.017020731046795846,
      "grad_norm": 0.08610329777002335,
      "learning_rate": 9.944730435093802e-06,
      "loss": 0.0182,
      "relation_loss": 0.001178442931268364,
      "selected_pairs": 27487.1,
      "state_query_norm": 1106.0921142578125,
      "state_relation_loss": 0.00027295281470287593,
      "step": 24180,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.001178442931268364,
      "weighted_relation_loss": 0.001178442931268364
    },
    {
      "action_query_norm": 625.1649658203125,
      "action_relation_loss": 0.0009434264153242111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.628624883068287,
      "flow_matching_loss": 0.01554007399827242,
      "grad_norm": 0.09247126430273056,
      "learning_rate": 9.911766769657115e-06,
      "loss": 0.0168,
      "relation_loss": 0.001237809972371906,
      "selected_pairs": 26993.2,
      "state_query_norm": 1107.7980834960938,
      "state_relation_loss": 0.0002943835657788441,
      "step": 24190,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001237809972371906,
      "weighted_relation_loss": 0.001237809972371906
    },
    {
      "action_query_norm": 627.9033447265625,
      "action_relation_loss": 0.0009256146848201751,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.63797942001871,
      "flow_matching_loss": 0.016769327875226735,
      "grad_norm": 0.08743304759263992,
      "learning_rate": 9.878851815203882e-06,
      "loss": 0.018,
      "relation_loss": 0.0012038657325319946,
      "selected_pairs": 27096.7,
      "state_query_norm": 1106.5748291015625,
      "state_relation_loss": 0.00027825104916701093,
      "step": 24200,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012038657325319946,
      "weighted_relation_loss": 0.0012038657325319946
    },
    {
      "action_query_norm": 625.1506713867187,
      "action_relation_loss": 0.0009114625747315586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.64733395696913,
      "flow_matching_loss": 0.015935923904180527,
      "grad_norm": 0.08738727867603302,
      "learning_rate": 9.845985611728864e-06,
      "loss": 0.0171,
      "relation_loss": 0.0011902270722202956,
      "selected_pairs": 26617.4,
      "state_query_norm": 1112.3725341796876,
      "state_relation_loss": 0.000278764491667971,
      "step": 24210,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011902270722202956,
      "weighted_relation_loss": 0.0011902270722202956
    },
    {
      "action_query_norm": 624.2582641601563,
      "action_relation_loss": 0.0009743596776388586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.65668849391955,
      "flow_matching_loss": 0.01808887403458357,
      "grad_norm": 0.08613809198141098,
      "learning_rate": 9.813168199167606e-06,
      "loss": 0.0193,
      "relation_loss": 0.001252589374780655,
      "selected_pairs": 26825.5,
      "state_query_norm": 1111.5634643554688,
      "state_relation_loss": 0.00027822970878332853,
      "step": 24220,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001252589374780655,
      "weighted_relation_loss": 0.001252589374780655
    },
    {
      "action_query_norm": 618.7889953613281,
      "action_relation_loss": 0.000903220137115568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.66604303086997,
      "flow_matching_loss": 0.018140569049865007,
      "grad_norm": 0.10535567253828049,
      "learning_rate": 9.780399617396341e-06,
      "loss": 0.0193,
      "relation_loss": 0.0011771070072427391,
      "selected_pairs": 26780.5,
      "state_query_norm": 1109.674267578125,
      "state_relation_loss": 0.00027388686139602215,
      "step": 24230,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011771070072427391,
      "weighted_relation_loss": 0.0011771070072427391
    },
    {
      "action_query_norm": 622.1361450195312,
      "action_relation_loss": 0.0009394775668624788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.675397567820394,
      "flow_matching_loss": 0.017437940649688243,
      "grad_norm": 0.09522363543510437,
      "learning_rate": 9.747679906232016e-06,
      "loss": 0.0187,
      "relation_loss": 0.0012164993560872972,
      "selected_pairs": 26877.4,
      "state_query_norm": 1122.447265625,
      "state_relation_loss": 0.0002770218008663505,
      "step": 24240,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012164993560872972,
      "weighted_relation_loss": 0.0012164993560872972
    },
    {
      "action_query_norm": 618.5188232421875,
      "action_relation_loss": 0.0009689096303191036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.684752104770816,
      "flow_matching_loss": 0.016539743915200234,
      "grad_norm": 0.07887603342533112,
      "learning_rate": 9.715009105432138e-06,
      "loss": 0.0178,
      "relation_loss": 0.0012335013248957693,
      "selected_pairs": 27467.2,
      "state_query_norm": 1112.0448364257813,
      "state_relation_loss": 0.00026459171203896403,
      "step": 24250,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0012335013248957693,
      "weighted_relation_loss": 0.0012335013248957693
    },
    {
      "action_query_norm": 627.6493713378907,
      "action_relation_loss": 0.0009081779513508082,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.694106641721234,
      "flow_matching_loss": 0.016577098984271288,
      "grad_norm": 0.07279346883296967,
      "learning_rate": 9.682387254694835e-06,
      "loss": 0.0178,
      "relation_loss": 0.0011781352804973722,
      "selected_pairs": 26482.2,
      "state_query_norm": 1125.3159912109375,
      "state_relation_loss": 0.00026995732623618095,
      "step": 24260,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011781352804973722,
      "weighted_relation_loss": 0.0011781352804973722
    },
    {
      "action_query_norm": 620.6500549316406,
      "action_relation_loss": 0.0008946229587309063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.703461178671656,
      "flow_matching_loss": 0.01684812447056174,
      "grad_norm": 0.09319277852773666,
      "learning_rate": 9.649814393658723e-06,
      "loss": 0.018,
      "relation_loss": 0.0011663399054668844,
      "selected_pairs": 26526.5,
      "state_query_norm": 1116.31328125,
      "state_relation_loss": 0.00027171694382559507,
      "step": 24270,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0011663399054668844,
      "weighted_relation_loss": 0.0011663399054668844
    },
    {
      "action_query_norm": 626.7356323242187,
      "action_relation_loss": 0.0009128823992796242,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.712815715622078,
      "flow_matching_loss": 0.01741507798433304,
      "grad_norm": 0.09057314693927765,
      "learning_rate": 9.617290561902881e-06,
      "loss": 0.0186,
      "relation_loss": 0.0011886966298334301,
      "selected_pairs": 27218.8,
      "state_query_norm": 1110.9777465820312,
      "state_relation_loss": 0.0002758142160018906,
      "step": 24280,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011886966298334301,
      "weighted_relation_loss": 0.0011886966298334301
    },
    {
      "action_query_norm": 623.8685485839844,
      "action_relation_loss": 0.0009287794993724674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.722170252572496,
      "flow_matching_loss": 0.017960681673139335,
      "grad_norm": 0.08694774657487869,
      "learning_rate": 9.584815798946861e-06,
      "loss": 0.0192,
      "relation_loss": 0.001200340932700783,
      "selected_pairs": 26989.9,
      "state_query_norm": 1111.9853759765624,
      "state_relation_loss": 0.00027156143041793255,
      "step": 24290,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001200340932700783,
      "weighted_relation_loss": 0.001200340932700783
    },
    {
      "action_query_norm": 625.9180297851562,
      "action_relation_loss": 0.0009101367671974003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.73152478952292,
      "flow_matching_loss": 0.0152638491243124,
      "grad_norm": 0.07847467064857483,
      "learning_rate": 9.55239014425055e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011820287560112774,
      "selected_pairs": 26968.6,
      "state_query_norm": 1115.4712524414062,
      "state_relation_loss": 0.0002718920004554093,
      "step": 24300,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011820287560112774,
      "weighted_relation_loss": 0.0011820287560112774
    },
    {
      "action_query_norm": 624.549609375,
      "action_relation_loss": 0.0009110008308198303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.74087932647334,
      "flow_matching_loss": 0.01757034193724394,
      "grad_norm": 0.09555621445178986,
      "learning_rate": 9.520013637214176e-06,
      "loss": 0.0188,
      "relation_loss": 0.0011797063867561518,
      "selected_pairs": 26895.7,
      "state_query_norm": 1114.7474243164063,
      "state_relation_loss": 0.00026870556175708773,
      "step": 24310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011797063867561518,
      "weighted_relation_loss": 0.0011797063867561518
    },
    {
      "action_query_norm": 627.0669250488281,
      "action_relation_loss": 0.0009280446101911366,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.75023386342376,
      "flow_matching_loss": 0.01624816581606865,
      "grad_norm": 0.13225051760673523,
      "learning_rate": 9.48768631717824e-06,
      "loss": 0.0174,
      "relation_loss": 0.0011972368229180574,
      "selected_pairs": 27479.5,
      "state_query_norm": 1120.3214599609375,
      "state_relation_loss": 0.0002691921952646226,
      "step": 24320,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0011972368229180574,
      "weighted_relation_loss": 0.0011972368229180574
    },
    {
      "action_query_norm": 625.0071533203125,
      "action_relation_loss": 0.000920135824708268,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.75958840037418,
      "flow_matching_loss": 0.016037026420235634,
      "grad_norm": 0.08240415155887604,
      "learning_rate": 9.455408223423496e-06,
      "loss": 0.0172,
      "relation_loss": 0.00118791387649253,
      "selected_pairs": 26052.3,
      "state_query_norm": 1127.0805786132812,
      "state_relation_loss": 0.00026777806924656035,
      "step": 24330,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.00118791387649253,
      "weighted_relation_loss": 0.00118791387649253
    },
    {
      "action_query_norm": 621.0952453613281,
      "action_relation_loss": 0.0009072835324332118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.768942937324603,
      "flow_matching_loss": 0.01780458688735962,
      "grad_norm": 0.08156043291091919,
      "learning_rate": 9.423179395170845e-06,
      "loss": 0.019,
      "relation_loss": 0.0011893147369846702,
      "selected_pairs": 26851.1,
      "state_query_norm": 1119.1237426757812,
      "state_relation_loss": 0.00028203118999954314,
      "step": 24340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011893147369846702,
      "weighted_relation_loss": 0.0011893147369846702
    },
    {
      "action_query_norm": 626.2506469726562,
      "action_relation_loss": 0.0009433712752070278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.778297474275025,
      "flow_matching_loss": 0.01681095650419593,
      "grad_norm": 0.08209434896707535,
      "learning_rate": 9.390999871581391e-06,
      "loss": 0.018,
      "relation_loss": 0.0012176341144368053,
      "selected_pairs": 27927.1,
      "state_query_norm": 1117.6961181640625,
      "state_relation_loss": 0.00027426285960245876,
      "step": 24350,
      "valid_action_samples": 236.8,
      "weighted_auxiliary_loss": 0.0012176341144368053,
      "weighted_relation_loss": 0.0012176341144368053
    },
    {
      "action_query_norm": 629.9229919433594,
      "action_relation_loss": 0.0009175916959065944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.787652011225443,
      "flow_matching_loss": 0.017504081968218087,
      "grad_norm": 0.09343820810317993,
      "learning_rate": 9.358869691756273e-06,
      "loss": 0.0187,
      "relation_loss": 0.0012221463141031564,
      "selected_pairs": 26719.0,
      "state_query_norm": 1111.1854370117187,
      "state_relation_loss": 0.00030455461237579586,
      "step": 24360,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0012221463141031564,
      "weighted_relation_loss": 0.0012221463141031564
    },
    {
      "action_query_norm": 619.8075927734375,
      "action_relation_loss": 0.0009170328732579947,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.797006548175865,
      "flow_matching_loss": 0.01710319984704256,
      "grad_norm": 0.09375317394733429,
      "learning_rate": 9.326788894736688e-06,
      "loss": 0.0183,
      "relation_loss": 0.0011940348078496755,
      "selected_pairs": 26320.1,
      "state_query_norm": 1117.39755859375,
      "state_relation_loss": 0.00027700191421899947,
      "step": 24370,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0011940348078496755,
      "weighted_relation_loss": 0.0011940348078496755
    },
    {
      "action_query_norm": 619.8251525878907,
      "action_relation_loss": 0.000896110909525305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.806361085126287,
      "flow_matching_loss": 0.015857155621051788,
      "grad_norm": 0.09375438839197159,
      "learning_rate": 9.294757519503811e-06,
      "loss": 0.017,
      "relation_loss": 0.0011727308155968785,
      "selected_pairs": 26846.6,
      "state_query_norm": 1093.992822265625,
      "state_relation_loss": 0.00027661991625791413,
      "step": 24380,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011727308155968785,
      "weighted_relation_loss": 0.0011727308155968785
    },
    {
      "action_query_norm": 616.3810119628906,
      "action_relation_loss": 0.0008860638248734176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.815715622076706,
      "flow_matching_loss": 0.015911099221557378,
      "grad_norm": 0.0909951850771904,
      "learning_rate": 9.262775604978819e-06,
      "loss": 0.0171,
      "relation_loss": 0.001158116734586656,
      "selected_pairs": 26919.4,
      "state_query_norm": 1119.3882568359375,
      "state_relation_loss": 0.0002720529184443876,
      "step": 24390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001158116734586656,
      "weighted_relation_loss": 0.001158116734586656
    },
    {
      "action_query_norm": 624.4159851074219,
      "action_relation_loss": 0.0009454931423533708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.825070159027128,
      "flow_matching_loss": 0.016941223107278348,
      "grad_norm": 0.09199453145265579,
      "learning_rate": 9.230843190022725e-06,
      "loss": 0.0182,
      "relation_loss": 0.001216557319276035,
      "selected_pairs": 26739.8,
      "state_query_norm": 1108.240380859375,
      "state_relation_loss": 0.00027106415800517427,
      "step": 24400,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001216557319276035,
      "weighted_relation_loss": 0.001216557319276035
    },
    {
      "action_query_norm": 617.8299926757812,
      "action_relation_loss": 0.000930997944669798,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.83442469597755,
      "flow_matching_loss": 0.01622037524357438,
      "grad_norm": 0.093663290143013,
      "learning_rate": 9.198960313436445e-06,
      "loss": 0.0174,
      "relation_loss": 0.0012170742847956717,
      "selected_pairs": 26754.1,
      "state_query_norm": 1112.4754760742187,
      "state_relation_loss": 0.00028607634012587366,
      "step": 24410,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012170742847956717,
      "weighted_relation_loss": 0.0012170742847956717
    },
    {
      "action_query_norm": 628.9302673339844,
      "action_relation_loss": 0.0009359479590784758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.84377923292797,
      "flow_matching_loss": 0.016409194376319648,
      "grad_norm": 0.0858210101723671,
      "learning_rate": 9.16712701396067e-06,
      "loss": 0.0176,
      "relation_loss": 0.0012156171724200249,
      "selected_pairs": 27244.9,
      "state_query_norm": 1117.7039672851563,
      "state_relation_loss": 0.0002796691987896338,
      "step": 24420,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0012156171724200249,
      "weighted_relation_loss": 0.0012156171724200249
    },
    {
      "action_query_norm": 622.6022644042969,
      "action_relation_loss": 0.0009346320352051407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.85313376987839,
      "flow_matching_loss": 0.015157936979085207,
      "grad_norm": 0.08698903769254684,
      "learning_rate": 9.135343330275864e-06,
      "loss": 0.0164,
      "relation_loss": 0.0012078322004526853,
      "selected_pairs": 26730.6,
      "state_query_norm": 1123.1325439453126,
      "state_relation_loss": 0.0002732001608819701,
      "step": 24430,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012078322004526853,
      "weighted_relation_loss": 0.0012078322004526853
    },
    {
      "action_query_norm": 622.8950927734375,
      "action_relation_loss": 0.0009033191134221851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.862488306828812,
      "flow_matching_loss": 0.015270424354821444,
      "grad_norm": 0.0825263112783432,
      "learning_rate": 9.10360930100218e-06,
      "loss": 0.0165,
      "relation_loss": 0.0011822815402410924,
      "selected_pairs": 27199.3,
      "state_query_norm": 1103.9465942382812,
      "state_relation_loss": 0.00027896243263967333,
      "step": 24440,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011822815402410924,
      "weighted_relation_loss": 0.0011822815402410924
    },
    {
      "action_query_norm": 629.5935180664062,
      "action_relation_loss": 0.0008942122105509043,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.871842843779234,
      "flow_matching_loss": 0.016219641733914613,
      "grad_norm": 0.08236978948116302,
      "learning_rate": 9.07192496469949e-06,
      "loss": 0.0174,
      "relation_loss": 0.0011814710218459369,
      "selected_pairs": 26649.6,
      "state_query_norm": 1135.2120483398437,
      "state_relation_loss": 0.00028725880838464946,
      "step": 24450,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0011814710218459369,
      "weighted_relation_loss": 0.0011814710218459369
    },
    {
      "action_query_norm": 625.4584106445312,
      "action_relation_loss": 0.0009098835580516607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.881197380729652,
      "flow_matching_loss": 0.014652752503752708,
      "grad_norm": 0.0858851745724678,
      "learning_rate": 9.040290359867231e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011825373745523392,
      "selected_pairs": 26925.1,
      "state_query_norm": 1100.9650634765626,
      "state_relation_loss": 0.0002726538252318278,
      "step": 24460,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011825373745523392,
      "weighted_relation_loss": 0.0011825373745523392
    },
    {
      "action_query_norm": 622.0926025390625,
      "action_relation_loss": 0.0008902704168576748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.890551917680074,
      "flow_matching_loss": 0.017050751578062774,
      "grad_norm": 0.08978137373924255,
      "learning_rate": 9.008705524944438e-06,
      "loss": 0.0182,
      "relation_loss": 0.0011615013005211948,
      "selected_pairs": 27174.9,
      "state_query_norm": 1114.109375,
      "state_relation_loss": 0.00027123088511871174,
      "step": 24470,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0011615013005211948,
      "weighted_relation_loss": 0.0011615013005211948
    },
    {
      "action_query_norm": 615.3507202148437,
      "action_relation_loss": 0.0008919690677430481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.899906454630496,
      "flow_matching_loss": 0.0161358131095767,
      "grad_norm": 0.06795226037502289,
      "learning_rate": 8.977170498309651e-06,
      "loss": 0.0173,
      "relation_loss": 0.00117580930236727,
      "selected_pairs": 26758.1,
      "state_query_norm": 1111.8300659179688,
      "state_relation_loss": 0.00028384024917613713,
      "step": 24480,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.00117580930236727,
      "weighted_relation_loss": 0.00117580930236727
    },
    {
      "action_query_norm": 617.25673828125,
      "action_relation_loss": 0.0009355209360364825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.90926099158092,
      "flow_matching_loss": 0.015974684804677963,
      "grad_norm": 0.10042708367109299,
      "learning_rate": 8.945685318280916e-06,
      "loss": 0.0172,
      "relation_loss": 0.0012072237790562213,
      "selected_pairs": 26942.2,
      "state_query_norm": 1120.875927734375,
      "state_relation_loss": 0.00027170286339242013,
      "step": 24490,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0012072237790562213,
      "weighted_relation_loss": 0.0012072237790562213
    },
    {
      "action_query_norm": 623.5231506347657,
      "action_relation_loss": 0.0009676383750047534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.918615528531337,
      "flow_matching_loss": 0.016880141105502844,
      "grad_norm": 0.0889635905623436,
      "learning_rate": 8.914250023115672e-06,
      "loss": 0.0181,
      "relation_loss": 0.0012421263265423477,
      "selected_pairs": 27293.2,
      "state_query_norm": 1104.2764892578125,
      "state_relation_loss": 0.00027448796754470095,
      "step": 24500,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0012421263265423477,
      "weighted_relation_loss": 0.0012421263265423477
    },
    {
      "action_query_norm": 623.3124084472656,
      "action_relation_loss": 0.0008950951567385345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.92797006548176,
      "flow_matching_loss": 0.014881814736872912,
      "grad_norm": 0.08887587487697601,
      "learning_rate": 8.882864651010798e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011782227898947894,
      "selected_pairs": 26600.5,
      "state_query_norm": 1124.1951049804688,
      "state_relation_loss": 0.0002831276273354888,
      "step": 24510,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011782227898947894,
      "weighted_relation_loss": 0.0011782227898947894
    },
    {
      "action_query_norm": 619.7880859375,
      "action_relation_loss": 0.000898781040450558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.93732460243218,
      "flow_matching_loss": 0.016136032808572055,
      "grad_norm": 0.0852581113576889,
      "learning_rate": 8.851529240102462e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011734928819350898,
      "selected_pairs": 26551.2,
      "state_query_norm": 1109.2405029296874,
      "state_relation_loss": 0.0002747118385741487,
      "step": 24520,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011734928819350898,
      "weighted_relation_loss": 0.0011734928819350898
    },
    {
      "action_query_norm": 620.8747924804687,
      "action_relation_loss": 0.0008931992808356882,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.9466791393826,
      "flow_matching_loss": 0.016910208389163017,
      "grad_norm": 0.09979775547981262,
      "learning_rate": 8.820243828466134e-06,
      "loss": 0.0181,
      "relation_loss": 0.0011618332820944488,
      "selected_pairs": 26626.4,
      "state_query_norm": 1108.9876831054687,
      "state_relation_loss": 0.00026863400998990985,
      "step": 24530,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011618332820944488,
      "weighted_relation_loss": 0.0011618332820944488
    },
    {
      "action_query_norm": 625.8627380371094,
      "action_relation_loss": 0.0009459350723773241,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.95603367633302,
      "flow_matching_loss": 0.015706556290388106,
      "grad_norm": 0.08015875518321991,
      "learning_rate": 8.789008454116566e-06,
      "loss": 0.0169,
      "relation_loss": 0.001219566399231553,
      "selected_pairs": 27127.0,
      "state_query_norm": 1119.3735229492188,
      "state_relation_loss": 0.00027363133267499506,
      "step": 24540,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.001219566399231553,
      "weighted_relation_loss": 0.001219566399231553
    },
    {
      "action_query_norm": 622.4143188476562,
      "action_relation_loss": 0.0009156638698186725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.965388213283443,
      "flow_matching_loss": 0.015507622994482518,
      "grad_norm": 0.08824001997709274,
      "learning_rate": 8.757823155007655e-06,
      "loss": 0.0167,
      "relation_loss": 0.001200847188010812,
      "selected_pairs": 26574.6,
      "state_query_norm": 1119.9877685546876,
      "state_relation_loss": 0.00028518329781945795,
      "step": 24550,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.001200847188010812,
      "weighted_relation_loss": 0.001200847188010812
    },
    {
      "action_query_norm": 619.2166625976563,
      "action_relation_loss": 0.0009132474544458092,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.974742750233865,
      "flow_matching_loss": 0.01451640399172902,
      "grad_norm": 0.10637693107128143,
      "learning_rate": 8.72668796903251e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011933814035728573,
      "selected_pairs": 26961.0,
      "state_query_norm": 1110.6894897460938,
      "state_relation_loss": 0.00028013394912704823,
      "step": 24560,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011933814035728573,
      "weighted_relation_loss": 0.0011933814035728573
    },
    {
      "action_query_norm": 620.6001281738281,
      "action_relation_loss": 0.0008824851363897323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.984097287184284,
      "flow_matching_loss": 0.016195790003985168,
      "grad_norm": 0.08999781310558319,
      "learning_rate": 8.695602934023301e-06,
      "loss": 0.0174,
      "relation_loss": 0.0011647137347608806,
      "selected_pairs": 26544.5,
      "state_query_norm": 1115.3982788085937,
      "state_relation_loss": 0.00028222858818480743,
      "step": 24570,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011647137347608806,
      "weighted_relation_loss": 0.0011647137347608806
    },
    {
      "action_query_norm": 622.56201171875,
      "action_relation_loss": 0.0009459938097279519,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.993451824134706,
      "flow_matching_loss": 0.014953601360321044,
      "grad_norm": 0.08130466192960739,
      "learning_rate": 8.664568087751274e-06,
      "loss": 0.0162,
      "relation_loss": 0.0012240541051141917,
      "selected_pairs": 26387.5,
      "state_query_norm": 1139.27001953125,
      "state_relation_loss": 0.00027806029247585685,
      "step": 24580,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0012240541051141917,
      "weighted_relation_loss": 0.0012240541051141917
    },
    {
      "action_query_norm": 625.8769592285156,
      "action_relation_loss": 0.0009134112682659179,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.002806361085128,
      "flow_matching_loss": 0.015860987175256015,
      "grad_norm": 0.09291727095842361,
      "learning_rate": 8.633583467926697e-06,
      "loss": 0.017,
      "relation_loss": 0.0011801236192695796,
      "selected_pairs": 24489.3,
      "state_query_norm": 1113.6604248046874,
      "state_relation_loss": 0.00026671234954847025,
      "step": 24590,
      "valid_action_samples": 214.7,
      "weighted_auxiliary_loss": 0.0011801236192695796,
      "weighted_relation_loss": 0.0011801236192695796
    },
    {
      "action_query_norm": 621.5218017578125,
      "action_relation_loss": 0.0009215488622430712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.012160898035546,
      "flow_matching_loss": 0.016447700466960667,
      "grad_norm": 0.10640475898981094,
      "learning_rate": 8.602649112198796e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011901857797056437,
      "selected_pairs": 26872.9,
      "state_query_norm": 1113.4421630859374,
      "state_relation_loss": 0.00026863691600738094,
      "step": 24600,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011901857797056437,
      "weighted_relation_loss": 0.0011901857797056437
    },
    {
      "action_query_norm": 623.8995544433594,
      "action_relation_loss": 0.0009242206171620638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.021515434985968,
      "flow_matching_loss": 0.016730991285294296,
      "grad_norm": 0.09595052897930145,
      "learning_rate": 8.571765058155745e-06,
      "loss": 0.0179,
      "relation_loss": 0.0011986115016043187,
      "selected_pairs": 26931.1,
      "state_query_norm": 1121.7419799804688,
      "state_relation_loss": 0.00027439086989033967,
      "step": 24610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011986115016043187,
      "weighted_relation_loss": 0.0011986115016043187
    },
    {
      "action_query_norm": 616.1539794921875,
      "action_relation_loss": 0.000955081795109436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.03086997193639,
      "flow_matching_loss": 0.0159840002655983,
      "grad_norm": 0.0861804410815239,
      "learning_rate": 8.540931343324583e-06,
      "loss": 0.0172,
      "relation_loss": 0.001227953948546201,
      "selected_pairs": 26990.8,
      "state_query_norm": 1106.520166015625,
      "state_relation_loss": 0.0002728721534367651,
      "step": 24620,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001227953948546201,
      "weighted_relation_loss": 0.001227953948546201
    },
    {
      "action_query_norm": 624.680078125,
      "action_relation_loss": 0.0008628042181953788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.04022450888681,
      "flow_matching_loss": 0.016040533035993575,
      "grad_norm": 0.07867874950170517,
      "learning_rate": 8.51014800517117e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011347624473273753,
      "selected_pairs": 26803.9,
      "state_query_norm": 1118.8493286132812,
      "state_relation_loss": 0.0002719582262216136,
      "step": 24630,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011347624473273753,
      "weighted_relation_loss": 0.0011347624473273753
    },
    {
      "action_query_norm": 628.1125610351562,
      "action_relation_loss": 0.0009003599116113037,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.04957904583723,
      "flow_matching_loss": 0.015283575095236302,
      "grad_norm": 0.07113049924373627,
      "learning_rate": 8.479415081100167e-06,
      "loss": 0.0165,
      "relation_loss": 0.001178681361488998,
      "selected_pairs": 27299.5,
      "state_query_norm": 1110.6216552734375,
      "state_relation_loss": 0.00027832143241539596,
      "step": 24640,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.001178681361488998,
      "weighted_relation_loss": 0.001178681361488998
    },
    {
      "action_query_norm": 624.073388671875,
      "action_relation_loss": 0.0008841004106216133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.058933582787652,
      "flow_matching_loss": 0.01652136417105794,
      "grad_norm": 0.08131279051303864,
      "learning_rate": 8.448732608454967e-06,
      "loss": 0.0177,
      "relation_loss": 0.0011619178694672883,
      "selected_pairs": 27127.9,
      "state_query_norm": 1109.1705322265625,
      "state_relation_loss": 0.0002778174733975902,
      "step": 24650,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011619178694672883,
      "weighted_relation_loss": 0.0011619178694672883
    },
    {
      "action_query_norm": 617.3503234863281,
      "action_relation_loss": 0.0009597852302249521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.068288119738074,
      "flow_matching_loss": 0.01667954623699188,
      "grad_norm": 0.0928587093949318,
      "learning_rate": 8.418100624517688e-06,
      "loss": 0.0179,
      "relation_loss": 0.0012401699787005782,
      "selected_pairs": 26635.6,
      "state_query_norm": 1108.509326171875,
      "state_relation_loss": 0.0002803847542963922,
      "step": 24660,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0012401699787005782,
      "weighted_relation_loss": 0.0012401699787005782
    },
    {
      "action_query_norm": 619.0235107421875,
      "action_relation_loss": 0.0009231972682755441,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.077642656688493,
      "flow_matching_loss": 0.016211727727204563,
      "grad_norm": 0.09953274577856064,
      "learning_rate": 8.387519166509062e-06,
      "loss": 0.0174,
      "relation_loss": 0.0012038101791404187,
      "selected_pairs": 26813.1,
      "state_query_norm": 1105.0210083007812,
      "state_relation_loss": 0.0002806129137752578,
      "step": 24670,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012038101791404187,
      "weighted_relation_loss": 0.0012038101791404187
    },
    {
      "action_query_norm": 621.8672119140625,
      "action_relation_loss": 0.0009245680761523545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.086997193638915,
      "flow_matching_loss": 0.015725597087293864,
      "grad_norm": 0.10605961829423904,
      "learning_rate": 8.356988271588445e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011942085810005664,
      "selected_pairs": 27224.4,
      "state_query_norm": 1112.747265625,
      "state_relation_loss": 0.00026964050775859507,
      "step": 24680,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011942085810005664,
      "weighted_relation_loss": 0.0011942085810005664
    },
    {
      "action_query_norm": 626.4321411132812,
      "action_relation_loss": 0.000865386676741764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.096351730589337,
      "flow_matching_loss": 0.01702880412340164,
      "grad_norm": 0.08751671761274338,
      "learning_rate": 8.326507976853765e-06,
      "loss": 0.0182,
      "relation_loss": 0.0011401902185752988,
      "selected_pairs": 26491.7,
      "state_query_norm": 1109.7334716796875,
      "state_relation_loss": 0.00027480356220621616,
      "step": 24690,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011401902185752988,
      "weighted_relation_loss": 0.0011401902185752988
    },
    {
      "action_query_norm": 621.1516845703125,
      "action_relation_loss": 0.0008911475306376815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.105706267539755,
      "flow_matching_loss": 0.015933684073388577,
      "grad_norm": 0.0922563225030899,
      "learning_rate": 8.296078319341443e-06,
      "loss": 0.0171,
      "relation_loss": 0.0011657510069198907,
      "selected_pairs": 27105.7,
      "state_query_norm": 1106.3175659179688,
      "state_relation_loss": 0.00027460347191663457,
      "step": 24700,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011657510069198907,
      "weighted_relation_loss": 0.0011657510069198907
    },
    {
      "action_query_norm": 625.8445495605469,
      "action_relation_loss": 0.0008962241350673139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.115060804490177,
      "flow_matching_loss": 0.015896189492195845,
      "grad_norm": 0.09550672024488449,
      "learning_rate": 8.265699336026383e-06,
      "loss": 0.0171,
      "relation_loss": 0.001165277964901179,
      "selected_pairs": 26964.4,
      "state_query_norm": 1131.4564575195313,
      "state_relation_loss": 0.0002690538385650143,
      "step": 24710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001165277964901179,
      "weighted_relation_loss": 0.001165277964901179
    },
    {
      "action_query_norm": 622.1709594726562,
      "action_relation_loss": 0.0009031343914102763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.1244153414406,
      "flow_matching_loss": 0.015874268114566804,
      "grad_norm": 0.08194174617528915,
      "learning_rate": 8.235371063821923e-06,
      "loss": 0.0171,
      "relation_loss": 0.0011803998844698072,
      "selected_pairs": 26925.7,
      "state_query_norm": 1109.7080444335938,
      "state_relation_loss": 0.0002772654726868495,
      "step": 24720,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011803998844698072,
      "weighted_relation_loss": 0.0011803998844698072
    },
    {
      "action_query_norm": 624.3716491699219,
      "action_relation_loss": 0.0009034375369083136,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.13376987839102,
      "flow_matching_loss": 0.015409342758357526,
      "grad_norm": 0.07605388015508652,
      "learning_rate": 8.205093539579767e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011744956369511782,
      "selected_pairs": 27083.4,
      "state_query_norm": 1103.5962646484375,
      "state_relation_loss": 0.00027105810586363076,
      "step": 24730,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011744956369511782,
      "weighted_relation_loss": 0.0011744956369511782
    },
    {
      "action_query_norm": 625.185986328125,
      "action_relation_loss": 0.0009021906123962253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.14312441534144,
      "flow_matching_loss": 0.016836373414844275,
      "grad_norm": 0.09398201107978821,
      "learning_rate": 8.174866800089963e-06,
      "loss": 0.018,
      "relation_loss": 0.0011736214626580476,
      "selected_pairs": 27128.5,
      "state_query_norm": 1106.9516479492188,
      "state_relation_loss": 0.0002714308415306732,
      "step": 24740,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011736214626580476,
      "weighted_relation_loss": 0.0011736214626580476
    },
    {
      "action_query_norm": 626.1563659667969,
      "action_relation_loss": 0.000870727829169482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.15247895229186,
      "flow_matching_loss": 0.016156031005084514,
      "grad_norm": 0.0934443473815918,
      "learning_rate": 8.144690882080853e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011329182074405254,
      "selected_pairs": 26945.2,
      "state_query_norm": 1119.616064453125,
      "state_relation_loss": 0.00026219037536066023,
      "step": 24750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011329182074405254,
      "weighted_relation_loss": 0.0011329182074405254
    },
    {
      "action_query_norm": 619.6306396484375,
      "action_relation_loss": 0.0008773476642090827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.161833489242284,
      "flow_matching_loss": 0.01551433065906167,
      "grad_norm": 0.07491499185562134,
      "learning_rate": 8.114565822219006e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011614767368882894,
      "selected_pairs": 26463.0,
      "state_query_norm": 1123.4702758789062,
      "state_relation_loss": 0.0002841290843207389,
      "step": 24760,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011614767368882894,
      "weighted_relation_loss": 0.0011614767368882894
    },
    {
      "action_query_norm": 624.2089294433594,
      "action_relation_loss": 0.0008965764252934604,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.171188026192702,
      "flow_matching_loss": 0.015581388771533967,
      "grad_norm": 0.0970882847905159,
      "learning_rate": 8.084491657109233e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011865498963743448,
      "selected_pairs": 26872.1,
      "state_query_norm": 1101.6314331054687,
      "state_relation_loss": 0.00028997345361858606,
      "step": 24770,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011865498963743448,
      "weighted_relation_loss": 0.0011865498963743448
    },
    {
      "action_query_norm": 619.1877990722656,
      "action_relation_loss": 0.0008868593198712915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.180542563143124,
      "flow_matching_loss": 0.015056103188544511,
      "grad_norm": 0.09487228095531464,
      "learning_rate": 8.054468423294476e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011593182571232319,
      "selected_pairs": 26274.3,
      "state_query_norm": 1120.0314575195312,
      "state_relation_loss": 0.00027245894307270646,
      "step": 24780,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0011593182571232319,
      "weighted_relation_loss": 0.0011593182571232319
    },
    {
      "action_query_norm": 624.9968872070312,
      "action_relation_loss": 0.0008889817108865827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.189897100093546,
      "flow_matching_loss": 0.015143971517682075,
      "grad_norm": 0.09546031802892685,
      "learning_rate": 8.024496157255785e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011610527057200671,
      "selected_pairs": 26669.2,
      "state_query_norm": 1108.4308715820312,
      "state_relation_loss": 0.00027207098610233517,
      "step": 24790,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011610527057200671,
      "weighted_relation_loss": 0.0011610527057200671
    },
    {
      "action_query_norm": 621.6230224609375,
      "action_relation_loss": 0.0008966881548985838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.199251637043965,
      "flow_matching_loss": 0.016342039685696363,
      "grad_norm": 0.10133738815784454,
      "learning_rate": 7.994574895412293e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011670790263451636,
      "selected_pairs": 26687.5,
      "state_query_norm": 1124.7777465820313,
      "state_relation_loss": 0.0002703908859984949,
      "step": 24800,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011670790263451636,
      "weighted_relation_loss": 0.0011670790263451636
    },
    {
      "action_query_norm": 622.6126098632812,
      "action_relation_loss": 0.000873993750428781,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.208606173994387,
      "flow_matching_loss": 0.01637019133195281,
      "grad_norm": 0.09264910221099854,
      "learning_rate": 7.964704674121149e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011425585602410139,
      "selected_pairs": 26806.3,
      "state_query_norm": 1117.7187133789062,
      "state_relation_loss": 0.0002685648069018498,
      "step": 24810,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011425585602410139,
      "weighted_relation_loss": 0.0011425585602410139
    },
    {
      "action_query_norm": 626.8655456542969,
      "action_relation_loss": 0.0008570405945647508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.21796071094481,
      "flow_matching_loss": 0.014549459517002105,
      "grad_norm": 0.08597612380981445,
      "learning_rate": 7.934885529677505e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011214945698156953,
      "selected_pairs": 26762.7,
      "state_query_norm": 1101.018115234375,
      "state_relation_loss": 0.00026445397816132754,
      "step": 24820,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011214945698156953,
      "weighted_relation_loss": 0.0011214945698156953
    },
    {
      "action_query_norm": 622.2939697265625,
      "action_relation_loss": 0.0008997381431981921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.22731524789523,
      "flow_matching_loss": 0.01770758591592312,
      "grad_norm": 0.08948367834091187,
      "learning_rate": 7.905117498314412e-06,
      "loss": 0.0189,
      "relation_loss": 0.001176971511449665,
      "selected_pairs": 26683.5,
      "state_query_norm": 1109.9159301757813,
      "state_relation_loss": 0.00027723336243070664,
      "step": 24830,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.001176971511449665,
      "weighted_relation_loss": 0.001176971511449665
    },
    {
      "action_query_norm": 624.3901306152344,
      "action_relation_loss": 0.0008909569703973829,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.23666978484565,
      "flow_matching_loss": 0.01579159256070852,
      "grad_norm": 0.09184573590755463,
      "learning_rate": 7.87540061620286e-06,
      "loss": 0.017,
      "relation_loss": 0.0011644416954368353,
      "selected_pairs": 27226.3,
      "state_query_norm": 1115.0359252929688,
      "state_relation_loss": 0.0002734847366809845,
      "step": 24840,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011644416954368353,
      "weighted_relation_loss": 0.0011644416954368353
    },
    {
      "action_query_norm": 624.8112731933594,
      "action_relation_loss": 0.0008864772971719503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.24602432179607,
      "flow_matching_loss": 0.016935106925666334,
      "grad_norm": 0.10093779116868973,
      "learning_rate": 7.845734919451647e-06,
      "loss": 0.0181,
      "relation_loss": 0.0011690645129419862,
      "selected_pairs": 26690.9,
      "state_query_norm": 1115.5759399414062,
      "state_relation_loss": 0.0002825872128596529,
      "step": 24850,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011690645129419862,
      "weighted_relation_loss": 0.0011690645129419862
    },
    {
      "action_query_norm": 625.9025268554688,
      "action_relation_loss": 0.000921163585735485,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.255378858746493,
      "flow_matching_loss": 0.015120192058384418,
      "grad_norm": 0.09060823917388916,
      "learning_rate": 7.816120444107383e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011917942319996655,
      "selected_pairs": 26548.6,
      "state_query_norm": 1113.5000122070312,
      "state_relation_loss": 0.0002706306549953297,
      "step": 24860,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011917942319996655,
      "weighted_relation_loss": 0.0011917942319996655
    },
    {
      "action_query_norm": 623.4068237304688,
      "action_relation_loss": 0.0008934402896557004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.26473339569691,
      "flow_matching_loss": 0.016674684174358845,
      "grad_norm": 0.09501081705093384,
      "learning_rate": 7.786557226154473e-06,
      "loss": 0.0178,
      "relation_loss": 0.0011572587653063237,
      "selected_pairs": 26778.5,
      "state_query_norm": 1117.0080078125,
      "state_relation_loss": 0.00026381846982985734,
      "step": 24870,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011572587653063237,
      "weighted_relation_loss": 0.0011572587653063237
    },
    {
      "action_query_norm": 621.8539916992188,
      "action_relation_loss": 0.0008932030876167118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.274087932647333,
      "flow_matching_loss": 0.017250045947730542,
      "grad_norm": 0.08954363316297531,
      "learning_rate": 7.757045301514998e-06,
      "loss": 0.0184,
      "relation_loss": 0.001169303327333182,
      "selected_pairs": 26249.9,
      "state_query_norm": 1124.8843505859375,
      "state_relation_loss": 0.00027610022807493805,
      "step": 24880,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.001169303327333182,
      "weighted_relation_loss": 0.001169303327333182
    },
    {
      "action_query_norm": 623.699755859375,
      "action_relation_loss": 0.0009048888168763369,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.283442469597755,
      "flow_matching_loss": 0.016634228453040124,
      "grad_norm": 0.08553682267665863,
      "learning_rate": 7.727584706048735e-06,
      "loss": 0.0178,
      "relation_loss": 0.001177481038030237,
      "selected_pairs": 26849.3,
      "state_query_norm": 1113.2489624023438,
      "state_relation_loss": 0.0002725922386161983,
      "step": 24890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001177481038030237,
      "weighted_relation_loss": 0.001177481038030237
    },
    {
      "action_query_norm": 622.1630493164063,
      "action_relation_loss": 0.000895680271787569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.292797006548177,
      "flow_matching_loss": 0.016170684713870287,
      "grad_norm": 0.09893206506967545,
      "learning_rate": 7.698175475553077e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011689974460750818,
      "selected_pairs": 26786.7,
      "state_query_norm": 1104.474755859375,
      "state_relation_loss": 0.00027331715973559766,
      "step": 24900,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011689974460750818,
      "weighted_relation_loss": 0.0011689974460750818
    },
    {
      "action_query_norm": 628.6190124511719,
      "action_relation_loss": 0.0008722266415134072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.302151543498596,
      "flow_matching_loss": 0.017376094404608012,
      "grad_norm": 0.1098930686712265,
      "learning_rate": 7.668817645763021e-06,
      "loss": 0.0185,
      "relation_loss": 0.0011400151648558675,
      "selected_pairs": 27193.4,
      "state_query_norm": 1106.6691650390626,
      "state_relation_loss": 0.00026778853207360954,
      "step": 24910,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011400151648558675,
      "weighted_relation_loss": 0.0011400151648558675
    },
    {
      "action_query_norm": 623.2230224609375,
      "action_relation_loss": 0.0008801308111287653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.311506080449018,
      "flow_matching_loss": 0.01683241855353117,
      "grad_norm": 0.07704431563615799,
      "learning_rate": 7.639511252351088e-06,
      "loss": 0.018,
      "relation_loss": 0.001164922141470015,
      "selected_pairs": 26798.3,
      "state_query_norm": 1107.996337890625,
      "state_relation_loss": 0.0002847913303412497,
      "step": 24920,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001164922141470015,
      "weighted_relation_loss": 0.001164922141470015
    },
    {
      "action_query_norm": 622.4239501953125,
      "action_relation_loss": 0.0009019590506795794,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.32086061739944,
      "flow_matching_loss": 0.017204090021550655,
      "grad_norm": 0.08784591406583786,
      "learning_rate": 7.610256330927324e-06,
      "loss": 0.0184,
      "relation_loss": 0.0011720414622686803,
      "selected_pairs": 26691.1,
      "state_query_norm": 1104.816162109375,
      "state_relation_loss": 0.0002700824057683349,
      "step": 24930,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011720414622686803,
      "weighted_relation_loss": 0.0011720414622686803
    },
    {
      "action_query_norm": 613.2468627929687,
      "action_relation_loss": 0.0009398372552823275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.330215154349858,
      "flow_matching_loss": 0.015261399280279874,
      "grad_norm": 0.07629163563251495,
      "learning_rate": 7.581052917039211e-06,
      "loss": 0.0165,
      "relation_loss": 0.0012145789340138435,
      "selected_pairs": 26368.4,
      "state_query_norm": 1115.1715576171875,
      "state_relation_loss": 0.00027474167582113295,
      "step": 24940,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0012145789340138435,
      "weighted_relation_loss": 0.0012145789340138435
    },
    {
      "action_query_norm": 624.4331420898437,
      "action_relation_loss": 0.0009078319591935724,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.33956969130028,
      "flow_matching_loss": 0.01605643592774868,
      "grad_norm": 0.08533435314893723,
      "learning_rate": 7.551901046171645e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011799504747614265,
      "selected_pairs": 26924.3,
      "state_query_norm": 1107.8114624023438,
      "state_relation_loss": 0.00027211852720938624,
      "step": 24950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011799504747614265,
      "weighted_relation_loss": 0.0011799504747614265
    },
    {
      "action_query_norm": 620.8060424804687,
      "action_relation_loss": 0.0008952884585596621,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.348924228250702,
      "flow_matching_loss": 0.0154572113417089,
      "grad_norm": 0.09006008505821228,
      "learning_rate": 7.522800753746895e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011678338632918894,
      "selected_pairs": 26442.2,
      "state_query_norm": 1101.0295532226562,
      "state_relation_loss": 0.0002725454221945256,
      "step": 24960,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011678338632918894,
      "weighted_relation_loss": 0.0011678338632918894
    },
    {
      "action_query_norm": 624.8933471679687,
      "action_relation_loss": 0.0008983570034615696,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.358278765201124,
      "flow_matching_loss": 0.016336744371801614,
      "grad_norm": 0.06635038554668427,
      "learning_rate": 7.493752075124577e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011646539671346546,
      "selected_pairs": 26739.0,
      "state_query_norm": 1104.8251953125,
      "state_relation_loss": 0.000266296966583468,
      "step": 24970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011646539671346546,
      "weighted_relation_loss": 0.0011646539671346546
    },
    {
      "action_query_norm": 615.0729797363281,
      "action_relation_loss": 0.0008926005859393627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.367633302151543,
      "flow_matching_loss": 0.015571956057101488,
      "grad_norm": 0.06565286964178085,
      "learning_rate": 7.464755045601557e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011621612240560353,
      "selected_pairs": 26484.2,
      "state_query_norm": 1114.3830322265626,
      "state_relation_loss": 0.0002695606439374387,
      "step": 24980,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011621612240560353,
      "weighted_relation_loss": 0.0011621612240560353
    },
    {
      "action_query_norm": 625.1360656738282,
      "action_relation_loss": 0.000860922975698486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.376987839101965,
      "flow_matching_loss": 0.014639336802065372,
      "grad_norm": 0.09769979119300842,
      "learning_rate": 7.435809700411972e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011362401535734534,
      "selected_pairs": 26875.9,
      "state_query_norm": 1106.3670043945312,
      "state_relation_loss": 0.00027531716914381833,
      "step": 24990,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011362401535734534,
      "weighted_relation_loss": 0.0011362401535734534
    },
    {
      "action_query_norm": 623.5777404785156,
      "action_relation_loss": 0.0008642457134556026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.386342376052387,
      "flow_matching_loss": 0.015837033838033678,
      "grad_norm": 0.07728689163923264,
      "learning_rate": 7.40691607472715e-06,
      "loss": 0.017,
      "relation_loss": 0.0011327612679451703,
      "selected_pairs": 27314.7,
      "state_query_norm": 1112.8649047851563,
      "state_relation_loss": 0.0002685155690414831,
      "step": 25000,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0011327612679451703,
      "weighted_relation_loss": 0.0011327612679451703
    },
    {
      "action_query_norm": 611.904052734375,
      "action_relation_loss": 0.0009350408567115664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.395696913002805,
      "flow_matching_loss": 0.015266501810401678,
      "grad_norm": 0.10054911673069,
      "learning_rate": 7.3780742036555606e-06,
      "loss": 0.0165,
      "relation_loss": 0.0011997862369753421,
      "selected_pairs": 27018.6,
      "state_query_norm": 1121.2738891601562,
      "state_relation_loss": 0.0002647453686222434,
      "step": 25010,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011997862369753421,
      "weighted_relation_loss": 0.0011997862369753421
    },
    {
      "action_query_norm": 621.8531799316406,
      "action_relation_loss": 0.0008551071747206152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.405051449953227,
      "flow_matching_loss": 0.0157191950827837,
      "grad_norm": 0.09642587602138519,
      "learning_rate": 7.349284122242783e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011223316309042276,
      "selected_pairs": 27034.2,
      "state_query_norm": 1109.3677856445313,
      "state_relation_loss": 0.00026722446200437843,
      "step": 25020,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011223316309042276,
      "weighted_relation_loss": 0.0011223316309042276
    },
    {
      "action_query_norm": 620.0807861328125,
      "action_relation_loss": 0.0008945267065428198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.41440598690365,
      "flow_matching_loss": 0.014750638883560896,
      "grad_norm": 0.08079182356595993,
      "learning_rate": 7.320545865471512e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011685501202009618,
      "selected_pairs": 26298.4,
      "state_query_norm": 1104.3028442382813,
      "state_relation_loss": 0.00027402340492699295,
      "step": 25030,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0011685501202009618,
      "weighted_relation_loss": 0.0011685501202009618
    },
    {
      "action_query_norm": 617.8511474609375,
      "action_relation_loss": 0.000881581543944776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.42376052385407,
      "flow_matching_loss": 0.01550433961674571,
      "grad_norm": 0.1038389503955841,
      "learning_rate": 7.291859468261425e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011605467996560037,
      "selected_pairs": 27138.2,
      "state_query_norm": 1112.6942260742187,
      "state_relation_loss": 0.00027896526153199376,
      "step": 25040,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011605467996560037,
      "weighted_relation_loss": 0.0011605467996560037
    },
    {
      "action_query_norm": 623.1476196289062,
      "action_relation_loss": 0.0008406601729802788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.43311506080449,
      "flow_matching_loss": 0.015134033001959323,
      "grad_norm": 0.08691996335983276,
      "learning_rate": 7.263224965469195e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011170719168148934,
      "selected_pairs": 26711.3,
      "state_query_norm": 1111.0599853515625,
      "state_relation_loss": 0.00027641174383461473,
      "step": 25050,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011170719168148934,
      "weighted_relation_loss": 0.0011170719168148934
    },
    {
      "action_query_norm": 621.420556640625,
      "action_relation_loss": 0.0008696626464370638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.44246959775491,
      "flow_matching_loss": 0.015890552289783956,
      "grad_norm": 0.10421432554721832,
      "learning_rate": 7.234642391888446e-06,
      "loss": 0.017,
      "relation_loss": 0.0011343207326717675,
      "selected_pairs": 26623.0,
      "state_query_norm": 1112.3521850585937,
      "state_relation_loss": 0.00026465807459317146,
      "step": 25060,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011343207326717675,
      "weighted_relation_loss": 0.0011343207326717675
    },
    {
      "action_query_norm": 622.0316223144531,
      "action_relation_loss": 0.0009205716487485915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.451824134705333,
      "flow_matching_loss": 0.016496299393475057,
      "grad_norm": 0.08681535720825195,
      "learning_rate": 7.206111782249697e-06,
      "loss": 0.0177,
      "relation_loss": 0.0011924345861189067,
      "selected_pairs": 26804.4,
      "state_query_norm": 1119.8964965820312,
      "state_relation_loss": 0.0002718629373703152,
      "step": 25070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011924345861189067,
      "weighted_relation_loss": 0.0011924345861189067
    },
    {
      "action_query_norm": 625.3300048828125,
      "action_relation_loss": 0.0009064858546480537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.46117867165575,
      "flow_matching_loss": 0.016126951109617947,
      "grad_norm": 0.08096823841333389,
      "learning_rate": 7.1776331712203385e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011621613055467605,
      "selected_pairs": 26987.8,
      "state_query_norm": 1116.3774658203124,
      "state_relation_loss": 0.0002556754392571747,
      "step": 25080,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011621613055467605,
      "weighted_relation_loss": 0.0011621613055467605
    },
    {
      "action_query_norm": 626.3516296386719,
      "action_relation_loss": 0.0008740404387935996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.470533208606174,
      "flow_matching_loss": 0.01526741161942482,
      "grad_norm": 0.0766322910785675,
      "learning_rate": 7.149206593404561e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011382909025996924,
      "selected_pairs": 26871.2,
      "state_query_norm": 1118.52421875,
      "state_relation_loss": 0.00026425046380609276,
      "step": 25090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011382909025996924,
      "weighted_relation_loss": 0.0011382909025996924
    },
    {
      "action_query_norm": 621.6067626953125,
      "action_relation_loss": 0.000875162280863151,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.479887745556596,
      "flow_matching_loss": 0.015407169330865144,
      "grad_norm": 0.1032000184059143,
      "learning_rate": 7.120832083343337e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011443803552538157,
      "selected_pairs": 26639.2,
      "state_query_norm": 1104.4329956054687,
      "state_relation_loss": 0.0002692180656595156,
      "step": 25100,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0011443803552538157,
      "weighted_relation_loss": 0.0011443803552538157
    },
    {
      "action_query_norm": 625.0677795410156,
      "action_relation_loss": 0.0009084565099328756,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.489242282507014,
      "flow_matching_loss": 0.014222010876983405,
      "grad_norm": 0.08950918912887573,
      "learning_rate": 7.092509675514369e-06,
      "loss": 0.0154,
      "relation_loss": 0.0011695460067130624,
      "selected_pairs": 27055.6,
      "state_query_norm": 1118.3046264648438,
      "state_relation_loss": 0.0002610894895042293,
      "step": 25110,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011695460067130624,
      "weighted_relation_loss": 0.0011695460067130624
    },
    {
      "action_query_norm": 626.2313842773438,
      "action_relation_loss": 0.0009308646665886044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.498596819457436,
      "flow_matching_loss": 0.014987792633473874,
      "grad_norm": 0.08145830780267715,
      "learning_rate": 7.064239404332063e-06,
      "loss": 0.0162,
      "relation_loss": 0.0012142737861722708,
      "selected_pairs": 27012.7,
      "state_query_norm": 1101.17568359375,
      "state_relation_loss": 0.0002834091224940494,
      "step": 25120,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012142737861722708,
      "weighted_relation_loss": 0.0012142737861722708
    },
    {
      "action_query_norm": 625.490380859375,
      "action_relation_loss": 0.0008952087198849767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.507951356407858,
      "flow_matching_loss": 0.01670911945402622,
      "grad_norm": 0.07722915709018707,
      "learning_rate": 7.03602130414745e-06,
      "loss": 0.0179,
      "relation_loss": 0.0011621203972026705,
      "selected_pairs": 27225.8,
      "state_query_norm": 1100.9628173828125,
      "state_relation_loss": 0.0002669116598553956,
      "step": 25130,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011621203972026705,
      "weighted_relation_loss": 0.0011621203972026705
    },
    {
      "action_query_norm": 623.2728393554687,
      "action_relation_loss": 0.0008821440453175456,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.51730589335828,
      "flow_matching_loss": 0.01658750269562006,
      "grad_norm": 0.0948646292090416,
      "learning_rate": 7.007855409248198e-06,
      "loss": 0.0177,
      "relation_loss": 0.0011607949738390744,
      "selected_pairs": 27010.0,
      "state_query_norm": 1115.1197631835937,
      "state_relation_loss": 0.0002786509314319119,
      "step": 25140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011607949738390744,
      "weighted_relation_loss": 0.0011607949738390744
    },
    {
      "action_query_norm": 629.5527038574219,
      "action_relation_loss": 0.0008835504704620689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.5266604303087,
      "flow_matching_loss": 0.015124053973704577,
      "grad_norm": 0.08605680614709854,
      "learning_rate": 6.979741753858521e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011564122163690627,
      "selected_pairs": 27117.5,
      "state_query_norm": 1117.4221801757812,
      "state_relation_loss": 0.0002728617429966107,
      "step": 25150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011564122163690627,
      "weighted_relation_loss": 0.0011564122163690627
    },
    {
      "action_query_norm": 623.599560546875,
      "action_relation_loss": 0.0009086650039535016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.53601496725912,
      "flow_matching_loss": 0.01640971899032593,
      "grad_norm": 0.06748021394014359,
      "learning_rate": 6.951680372139158e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011845816625282168,
      "selected_pairs": 26711.1,
      "state_query_norm": 1116.4753662109374,
      "state_relation_loss": 0.00027591667603701355,
      "step": 25160,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011845816625282168,
      "weighted_relation_loss": 0.0011845816625282168
    },
    {
      "action_query_norm": 620.0955688476563,
      "action_relation_loss": 0.0008682588580995798,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.545369504209543,
      "flow_matching_loss": 0.016179758217185737,
      "grad_norm": 0.08065568655729294,
      "learning_rate": 6.923671298187334e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011393802938982843,
      "selected_pairs": 26988.1,
      "state_query_norm": 1125.0912475585938,
      "state_relation_loss": 0.0002711214357987046,
      "step": 25170,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011393802938982843,
      "weighted_relation_loss": 0.0011393802938982843
    },
    {
      "action_query_norm": 626.0973449707031,
      "action_relation_loss": 0.0009103758435230702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.55472404115996,
      "flow_matching_loss": 0.014812887087464333,
      "grad_norm": 0.07443507760763168,
      "learning_rate": 6.895714566036704e-06,
      "loss": 0.016,
      "relation_loss": 0.0011791065451689064,
      "selected_pairs": 26915.5,
      "state_query_norm": 1117.3664428710938,
      "state_relation_loss": 0.00026873070746660235,
      "step": 25180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011791065451689064,
      "weighted_relation_loss": 0.0011791065451689064
    },
    {
      "action_query_norm": 628.2592956542969,
      "action_relation_loss": 0.000876629602862522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.564078578110383,
      "flow_matching_loss": 0.014729328826069832,
      "grad_norm": 0.09597590565681458,
      "learning_rate": 6.86781020965736e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011589539004489779,
      "selected_pairs": 27198.2,
      "state_query_norm": 1110.4352294921875,
      "state_relation_loss": 0.000282324303407222,
      "step": 25190,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011589539004489779,
      "weighted_relation_loss": 0.0011589539004489779
    },
    {
      "action_query_norm": 625.6401611328125,
      "action_relation_loss": 0.0008811990381218493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.573433115060805,
      "flow_matching_loss": 0.0155077263712883,
      "grad_norm": 0.06686340272426605,
      "learning_rate": 6.839958262955709e-06,
      "loss": 0.0167,
      "relation_loss": 0.001149133255239576,
      "selected_pairs": 26762.5,
      "state_query_norm": 1107.8580322265625,
      "state_relation_loss": 0.0002679342229384929,
      "step": 25200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001149133255239576,
      "weighted_relation_loss": 0.001149133255239576
    },
    {
      "action_query_norm": 627.0356750488281,
      "action_relation_loss": 0.0008749322209041565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.582787652011227,
      "flow_matching_loss": 0.015142569690942765,
      "grad_norm": 0.0871729701757431,
      "learning_rate": 6.8121587597744894e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011398756527341903,
      "selected_pairs": 27359.0,
      "state_query_norm": 1109.0908935546875,
      "state_relation_loss": 0.0002649434289196506,
      "step": 25210,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0011398756527341903,
      "weighted_relation_loss": 0.0011398756527341903
    },
    {
      "action_query_norm": 625.3163269042968,
      "action_relation_loss": 0.0008653072989545762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.592142188961645,
      "flow_matching_loss": 0.015879018418490886,
      "grad_norm": 0.09959963709115982,
      "learning_rate": 6.784411733892732e-06,
      "loss": 0.017,
      "relation_loss": 0.0011322719394229353,
      "selected_pairs": 26344.1,
      "state_query_norm": 1107.7828735351563,
      "state_relation_loss": 0.00026696463755797594,
      "step": 25220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0011322719394229353,
      "weighted_relation_loss": 0.0011322719394229353
    },
    {
      "action_query_norm": 623.3122863769531,
      "action_relation_loss": 0.0009098852518945932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.601496725912067,
      "flow_matching_loss": 0.014474429283291101,
      "grad_norm": 0.11051998287439346,
      "learning_rate": 6.756717219025666e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011798126972280443,
      "selected_pairs": 27274.5,
      "state_query_norm": 1116.942138671875,
      "state_relation_loss": 0.0002699274366023019,
      "step": 25230,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0011798126972280443,
      "weighted_relation_loss": 0.0011798126972280443
    },
    {
      "action_query_norm": 621.7645446777344,
      "action_relation_loss": 0.0008949466107878834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.61085126286249,
      "flow_matching_loss": 0.015579223819077014,
      "grad_norm": 0.09528063237667084,
      "learning_rate": 6.729075248824762e-06,
      "loss": 0.0167,
      "relation_loss": 0.001157366146799177,
      "selected_pairs": 27363.7,
      "state_query_norm": 1103.9406982421874,
      "state_relation_loss": 0.00026241955056320875,
      "step": 25240,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001157366146799177,
      "weighted_relation_loss": 0.001157366146799177
    },
    {
      "action_query_norm": 623.3576721191406,
      "action_relation_loss": 0.0008951083698775619,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.620205799812908,
      "flow_matching_loss": 0.015433019585907459,
      "grad_norm": 0.06888675689697266,
      "learning_rate": 6.701485856877615e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011662979726679624,
      "selected_pairs": 27063.2,
      "state_query_norm": 1112.2171630859375,
      "state_relation_loss": 0.00027118959842482583,
      "step": 25250,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011662979726679624,
      "weighted_relation_loss": 0.0011662979726679624
    },
    {
      "action_query_norm": 621.7309814453125,
      "action_relation_loss": 0.0008516157162375748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.62956033676333,
      "flow_matching_loss": 0.016941537987440823,
      "grad_norm": 0.07440708577632904,
      "learning_rate": 6.673949076707925e-06,
      "loss": 0.0181,
      "relation_loss": 0.0011222573346458375,
      "selected_pairs": 26298.0,
      "state_query_norm": 1112.2719482421876,
      "state_relation_loss": 0.0002706416038563475,
      "step": 25260,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0011222573346458375,
      "weighted_relation_loss": 0.0011222573346458375
    },
    {
      "action_query_norm": 624.70966796875,
      "action_relation_loss": 0.0008652792195789516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.63891487371375,
      "flow_matching_loss": 0.01598390582948923,
      "grad_norm": 0.1034383773803711,
      "learning_rate": 6.646464941775499e-06,
      "loss": 0.0171,
      "relation_loss": 0.001125652075279504,
      "selected_pairs": 26523.9,
      "state_query_norm": 1113.7404174804688,
      "state_relation_loss": 0.00026037284842459483,
      "step": 25270,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001125652075279504,
      "weighted_relation_loss": 0.001125652075279504
    },
    {
      "action_query_norm": 621.4405151367188,
      "action_relation_loss": 0.0008803497534245252,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.648269410664174,
      "flow_matching_loss": 0.014162457920610905,
      "grad_norm": 0.09273792803287506,
      "learning_rate": 6.619033485476128e-06,
      "loss": 0.0153,
      "relation_loss": 0.0011474392027594148,
      "selected_pairs": 26828.6,
      "state_query_norm": 1123.0855590820313,
      "state_relation_loss": 0.0002670894464245066,
      "step": 25280,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011474392027594148,
      "weighted_relation_loss": 0.0011474392027594148
    },
    {
      "action_query_norm": 622.4466247558594,
      "action_relation_loss": 0.0009094830777030439,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.657623947614592,
      "flow_matching_loss": 0.015824285987764596,
      "grad_norm": 0.09549583494663239,
      "learning_rate": 6.59165474114164e-06,
      "loss": 0.017,
      "relation_loss": 0.0011832995456643402,
      "selected_pairs": 27009.3,
      "state_query_norm": 1107.1942504882813,
      "state_relation_loss": 0.0002738164679612964,
      "step": 25290,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011832995456643402,
      "weighted_relation_loss": 0.0011832995456643402
    },
    {
      "action_query_norm": 625.7966369628906,
      "action_relation_loss": 0.0008520736650098115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.666978484565014,
      "flow_matching_loss": 0.015313263703137635,
      "grad_norm": 0.09034368395805359,
      "learning_rate": 6.564328742039783e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011187030584551394,
      "selected_pairs": 26507.6,
      "state_query_norm": 1110.3535034179688,
      "state_relation_loss": 0.0002666293905349448,
      "step": 25300,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011187030584551394,
      "weighted_relation_loss": 0.0011187030584551394
    },
    {
      "action_query_norm": 620.9349792480468,
      "action_relation_loss": 0.0009166446805465966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.676333021515436,
      "flow_matching_loss": 0.015819916129112245,
      "grad_norm": 0.10719047486782074,
      "learning_rate": 6.537055521374219e-06,
      "loss": 0.017,
      "relation_loss": 0.0011956270784139634,
      "selected_pairs": 26662.2,
      "state_query_norm": 1116.21845703125,
      "state_relation_loss": 0.0002789823993225582,
      "step": 25310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011956270784139634,
      "weighted_relation_loss": 0.0011956270784139634
    },
    {
      "action_query_norm": 617.7419860839843,
      "action_relation_loss": 0.0008874505525454879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.685687558465855,
      "flow_matching_loss": 0.01663619801402092,
      "grad_norm": 0.08817963302135468,
      "learning_rate": 6.509835112284485e-06,
      "loss": 0.0178,
      "relation_loss": 0.00115306347142905,
      "selected_pairs": 26413.3,
      "state_query_norm": 1112.2361938476563,
      "state_relation_loss": 0.00026561292470432817,
      "step": 25320,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.00115306347142905,
      "weighted_relation_loss": 0.00115306347142905
    },
    {
      "action_query_norm": 617.5570861816407,
      "action_relation_loss": 0.0009173218626528978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.695042095416277,
      "flow_matching_loss": 0.015497546270489692,
      "grad_norm": 0.07179681956768036,
      "learning_rate": 6.482667547845944e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011857040808536112,
      "selected_pairs": 26928.2,
      "state_query_norm": 1120.3903198242188,
      "state_relation_loss": 0.00026838222984224556,
      "step": 25330,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011857040808536112,
      "weighted_relation_loss": 0.0011857040808536112
    },
    {
      "action_query_norm": 614.6661071777344,
      "action_relation_loss": 0.0008625028247479349,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.7043966323667,
      "flow_matching_loss": 0.015028067119419574,
      "grad_norm": 0.09546925872564316,
      "learning_rate": 6.455552861069736e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011294572497718037,
      "selected_pairs": 26608.2,
      "state_query_norm": 1107.8740966796875,
      "state_relation_loss": 0.0002669544395757839,
      "step": 25340,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011294572497718037,
      "weighted_relation_loss": 0.0011294572497718037
    },
    {
      "action_query_norm": 623.348046875,
      "action_relation_loss": 0.0008644019195344299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.71375116931712,
      "flow_matching_loss": 0.0161731893196702,
      "grad_norm": 0.09311506152153015,
      "learning_rate": 6.428491084902788e-06,
      "loss": 0.0173,
      "relation_loss": 0.001130060653667897,
      "selected_pairs": 26758.3,
      "state_query_norm": 1111.2921630859375,
      "state_relation_loss": 0.000265658728312701,
      "step": 25350,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001130060653667897,
      "weighted_relation_loss": 0.001130060653667897
    },
    {
      "action_query_norm": 621.1876220703125,
      "action_relation_loss": 0.0008948668255470693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.72310570626754,
      "flow_matching_loss": 0.0148196030408144,
      "grad_norm": 0.07263768464326859,
      "learning_rate": 6.401482252227698e-06,
      "loss": 0.016,
      "relation_loss": 0.0011584802530705928,
      "selected_pairs": 27056.5,
      "state_query_norm": 1115.3486450195312,
      "state_relation_loss": 0.0002636134275235236,
      "step": 25360,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011584802530705928,
      "weighted_relation_loss": 0.0011584802530705928
    },
    {
      "action_query_norm": 626.8890625,
      "action_relation_loss": 0.0008933091827202589,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.73246024321796,
      "flow_matching_loss": 0.014155689161270857,
      "grad_norm": 0.09050557017326355,
      "learning_rate": 6.37452639586274e-06,
      "loss": 0.0153,
      "relation_loss": 0.00116718087811023,
      "selected_pairs": 26917.3,
      "state_query_norm": 1113.5423583984375,
      "state_relation_loss": 0.0002738716983003542,
      "step": 25370,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00116718087811023,
      "weighted_relation_loss": 0.00116718087811023
    },
    {
      "action_query_norm": 627.1521423339843,
      "action_relation_loss": 0.0009068886924069375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.741814780168383,
      "flow_matching_loss": 0.01514038685709238,
      "grad_norm": 0.09976021200418472,
      "learning_rate": 6.347623548561826e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011757136322557925,
      "selected_pairs": 26919.8,
      "state_query_norm": 1101.6709106445312,
      "state_relation_loss": 0.0002688249631319195,
      "step": 25380,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011757136322557925,
      "weighted_relation_loss": 0.0011757136322557925
    },
    {
      "action_query_norm": 624.4281066894531,
      "action_relation_loss": 0.0008465120219625533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.7511693171188,
      "flow_matching_loss": 0.0155521254055202,
      "grad_norm": 0.06961560994386673,
      "learning_rate": 6.320773743014441e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011141716851852835,
      "selected_pairs": 26946.2,
      "state_query_norm": 1114.3341918945312,
      "state_relation_loss": 0.00026765964285004886,
      "step": 25390,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011141716851852835,
      "weighted_relation_loss": 0.0011141716851852835
    },
    {
      "action_query_norm": 626.0819885253907,
      "action_relation_loss": 0.0008680225175339729,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.760523854069223,
      "flow_matching_loss": 0.015768569335341452,
      "grad_norm": 0.08221238851547241,
      "learning_rate": 6.293977011845648e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011424531461670996,
      "selected_pairs": 26923.0,
      "state_query_norm": 1095.5997436523437,
      "state_relation_loss": 0.00027443061117082834,
      "step": 25400,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011424531461670996,
      "weighted_relation_loss": 0.0011424531461670996
    },
    {
      "action_query_norm": 618.2214904785156,
      "action_relation_loss": 0.0008618466497864574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.769878391019645,
      "flow_matching_loss": 0.014575906749814748,
      "grad_norm": 0.08983170241117477,
      "learning_rate": 6.2672333876159835e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011339376913383603,
      "selected_pairs": 26533.1,
      "state_query_norm": 1107.36904296875,
      "state_relation_loss": 0.00027209105901420115,
      "step": 25410,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0011339376913383603,
      "weighted_relation_loss": 0.0011339376913383603
    },
    {
      "action_query_norm": 620.7204833984375,
      "action_relation_loss": 0.0009084073943085968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.779232927970064,
      "flow_matching_loss": 0.016353472135961056,
      "grad_norm": 0.07917851954698563,
      "learning_rate": 6.2405429028215e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011828026152215898,
      "selected_pairs": 27150.1,
      "state_query_norm": 1118.6772827148438,
      "state_relation_loss": 0.0002743952296441421,
      "step": 25420,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011828026152215898,
      "weighted_relation_loss": 0.0011828026152215898
    },
    {
      "action_query_norm": 619.1393859863281,
      "action_relation_loss": 0.0008658882637973875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.788587464920486,
      "flow_matching_loss": 0.01619669981300831,
      "grad_norm": 0.09255403280258179,
      "learning_rate": 6.213905589893631e-06,
      "loss": 0.0173,
      "relation_loss": 0.0011331311543472111,
      "selected_pairs": 26458.2,
      "state_query_norm": 1110.9893676757813,
      "state_relation_loss": 0.00026724287599790844,
      "step": 25430,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011331311543472111,
      "weighted_relation_loss": 0.0011331311543472111
    },
    {
      "action_query_norm": 629.301953125,
      "action_relation_loss": 0.0008620527980383486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.797942001870908,
      "flow_matching_loss": 0.014901026152074337,
      "grad_norm": 0.09629639983177185,
      "learning_rate": 6.187321481199221e-06,
      "loss": 0.016,
      "relation_loss": 0.0011362065677531063,
      "selected_pairs": 26967.1,
      "state_query_norm": 1109.9047485351562,
      "state_relation_loss": 0.0002741537697147578,
      "step": 25440,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011362065677531063,
      "weighted_relation_loss": 0.0011362065677531063
    },
    {
      "action_query_norm": 624.9529602050782,
      "action_relation_loss": 0.0008458147058263421,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.80729653882133,
      "flow_matching_loss": 0.017085317987948656,
      "grad_norm": 0.0893702283501625,
      "learning_rate": 6.1607906090404895e-06,
      "loss": 0.0182,
      "relation_loss": 0.0011053007561713456,
      "selected_pairs": 26824.3,
      "state_query_norm": 1119.9120727539062,
      "state_relation_loss": 0.0002594860605313443,
      "step": 25450,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011053007561713456,
      "weighted_relation_loss": 0.0011053007561713456
    },
    {
      "action_query_norm": 622.4546020507812,
      "action_relation_loss": 0.000860201136674732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.81665107577175,
      "flow_matching_loss": 0.016870068944990636,
      "grad_norm": 0.08211370557546616,
      "learning_rate": 6.134313005654929e-06,
      "loss": 0.018,
      "relation_loss": 0.0011275962577201426,
      "selected_pairs": 27002.4,
      "state_query_norm": 1128.2145263671875,
      "state_relation_loss": 0.0002673951239557937,
      "step": 25460,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011275962577201426,
      "weighted_relation_loss": 0.0011275962577201426
    },
    {
      "action_query_norm": 618.9166442871094,
      "action_relation_loss": 0.0009118433110415936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.82600561272217,
      "flow_matching_loss": 0.016408311296254398,
      "grad_norm": 0.0813300758600235,
      "learning_rate": 6.107888703215337e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011801538523286581,
      "selected_pairs": 26602.2,
      "state_query_norm": 1100.0952270507812,
      "state_relation_loss": 0.00026831054128706455,
      "step": 25470,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011801538523286581,
      "weighted_relation_loss": 0.0011801538523286581
    },
    {
      "action_query_norm": 620.5409973144531,
      "action_relation_loss": 0.0009451648045796901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.835360149672592,
      "flow_matching_loss": 0.015278113540261984,
      "grad_norm": 0.09552920609712601,
      "learning_rate": 6.081517733829722e-06,
      "loss": 0.0165,
      "relation_loss": 0.001214339048601687,
      "selected_pairs": 26549.8,
      "state_query_norm": 1119.860986328125,
      "state_relation_loss": 0.00026917424402199684,
      "step": 25480,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001214339048601687,
      "weighted_relation_loss": 0.001214339048601687
    },
    {
      "action_query_norm": 625.466943359375,
      "action_relation_loss": 0.0008685397624503821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.84471468662301,
      "flow_matching_loss": 0.015523834619671106,
      "grad_norm": 0.08655230700969696,
      "learning_rate": 6.055200129541294e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011415815912187099,
      "selected_pairs": 27551.4,
      "state_query_norm": 1112.3355346679687,
      "state_relation_loss": 0.00027304182876832785,
      "step": 25490,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0011415815912187099,
      "weighted_relation_loss": 0.0011415815912187099
    },
    {
      "action_query_norm": 617.7454162597656,
      "action_relation_loss": 0.0009175748564302921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.854069223573433,
      "flow_matching_loss": 0.015207334235310555,
      "grad_norm": 0.07826322317123413,
      "learning_rate": 6.0289359223284434e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011800609878264367,
      "selected_pairs": 26616.7,
      "state_query_norm": 1091.795703125,
      "state_relation_loss": 0.0002624861226649955,
      "step": 25500,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011800609878264367,
      "weighted_relation_loss": 0.0011800609878264367
    },
    {
      "action_query_norm": 626.6554321289062,
      "action_relation_loss": 0.000893063162220642,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.863423760523855,
      "flow_matching_loss": 0.015430455096065998,
      "grad_norm": 0.0918932780623436,
      "learning_rate": 6.002725144104648e-06,
      "loss": 0.0166,
      "relation_loss": 0.001163787639234215,
      "selected_pairs": 27244.8,
      "state_query_norm": 1124.7038940429688,
      "state_relation_loss": 0.0002707244653720409,
      "step": 25510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001163787639234215,
      "weighted_relation_loss": 0.001163787639234215
    },
    {
      "action_query_norm": 619.8736755371094,
      "action_relation_loss": 0.0008827470708638429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.872778297474277,
      "flow_matching_loss": 0.01608160138130188,
      "grad_norm": 0.13320516049861908,
      "learning_rate": 5.976567826718477e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011582195293158293,
      "selected_pairs": 26802.6,
      "state_query_norm": 1109.956787109375,
      "state_relation_loss": 0.0002754724642727524,
      "step": 25520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011582195293158293,
      "weighted_relation_loss": 0.0011582195293158293
    },
    {
      "action_query_norm": 626.2157958984375,
      "action_relation_loss": 0.000870306056458503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.882132834424695,
      "flow_matching_loss": 0.016566792223602533,
      "grad_norm": 0.08979816734790802,
      "learning_rate": 5.9504640019535315e-06,
      "loss": 0.0177,
      "relation_loss": 0.001151061593554914,
      "selected_pairs": 27007.7,
      "state_query_norm": 1100.7258666992188,
      "state_relation_loss": 0.000280755537096411,
      "step": 25530,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001151061593554914,
      "weighted_relation_loss": 0.001151061593554914
    },
    {
      "action_query_norm": 625.1548889160156,
      "action_relation_loss": 0.0009021017991472036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.891487371375117,
      "flow_matching_loss": 0.015491019561886787,
      "grad_norm": 0.135057270526886,
      "learning_rate": 5.924413701528414e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011716171400621533,
      "selected_pairs": 26937.4,
      "state_query_norm": 1113.0418823242187,
      "state_relation_loss": 0.00026951533218380066,
      "step": 25540,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011716171400621533,
      "weighted_relation_loss": 0.0011716171400621533
    },
    {
      "action_query_norm": 625.1606506347656,
      "action_relation_loss": 0.0008475000446196646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.90084190832554,
      "flow_matching_loss": 0.016476218495517968,
      "grad_norm": 0.0709003135561943,
      "learning_rate": 5.898416957096703e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011155587388202548,
      "selected_pairs": 26711.8,
      "state_query_norm": 1125.5688110351562,
      "state_relation_loss": 0.0002680586912902072,
      "step": 25550,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011155587388202548,
      "weighted_relation_loss": 0.0011155587388202548
    },
    {
      "action_query_norm": 623.5550476074219,
      "action_relation_loss": 0.0008980625949334353,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.910196445275957,
      "flow_matching_loss": 0.014205993432551623,
      "grad_norm": 0.07629529386758804,
      "learning_rate": 5.872473800246913e-06,
      "loss": 0.0154,
      "relation_loss": 0.0011690171319060027,
      "selected_pairs": 27112.2,
      "state_query_norm": 1094.0620239257812,
      "state_relation_loss": 0.0002709545398829505,
      "step": 25560,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011690171319060027,
      "weighted_relation_loss": 0.0011690171319060027
    },
    {
      "action_query_norm": 616.3423706054688,
      "action_relation_loss": 0.0009129572485107928,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.91955098222638,
      "flow_matching_loss": 0.013821832556277513,
      "grad_norm": 0.08470442146062851,
      "learning_rate": 5.846584262502402e-06,
      "loss": 0.015,
      "relation_loss": 0.001183503319043666,
      "selected_pairs": 26278.7,
      "state_query_norm": 1120.7106323242188,
      "state_relation_loss": 0.0002705460763536394,
      "step": 25570,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.001183503319043666,
      "weighted_relation_loss": 0.001183503319043666
    },
    {
      "action_query_norm": 623.2944091796875,
      "action_relation_loss": 0.0008801064454019069,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.9289055191768,
      "flow_matching_loss": 0.014856788981705903,
      "grad_norm": 0.11511258035898209,
      "learning_rate": 5.820748375321411e-06,
      "loss": 0.016,
      "relation_loss": 0.0011575584416277707,
      "selected_pairs": 27037.7,
      "state_query_norm": 1118.6030639648438,
      "state_relation_loss": 0.0002774520035018213,
      "step": 25580,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011575584416277707,
      "weighted_relation_loss": 0.0011575584416277707
    },
    {
      "action_query_norm": 621.5482482910156,
      "action_relation_loss": 0.000906092586228624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.938260056127223,
      "flow_matching_loss": 0.014850137289613485,
      "grad_norm": 0.09431585669517517,
      "learning_rate": 5.794966170096977e-06,
      "loss": 0.016,
      "relation_loss": 0.001175215095281601,
      "selected_pairs": 27087.0,
      "state_query_norm": 1114.7559692382813,
      "state_relation_loss": 0.00026912251341855155,
      "step": 25590,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001175215095281601,
      "weighted_relation_loss": 0.001175215095281601
    },
    {
      "action_query_norm": 628.2786010742187,
      "action_relation_loss": 0.0008550823375117033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.947614593077642,
      "flow_matching_loss": 0.015461244247853756,
      "grad_norm": 0.07354343682527542,
      "learning_rate": 5.769237678156897e-06,
      "loss": 0.0166,
      "relation_loss": 0.001115980069153011,
      "selected_pairs": 27060.3,
      "state_query_norm": 1114.297802734375,
      "state_relation_loss": 0.0002608977403724566,
      "step": 25600,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001115980069153011,
      "weighted_relation_loss": 0.001115980069153011
    },
    {
      "action_query_norm": 628.1290710449218,
      "action_relation_loss": 0.0008662173349875957,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.956969130028064,
      "flow_matching_loss": 0.014632635563611985,
      "grad_norm": 0.07605783641338348,
      "learning_rate": 5.743562930763735e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011310358881019057,
      "selected_pairs": 27312.3,
      "state_query_norm": 1106.6083984375,
      "state_relation_loss": 0.00026481854147277775,
      "step": 25610,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0011310358881019057,
      "weighted_relation_loss": 0.0011310358881019057
    },
    {
      "action_query_norm": 622.7638122558594,
      "action_relation_loss": 0.000862963154213503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.966323666978486,
      "flow_matching_loss": 0.01556759374216199,
      "grad_norm": 0.07595288753509521,
      "learning_rate": 5.717941959114726e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011239319923333825,
      "selected_pairs": 26976.4,
      "state_query_norm": 1116.12353515625,
      "state_relation_loss": 0.00026096885412698614,
      "step": 25620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011239319923333825,
      "weighted_relation_loss": 0.0011239319923333825
    },
    {
      "action_query_norm": 621.8678955078125,
      "action_relation_loss": 0.0008649563125800342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.975678203928904,
      "flow_matching_loss": 0.015531861502677203,
      "grad_norm": 0.08735304325819016,
      "learning_rate": 5.69237479434176e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011361142154783011,
      "selected_pairs": 26990.2,
      "state_query_norm": 1112.3678588867188,
      "state_relation_loss": 0.00027115789271192625,
      "step": 25630,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011361142154783011,
      "weighted_relation_loss": 0.0011361142154783011
    },
    {
      "action_query_norm": 628.2569030761719,
      "action_relation_loss": 0.0008723639999516309,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.985032740879326,
      "flow_matching_loss": 0.01809549517929554,
      "grad_norm": 0.09465105831623077,
      "learning_rate": 5.666861467511353e-06,
      "loss": 0.0192,
      "relation_loss": 0.0011346943327225744,
      "selected_pairs": 27268.8,
      "state_query_norm": 1129.3988403320313,
      "state_relation_loss": 0.00026233032112941144,
      "step": 25640,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0011346943327225744,
      "weighted_relation_loss": 0.0011346943327225744
    },
    {
      "action_query_norm": 623.5359252929687,
      "action_relation_loss": 0.000880606152350083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.99438727782975,
      "flow_matching_loss": 0.01664246479049325,
      "grad_norm": 0.11086855083703995,
      "learning_rate": 5.641402009624591e-06,
      "loss": 0.0178,
      "relation_loss": 0.0011487090960144997,
      "selected_pairs": 26721.6,
      "state_query_norm": 1100.5597412109375,
      "state_relation_loss": 0.00026810292911250144,
      "step": 25650,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011487090960144997,
      "weighted_relation_loss": 0.0011487090960144997
    },
    {
      "action_query_norm": 623.5894714355469,
      "action_relation_loss": 0.0008833632746245712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.003741814780167,
      "flow_matching_loss": 0.01671924041584134,
      "grad_norm": 0.0959775522351265,
      "learning_rate": 5.615996451617145e-06,
      "loss": 0.0179,
      "relation_loss": 0.0011478943866677581,
      "selected_pairs": 24154.5,
      "state_query_norm": 1108.0922485351562,
      "state_relation_loss": 0.0002645311105879955,
      "step": 25660,
      "valid_action_samples": 213.5,
      "weighted_auxiliary_loss": 0.0011478943866677581,
      "weighted_relation_loss": 0.0011478943866677581
    },
    {
      "action_query_norm": 626.7326782226562,
      "action_relation_loss": 0.0008342420449480414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.01309635173059,
      "flow_matching_loss": 0.016469197068363427,
      "grad_norm": 0.1038411483168602,
      "learning_rate": 5.590644824359148e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010995766439009457,
      "selected_pairs": 26990.7,
      "state_query_norm": 1107.5184814453125,
      "state_relation_loss": 0.0002653345858561806,
      "step": 25670,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010995766439009457,
      "weighted_relation_loss": 0.0010995766439009457
    },
    {
      "action_query_norm": 618.7317626953125,
      "action_relation_loss": 0.0009017241885885596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.02245088868101,
      "flow_matching_loss": 0.015209415834397078,
      "grad_norm": 0.08150086551904678,
      "learning_rate": 5.56534715865521e-06,
      "loss": 0.0164,
      "relation_loss": 0.001173088524956256,
      "selected_pairs": 26950.9,
      "state_query_norm": 1094.729638671875,
      "state_relation_loss": 0.00027136434509884567,
      "step": 25680,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001173088524956256,
      "weighted_relation_loss": 0.001173088524956256
    },
    {
      "action_query_norm": 621.536083984375,
      "action_relation_loss": 0.0008867619209922851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.031805425631433,
      "flow_matching_loss": 0.015732535254210235,
      "grad_norm": 0.0854961946606636,
      "learning_rate": 5.5401034852443965e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011591302463784814,
      "selected_pairs": 26873.9,
      "state_query_norm": 1108.5628173828125,
      "state_relation_loss": 0.00027236833120696246,
      "step": 25690,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011591302463784814,
      "weighted_relation_loss": 0.0011591302463784814
    },
    {
      "action_query_norm": 625.4586669921875,
      "action_relation_loss": 0.0008648576156701892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.04115996258185,
      "flow_matching_loss": 0.01496082916855812,
      "grad_norm": 0.08245410025119781,
      "learning_rate": 5.514913834800134e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011378052178770304,
      "selected_pairs": 26285.0,
      "state_query_norm": 1113.702294921875,
      "state_relation_loss": 0.0002729476022068411,
      "step": 25700,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0011378052178770304,
      "weighted_relation_loss": 0.0011378052178770304
    },
    {
      "action_query_norm": 628.3027465820312,
      "action_relation_loss": 0.000866570498328656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.050514499532273,
      "flow_matching_loss": 0.0167111050337553,
      "grad_norm": 0.0897325649857521,
      "learning_rate": 5.489778237930238e-06,
      "loss": 0.0178,
      "relation_loss": 0.001132609392516315,
      "selected_pairs": 27012.2,
      "state_query_norm": 1103.4086181640625,
      "state_relation_loss": 0.00026603889127727596,
      "step": 25710,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001132609392516315,
      "weighted_relation_loss": 0.001132609392516315
    },
    {
      "action_query_norm": 622.3525512695312,
      "action_relation_loss": 0.0008264122297987342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.059869036482695,
      "flow_matching_loss": 0.01471123807132244,
      "grad_norm": 0.07604436576366425,
      "learning_rate": 5.464696725176827e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010928182280622422,
      "selected_pairs": 26459.1,
      "state_query_norm": 1105.1359252929688,
      "state_relation_loss": 0.00026640600408427416,
      "step": 25720,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0010928182280622422,
      "weighted_relation_loss": 0.0010928182280622422
    },
    {
      "action_query_norm": 629.4636901855469,
      "action_relation_loss": 0.0008508414262905717,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.069223573433113,
      "flow_matching_loss": 0.014347197394818068,
      "grad_norm": 0.09005078673362732,
      "learning_rate": 5.439669327016294e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011186046642251313,
      "selected_pairs": 27454.3,
      "state_query_norm": 1100.5331665039062,
      "state_relation_loss": 0.0002677632262930274,
      "step": 25730,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0011186046642251313,
      "weighted_relation_loss": 0.0011186046642251313
    },
    {
      "action_query_norm": 622.4290222167969,
      "action_relation_loss": 0.0009111693070735782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.078578110383535,
      "flow_matching_loss": 0.014921930525451899,
      "grad_norm": 0.0862506851553917,
      "learning_rate": 5.4146960738592985e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011787678115069867,
      "selected_pairs": 27090.9,
      "state_query_norm": 1106.7312866210937,
      "state_relation_loss": 0.0002675985073437914,
      "step": 25740,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011787678115069867,
      "weighted_relation_loss": 0.0011787678115069867
    },
    {
      "action_query_norm": 626.2740661621094,
      "action_relation_loss": 0.0008446361811365932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.087932647333957,
      "flow_matching_loss": 0.014889567624777555,
      "grad_norm": 0.08572740852832794,
      "learning_rate": 5.389776996050694e-06,
      "loss": 0.016,
      "relation_loss": 0.0011138216592371463,
      "selected_pairs": 26375.8,
      "state_query_norm": 1122.779443359375,
      "state_relation_loss": 0.00026918546645902095,
      "step": 25750,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0011138216592371463,
      "weighted_relation_loss": 0.0011138216592371463
    },
    {
      "action_query_norm": 621.4710205078125,
      "action_relation_loss": 0.0008680689032189548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.09728718428438,
      "flow_matching_loss": 0.015091092232614756,
      "grad_norm": 0.07683993130922318,
      "learning_rate": 5.364912123869492e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011306503205560149,
      "selected_pairs": 26707.1,
      "state_query_norm": 1114.7296264648437,
      "state_relation_loss": 0.00026258140278514477,
      "step": 25760,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011306503205560149,
      "weighted_relation_loss": 0.0011306503205560149
    },
    {
      "action_query_norm": 615.7569519042969,
      "action_relation_loss": 0.0008899063570424915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.106641721234798,
      "flow_matching_loss": 0.013264682609587908,
      "grad_norm": 0.06859380006790161,
      "learning_rate": 5.3401014875288864e-06,
      "loss": 0.0144,
      "relation_loss": 0.0011603859369643033,
      "selected_pairs": 26504.7,
      "state_query_norm": 1111.2421997070312,
      "state_relation_loss": 0.0002704795653698966,
      "step": 25770,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011603859369643033,
      "weighted_relation_loss": 0.0011603859369643033
    },
    {
      "action_query_norm": 625.6939208984375,
      "action_relation_loss": 0.0008552352373953909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.11599625818522,
      "flow_matching_loss": 0.017222390696406364,
      "grad_norm": 0.09120902419090271,
      "learning_rate": 5.315345117176129e-06,
      "loss": 0.0183,
      "relation_loss": 0.0011171394027769566,
      "selected_pairs": 27503.5,
      "state_query_norm": 1120.6209228515625,
      "state_relation_loss": 0.0002619041566504166,
      "step": 25780,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0011171394027769566,
      "weighted_relation_loss": 0.0011171394027769566
    },
    {
      "action_query_norm": 621.8490600585938,
      "action_relation_loss": 0.0008414388576056808,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.125350795135642,
      "flow_matching_loss": 0.015790259838104247,
      "grad_norm": 0.08971169590950012,
      "learning_rate": 5.2906430428925405e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011135949986055492,
      "selected_pairs": 26430.9,
      "state_query_norm": 1105.7046875,
      "state_relation_loss": 0.0002721561322687194,
      "step": 25790,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011135949986055492,
      "weighted_relation_loss": 0.0011135949986055492
    },
    {
      "action_query_norm": 622.0622009277344,
      "action_relation_loss": 0.0008866251329891384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.13470533208606,
      "flow_matching_loss": 0.015225858613848686,
      "grad_norm": 0.09019879251718521,
      "learning_rate": 5.265995294693471e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011568641988560558,
      "selected_pairs": 27016.5,
      "state_query_norm": 1112.148876953125,
      "state_relation_loss": 0.00027023904549423606,
      "step": 25800,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011568641988560558,
      "weighted_relation_loss": 0.0011568641988560558
    },
    {
      "action_query_norm": 617.9769348144531,
      "action_relation_loss": 0.0008515422057826072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.144059869036482,
      "flow_matching_loss": 0.014057893864810466,
      "grad_norm": 0.09212075918912888,
      "learning_rate": 5.241401902528254e-06,
      "loss": 0.0152,
      "relation_loss": 0.0011068168911151587,
      "selected_pairs": 26362.7,
      "state_query_norm": 1113.5228149414063,
      "state_relation_loss": 0.00025527468387736006,
      "step": 25810,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0011068168911151587,
      "weighted_relation_loss": 0.0011068168911151587
    },
    {
      "action_query_norm": 625.226416015625,
      "action_relation_loss": 0.000892803375609219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.153414405986904,
      "flow_matching_loss": 0.0166028443723917,
      "grad_norm": 0.10383977741003036,
      "learning_rate": 5.216862896280189e-06,
      "loss": 0.0178,
      "relation_loss": 0.0011602357029914857,
      "selected_pairs": 26848.0,
      "state_query_norm": 1107.7572265625,
      "state_relation_loss": 0.0002674323070095852,
      "step": 25820,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011602357029914857,
      "weighted_relation_loss": 0.0011602357029914857
    },
    {
      "action_query_norm": 620.9184692382812,
      "action_relation_loss": 0.00085463454015553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.162768942937326,
      "flow_matching_loss": 0.015457954816520214,
      "grad_norm": 0.09073159098625183,
      "learning_rate": 5.192378305766471e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011345434701070189,
      "selected_pairs": 26735.3,
      "state_query_norm": 1109.1786010742187,
      "state_relation_loss": 0.00027990892849629747,
      "step": 25830,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011345434701070189,
      "weighted_relation_loss": 0.0011345434701070189
    },
    {
      "action_query_norm": 620.245751953125,
      "action_relation_loss": 0.0008823455136734992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.172123479887745,
      "flow_matching_loss": 0.016711931116878985,
      "grad_norm": 0.10588806122541428,
      "learning_rate": 5.1679481607382065e-06,
      "loss": 0.0179,
      "relation_loss": 0.0011533729266375303,
      "selected_pairs": 26625.9,
      "state_query_norm": 1109.107763671875,
      "state_relation_loss": 0.0002710274246055633,
      "step": 25840,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011533729266375303,
      "weighted_relation_loss": 0.0011533729266375303
    },
    {
      "action_query_norm": 620.8557495117187,
      "action_relation_loss": 0.0008563460898585618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.181478016838167,
      "flow_matching_loss": 0.014170154463499784,
      "grad_norm": 0.07348302751779556,
      "learning_rate": 5.1435724908803105e-06,
      "loss": 0.0153,
      "relation_loss": 0.0011190721765160561,
      "selected_pairs": 26507.2,
      "state_query_norm": 1118.70107421875,
      "state_relation_loss": 0.0002627260822919197,
      "step": 25850,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011190721765160561,
      "weighted_relation_loss": 0.0011190721765160561
    },
    {
      "action_query_norm": 617.0719177246094,
      "action_relation_loss": 0.0008645336667541415,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.19083255378859,
      "flow_matching_loss": 0.01611492121592164,
      "grad_norm": 0.09521342813968658,
      "learning_rate": 5.119251325811514e-06,
      "loss": 0.0172,
      "relation_loss": 0.001134721946436912,
      "selected_pairs": 26467.7,
      "state_query_norm": 1117.2679077148437,
      "state_relation_loss": 0.0002701883058762178,
      "step": 25860,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001134721946436912,
      "weighted_relation_loss": 0.001134721946436912
    },
    {
      "action_query_norm": 628.1066528320313,
      "action_relation_loss": 0.0008596450265031308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.200187090739007,
      "flow_matching_loss": 0.014663718547672034,
      "grad_norm": 0.0747852697968483,
      "learning_rate": 5.094984695084348e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011258029262535274,
      "selected_pairs": 27246.6,
      "state_query_norm": 1121.3577392578125,
      "state_relation_loss": 0.0002661578851984814,
      "step": 25870,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011258029262535274,
      "weighted_relation_loss": 0.0011258029262535274
    },
    {
      "action_query_norm": 626.9454956054688,
      "action_relation_loss": 0.0008326671610120684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.20954162768943,
      "flow_matching_loss": 0.01655182857066393,
      "grad_norm": 0.10053875297307968,
      "learning_rate": 5.070772628185039e-06,
      "loss": 0.0177,
      "relation_loss": 0.0011064075282774866,
      "selected_pairs": 26524.4,
      "state_query_norm": 1126.3616455078125,
      "state_relation_loss": 0.00027374037308618425,
      "step": 25880,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0011064075282774866,
      "weighted_relation_loss": 0.0011064075282774866
    },
    {
      "action_query_norm": 624.9973815917969,
      "action_relation_loss": 0.0008419235644396394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.21889616463985,
      "flow_matching_loss": 0.015712182316929103,
      "grad_norm": 0.07836343348026276,
      "learning_rate": 5.046615154533535e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011058790842071176,
      "selected_pairs": 26999.5,
      "state_query_norm": 1128.1584838867188,
      "state_relation_loss": 0.0002639555197674781,
      "step": 25890,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011058790842071176,
      "weighted_relation_loss": 0.0011058790842071176
    },
    {
      "action_query_norm": 625.6135314941406,
      "action_relation_loss": 0.0008890289696864784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.22825070159027,
      "flow_matching_loss": 0.01519884867593646,
      "grad_norm": 0.08678079396486282,
      "learning_rate": 5.022512303483451e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011691008461639286,
      "selected_pairs": 26480.8,
      "state_query_norm": 1115.8431030273437,
      "state_relation_loss": 0.00028007188520859925,
      "step": 25900,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011691008461639286,
      "weighted_relation_loss": 0.0011691008461639286
    },
    {
      "action_query_norm": 619.3649841308594,
      "action_relation_loss": 0.0008799279341474176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.23760523854069,
      "flow_matching_loss": 0.01565448958426714,
      "grad_norm": 0.09492599219083786,
      "learning_rate": 4.998464104322015e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011549338814802468,
      "selected_pairs": 26760.8,
      "state_query_norm": 1102.732275390625,
      "state_relation_loss": 0.00027500596770551055,
      "step": 25910,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011549338814802468,
      "weighted_relation_loss": 0.0011549338814802468
    },
    {
      "action_query_norm": 624.4405700683594,
      "action_relation_loss": 0.0008471945533528924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.246959775491113,
      "flow_matching_loss": 0.014339418895542622,
      "grad_norm": 0.1054888665676117,
      "learning_rate": 4.974470586270047e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011295197415165602,
      "selected_pairs": 26667.5,
      "state_query_norm": 1111.7216064453125,
      "state_relation_loss": 0.0002823251736117527,
      "step": 25920,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011295197415165602,
      "weighted_relation_loss": 0.0011295197415165602
    },
    {
      "action_query_norm": 623.8150024414062,
      "action_relation_loss": 0.0008699917758349329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.256314312441535,
      "flow_matching_loss": 0.015013782680034638,
      "grad_norm": 0.08709219098091125,
      "learning_rate": 4.950531778481965e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011446982040069996,
      "selected_pairs": 27333.7,
      "state_query_norm": 1112.8068115234375,
      "state_relation_loss": 0.0002747064281720668,
      "step": 25930,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0011446982040069996,
      "weighted_relation_loss": 0.0011446982040069996
    },
    {
      "action_query_norm": 620.069775390625,
      "action_relation_loss": 0.0008882881316822022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.265668849391954,
      "flow_matching_loss": 0.015651124995201826,
      "grad_norm": 0.09179501980543137,
      "learning_rate": 4.926647710045651e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011706347926519812,
      "selected_pairs": 27007.0,
      "state_query_norm": 1098.60810546875,
      "state_relation_loss": 0.00028234664787305517,
      "step": 25940,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011706347926519812,
      "weighted_relation_loss": 0.0011706347926519812
    },
    {
      "action_query_norm": 619.3699645996094,
      "action_relation_loss": 0.0008877088141161949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.275023386342376,
      "flow_matching_loss": 0.0154386087320745,
      "grad_norm": 0.07130111753940582,
      "learning_rate": 4.9028184099825125e-06,
      "loss": 0.0166,
      "relation_loss": 0.0011612009606324137,
      "selected_pairs": 26547.5,
      "state_query_norm": 1121.9713012695313,
      "state_relation_loss": 0.0002734921523369849,
      "step": 25950,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011612009606324137,
      "weighted_relation_loss": 0.0011612009606324137
    },
    {
      "action_query_norm": 618.7507873535156,
      "action_relation_loss": 0.000846849026856944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.284377923292798,
      "flow_matching_loss": 0.014906090311706066,
      "grad_norm": 0.0728854164481163,
      "learning_rate": 4.879043907247382e-06,
      "loss": 0.016,
      "relation_loss": 0.0011233813944272696,
      "selected_pairs": 26857.4,
      "state_query_norm": 1111.5863403320313,
      "state_relation_loss": 0.00027653236465994266,
      "step": 25960,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011233813944272696,
      "weighted_relation_loss": 0.0011233813944272696
    },
    {
      "action_query_norm": 620.4484008789062,
      "action_relation_loss": 0.0008662772423122078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.293732460243216,
      "flow_matching_loss": 0.01608681445941329,
      "grad_norm": 0.08703858405351639,
      "learning_rate": 4.8553242307285414e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011368763516657054,
      "selected_pairs": 26987.7,
      "state_query_norm": 1122.6758422851562,
      "state_relation_loss": 0.00027059912390541284,
      "step": 25970,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011368763516657054,
      "weighted_relation_loss": 0.0011368763516657054
    },
    {
      "action_query_norm": 624.9376098632813,
      "action_relation_loss": 0.0008918554289266467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.30308699719364,
      "flow_matching_loss": 0.01404744703322649,
      "grad_norm": 0.09407204389572144,
      "learning_rate": 4.8316594092476185e-06,
      "loss": 0.0152,
      "relation_loss": 0.0011726101278327406,
      "selected_pairs": 26414.4,
      "state_query_norm": 1101.3510131835938,
      "state_relation_loss": 0.000280754707637243,
      "step": 25980,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0011726101278327406,
      "weighted_relation_loss": 0.0011726101278327406
    },
    {
      "action_query_norm": 625.3806640625,
      "action_relation_loss": 0.0008526860096026212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.31244153414406,
      "flow_matching_loss": 0.01488763215020299,
      "grad_norm": 0.08364809304475784,
      "learning_rate": 4.808049471559617e-06,
      "loss": 0.016,
      "relation_loss": 0.0011189140845090152,
      "selected_pairs": 27194.8,
      "state_query_norm": 1124.1517822265625,
      "state_relation_loss": 0.0002662280632648617,
      "step": 25990,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011189140845090152,
      "weighted_relation_loss": 0.0011189140845090152
    },
    {
      "action_query_norm": 616.2747985839844,
      "action_relation_loss": 0.000913976231822744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.321796071094482,
      "flow_matching_loss": 0.015153794735670089,
      "grad_norm": 0.08939162641763687,
      "learning_rate": 4.784494446352833e-06,
      "loss": 0.0163,
      "relation_loss": 0.001188819552771747,
      "selected_pairs": 27298.0,
      "state_query_norm": 1103.4285278320312,
      "state_relation_loss": 0.0002748433267697692,
      "step": 26000,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.001188819552771747,
      "weighted_relation_loss": 0.001188819552771747
    },
    {
      "action_query_norm": 618.7739074707031,
      "action_relation_loss": 0.0008597065869253128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.3311506080449,
      "flow_matching_loss": 0.015610856749117375,
      "grad_norm": 0.08445584774017334,
      "learning_rate": 4.760994362248833e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011328261345624924,
      "selected_pairs": 26464.2,
      "state_query_norm": 1116.647216796875,
      "state_relation_loss": 0.00027311953308526424,
      "step": 26010,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011328261345624924,
      "weighted_relation_loss": 0.0011328261345624924
    },
    {
      "action_query_norm": 621.8725708007812,
      "action_relation_loss": 0.0008421855338383466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.340505144995323,
      "flow_matching_loss": 0.013889276888221502,
      "grad_norm": 0.07854462414979935,
      "learning_rate": 4.737549247802442e-06,
      "loss": 0.015,
      "relation_loss": 0.0011164182564243675,
      "selected_pairs": 27081.7,
      "state_query_norm": 1091.1043701171875,
      "state_relation_loss": 0.00027423272404121236,
      "step": 26020,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011164182564243675,
      "weighted_relation_loss": 0.0011164182564243675
    },
    {
      "action_query_norm": 629.2702392578125,
      "action_relation_loss": 0.0008723573759198189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.349859681945745,
      "flow_matching_loss": 0.015255451202392578,
      "grad_norm": 0.09549691528081894,
      "learning_rate": 4.714159131501689e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011445910553447903,
      "selected_pairs": 26393.5,
      "state_query_norm": 1109.2339721679687,
      "state_relation_loss": 0.0002722336881561205,
      "step": 26030,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0011445910553447903,
      "weighted_relation_loss": 0.0011445910553447903
    },
    {
      "action_query_norm": 626.905908203125,
      "action_relation_loss": 0.0008796579495538026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.359214218896163,
      "flow_matching_loss": 0.014277285151183605,
      "grad_norm": 0.06870570778846741,
      "learning_rate": 4.690824041767766e-06,
      "loss": 0.0154,
      "relation_loss": 0.0011414758511818946,
      "selected_pairs": 26807.1,
      "state_query_norm": 1113.3470703125,
      "state_relation_loss": 0.0002618178928969428,
      "step": 26040,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011414758511818946,
      "weighted_relation_loss": 0.0011414758511818946
    },
    {
      "action_query_norm": 628.8514587402344,
      "action_relation_loss": 0.0008370496332645416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.368568755846585,
      "flow_matching_loss": 0.014093043096363545,
      "grad_norm": 0.07660029083490372,
      "learning_rate": 4.6675440069550105e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010966350673697889,
      "selected_pairs": 27058.9,
      "state_query_norm": 1099.9955322265625,
      "state_relation_loss": 0.00025958542973967267,
      "step": 26050,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010966350673697889,
      "weighted_relation_loss": 0.0010966350673697889
    },
    {
      "action_query_norm": 623.9452331542968,
      "action_relation_loss": 0.0008639345935080201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.377923292797007,
      "flow_matching_loss": 0.015167580358684063,
      "grad_norm": 0.08387059718370438,
      "learning_rate": 4.64431905535086e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011314068222418427,
      "selected_pairs": 27034.8,
      "state_query_norm": 1118.9463256835938,
      "state_relation_loss": 0.0002674722141819075,
      "step": 26060,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011314068222418427,
      "weighted_relation_loss": 0.0011314068222418427
    },
    {
      "action_query_norm": 621.2260559082031,
      "action_relation_loss": 0.0008567508310079574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.38727782974743,
      "flow_matching_loss": 0.01565526146441698,
      "grad_norm": 0.08714143931865692,
      "learning_rate": 4.621149215175808e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011359161464497447,
      "selected_pairs": 27197.5,
      "state_query_norm": 1113.83212890625,
      "state_relation_loss": 0.0002791653241729364,
      "step": 26070,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011359161464497447,
      "weighted_relation_loss": 0.0011359161464497447
    },
    {
      "action_query_norm": 627.5734375,
      "action_relation_loss": 0.0008109562448225915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.396632366697848,
      "flow_matching_loss": 0.014600569289177657,
      "grad_norm": 0.08669241517782211,
      "learning_rate": 4.5980345145834154e-06,
      "loss": 0.0157,
      "relation_loss": 0.0010859933216124773,
      "selected_pairs": 27361.6,
      "state_query_norm": 1097.5501098632812,
      "state_relation_loss": 0.000275037060782779,
      "step": 26080,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0010859933216124773,
      "weighted_relation_loss": 0.0010859933216124773
    },
    {
      "action_query_norm": 618.1953308105469,
      "action_relation_loss": 0.000844766286900267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.40598690364827,
      "flow_matching_loss": 0.015583532396703959,
      "grad_norm": 0.11521348357200623,
      "learning_rate": 4.574974981660213e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011090850108303129,
      "selected_pairs": 26965.4,
      "state_query_norm": 1109.5741455078125,
      "state_relation_loss": 0.00026431872393004594,
      "step": 26090,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011090850108303129,
      "weighted_relation_loss": 0.0011090850108303129
    },
    {
      "action_query_norm": 622.5276977539063,
      "action_relation_loss": 0.0008308836549986154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.41534144059869,
      "flow_matching_loss": 0.014507838618010282,
      "grad_norm": 0.10205892473459244,
      "learning_rate": 4.551970644425707e-06,
      "loss": 0.0156,
      "relation_loss": 0.0010902612353675067,
      "selected_pairs": 27200.8,
      "state_query_norm": 1110.8671020507813,
      "state_relation_loss": 0.0002593775701825507,
      "step": 26100,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0010902612353675067,
      "weighted_relation_loss": 0.0010902612353675067
    },
    {
      "action_query_norm": 620.7785278320313,
      "action_relation_loss": 0.000847357523161918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.42469597754911,
      "flow_matching_loss": 0.013828863110393285,
      "grad_norm": 0.07639946043491364,
      "learning_rate": 4.529021530832328e-06,
      "loss": 0.0149,
      "relation_loss": 0.0011015190044417976,
      "selected_pairs": 27288.5,
      "state_query_norm": 1114.1551635742187,
      "state_relation_loss": 0.0002541614856454544,
      "step": 26110,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011015190044417976,
      "weighted_relation_loss": 0.0011015190044417976
    },
    {
      "action_query_norm": 625.9575378417969,
      "action_relation_loss": 0.0008595542400144041,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.434050514499532,
      "flow_matching_loss": 0.015659529063850642,
      "grad_norm": 0.08137667179107666,
      "learning_rate": 4.5061276687654306e-06,
      "loss": 0.0168,
      "relation_loss": 0.001144338073208928,
      "selected_pairs": 26460.5,
      "state_query_norm": 1115.3572998046875,
      "state_relation_loss": 0.000284783830284141,
      "step": 26120,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001144338073208928,
      "weighted_relation_loss": 0.001144338073208928
    },
    {
      "action_query_norm": 616.0450988769531,
      "action_relation_loss": 0.0008768567990045994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.443405051449954,
      "flow_matching_loss": 0.014737803395837546,
      "grad_norm": 0.08351447433233261,
      "learning_rate": 4.4832890860431965e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011479742010124029,
      "selected_pairs": 26351.5,
      "state_query_norm": 1102.21845703125,
      "state_relation_loss": 0.00027111739036627113,
      "step": 26130,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0011479742010124029,
      "weighted_relation_loss": 0.0011479742010124029
    },
    {
      "action_query_norm": 617.6278015136719,
      "action_relation_loss": 0.0008551432227250188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.452759588400376,
      "flow_matching_loss": 0.015328174643218517,
      "grad_norm": 0.11376316845417023,
      "learning_rate": 4.460505810416682e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011175996507517994,
      "selected_pairs": 26860.1,
      "state_query_norm": 1122.07568359375,
      "state_relation_loss": 0.000262456426571589,
      "step": 26140,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011175996507517994,
      "weighted_relation_loss": 0.0011175996507517994
    },
    {
      "action_query_norm": 623.6560607910156,
      "action_relation_loss": 0.000861837452976033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.462114125350794,
      "flow_matching_loss": 0.014782131183892488,
      "grad_norm": 0.08087724447250366,
      "learning_rate": 4.4377778695696984e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011198833934031426,
      "selected_pairs": 27406.5,
      "state_query_norm": 1115.3718505859374,
      "state_relation_loss": 0.0002580459404271096,
      "step": 26150,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0011198833934031426,
      "weighted_relation_loss": 0.0011198833934031426
    },
    {
      "action_query_norm": 625.8588806152344,
      "action_relation_loss": 0.0008388693444430828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.471468662301216,
      "flow_matching_loss": 0.01526702642440796,
      "grad_norm": 0.09150870144367218,
      "learning_rate": 4.415105291118843e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011066218605265022,
      "selected_pairs": 27360.8,
      "state_query_norm": 1097.9783081054688,
      "state_relation_loss": 0.0002677525131730363,
      "step": 26160,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0011066218605265022,
      "weighted_relation_loss": 0.0011066218605265022
    },
    {
      "action_query_norm": 620.9327331542969,
      "action_relation_loss": 0.0008694982621818781,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.48082319925164,
      "flow_matching_loss": 0.01397705813869834,
      "grad_norm": 0.08980591595172882,
      "learning_rate": 4.392488102613435e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011347766849212348,
      "selected_pairs": 27388.4,
      "state_query_norm": 1111.6666381835937,
      "state_relation_loss": 0.00026527843438088893,
      "step": 26170,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0011347766849212348,
      "weighted_relation_loss": 0.0011347766849212348
    },
    {
      "action_query_norm": 624.5740417480469,
      "action_relation_loss": 0.0008287956530693919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.490177736202057,
      "flow_matching_loss": 0.013508915808051825,
      "grad_norm": 0.1132093220949173,
      "learning_rate": 4.369926331535473e-06,
      "loss": 0.0146,
      "relation_loss": 0.0010965318186208605,
      "selected_pairs": 26666.3,
      "state_query_norm": 1102.2951171875,
      "state_relation_loss": 0.00026773615973070265,
      "step": 26180,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010965318186208605,
      "weighted_relation_loss": 0.0010965318186208605
    },
    {
      "action_query_norm": 617.2390869140625,
      "action_relation_loss": 0.0008751731540542096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.49953227315248,
      "flow_matching_loss": 0.015162260830402374,
      "grad_norm": 0.07177304476499557,
      "learning_rate": 4.347420005299668e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011503217974677682,
      "selected_pairs": 27031.0,
      "state_query_norm": 1096.58544921875,
      "state_relation_loss": 0.0002751486550550908,
      "step": 26190,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011503217974677682,
      "weighted_relation_loss": 0.0011503217974677682
    },
    {
      "action_query_norm": 626.0659729003906,
      "action_relation_loss": 0.0008351555268745869,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.5088868101029,
      "flow_matching_loss": 0.01649250602349639,
      "grad_norm": 0.1240607276558876,
      "learning_rate": 4.324969151253317e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011014864547178149,
      "selected_pairs": 27108.0,
      "state_query_norm": 1120.26416015625,
      "state_relation_loss": 0.0002663309162016958,
      "step": 26200,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011014864547178149,
      "weighted_relation_loss": 0.0011014864547178149
    },
    {
      "action_query_norm": 617.5160827636719,
      "action_relation_loss": 0.0008775891503319145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.51824134705332,
      "flow_matching_loss": 0.015003903675824404,
      "grad_norm": 0.08087248355150223,
      "learning_rate": 4.302573796676312e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011498493608087302,
      "selected_pairs": 26549.1,
      "state_query_norm": 1093.4485229492188,
      "state_relation_loss": 0.0002722601988352835,
      "step": 26210,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011498493608087302,
      "weighted_relation_loss": 0.0011498493608087302
    },
    {
      "action_query_norm": 623.4134948730468,
      "action_relation_loss": 0.0008311726618558168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.52759588400374,
      "flow_matching_loss": 0.013970237132161856,
      "grad_norm": 0.08805396407842636,
      "learning_rate": 4.2802339687811385e-06,
      "loss": 0.0151,
      "relation_loss": 0.0010924733243882657,
      "selected_pairs": 27013.8,
      "state_query_norm": 1107.1283447265625,
      "state_relation_loss": 0.00026130066107725723,
      "step": 26220,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010924733243882657,
      "weighted_relation_loss": 0.0010924733243882657
    },
    {
      "action_query_norm": 619.8656799316407,
      "action_relation_loss": 0.0008662838896270842,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.536950420954163,
      "flow_matching_loss": 0.015043631754815578,
      "grad_norm": 0.09929444640874863,
      "learning_rate": 4.257949694712771e-06,
      "loss": 0.0162,
      "relation_loss": 0.001140241476241499,
      "selected_pairs": 26383.0,
      "state_query_norm": 1106.252880859375,
      "state_relation_loss": 0.0002739575778832659,
      "step": 26230,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001140241476241499,
      "weighted_relation_loss": 0.001140241476241499
    },
    {
      "action_query_norm": 622.0199951171875,
      "action_relation_loss": 0.0008989289286546409,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.546304957904585,
      "flow_matching_loss": 0.015756415855139493,
      "grad_norm": 0.08695920556783676,
      "learning_rate": 4.235721001548726e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011638745898380875,
      "selected_pairs": 26667.5,
      "state_query_norm": 1114.7541015625,
      "state_relation_loss": 0.00026494567573536186,
      "step": 26240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011638745898380875,
      "weighted_relation_loss": 0.0011638745898380875
    },
    {
      "action_query_norm": 623.60673828125,
      "action_relation_loss": 0.0008393764437641948,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.555659494855004,
      "flow_matching_loss": 0.0157340575940907,
      "grad_norm": 0.08072800189256668,
      "learning_rate": 4.21354791629896e-06,
      "loss": 0.0168,
      "relation_loss": 0.001106445596087724,
      "selected_pairs": 27389.7,
      "state_query_norm": 1109.277392578125,
      "state_relation_loss": 0.00026706914941314607,
      "step": 26250,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.001106445596087724,
      "weighted_relation_loss": 0.001106445596087724
    },
    {
      "action_query_norm": 614.0228271484375,
      "action_relation_loss": 0.0008672819123603404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.565014031805426,
      "flow_matching_loss": 0.014745626226067543,
      "grad_norm": 0.07745279371738434,
      "learning_rate": 4.191430465905843e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011345660546794533,
      "selected_pairs": 26804.6,
      "state_query_norm": 1101.5420654296875,
      "state_relation_loss": 0.0002672841539606452,
      "step": 26260,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011345660546794533,
      "weighted_relation_loss": 0.0011345660546794533
    },
    {
      "action_query_norm": 615.2474487304687,
      "action_relation_loss": 0.0008746963168960065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.574368568755848,
      "flow_matching_loss": 0.017150308843702076,
      "grad_norm": 0.08394525945186615,
      "learning_rate": 4.169368677244184e-06,
      "loss": 0.0183,
      "relation_loss": 0.0011512923403643072,
      "selected_pairs": 26458.5,
      "state_query_norm": 1104.3459228515626,
      "state_relation_loss": 0.0002765960467513651,
      "step": 26270,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011512923403643072,
      "weighted_relation_loss": 0.0011512923403643072
    },
    {
      "action_query_norm": 624.0553161621094,
      "action_relation_loss": 0.0008557853056117892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.583723105706266,
      "flow_matching_loss": 0.014085239171981812,
      "grad_norm": 0.09234883636236191,
      "learning_rate": 4.147362577121122e-06,
      "loss": 0.0152,
      "relation_loss": 0.0011220745043829083,
      "selected_pairs": 26782.5,
      "state_query_norm": 1126.7599731445312,
      "state_relation_loss": 0.00026628921186784283,
      "step": 26280,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011220745043829083,
      "weighted_relation_loss": 0.0011220745043829083
    },
    {
      "action_query_norm": 619.696142578125,
      "action_relation_loss": 0.0008613356214482338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.593077642656688,
      "flow_matching_loss": 0.014646556787192821,
      "grad_norm": 0.09561392664909363,
      "learning_rate": 4.125412192276157e-06,
      "loss": 0.0158,
      "relation_loss": 0.00112642525928095,
      "selected_pairs": 26708.4,
      "state_query_norm": 1101.0512329101562,
      "state_relation_loss": 0.00026508963201195,
      "step": 26290,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00112642525928095,
      "weighted_relation_loss": 0.00112642525928095
    },
    {
      "action_query_norm": 623.5146179199219,
      "action_relation_loss": 0.000831718131667003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.60243217960711,
      "flow_matching_loss": 0.015977890230715275,
      "grad_norm": 0.07716967165470123,
      "learning_rate": 4.10351754938107e-06,
      "loss": 0.0171,
      "relation_loss": 0.0010915952967479824,
      "selected_pairs": 27131.7,
      "state_query_norm": 1111.4353881835937,
      "state_relation_loss": 0.0002598771679913625,
      "step": 26300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010915952967479824,
      "weighted_relation_loss": 0.0010915952967479824
    },
    {
      "action_query_norm": 618.4845397949218,
      "action_relation_loss": 0.0008300644753035158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.611786716557532,
      "flow_matching_loss": 0.014379344880580902,
      "grad_norm": 0.08284686505794525,
      "learning_rate": 4.081678675039913e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011018708581104874,
      "selected_pairs": 27249.3,
      "state_query_norm": 1110.4393432617187,
      "state_relation_loss": 0.00027180638280697165,
      "step": 26310,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011018708581104874,
      "weighted_relation_loss": 0.0011018708581104874
    },
    {
      "action_query_norm": 630.2207153320312,
      "action_relation_loss": 0.0008419486752245575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.62114125350795,
      "flow_matching_loss": 0.017373879719525576,
      "grad_norm": 0.06836239993572235,
      "learning_rate": 4.0598955957889876e-06,
      "loss": 0.0185,
      "relation_loss": 0.001103404932655394,
      "selected_pairs": 27407.5,
      "state_query_norm": 1107.684423828125,
      "state_relation_loss": 0.00026145625015487893,
      "step": 26320,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001103404932655394,
      "weighted_relation_loss": 0.001103404932655394
    },
    {
      "action_query_norm": 612.1756652832031,
      "action_relation_loss": 0.00084846310201101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.630495790458372,
      "flow_matching_loss": 0.014428088814020157,
      "grad_norm": 0.08563816547393799,
      "learning_rate": 4.038168338096776e-06,
      "loss": 0.0155,
      "relation_loss": 0.001115736470092088,
      "selected_pairs": 27360.3,
      "state_query_norm": 1098.8414306640625,
      "state_relation_loss": 0.0002672733797226101,
      "step": 26330,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001115736470092088,
      "weighted_relation_loss": 0.001115736470092088
    },
    {
      "action_query_norm": 620.5043518066407,
      "action_relation_loss": 0.0008748399268370122,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.639850327408794,
      "flow_matching_loss": 0.015267306473106146,
      "grad_norm": 0.0924636721611023,
      "learning_rate": 4.016496928363944e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011314927134662867,
      "selected_pairs": 26462.4,
      "state_query_norm": 1103.8449584960938,
      "state_relation_loss": 0.0002566527939052321,
      "step": 26340,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011314927134662867,
      "weighted_relation_loss": 0.0011314927134662867
    },
    {
      "action_query_norm": 622.9174011230468,
      "action_relation_loss": 0.0008590028912294656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.649204864359213,
      "flow_matching_loss": 0.014441240578889847,
      "grad_norm": 0.08380909264087677,
      "learning_rate": 3.994881392923317e-06,
      "loss": 0.0156,
      "relation_loss": 0.0011233309749513864,
      "selected_pairs": 27248.1,
      "state_query_norm": 1116.6994018554688,
      "state_relation_loss": 0.0002643280677148141,
      "step": 26350,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011233309749513864,
      "weighted_relation_loss": 0.0011233309749513864
    },
    {
      "action_query_norm": 621.238720703125,
      "action_relation_loss": 0.0008757765754126012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.658559401309635,
      "flow_matching_loss": 0.015764986723661424,
      "grad_norm": 0.10761420428752899,
      "learning_rate": 3.9733217580397945e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011274799471721053,
      "selected_pairs": 26503.0,
      "state_query_norm": 1098.4478149414062,
      "state_relation_loss": 0.0002517033921321854,
      "step": 26360,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011274799471721053,
      "weighted_relation_loss": 0.0011274799471721053
    },
    {
      "action_query_norm": 618.9785400390625,
      "action_relation_loss": 0.0009279523568693548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.667913938260057,
      "flow_matching_loss": 0.015986524801701308,
      "grad_norm": 0.08573933690786362,
      "learning_rate": 3.951818049910366e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011977955466136337,
      "selected_pairs": 26573.6,
      "state_query_norm": 1109.5611572265625,
      "state_relation_loss": 0.0002698431897442788,
      "step": 26370,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011977955466136337,
      "weighted_relation_loss": 0.0011977955466136337
    },
    {
      "action_query_norm": 616.3029296875,
      "action_relation_loss": 0.0008357882907148451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.67726847521048,
      "flow_matching_loss": 0.013645015098154545,
      "grad_norm": 0.08380018174648285,
      "learning_rate": 3.930370294664071e-06,
      "loss": 0.0147,
      "relation_loss": 0.0011017852928489446,
      "selected_pairs": 26597.9,
      "state_query_norm": 1108.7633422851563,
      "state_relation_loss": 0.00026599699340295045,
      "step": 26380,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011017852928489446,
      "weighted_relation_loss": 0.0011017852928489446
    },
    {
      "action_query_norm": 621.8196411132812,
      "action_relation_loss": 0.0008597893815021962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.686623012160897,
      "flow_matching_loss": 0.013717749062925578,
      "grad_norm": 0.09029699862003326,
      "learning_rate": 3.908978518361938e-06,
      "loss": 0.0149,
      "relation_loss": 0.0011347652762196958,
      "selected_pairs": 27296.2,
      "state_query_norm": 1099.5109985351562,
      "state_relation_loss": 0.00027497588307596744,
      "step": 26390,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011347652762196958,
      "weighted_relation_loss": 0.0011347652762196958
    },
    {
      "action_query_norm": 622.1710144042969,
      "action_relation_loss": 0.0008047867740970105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.69597754911132,
      "flow_matching_loss": 0.016060423757880925,
      "grad_norm": 0.11171458661556244,
      "learning_rate": 3.887642746997017e-06,
      "loss": 0.0171,
      "relation_loss": 0.0010681028477847576,
      "selected_pairs": 26890.6,
      "state_query_norm": 1101.0224731445312,
      "state_relation_loss": 0.00026331608241889624,
      "step": 26400,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010681028477847576,
      "weighted_relation_loss": 0.0010681028477847576
    },
    {
      "action_query_norm": 619.9687194824219,
      "action_relation_loss": 0.0008703065861482174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.70533208606174,
      "flow_matching_loss": 0.014942276570945979,
      "grad_norm": 0.10936913639307022,
      "learning_rate": 3.866363006494255e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011432238155975939,
      "selected_pairs": 26556.2,
      "state_query_norm": 1121.6029418945313,
      "state_relation_loss": 0.0002729172236286104,
      "step": 26410,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011432238155975939,
      "weighted_relation_loss": 0.0011432238155975939
    },
    {
      "action_query_norm": 622.5294799804688,
      "action_relation_loss": 0.000859342998592183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.71468662301216,
      "flow_matching_loss": 0.015659889671951532,
      "grad_norm": 0.10457547754049301,
      "learning_rate": 3.8451393227105725e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011281556566245855,
      "selected_pairs": 27245.9,
      "state_query_norm": 1105.8440307617188,
      "state_relation_loss": 0.0002688126842258498,
      "step": 26420,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011281556566245855,
      "weighted_relation_loss": 0.0011281556566245855
    },
    {
      "action_query_norm": 620.5953186035156,
      "action_relation_loss": 0.000841630797367543,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.72404115996258,
      "flow_matching_loss": 0.014291059505194426,
      "grad_norm": 0.08012765645980835,
      "learning_rate": 3.823971721434727e-06,
      "loss": 0.0154,
      "relation_loss": 0.0011141754337586462,
      "selected_pairs": 26847.8,
      "state_query_norm": 1098.0408203125,
      "state_relation_loss": 0.00027254464221186935,
      "step": 26430,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011141754337586462,
      "weighted_relation_loss": 0.0011141754337586462
    },
    {
      "action_query_norm": 617.4117553710937,
      "action_relation_loss": 0.0008486116130370647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.733395696913004,
      "flow_matching_loss": 0.013959089014679193,
      "grad_norm": 0.06967522948980331,
      "learning_rate": 3.8028602283873505e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011261561536230146,
      "selected_pairs": 26970.7,
      "state_query_norm": 1101.0835815429687,
      "state_relation_loss": 0.0002775445464067161,
      "step": 26440,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011261561536230146,
      "weighted_relation_loss": 0.0011261561536230146
    },
    {
      "action_query_norm": 624.0335144042969,
      "action_relation_loss": 0.0008978565165307373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.742750233863426,
      "flow_matching_loss": 0.015796677395701408,
      "grad_norm": 0.0999443456530571,
      "learning_rate": 3.781804869220912e-06,
      "loss": 0.017,
      "relation_loss": 0.0011685260804370045,
      "selected_pairs": 26505.4,
      "state_query_norm": 1121.3808837890624,
      "state_relation_loss": 0.0002706695726374164,
      "step": 26450,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011685260804370045,
      "weighted_relation_loss": 0.0011685260804370045
    },
    {
      "action_query_norm": 621.5495422363281,
      "action_relation_loss": 0.0008808905375190079,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.752104770813844,
      "flow_matching_loss": 0.014229181222617626,
      "grad_norm": 0.08472122997045517,
      "learning_rate": 3.760805669519646e-06,
      "loss": 0.0154,
      "relation_loss": 0.0011516144732013345,
      "selected_pairs": 26966.3,
      "state_query_norm": 1120.7353881835938,
      "state_relation_loss": 0.0002707239298615605,
      "step": 26460,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011516144732013345,
      "weighted_relation_loss": 0.0011516144732013345
    },
    {
      "action_query_norm": 612.1471618652344,
      "action_relation_loss": 0.0008726656320504844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.761459307764266,
      "flow_matching_loss": 0.01551765538752079,
      "grad_norm": 0.08812130242586136,
      "learning_rate": 3.7398626547995587e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011352292844094337,
      "selected_pairs": 26209.0,
      "state_query_norm": 1105.655859375,
      "state_relation_loss": 0.00026256366982124746,
      "step": 26470,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0011352292844094337,
      "weighted_relation_loss": 0.0011352292844094337
    },
    {
      "action_query_norm": 615.795947265625,
      "action_relation_loss": 0.0008566400094423444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.770813844714688,
      "flow_matching_loss": 0.013880566786974668,
      "grad_norm": 0.08517419546842575,
      "learning_rate": 3.7189758505083995e-06,
      "loss": 0.015,
      "relation_loss": 0.0011266848887316883,
      "selected_pairs": 27201.9,
      "state_query_norm": 1098.5359252929688,
      "state_relation_loss": 0.000270044882199727,
      "step": 26480,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011266848887316883,
      "weighted_relation_loss": 0.0011266848887316883
    },
    {
      "action_query_norm": 615.9254333496094,
      "action_relation_loss": 0.0008779231400694698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.780168381665106,
      "flow_matching_loss": 0.015862400457262994,
      "grad_norm": 0.09249447286128998,
      "learning_rate": 3.6981452820255837e-06,
      "loss": 0.017,
      "relation_loss": 0.0011463871807791293,
      "selected_pairs": 26734.9,
      "state_query_norm": 1093.2369506835937,
      "state_relation_loss": 0.00026846404653042556,
      "step": 26490,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011463871807791293,
      "weighted_relation_loss": 0.0011463871807791293
    },
    {
      "action_query_norm": 618.385498046875,
      "action_relation_loss": 0.0008446041494607926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.78952291861553,
      "flow_matching_loss": 0.01565442830324173,
      "grad_norm": 0.07872694730758667,
      "learning_rate": 3.6773709746622363e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011080786469392478,
      "selected_pairs": 26624.0,
      "state_query_norm": 1104.6086303710938,
      "state_relation_loss": 0.0002634744989336468,
      "step": 26500,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011080786469392478,
      "weighted_relation_loss": 0.0011080786469392478
    },
    {
      "action_query_norm": 618.0748046875,
      "action_relation_loss": 0.0008532606589142233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.79887745556595,
      "flow_matching_loss": 0.013668750878423452,
      "grad_norm": 0.0936577245593071,
      "learning_rate": 3.6566529536610995e-06,
      "loss": 0.0148,
      "relation_loss": 0.0011193533195182682,
      "selected_pairs": 26500.9,
      "state_query_norm": 1123.4305786132813,
      "state_relation_loss": 0.0002660926431417465,
      "step": 26510,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011193533195182682,
      "weighted_relation_loss": 0.0011193533195182682
    },
    {
      "action_query_norm": 622.7977661132812,
      "action_relation_loss": 0.0008519402763340622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.80823199251637,
      "flow_matching_loss": 0.013320111390203237,
      "grad_norm": 0.09752153605222702,
      "learning_rate": 3.635991244196513e-06,
      "loss": 0.0144,
      "relation_loss": 0.0011146875098347665,
      "selected_pairs": 26987.8,
      "state_query_norm": 1118.71826171875,
      "state_relation_loss": 0.0002627472538733855,
      "step": 26520,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011146875098347665,
      "weighted_relation_loss": 0.0011146875098347665
    },
    {
      "action_query_norm": 620.2897033691406,
      "action_relation_loss": 0.0008402396342717111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.81758652946679,
      "flow_matching_loss": 0.018047502636909483,
      "grad_norm": 0.08835780620574951,
      "learning_rate": 3.615385871374405e-06,
      "loss": 0.0191,
      "relation_loss": 0.001100877090357244,
      "selected_pairs": 26823.8,
      "state_query_norm": 1125.5133056640625,
      "state_relation_loss": 0.0002606374735478312,
      "step": 26530,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001100877090357244,
      "weighted_relation_loss": 0.001100877090357244
    },
    {
      "action_query_norm": 622.4894592285157,
      "action_relation_loss": 0.00086198088247329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.826941066417213,
      "flow_matching_loss": 0.016296823043376207,
      "grad_norm": 0.1101670041680336,
      "learning_rate": 3.5948368602322367e-06,
      "loss": 0.0174,
      "relation_loss": 0.0011338778771460057,
      "selected_pairs": 27078.5,
      "state_query_norm": 1123.0240966796875,
      "state_relation_loss": 0.00027189700631424785,
      "step": 26540,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011338778771460057,
      "weighted_relation_loss": 0.0011338778771460057
    },
    {
      "action_query_norm": 625.3525512695312,
      "action_relation_loss": 0.0008504561439622194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.836295603367635,
      "flow_matching_loss": 0.015966959297657013,
      "grad_norm": 0.09999266266822815,
      "learning_rate": 3.5743442357390187e-06,
      "loss": 0.0171,
      "relation_loss": 0.0011070040636695921,
      "selected_pairs": 26732.4,
      "state_query_norm": 1107.829150390625,
      "state_relation_loss": 0.000256547903700266,
      "step": 26550,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011070040636695921,
      "weighted_relation_loss": 0.0011070040636695921
    },
    {
      "action_query_norm": 617.0553100585937,
      "action_relation_loss": 0.0008613216457888484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.845650140318053,
      "flow_matching_loss": 0.015193265210837126,
      "grad_norm": 0.07303013652563095,
      "learning_rate": 3.5539080227951937e-06,
      "loss": 0.0163,
      "relation_loss": 0.001126936194486916,
      "selected_pairs": 27386.1,
      "state_query_norm": 1112.4691162109375,
      "state_relation_loss": 0.00026561454142211006,
      "step": 26560,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.001126936194486916,
      "weighted_relation_loss": 0.001126936194486916
    },
    {
      "action_query_norm": 617.1000244140625,
      "action_relation_loss": 0.0008759096555877477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.855004677268475,
      "flow_matching_loss": 0.014998175296932459,
      "grad_norm": 0.07125753164291382,
      "learning_rate": 3.533528246232709e-06,
      "loss": 0.0161,
      "relation_loss": 0.001144461624789983,
      "selected_pairs": 26621.6,
      "state_query_norm": 1110.3845581054688,
      "state_relation_loss": 0.00026855196920223535,
      "step": 26570,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001144461624789983,
      "weighted_relation_loss": 0.001144461624789983
    },
    {
      "action_query_norm": 615.6782409667969,
      "action_relation_loss": 0.0008684612053912133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.864359214218897,
      "flow_matching_loss": 0.012999094743281602,
      "grad_norm": 0.11014366894960403,
      "learning_rate": 3.5132049308149117e-06,
      "loss": 0.0141,
      "relation_loss": 0.0011379956966266036,
      "selected_pairs": 27155.4,
      "state_query_norm": 1125.674951171875,
      "state_relation_loss": 0.0002695344970561564,
      "step": 26580,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011379956966266036,
      "weighted_relation_loss": 0.0011379956966266036
    },
    {
      "action_query_norm": 623.8919311523438,
      "action_relation_loss": 0.000873606459936127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.873713751169316,
      "flow_matching_loss": 0.01473593758419156,
      "grad_norm": 0.08578374236822128,
      "learning_rate": 3.4929381012365425e-06,
      "loss": 0.0159,
      "relation_loss": 0.001142490073107183,
      "selected_pairs": 26755.4,
      "state_query_norm": 1118.0959838867188,
      "state_relation_loss": 0.0002688836015295237,
      "step": 26590,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001142490073107183,
      "weighted_relation_loss": 0.001142490073107183
    },
    {
      "action_query_norm": 618.7734741210937,
      "action_relation_loss": 0.0008220824354793876,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.883068288119738,
      "flow_matching_loss": 0.015162255242466926,
      "grad_norm": 0.08863966166973114,
      "learning_rate": 3.472727782123697e-06,
      "loss": 0.0163,
      "relation_loss": 0.0010934932855889202,
      "selected_pairs": 27486.0,
      "state_query_norm": 1119.3389038085938,
      "state_relation_loss": 0.00027141086175106464,
      "step": 26600,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0010934932855889202,
      "weighted_relation_loss": 0.0010934932855889202
    },
    {
      "action_query_norm": 617.763818359375,
      "action_relation_loss": 0.000891543721081689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.89242282507016,
      "flow_matching_loss": 0.013866357784718276,
      "grad_norm": 0.10623190551996231,
      "learning_rate": 3.4525739980338425e-06,
      "loss": 0.015,
      "relation_loss": 0.0011493543745018542,
      "selected_pairs": 26988.9,
      "state_query_norm": 1107.8560668945313,
      "state_relation_loss": 0.0002578106679720804,
      "step": 26610,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011493543745018542,
      "weighted_relation_loss": 0.0011493543745018542
    },
    {
      "action_query_norm": 615.6385070800782,
      "action_relation_loss": 0.0008697319368366152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.90177736202058,
      "flow_matching_loss": 0.014622835349291563,
      "grad_norm": 0.10093614459037781,
      "learning_rate": 3.432476773455712e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011361113167367876,
      "selected_pairs": 27061.7,
      "state_query_norm": 1104.8865478515625,
      "state_relation_loss": 0.00026637937262421476,
      "step": 26620,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011361113167367876,
      "weighted_relation_loss": 0.0011361113167367876
    },
    {
      "action_query_norm": 623.7913024902343,
      "action_relation_loss": 0.0008468083629850298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.911131898971,
      "flow_matching_loss": 0.014558338932693005,
      "grad_norm": 0.08172973990440369,
      "learning_rate": 3.4124361328093378e-06,
      "loss": 0.0157,
      "relation_loss": 0.001116680703125894,
      "selected_pairs": 26921.1,
      "state_query_norm": 1077.3539184570313,
      "state_relation_loss": 0.00026987232849933206,
      "step": 26630,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001116680703125894,
      "weighted_relation_loss": 0.001116680703125894
    },
    {
      "action_query_norm": 619.5310668945312,
      "action_relation_loss": 0.0008274804509710521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.920486435921422,
      "flow_matching_loss": 0.016565672867000104,
      "grad_norm": 0.12465187907218933,
      "learning_rate": 3.392452100445975e-06,
      "loss": 0.0177,
      "relation_loss": 0.001088724925648421,
      "selected_pairs": 26167.9,
      "state_query_norm": 1105.274951171875,
      "state_relation_loss": 0.00026124447758775206,
      "step": 26640,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.001088724925648421,
      "weighted_relation_loss": 0.001088724925648421
    },
    {
      "action_query_norm": 621.5736694335938,
      "action_relation_loss": 0.0008446938882116228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.929840972871844,
      "flow_matching_loss": 0.01653260290622711,
      "grad_norm": 0.09236866980791092,
      "learning_rate": 3.3725247006481163e-06,
      "loss": 0.0176,
      "relation_loss": 0.0011091037653386594,
      "selected_pairs": 26805.8,
      "state_query_norm": 1107.61669921875,
      "state_relation_loss": 0.00026440989167895166,
      "step": 26650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011091037653386594,
      "weighted_relation_loss": 0.0011091037653386594
    },
    {
      "action_query_norm": 621.694189453125,
      "action_relation_loss": 0.0008560983580537141,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.939195509822262,
      "flow_matching_loss": 0.015747634693980217,
      "grad_norm": 0.06965207308530807,
      "learning_rate": 3.3526539576294313e-06,
      "loss": 0.0169,
      "relation_loss": 0.001130090153310448,
      "selected_pairs": 26824.6,
      "state_query_norm": 1106.5323608398437,
      "state_relation_loss": 0.0002739917807048187,
      "step": 26660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001130090153310448,
      "weighted_relation_loss": 0.001130090153310448
    },
    {
      "action_query_norm": 616.250732421875,
      "action_relation_loss": 0.0008580391760915518,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.948550046772684,
      "flow_matching_loss": 0.013650638237595557,
      "grad_norm": 0.08419174700975418,
      "learning_rate": 3.332839895534745e-06,
      "loss": 0.0148,
      "relation_loss": 0.0011200213222764433,
      "selected_pairs": 26559.0,
      "state_query_norm": 1122.4953002929688,
      "state_relation_loss": 0.00026198214327450843,
      "step": 26670,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011200213222764433,
      "weighted_relation_loss": 0.0011200213222764433
    },
    {
      "action_query_norm": 621.9929809570312,
      "action_relation_loss": 0.0008267969416920096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.957904583723106,
      "flow_matching_loss": 0.01420669788494706,
      "grad_norm": 0.0748620480298996,
      "learning_rate": 3.3130825384400155e-06,
      "loss": 0.0153,
      "relation_loss": 0.001094485993962735,
      "selected_pairs": 27244.6,
      "state_query_norm": 1127.001513671875,
      "state_relation_loss": 0.0002676890493603423,
      "step": 26680,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001094485993962735,
      "weighted_relation_loss": 0.001094485993962735
    },
    {
      "action_query_norm": 621.977392578125,
      "action_relation_loss": 0.0008228513353969902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.96725912067353,
      "flow_matching_loss": 0.01533409571275115,
      "grad_norm": 0.08050144463777542,
      "learning_rate": 3.293381910352278e-06,
      "loss": 0.0164,
      "relation_loss": 0.0010936534497886896,
      "selected_pairs": 27035.9,
      "state_query_norm": 1109.509130859375,
      "state_relation_loss": 0.0002708021173020825,
      "step": 26690,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010936534497886896,
      "weighted_relation_loss": 0.0010936534497886896
    },
    {
      "action_query_norm": 623.7440734863281,
      "action_relation_loss": 0.0008060575113631785,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.976613657623947,
      "flow_matching_loss": 0.016107097361236812,
      "grad_norm": 0.09186083823442459,
      "learning_rate": 3.273738035209678e-06,
      "loss": 0.0172,
      "relation_loss": 0.00108264097943902,
      "selected_pairs": 26709.1,
      "state_query_norm": 1104.3739501953125,
      "state_relation_loss": 0.0002765834651654586,
      "step": 26700,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00108264097943902,
      "weighted_relation_loss": 0.00108264097943902
    },
    {
      "action_query_norm": 621.9758361816406,
      "action_relation_loss": 0.0008497033675666898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.98596819457437,
      "flow_matching_loss": 0.015556726232171058,
      "grad_norm": 0.07304549962282181,
      "learning_rate": 3.254150936881356e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011099681840278208,
      "selected_pairs": 26479.0,
      "state_query_norm": 1119.49794921875,
      "state_relation_loss": 0.0002602648382890038,
      "step": 26710,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0011099681840278208,
      "weighted_relation_loss": 0.0011099681840278208
    },
    {
      "action_query_norm": 620.1849304199219,
      "action_relation_loss": 0.0008473698282614351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.99532273152479,
      "flow_matching_loss": 0.013980212341994046,
      "grad_norm": 0.08948294073343277,
      "learning_rate": 3.234620639167496e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011041545658372343,
      "selected_pairs": 26271.5,
      "state_query_norm": 1114.2765380859375,
      "state_relation_loss": 0.00025678474048618227,
      "step": 26720,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0011041545658372343,
      "weighted_relation_loss": 0.0011041545658372343
    },
    {
      "action_query_norm": 623.2447692871094,
      "action_relation_loss": 0.0008478710835333914,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.00467726847521,
      "flow_matching_loss": 0.014424287527799607,
      "grad_norm": 0.08626634627580643,
      "learning_rate": 3.2151471657992482e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011188962322194128,
      "selected_pairs": 24424.8,
      "state_query_norm": 1102.6457397460938,
      "state_relation_loss": 0.00027102515887236225,
      "step": 26730,
      "valid_action_samples": 214.9,
      "weighted_auxiliary_loss": 0.0011188962322194128,
      "weighted_relation_loss": 0.0011188962322194128
    },
    {
      "action_query_norm": 624.1746948242187,
      "action_relation_loss": 0.0008460152079351246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.01403180542563,
      "flow_matching_loss": 0.016117136925458908,
      "grad_norm": 0.07269493490457535,
      "learning_rate": 3.195730540438718e-06,
      "loss": 0.0172,
      "relation_loss": 0.001114714110735804,
      "selected_pairs": 26532.7,
      "state_query_norm": 1117.6489624023438,
      "state_relation_loss": 0.000268698888248764,
      "step": 26740,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001114714110735804,
      "weighted_relation_loss": 0.001114714110735804
    },
    {
      "action_query_norm": 616.4843383789063,
      "action_relation_loss": 0.0008714458032045514,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.023386342376053,
      "flow_matching_loss": 0.017526574712246657,
      "grad_norm": 0.09033285826444626,
      "learning_rate": 3.1763707866789336e-06,
      "loss": 0.0187,
      "relation_loss": 0.0011376855312846602,
      "selected_pairs": 26872.2,
      "state_query_norm": 1125.4455322265626,
      "state_relation_loss": 0.0002662397106178105,
      "step": 26750,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011376855312846602,
      "weighted_relation_loss": 0.0011376855312846602
    },
    {
      "action_query_norm": 624.352978515625,
      "action_relation_loss": 0.0008008048869669437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.03274087932647,
      "flow_matching_loss": 0.014780216850340366,
      "grad_norm": 0.08901500701904297,
      "learning_rate": 3.1570679280438165e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010587944532744586,
      "selected_pairs": 27334.5,
      "state_query_norm": 1122.3936645507813,
      "state_relation_loss": 0.0002579895779490471,
      "step": 26760,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010587944532744586,
      "weighted_relation_loss": 0.0010587944532744586
    },
    {
      "action_query_norm": 621.3048583984375,
      "action_relation_loss": 0.0008520192466676235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.042095416276894,
      "flow_matching_loss": 0.015055415034294129,
      "grad_norm": 0.10196426510810852,
      "learning_rate": 3.1378219879881767e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011099335388280451,
      "selected_pairs": 26759.2,
      "state_query_norm": 1122.4033569335938,
      "state_relation_loss": 0.0002579142834292725,
      "step": 26770,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011099335388280451,
      "weighted_relation_loss": 0.0011099335388280451
    },
    {
      "action_query_norm": 624.6516906738282,
      "action_relation_loss": 0.0008408660651184618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.051449953227316,
      "flow_matching_loss": 0.01684486847370863,
      "grad_norm": 0.09198548644781113,
      "learning_rate": 3.11863298989763e-06,
      "loss": 0.018,
      "relation_loss": 0.0011053615133278071,
      "selected_pairs": 27291.6,
      "state_query_norm": 1100.9785766601562,
      "state_relation_loss": 0.0002644954452989623,
      "step": 26780,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011053615133278071,
      "weighted_relation_loss": 0.0011053615133278071
    },
    {
      "action_query_norm": 627.8452453613281,
      "action_relation_loss": 0.0008373435935936868,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.060804490177738,
      "flow_matching_loss": 0.015237871930003167,
      "grad_norm": 0.09257777035236359,
      "learning_rate": 3.0995009570886303e-06,
      "loss": 0.0163,
      "relation_loss": 0.001110418001189828,
      "selected_pairs": 27340.9,
      "state_query_norm": 1131.7843994140626,
      "state_relation_loss": 0.0002730743944994174,
      "step": 26790,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.001110418001189828,
      "weighted_relation_loss": 0.001110418001189828
    },
    {
      "action_query_norm": 619.7232788085937,
      "action_relation_loss": 0.0008614614082034678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.070159027128156,
      "flow_matching_loss": 0.01393819972872734,
      "grad_norm": 0.08681193739175797,
      "learning_rate": 3.0804259128083936e-06,
      "loss": 0.0151,
      "relation_loss": 0.001128754927776754,
      "selected_pairs": 27036.1,
      "state_query_norm": 1118.6597900390625,
      "state_relation_loss": 0.0002672935224836692,
      "step": 26800,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001128754927776754,
      "weighted_relation_loss": 0.001128754927776754
    },
    {
      "action_query_norm": 623.0852844238282,
      "action_relation_loss": 0.0008628851384855807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.079513564078578,
      "flow_matching_loss": 0.015281097404658795,
      "grad_norm": 0.10208888351917267,
      "learning_rate": 3.06140788023489e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011214868165552617,
      "selected_pairs": 27618.5,
      "state_query_norm": 1107.554052734375,
      "state_relation_loss": 0.0002586016751592979,
      "step": 26810,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0011214868165552617,
      "weighted_relation_loss": 0.0011214868165552617
    },
    {
      "action_query_norm": 618.7669860839844,
      "action_relation_loss": 0.0008603846596088261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.088868101029,
      "flow_matching_loss": 0.014549269806593657,
      "grad_norm": 0.08030060678720474,
      "learning_rate": 3.042446882476846e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011198868276551365,
      "selected_pairs": 27602.7,
      "state_query_norm": 1116.7634399414062,
      "state_relation_loss": 0.0002595021534943953,
      "step": 26820,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.0011198868276551365,
      "weighted_relation_loss": 0.0011198868276551365
    },
    {
      "action_query_norm": 611.451025390625,
      "action_relation_loss": 0.0008377799123991281,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.09822263797942,
      "flow_matching_loss": 0.014359257090836764,
      "grad_norm": 0.09887140244245529,
      "learning_rate": 3.023542942573643e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011062113684602083,
      "selected_pairs": 26807.6,
      "state_query_norm": 1100.3466674804688,
      "state_relation_loss": 0.0002684314589714631,
      "step": 26830,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011062113684602083,
      "weighted_relation_loss": 0.0011062113684602083
    },
    {
      "action_query_norm": 625.442578125,
      "action_relation_loss": 0.0008567031531129032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.10757717492984,
      "flow_matching_loss": 0.014070030581206084,
      "grad_norm": 0.08474478870630264,
      "learning_rate": 3.004696083495351e-06,
      "loss": 0.0152,
      "relation_loss": 0.0011123832082375884,
      "selected_pairs": 27126.7,
      "state_query_norm": 1119.8406005859374,
      "state_relation_loss": 0.00025568004202796145,
      "step": 26840,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011123832082375884,
      "weighted_relation_loss": 0.0011123832082375884
    },
    {
      "action_query_norm": 620.6744201660156,
      "action_relation_loss": 0.0008672404219396412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.116931711880262,
      "flow_matching_loss": 0.01575917610898614,
      "grad_norm": 0.08624887466430664,
      "learning_rate": 2.98590632814269e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011370894848369063,
      "selected_pairs": 27054.0,
      "state_query_norm": 1108.3222534179688,
      "state_relation_loss": 0.00026984907017322255,
      "step": 26850,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011370894848369063,
      "weighted_relation_loss": 0.0011370894848369063
    },
    {
      "action_query_norm": 615.86787109375,
      "action_relation_loss": 0.0008579044137150049,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.126286248830684,
      "flow_matching_loss": 0.01586234150454402,
      "grad_norm": 0.0710822120308876,
      "learning_rate": 2.9671736993469702e-06,
      "loss": 0.017,
      "relation_loss": 0.0011234489269554616,
      "selected_pairs": 26206.7,
      "state_query_norm": 1109.6932983398438,
      "state_relation_loss": 0.0002655445146956481,
      "step": 26860,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0011234489269554616,
      "weighted_relation_loss": 0.0011234489269554616
    },
    {
      "action_query_norm": 620.8005493164062,
      "action_relation_loss": 0.0008329005853738636,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.135640785781103,
      "flow_matching_loss": 0.013466002605855464,
      "grad_norm": 0.08250798285007477,
      "learning_rate": 2.9484982198701216e-06,
      "loss": 0.0146,
      "relation_loss": 0.00110721344826743,
      "selected_pairs": 26733.7,
      "state_query_norm": 1112.5666015625,
      "state_relation_loss": 0.000274312861438375,
      "step": 26870,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00110721344826743,
      "weighted_relation_loss": 0.00110721344826743
    },
    {
      "action_query_norm": 614.2699951171875,
      "action_relation_loss": 0.000838346075033769,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.144995322731525,
      "flow_matching_loss": 0.014937569573521613,
      "grad_norm": 0.0850367471575737,
      "learning_rate": 2.929879912404604e-06,
      "loss": 0.016,
      "relation_loss": 0.0011012707720510662,
      "selected_pairs": 27156.1,
      "state_query_norm": 1111.7571899414063,
      "state_relation_loss": 0.0002629247115692124,
      "step": 26880,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011012707720510662,
      "weighted_relation_loss": 0.0011012707720510662
    },
    {
      "action_query_norm": 628.8034729003906,
      "action_relation_loss": 0.0008178102259989827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.154349859681947,
      "flow_matching_loss": 0.014827671740204097,
      "grad_norm": 0.09628738462924957,
      "learning_rate": 2.9113187995734147e-06,
      "loss": 0.0159,
      "relation_loss": 0.0010808360297232867,
      "selected_pairs": 27372.3,
      "state_query_norm": 1110.7924682617188,
      "state_relation_loss": 0.00026302578480681406,
      "step": 26890,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0010808360297232867,
      "weighted_relation_loss": 0.0010808360297232867
    },
    {
      "action_query_norm": 624.3474670410156,
      "action_relation_loss": 0.0008561426133383065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.163704396632365,
      "flow_matching_loss": 0.013916244171559811,
      "grad_norm": 0.1414981335401535,
      "learning_rate": 2.8928149039300524e-06,
      "loss": 0.015,
      "relation_loss": 0.0011293996358290315,
      "selected_pairs": 27436.6,
      "state_query_norm": 1094.452099609375,
      "state_relation_loss": 0.00027325701084919275,
      "step": 26900,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0011293996358290315,
      "weighted_relation_loss": 0.0011293996358290315
    },
    {
      "action_query_norm": 620.1252319335938,
      "action_relation_loss": 0.0008538177993614226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.173058933582787,
      "flow_matching_loss": 0.014338727295398711,
      "grad_norm": 0.08794771134853363,
      "learning_rate": 2.8743682479584975e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011141102062538267,
      "selected_pairs": 27062.2,
      "state_query_norm": 1093.2360961914062,
      "state_relation_loss": 0.00026029241853393615,
      "step": 26910,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011141102062538267,
      "weighted_relation_loss": 0.0011141102062538267
    },
    {
      "action_query_norm": 621.0865966796875,
      "action_relation_loss": 0.0008239582122769207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.18241347053321,
      "flow_matching_loss": 0.01576742557808757,
      "grad_norm": 0.09826472401618958,
      "learning_rate": 2.8559788540731824e-06,
      "loss": 0.0169,
      "relation_loss": 0.0010894746636040508,
      "selected_pairs": 26963.9,
      "state_query_norm": 1106.1193237304688,
      "state_relation_loss": 0.00026551644259598106,
      "step": 26920,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010894746636040508,
      "weighted_relation_loss": 0.0010894746636040508
    },
    {
      "action_query_norm": 620.8786682128906,
      "action_relation_loss": 0.0008476073271594942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.19176800748363,
      "flow_matching_loss": 0.014719638228416442,
      "grad_norm": 0.08591114729642868,
      "learning_rate": 2.8376467446189493e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011201455490663648,
      "selected_pairs": 26874.1,
      "state_query_norm": 1106.5381225585938,
      "state_relation_loss": 0.00027253821899648757,
      "step": 26930,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011201455490663648,
      "weighted_relation_loss": 0.0011201455490663648
    },
    {
      "action_query_norm": 620.7487426757813,
      "action_relation_loss": 0.0008592713100370019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.20112254443405,
      "flow_matching_loss": 0.014503728598356247,
      "grad_norm": 0.10211198776960373,
      "learning_rate": 2.8193719418710406e-06,
      "loss": 0.0156,
      "relation_loss": 0.0011325234081596136,
      "selected_pairs": 26698.5,
      "state_query_norm": 1094.1881103515625,
      "state_relation_loss": 0.00027325208648107944,
      "step": 26940,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011325234081596136,
      "weighted_relation_loss": 0.0011325234081596136
    },
    {
      "action_query_norm": 616.6353698730469,
      "action_relation_loss": 0.0008482538221869617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.21047708138447,
      "flow_matching_loss": 0.015349163394421338,
      "grad_norm": 0.07760624587535858,
      "learning_rate": 2.8011544680350667e-06,
      "loss": 0.0165,
      "relation_loss": 0.0011141123599372805,
      "selected_pairs": 26916.6,
      "state_query_norm": 1092.5336181640625,
      "state_relation_loss": 0.0002658585261087865,
      "step": 26950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011141123599372805,
      "weighted_relation_loss": 0.0011141123599372805
    },
    {
      "action_query_norm": 618.3649719238281,
      "action_relation_loss": 0.0008436296076979489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.219831618334894,
      "flow_matching_loss": 0.015137192606925965,
      "grad_norm": 0.0841633528470993,
      "learning_rate": 2.7829943452469753e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011021176469512284,
      "selected_pairs": 26252.4,
      "state_query_norm": 1108.0084716796875,
      "state_relation_loss": 0.0002584880552603863,
      "step": 26960,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0011021176469512284,
      "weighted_relation_loss": 0.0011021176469512284
    },
    {
      "action_query_norm": 619.065283203125,
      "action_relation_loss": 0.0008692929812241345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.229186155285312,
      "flow_matching_loss": 0.014290361665189266,
      "grad_norm": 0.0900050550699234,
      "learning_rate": 2.7648915955730213e-06,
      "loss": 0.0154,
      "relation_loss": 0.001140089810360223,
      "selected_pairs": 27243.6,
      "state_query_norm": 1099.8826782226563,
      "state_relation_loss": 0.00027079681749455633,
      "step": 26970,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001140089810360223,
      "weighted_relation_loss": 0.001140089810360223
    },
    {
      "action_query_norm": 622.1859619140625,
      "action_relation_loss": 0.0008400397375226021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.238540692235734,
      "flow_matching_loss": 0.013439098652452231,
      "grad_norm": 0.11347765475511551,
      "learning_rate": 2.746846241009765e-06,
      "loss": 0.0145,
      "relation_loss": 0.0011076927999965846,
      "selected_pairs": 27429.0,
      "state_query_norm": 1113.6949584960937,
      "state_relation_loss": 0.0002676530712051317,
      "step": 26980,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0011076927999965846,
      "weighted_relation_loss": 0.0011076927999965846
    },
    {
      "action_query_norm": 621.6621887207032,
      "action_relation_loss": 0.0008458772441372276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.247895229186156,
      "flow_matching_loss": 0.014809301309287547,
      "grad_norm": 0.08185157179832458,
      "learning_rate": 2.7288583034839945e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011103635770268738,
      "selected_pairs": 26765.5,
      "state_query_norm": 1120.045068359375,
      "state_relation_loss": 0.00026448633580002936,
      "step": 26990,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011103635770268738,
      "weighted_relation_loss": 0.0011103635770268738
    },
    {
      "action_query_norm": 623.01689453125,
      "action_relation_loss": 0.0007981223985552788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.257249766136574,
      "flow_matching_loss": 0.015834769513458014,
      "grad_norm": 0.09388093650341034,
      "learning_rate": 2.7109278048527752e-06,
      "loss": 0.0169,
      "relation_loss": 0.0010694600408896805,
      "selected_pairs": 27078.4,
      "state_query_norm": 1108.7892944335938,
      "state_relation_loss": 0.0002713376365136355,
      "step": 27000,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010694600408896805,
      "weighted_relation_loss": 0.0010694600408896805
    },
    {
      "action_query_norm": 615.9682067871094,
      "action_relation_loss": 0.0008970037684775889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.266604303086996,
      "flow_matching_loss": 0.014479727577418089,
      "grad_norm": 0.07638867944478989,
      "learning_rate": 2.6930547669033413e-06,
      "loss": 0.0156,
      "relation_loss": 0.0011595641146413982,
      "selected_pairs": 26503.1,
      "state_query_norm": 1098.356494140625,
      "state_relation_loss": 0.00026256034179823474,
      "step": 27010,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011595641146413982,
      "weighted_relation_loss": 0.0011595641146413982
    },
    {
      "action_query_norm": 620.514990234375,
      "action_relation_loss": 0.0008477126015350222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.27595884003742,
      "flow_matching_loss": 0.015219263639301062,
      "grad_norm": 0.08983445167541504,
      "learning_rate": 2.675239211353109e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011193162528797985,
      "selected_pairs": 26916.3,
      "state_query_norm": 1107.6533813476562,
      "state_relation_loss": 0.00027160366298630836,
      "step": 27020,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011193162528797985,
      "weighted_relation_loss": 0.0011193162528797985
    },
    {
      "action_query_norm": 628.6657531738281,
      "action_relation_loss": 0.0008375052770134061,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.28531337698784,
      "flow_matching_loss": 0.01410566121339798,
      "grad_norm": 0.09573953598737717,
      "learning_rate": 2.6574811598496786e-06,
      "loss": 0.0152,
      "relation_loss": 0.00110564831411466,
      "selected_pairs": 26648.8,
      "state_query_norm": 1111.3533935546875,
      "state_relation_loss": 0.0002681430545635521,
      "step": 27030,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.00110564831411466,
      "weighted_relation_loss": 0.00110564831411466
    },
    {
      "action_query_norm": 622.9689758300781,
      "action_relation_loss": 0.0008313684200402349,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.29466791393826,
      "flow_matching_loss": 0.015829831082373857,
      "grad_norm": 0.07489251345396042,
      "learning_rate": 2.6397806339707456e-06,
      "loss": 0.0169,
      "relation_loss": 0.0011010568821802735,
      "selected_pairs": 26339.6,
      "state_query_norm": 1100.2174560546875,
      "state_relation_loss": 0.0002696884592296556,
      "step": 27040,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0011010568821802735,
      "weighted_relation_loss": 0.0011010568821802735
    },
    {
      "action_query_norm": 617.0344787597656,
      "action_relation_loss": 0.0008405304688494653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.30402245088868,
      "flow_matching_loss": 0.014427508972585202,
      "grad_norm": 0.09201066195964813,
      "learning_rate": 2.6221376552241217e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011100748903118073,
      "selected_pairs": 26583.7,
      "state_query_norm": 1099.7848754882812,
      "state_relation_loss": 0.0002695444243727252,
      "step": 27050,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011100748903118073,
      "weighted_relation_loss": 0.0011100748903118073
    },
    {
      "action_query_norm": 614.8837829589844,
      "action_relation_loss": 0.0008705253363586962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.313376987839103,
      "flow_matching_loss": 0.013920343294739724,
      "grad_norm": 0.10064602643251419,
      "learning_rate": 2.604552245047681e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011411939049139619,
      "selected_pairs": 26685.0,
      "state_query_norm": 1090.8078735351562,
      "state_relation_loss": 0.00027066857146564873,
      "step": 27060,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011411939049139619,
      "weighted_relation_loss": 0.0011411939049139619
    },
    {
      "action_query_norm": 610.87451171875,
      "action_relation_loss": 0.0008304760151077062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.32273152478952,
      "flow_matching_loss": 0.014644010551273822,
      "grad_norm": 0.07633747905492783,
      "learning_rate": 2.587024424809359e-06,
      "loss": 0.0157,
      "relation_loss": 0.0010843515628948808,
      "selected_pairs": 26525.5,
      "state_query_norm": 1114.3045288085937,
      "state_relation_loss": 0.0002538755434216,
      "step": 27070,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010843515628948808,
      "weighted_relation_loss": 0.0010843515628948808
    },
    {
      "action_query_norm": 613.4111938476562,
      "action_relation_loss": 0.0008377057209145277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.332086061739943,
      "flow_matching_loss": 0.014483612962067127,
      "grad_norm": 0.0896272361278534,
      "learning_rate": 2.5695542158071186e-06,
      "loss": 0.0156,
      "relation_loss": 0.0010999129386618733,
      "selected_pairs": 26923.0,
      "state_query_norm": 1106.839599609375,
      "state_relation_loss": 0.00026220723084406926,
      "step": 27080,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010999129386618733,
      "weighted_relation_loss": 0.0010999129386618733
    },
    {
      "action_query_norm": 616.5421813964844,
      "action_relation_loss": 0.0008622033230494708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.341440598690365,
      "flow_matching_loss": 0.014341971278190613,
      "grad_norm": 0.08928890526294708,
      "learning_rate": 2.5521416392689068e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011409265222027897,
      "selected_pairs": 26053.7,
      "state_query_norm": 1083.4650390625,
      "state_relation_loss": 0.0002787231904221699,
      "step": 27090,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0011409265222027897,
      "weighted_relation_loss": 0.0011409265222027897
    },
    {
      "action_query_norm": 617.1146911621094,
      "action_relation_loss": 0.0008369709074031561,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.350795135640787,
      "flow_matching_loss": 0.01385922748595476,
      "grad_norm": 0.08857274055480957,
      "learning_rate": 2.5347867163526384e-06,
      "loss": 0.015,
      "relation_loss": 0.00109700420871377,
      "selected_pairs": 26646.4,
      "state_query_norm": 1110.1190551757813,
      "state_relation_loss": 0.0002600332940346561,
      "step": 27100,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.00109700420871377,
      "weighted_relation_loss": 0.00109700420871377
    },
    {
      "action_query_norm": 611.997607421875,
      "action_relation_loss": 0.0008776761882472783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.360149672591206,
      "flow_matching_loss": 0.014954991824924945,
      "grad_norm": 0.09578970074653625,
      "learning_rate": 2.517489468146189e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011410171398892998,
      "selected_pairs": 26331.1,
      "state_query_norm": 1098.24736328125,
      "state_relation_loss": 0.00026334095746278764,
      "step": 27110,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0011410171398892998,
      "weighted_relation_loss": 0.0011410171398892998
    },
    {
      "action_query_norm": 622.064453125,
      "action_relation_loss": 0.0008447119442280381,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.369504209541628,
      "flow_matching_loss": 0.015391420479863882,
      "grad_norm": 0.09778299927711487,
      "learning_rate": 2.500249915667341e-06,
      "loss": 0.0165,
      "relation_loss": 0.0011085076024755836,
      "selected_pairs": 27032.1,
      "state_query_norm": 1119.7743408203125,
      "state_relation_loss": 0.0002637956597027369,
      "step": 27120,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011085076024755836,
      "weighted_relation_loss": 0.0011085076024755836
    },
    {
      "action_query_norm": 621.7861877441406,
      "action_relation_loss": 0.0008473300316836685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.37885874649205,
      "flow_matching_loss": 0.014062139298766851,
      "grad_norm": 0.10344526916742325,
      "learning_rate": 2.4830680798637817e-06,
      "loss": 0.0152,
      "relation_loss": 0.0011099968338385223,
      "selected_pairs": 26896.3,
      "state_query_norm": 1103.8806274414062,
      "state_relation_loss": 0.0002626668181619607,
      "step": 27130,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011099968338385223,
      "weighted_relation_loss": 0.0011099968338385223
    },
    {
      "action_query_norm": 617.7472595214844,
      "action_relation_loss": 0.000817741989158094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.388213283442468,
      "flow_matching_loss": 0.015102669224143028,
      "grad_norm": 0.06724544614553452,
      "learning_rate": 2.465943981613056e-06,
      "loss": 0.0162,
      "relation_loss": 0.001092932850588113,
      "selected_pairs": 27339.1,
      "state_query_norm": 1103.2255249023438,
      "state_relation_loss": 0.0002751908657955937,
      "step": 27140,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.001092932850588113,
      "weighted_relation_loss": 0.001092932850588113
    },
    {
      "action_query_norm": 617.7854370117187,
      "action_relation_loss": 0.00084487798740156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.39756782039289,
      "flow_matching_loss": 0.01524141076952219,
      "grad_norm": 0.07112696021795273,
      "learning_rate": 2.448877641722569e-06,
      "loss": 0.0164,
      "relation_loss": 0.001118417805992067,
      "selected_pairs": 27184.9,
      "state_query_norm": 1106.9964477539063,
      "state_relation_loss": 0.0002735398069489747,
      "step": 27150,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001118417805992067,
      "weighted_relation_loss": 0.001118417805992067
    },
    {
      "action_query_norm": 617.8009643554688,
      "action_relation_loss": 0.0008526097051799297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.406922357343312,
      "flow_matching_loss": 0.015745253302156925,
      "grad_norm": 0.08304695785045624,
      "learning_rate": 2.431869080929522e-06,
      "loss": 0.0169,
      "relation_loss": 0.001116110838484019,
      "selected_pairs": 27335.9,
      "state_query_norm": 1122.7280029296876,
      "state_relation_loss": 0.0002635011405800469,
      "step": 27160,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.001116110838484019,
      "weighted_relation_loss": 0.001116110838484019
    },
    {
      "action_query_norm": 625.1933166503907,
      "action_relation_loss": 0.0008182513527572155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.416276894293734,
      "flow_matching_loss": 0.014354570023715495,
      "grad_norm": 0.08991743624210358,
      "learning_rate": 2.4149183199009216e-06,
      "loss": 0.0154,
      "relation_loss": 0.0010864346637390553,
      "selected_pairs": 26875.0,
      "state_query_norm": 1115.6587158203124,
      "state_relation_loss": 0.0002681833109818399,
      "step": 27170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010864346637390553,
      "weighted_relation_loss": 0.0010864346637390553
    },
    {
      "action_query_norm": 618.7008544921875,
      "action_relation_loss": 0.0008256579225417226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.425631431244152,
      "flow_matching_loss": 0.014065908174961806,
      "grad_norm": 0.08612947165966034,
      "learning_rate": 2.3980253792335427e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010916679515503348,
      "selected_pairs": 26785.0,
      "state_query_norm": 1126.9676635742187,
      "state_relation_loss": 0.00026601004356052727,
      "step": 27180,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0010916679515503348,
      "weighted_relation_loss": 0.0010916679515503348
    },
    {
      "action_query_norm": 616.7744995117188,
      "action_relation_loss": 0.0008403909974731505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.434985968194574,
      "flow_matching_loss": 0.014292053412646055,
      "grad_norm": 0.06373083591461182,
      "learning_rate": 2.381190279453899e-06,
      "loss": 0.0154,
      "relation_loss": 0.001101140771061182,
      "selected_pairs": 26754.7,
      "state_query_norm": 1108.4809692382812,
      "state_relation_loss": 0.00026074977358803153,
      "step": 27190,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001101140771061182,
      "weighted_relation_loss": 0.001101140771061182
    },
    {
      "action_query_norm": 612.7357727050781,
      "action_relation_loss": 0.000870634731836617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.444340505144996,
      "flow_matching_loss": 0.013875403348356486,
      "grad_norm": 0.07972758263349533,
      "learning_rate": 2.364413041018232e-06,
      "loss": 0.015,
      "relation_loss": 0.0011343164136633278,
      "selected_pairs": 26662.0,
      "state_query_norm": 1101.0160034179687,
      "state_relation_loss": 0.0002636816818267107,
      "step": 27200,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011343164136633278,
      "weighted_relation_loss": 0.0011343164136633278
    },
    {
      "action_query_norm": 620.9892395019531,
      "action_relation_loss": 0.0008562803850509226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.453695042095415,
      "flow_matching_loss": 0.014713576529175043,
      "grad_norm": 0.0894416868686676,
      "learning_rate": 2.347693684312463e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011229357216507197,
      "selected_pairs": 27317.7,
      "state_query_norm": 1106.26064453125,
      "state_relation_loss": 0.00026665533077903093,
      "step": 27210,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0011229357216507197,
      "weighted_relation_loss": 0.0011229357216507197
    },
    {
      "action_query_norm": 619.4724487304687,
      "action_relation_loss": 0.0008659687475301326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.463049579045837,
      "flow_matching_loss": 0.014486729353666305,
      "grad_norm": 0.10333539545536041,
      "learning_rate": 2.331032229652186e-06,
      "loss": 0.0156,
      "relation_loss": 0.001126803015358746,
      "selected_pairs": 26672.8,
      "state_query_norm": 1111.4578125,
      "state_relation_loss": 0.00026083428092533724,
      "step": 27220,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001126803015358746,
      "weighted_relation_loss": 0.001126803015358746
    },
    {
      "action_query_norm": 623.5675231933594,
      "action_relation_loss": 0.0008287010248750448,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.47240411599626,
      "flow_matching_loss": 0.014067403227090835,
      "grad_norm": 0.07335484772920609,
      "learning_rate": 2.314428697282628e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010963375330902635,
      "selected_pairs": 26569.9,
      "state_query_norm": 1117.9136596679687,
      "state_relation_loss": 0.0002676365096704103,
      "step": 27230,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0010963375330902635,
      "weighted_relation_loss": 0.0010963375330902635
    },
    {
      "action_query_norm": 617.355224609375,
      "action_relation_loss": 0.0008247769786976278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.48175865294668,
      "flow_matching_loss": 0.014099070336669683,
      "grad_norm": 0.06948163360357285,
      "learning_rate": 2.297883107378673e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010885714204050601,
      "selected_pairs": 26504.0,
      "state_query_norm": 1115.4907592773438,
      "state_relation_loss": 0.00026379441551398485,
      "step": 27240,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0010885714204050601,
      "weighted_relation_loss": 0.0010885714204050601
    },
    {
      "action_query_norm": 623.7289978027344,
      "action_relation_loss": 0.0008108186651952564,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.4911131898971,
      "flow_matching_loss": 0.014534734282642603,
      "grad_norm": 0.07062167674303055,
      "learning_rate": 2.2813954800447513e-06,
      "loss": 0.0156,
      "relation_loss": 0.0010687805886846035,
      "selected_pairs": 26848.6,
      "state_query_norm": 1089.7770263671875,
      "state_relation_loss": 0.00025796191912377255,
      "step": 27250,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010687805886846035,
      "weighted_relation_loss": 0.0010687805886846035
    },
    {
      "action_query_norm": 621.4369750976563,
      "action_relation_loss": 0.0008663368702400476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.50046772684752,
      "flow_matching_loss": 0.014302654843777417,
      "grad_norm": 0.09649889171123505,
      "learning_rate": 2.2649658353148974e-06,
      "loss": 0.0154,
      "relation_loss": 0.00113196880556643,
      "selected_pairs": 27200.8,
      "state_query_norm": 1114.788671875,
      "state_relation_loss": 0.0002656319207744673,
      "step": 27260,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.00113196880556643,
      "weighted_relation_loss": 0.00113196880556643
    },
    {
      "action_query_norm": 620.5954223632813,
      "action_relation_loss": 0.0007934017572551966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.509822263797943,
      "flow_matching_loss": 0.014580017421394587,
      "grad_norm": 0.0912742018699646,
      "learning_rate": 2.2485941931526645e-06,
      "loss": 0.0156,
      "relation_loss": 0.0010642255656421185,
      "selected_pairs": 26897.3,
      "state_query_norm": 1093.83720703125,
      "state_relation_loss": 0.00027082382002845405,
      "step": 27270,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010642255656421185,
      "weighted_relation_loss": 0.0010642255656421185
    },
    {
      "action_query_norm": 613.5706420898438,
      "action_relation_loss": 0.0008269937650766223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.51917680074836,
      "flow_matching_loss": 0.015058923792093993,
      "grad_norm": 0.08023173362016678,
      "learning_rate": 2.232280573451151e-06,
      "loss": 0.0161,
      "relation_loss": 0.0010853736428543925,
      "selected_pairs": 27030.8,
      "state_query_norm": 1105.6375854492187,
      "state_relation_loss": 0.0002583798879641108,
      "step": 27280,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010853736428543925,
      "weighted_relation_loss": 0.0010853736428543925
    },
    {
      "action_query_norm": 615.773583984375,
      "action_relation_loss": 0.0008414378855377436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.528531337698784,
      "flow_matching_loss": 0.015201420057564974,
      "grad_norm": 0.0716683492064476,
      "learning_rate": 2.216024996032945e-06,
      "loss": 0.0163,
      "relation_loss": 0.001110678748227656,
      "selected_pairs": 26465.1,
      "state_query_norm": 1103.6868896484375,
      "state_relation_loss": 0.0002692408597795293,
      "step": 27290,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001110678748227656,
      "weighted_relation_loss": 0.001110678748227656
    },
    {
      "action_query_norm": 623.979052734375,
      "action_relation_loss": 0.000865301612066105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.537885874649206,
      "flow_matching_loss": 0.01440973998978734,
      "grad_norm": 0.07726259529590607,
      "learning_rate": 2.1998274806501074e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011257510632276535,
      "selected_pairs": 27131.8,
      "state_query_norm": 1119.762890625,
      "state_relation_loss": 0.00026044945407193153,
      "step": 27300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011257510632276535,
      "weighted_relation_loss": 0.0011257510632276535
    },
    {
      "action_query_norm": 622.2267700195313,
      "action_relation_loss": 0.0008641599793918431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.547240411599624,
      "flow_matching_loss": 0.015205912198871374,
      "grad_norm": 0.07985740900039673,
      "learning_rate": 2.183688046984139e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011423899326473475,
      "selected_pairs": 26712.3,
      "state_query_norm": 1093.7512939453125,
      "state_relation_loss": 0.00027822994743473826,
      "step": 27310,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011423899326473475,
      "weighted_relation_loss": 0.0011423899326473475
    },
    {
      "action_query_norm": 613.1903259277344,
      "action_relation_loss": 0.0008565734606236219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.556594948550046,
      "flow_matching_loss": 0.015073620714247227,
      "grad_norm": 0.10020481795072556,
      "learning_rate": 2.167606714645981e-06,
      "loss": 0.0162,
      "relation_loss": 0.001118271180894226,
      "selected_pairs": 27203.6,
      "state_query_norm": 1097.19013671875,
      "state_relation_loss": 0.0002616977275465615,
      "step": 27320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001118271180894226,
      "weighted_relation_loss": 0.001118271180894226
    },
    {
      "action_query_norm": 618.7982055664063,
      "action_relation_loss": 0.0008821647090371699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.565949485500468,
      "flow_matching_loss": 0.015292725153267383,
      "grad_norm": 0.09718184918165207,
      "learning_rate": 2.151583503175958e-06,
      "loss": 0.0164,
      "relation_loss": 0.0011493314988911152,
      "selected_pairs": 26394.5,
      "state_query_norm": 1100.5849609375,
      "state_relation_loss": 0.00026716679567471144,
      "step": 27330,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0011493314988911152,
      "weighted_relation_loss": 0.0011493314988911152
    },
    {
      "action_query_norm": 618.9714416503906,
      "action_relation_loss": 0.0008550705329980701,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.57530402245089,
      "flow_matching_loss": 0.012806286104023457,
      "grad_norm": 0.07577740401029587,
      "learning_rate": 2.1356184320437957e-06,
      "loss": 0.0139,
      "relation_loss": 0.0011143436306156218,
      "selected_pairs": 26857.3,
      "state_query_norm": 1118.1541259765625,
      "state_relation_loss": 0.0002592730859760195,
      "step": 27340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011143436306156218,
      "weighted_relation_loss": 0.0011143436306156218
    },
    {
      "action_query_norm": 622.2773071289063,
      "action_relation_loss": 0.0008633618766907602,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.58465855940131,
      "flow_matching_loss": 0.013992677722126246,
      "grad_norm": 0.08493608236312866,
      "learning_rate": 2.119711520648554e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011255996534600854,
      "selected_pairs": 26735.5,
      "state_query_norm": 1110.3226196289063,
      "state_relation_loss": 0.0002622377913212404,
      "step": 27350,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011255996534600854,
      "weighted_relation_loss": 0.0011255996534600854
    },
    {
      "action_query_norm": 619.4238037109375,
      "action_relation_loss": 0.0008405998640228062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.59401309635173,
      "flow_matching_loss": 0.01463376646861434,
      "grad_norm": 0.08205113559961319,
      "learning_rate": 2.103862788318628e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011015307391062379,
      "selected_pairs": 26667.0,
      "state_query_norm": 1118.17265625,
      "state_relation_loss": 0.0002609308750834316,
      "step": 27360,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011015307391062379,
      "weighted_relation_loss": 0.0011015307391062379
    },
    {
      "action_query_norm": 620.570703125,
      "action_relation_loss": 0.0008380199782550335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.603367633302152,
      "flow_matching_loss": 0.014827607199549675,
      "grad_norm": 0.113624706864357,
      "learning_rate": 2.088072254311729e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011166627635248005,
      "selected_pairs": 27383.4,
      "state_query_norm": 1106.762255859375,
      "state_relation_loss": 0.000278642782359384,
      "step": 27370,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0011166627635248005,
      "weighted_relation_loss": 0.0011166627635248005
    },
    {
      "action_query_norm": 622.3095642089844,
      "action_relation_loss": 0.000820598896825686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.61272217025257,
      "flow_matching_loss": 0.013906501419842243,
      "grad_norm": 0.07142175734043121,
      "learning_rate": 2.0723399378148435e-06,
      "loss": 0.015,
      "relation_loss": 0.001086299039889127,
      "selected_pairs": 27645.2,
      "state_query_norm": 1115.6411254882812,
      "state_relation_loss": 0.00026570013869786635,
      "step": 27380,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.001086299039889127,
      "weighted_relation_loss": 0.001086299039889127
    },
    {
      "action_query_norm": 619.4595947265625,
      "action_relation_loss": 0.0007949183054734021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.622076707202993,
      "flow_matching_loss": 0.014745976496487855,
      "grad_norm": 0.08403363823890686,
      "learning_rate": 2.0566658579442065e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010620871558785438,
      "selected_pairs": 27403.9,
      "state_query_norm": 1105.6120971679688,
      "state_relation_loss": 0.00026716884749475864,
      "step": 27390,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0010620871558785438,
      "weighted_relation_loss": 0.0010620871558785438
    },
    {
      "action_query_norm": 617.7586242675782,
      "action_relation_loss": 0.0008527884841896593,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.631431244153415,
      "flow_matching_loss": 0.01480190958827734,
      "grad_norm": 0.08146634697914124,
      "learning_rate": 2.0410500337453176e-06,
      "loss": 0.0159,
      "relation_loss": 0.001123945089057088,
      "selected_pairs": 26668.3,
      "state_query_norm": 1106.4875122070312,
      "state_relation_loss": 0.0002711565946810879,
      "step": 27400,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001123945089057088,
      "weighted_relation_loss": 0.001123945089057088
    },
    {
      "action_query_norm": 618.1890441894532,
      "action_relation_loss": 0.0008269956684671343,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.640785781103837,
      "flow_matching_loss": 0.013794638589024544,
      "grad_norm": 0.08834493905305862,
      "learning_rate": 2.0254924841928645e-06,
      "loss": 0.0149,
      "relation_loss": 0.0010862756287679076,
      "selected_pairs": 26539.0,
      "state_query_norm": 1111.3124145507813,
      "state_relation_loss": 0.00025927996466634795,
      "step": 27410,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010862756287679076,
      "weighted_relation_loss": 0.0010862756287679076
    },
    {
      "action_query_norm": 618.0454223632812,
      "action_relation_loss": 0.0008291914185974747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.650140318054255,
      "flow_matching_loss": 0.013841443974524737,
      "grad_norm": 0.08569567650556564,
      "learning_rate": 2.009993228190754e-06,
      "loss": 0.0149,
      "relation_loss": 0.001094365236349404,
      "selected_pairs": 26538.6,
      "state_query_norm": 1123.1976928710938,
      "state_relation_loss": 0.00026517382357269523,
      "step": 27420,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001094365236349404,
      "weighted_relation_loss": 0.001094365236349404
    },
    {
      "action_query_norm": 619.8490905761719,
      "action_relation_loss": 0.0008247786608990282,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.659494855004677,
      "flow_matching_loss": 0.0143899486400187,
      "grad_norm": 0.10191923379898071,
      "learning_rate": 1.9945522845720323e-06,
      "loss": 0.0155,
      "relation_loss": 0.0010905794450081884,
      "selected_pairs": 25618.6,
      "state_query_norm": 1100.9811889648438,
      "state_relation_loss": 0.0002658007695572451,
      "step": 27430,
      "valid_action_samples": 226.8,
      "weighted_auxiliary_loss": 0.0010905794450081884,
      "weighted_relation_loss": 0.0010905794450081884
    },
    {
      "action_query_norm": 621.1149963378906,
      "action_relation_loss": 0.0008672296826262027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.6688493919551,
      "flow_matching_loss": 0.014007074851542711,
      "grad_norm": 0.08961936831474304,
      "learning_rate": 1.9791696720988963e-06,
      "loss": 0.0151,
      "relation_loss": 0.0011236826190724968,
      "selected_pairs": 27127.9,
      "state_query_norm": 1110.5427001953126,
      "state_relation_loss": 0.000256452921894379,
      "step": 27440,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011236826190724968,
      "weighted_relation_loss": 0.0011236826190724968
    },
    {
      "action_query_norm": 612.3512878417969,
      "action_relation_loss": 0.0008583630376961082,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.678203928905518,
      "flow_matching_loss": 0.014722455013543367,
      "grad_norm": 0.07795214653015137,
      "learning_rate": 1.9638454094626833e-06,
      "loss": 0.0158,
      "relation_loss": 0.001121153566055,
      "selected_pairs": 26978.1,
      "state_query_norm": 1125.9002807617187,
      "state_relation_loss": 0.00026279054291080686,
      "step": 27450,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001121153566055,
      "weighted_relation_loss": 0.001121153566055
    },
    {
      "action_query_norm": 621.1455749511719,
      "action_relation_loss": 0.000840314372908324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.68755846585594,
      "flow_matching_loss": 0.014573059137910604,
      "grad_norm": 0.0780014619231224,
      "learning_rate": 1.9485795152838104e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011080474825575948,
      "selected_pairs": 26769.2,
      "state_query_norm": 1110.1531005859374,
      "state_relation_loss": 0.00026773310819407927,
      "step": 27460,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011080474825575948,
      "weighted_relation_loss": 0.0011080474825575948
    },
    {
      "action_query_norm": 621.4176330566406,
      "action_relation_loss": 0.0008662241511046886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.69691300280636,
      "flow_matching_loss": 0.01316749006509781,
      "grad_norm": 0.08068397641181946,
      "learning_rate": 1.933372008111778e-06,
      "loss": 0.0143,
      "relation_loss": 0.0011391761363483965,
      "selected_pairs": 27462.2,
      "state_query_norm": 1098.340283203125,
      "state_relation_loss": 0.00027295198233332483,
      "step": 27470,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0011391761363483965,
      "weighted_relation_loss": 0.0011391761363483965
    },
    {
      "action_query_norm": 617.0148071289062,
      "action_relation_loss": 0.000876295130001381,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.706267539756784,
      "flow_matching_loss": 0.01539710508659482,
      "grad_norm": 0.11364281922578812,
      "learning_rate": 1.918222906425143e-06,
      "loss": 0.0165,
      "relation_loss": 0.0011494768667034804,
      "selected_pairs": 26966.5,
      "state_query_norm": 1095.1154296875,
      "state_relation_loss": 0.000273181738157291,
      "step": 27480,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011494768667034804,
      "weighted_relation_loss": 0.0011494768667034804
    },
    {
      "action_query_norm": 622.3991516113281,
      "action_relation_loss": 0.0008182982564903796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.715622076707202,
      "flow_matching_loss": 0.01501183956861496,
      "grad_norm": 0.06525632739067078,
      "learning_rate": 1.903132228631488e-06,
      "loss": 0.0161,
      "relation_loss": 0.0010910963639616967,
      "selected_pairs": 26807.4,
      "state_query_norm": 1105.110205078125,
      "state_relation_loss": 0.0002727980900090188,
      "step": 27490,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010910963639616967,
      "weighted_relation_loss": 0.0010910963639616967
    },
    {
      "action_query_norm": 613.4991149902344,
      "action_relation_loss": 0.000862502510426566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.724976613657624,
      "flow_matching_loss": 0.015717071201652288,
      "grad_norm": 0.10482598096132278,
      "learning_rate": 1.8880999930674215e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011279901256784797,
      "selected_pairs": 26761.1,
      "state_query_norm": 1103.63505859375,
      "state_relation_loss": 0.00026548762107267974,
      "step": 27500,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011279901256784797,
      "weighted_relation_loss": 0.0011279901256784797
    },
    {
      "action_query_norm": 610.64111328125,
      "action_relation_loss": 0.000858651997987181,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.734331150608046,
      "flow_matching_loss": 0.013850105181336402,
      "grad_norm": 0.06896355748176575,
      "learning_rate": 1.8731262179985166e-06,
      "loss": 0.015,
      "relation_loss": 0.001125494483858347,
      "selected_pairs": 27240.0,
      "state_query_norm": 1099.8676879882812,
      "state_relation_loss": 0.0002668424989678897,
      "step": 27510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001125494483858347,
      "weighted_relation_loss": 0.001125494483858347
    },
    {
      "action_query_norm": 618.4339965820312,
      "action_relation_loss": 0.0008759196440223605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.743685687558465,
      "flow_matching_loss": 0.0130305465310812,
      "grad_norm": 0.07690664380788803,
      "learning_rate": 1.8582109216193244e-06,
      "loss": 0.0142,
      "relation_loss": 0.0011431176797486841,
      "selected_pairs": 26778.9,
      "state_query_norm": 1100.4493896484375,
      "state_relation_loss": 0.0002671980211744085,
      "step": 27520,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011431176797486841,
      "weighted_relation_loss": 0.0011431176797486841
    },
    {
      "action_query_norm": 616.5231323242188,
      "action_relation_loss": 0.0008143602462951094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.753040224508887,
      "flow_matching_loss": 0.014676792360842229,
      "grad_norm": 0.080178402364254,
      "learning_rate": 1.8433541220533368e-06,
      "loss": 0.0158,
      "relation_loss": 0.001082539523486048,
      "selected_pairs": 26755.4,
      "state_query_norm": 1110.8736206054687,
      "state_relation_loss": 0.0002681792786461301,
      "step": 27530,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001082539523486048,
      "weighted_relation_loss": 0.001082539523486048
    },
    {
      "action_query_norm": 620.1043212890625,
      "action_relation_loss": 0.0008394130680244416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.76239476145931,
      "flow_matching_loss": 0.014625148475170135,
      "grad_norm": 0.08951215445995331,
      "learning_rate": 1.8285558373529577e-06,
      "loss": 0.0157,
      "relation_loss": 0.0011010123882442713,
      "selected_pairs": 27057.7,
      "state_query_norm": 1108.0451049804688,
      "state_relation_loss": 0.0002615993260405958,
      "step": 27540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011010123882442713,
      "weighted_relation_loss": 0.0011010123882442713
    },
    {
      "action_query_norm": 616.9431701660156,
      "action_relation_loss": 0.0008400982245802879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.77174929840973,
      "flow_matching_loss": 0.014421543292701244,
      "grad_norm": 0.11551634967327118,
      "learning_rate": 1.8138160854995145e-06,
      "loss": 0.0155,
      "relation_loss": 0.0011098543764092028,
      "selected_pairs": 27013.4,
      "state_query_norm": 1099.7911987304688,
      "state_relation_loss": 0.00026975615764968097,
      "step": 27550,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011098543764092028,
      "weighted_relation_loss": 0.0011098543764092028
    },
    {
      "action_query_norm": 614.8816040039062,
      "action_relation_loss": 0.0008640329819172621,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.78110383536015,
      "flow_matching_loss": 0.015315263625234365,
      "grad_norm": 0.08206379413604736,
      "learning_rate": 1.7991348844031864e-06,
      "loss": 0.0164,
      "relation_loss": 0.001124423136934638,
      "selected_pairs": 26875.3,
      "state_query_norm": 1101.357958984375,
      "state_relation_loss": 0.0002603901521069929,
      "step": 27560,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001124423136934638,
      "weighted_relation_loss": 0.001124423136934638
    },
    {
      "action_query_norm": 619.6514099121093,
      "action_relation_loss": 0.0008193427871447057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.79045837231057,
      "flow_matching_loss": 0.014032260235399007,
      "grad_norm": 0.09740372747182846,
      "learning_rate": 1.78451225190302e-06,
      "loss": 0.0151,
      "relation_loss": 0.0010886170726735144,
      "selected_pairs": 26593.5,
      "state_query_norm": 1098.3138916015625,
      "state_relation_loss": 0.0002692742957151495,
      "step": 27570,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0010886170726735144,
      "weighted_relation_loss": 0.0010886170726735144
    },
    {
      "action_query_norm": 614.4342956542969,
      "action_relation_loss": 0.0008279902453068644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.799812909260993,
      "flow_matching_loss": 0.013803744595497847,
      "grad_norm": 0.09669386595487595,
      "learning_rate": 1.7699482057668914e-06,
      "loss": 0.0149,
      "relation_loss": 0.0010862039518542588,
      "selected_pairs": 26303.4,
      "state_query_norm": 1112.5795654296876,
      "state_relation_loss": 0.00025821369781624525,
      "step": 27580,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0010862039518542588,
      "weighted_relation_loss": 0.0010862039518542588
    },
    {
      "action_query_norm": 619.2712829589843,
      "action_relation_loss": 0.0007957940106280148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.80916744621141,
      "flow_matching_loss": 0.014733076654374599,
      "grad_norm": 0.09098775684833527,
      "learning_rate": 1.7554427636914895e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010654408019036055,
      "selected_pairs": 26986.1,
      "state_query_norm": 1096.4553100585938,
      "state_relation_loss": 0.0002696467738132924,
      "step": 27590,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010654408019036055,
      "weighted_relation_loss": 0.0010654408019036055
    },
    {
      "action_query_norm": 620.6441345214844,
      "action_relation_loss": 0.0008345520473085344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.818521983161833,
      "flow_matching_loss": 0.014889911655336619,
      "grad_norm": 0.08079268783330917,
      "learning_rate": 1.7409959433022871e-06,
      "loss": 0.016,
      "relation_loss": 0.0010960946325212716,
      "selected_pairs": 26508.6,
      "state_query_norm": 1101.7574829101563,
      "state_relation_loss": 0.00026154257502639664,
      "step": 27600,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0010960946325212716,
      "weighted_relation_loss": 0.0010960946325212716
    },
    {
      "action_query_norm": 614.9087463378906,
      "action_relation_loss": 0.0008451630244962871,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.827876520112255,
      "flow_matching_loss": 0.01476196488365531,
      "grad_norm": 0.08419599384069443,
      "learning_rate": 1.7266077621535481e-06,
      "loss": 0.0159,
      "relation_loss": 0.0011076570372097194,
      "selected_pairs": 27127.3,
      "state_query_norm": 1112.541162109375,
      "state_relation_loss": 0.0002624940039822832,
      "step": 27610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011076570372097194,
      "weighted_relation_loss": 0.0011076570372097194
    },
    {
      "action_query_norm": 616.9364990234375,
      "action_relation_loss": 0.0008435177733190358,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.837231057062674,
      "flow_matching_loss": 0.014930600766092539,
      "grad_norm": 0.09249020367860794,
      "learning_rate": 1.7122782377282597e-06,
      "loss": 0.0161,
      "relation_loss": 0.0011218405794352292,
      "selected_pairs": 27129.8,
      "state_query_norm": 1098.7348022460938,
      "state_relation_loss": 0.00027832280611619355,
      "step": 27620,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011218405794352292,
      "weighted_relation_loss": 0.0011218405794352292
    },
    {
      "action_query_norm": 621.4532531738281,
      "action_relation_loss": 0.0007936693786177784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.846585594013096,
      "flow_matching_loss": 0.015312043204903602,
      "grad_norm": 0.09546992182731628,
      "learning_rate": 1.6980073874381497e-06,
      "loss": 0.0164,
      "relation_loss": 0.0010535013396292924,
      "selected_pairs": 26411.4,
      "state_query_norm": 1122.7774658203125,
      "state_relation_loss": 0.0002598319741082378,
      "step": 27630,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0010535013396292924,
      "weighted_relation_loss": 0.0010535013396292924
    },
    {
      "action_query_norm": 613.0180358886719,
      "action_relation_loss": 0.0008440967532806099,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.855940130963518,
      "flow_matching_loss": 0.01488339975476265,
      "grad_norm": 0.07721003890037537,
      "learning_rate": 1.6837952286236413e-06,
      "loss": 0.016,
      "relation_loss": 0.0011035108589567244,
      "selected_pairs": 26317.5,
      "state_query_norm": 1107.1462524414062,
      "state_relation_loss": 0.0002594140780274756,
      "step": 27640,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0011035108589567244,
      "weighted_relation_loss": 0.0011035108589567244
    },
    {
      "action_query_norm": 619.2357299804687,
      "action_relation_loss": 0.0007917814480606466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.86529466791394,
      "flow_matching_loss": 0.01370468745008111,
      "grad_norm": 0.07535801082849503,
      "learning_rate": 1.6696417785538488e-06,
      "loss": 0.0148,
      "relation_loss": 0.001056641194736585,
      "selected_pairs": 27108.6,
      "state_query_norm": 1107.5621215820313,
      "state_relation_loss": 0.00026485975104151294,
      "step": 27650,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001056641194736585,
      "weighted_relation_loss": 0.001056641194736585
    },
    {
      "action_query_norm": 617.5544982910156,
      "action_relation_loss": 0.0008395748503971845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.874649204864358,
      "flow_matching_loss": 0.01638067066669464,
      "grad_norm": 0.09453406184911728,
      "learning_rate": 1.6555470544265538e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011062570032663644,
      "selected_pairs": 26638.3,
      "state_query_norm": 1120.730419921875,
      "state_relation_loss": 0.00026668213977245616,
      "step": 27660,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0011062570032663644,
      "weighted_relation_loss": 0.0011062570032663644
    },
    {
      "action_query_norm": 622.1277221679687,
      "action_relation_loss": 0.0008480506774503737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.88400374181478,
      "flow_matching_loss": 0.013765308354049922,
      "grad_norm": 0.058434128761291504,
      "learning_rate": 1.6415110733681738e-06,
      "loss": 0.0149,
      "relation_loss": 0.001111062220297754,
      "selected_pairs": 26759.9,
      "state_query_norm": 1113.6951049804688,
      "state_relation_loss": 0.00026301155157852917,
      "step": 27670,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001111062220297754,
      "weighted_relation_loss": 0.001111062220297754
    },
    {
      "action_query_norm": 617.5506774902344,
      "action_relation_loss": 0.0007956790504977107,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.893358278765202,
      "flow_matching_loss": 0.013927221950143575,
      "grad_norm": 0.0853688046336174,
      "learning_rate": 1.6275338524337436e-06,
      "loss": 0.015,
      "relation_loss": 0.0010595428524538875,
      "selected_pairs": 27058.3,
      "state_query_norm": 1118.0086669921875,
      "state_relation_loss": 0.0002638637917698361,
      "step": 27680,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010595428524538875,
      "weighted_relation_loss": 0.0010595428524538875
    },
    {
      "action_query_norm": 622.4847839355468,
      "action_relation_loss": 0.0008341313165146858,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.90271281571562,
      "flow_matching_loss": 0.015336782205849886,
      "grad_norm": 0.10802653431892395,
      "learning_rate": 1.6136154086069056e-06,
      "loss": 0.0164,
      "relation_loss": 0.0010957812308333814,
      "selected_pairs": 26735.9,
      "state_query_norm": 1098.7010131835937,
      "state_relation_loss": 0.00026164991431869565,
      "step": 27690,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010957812308333814,
      "weighted_relation_loss": 0.0010957812308333814
    },
    {
      "action_query_norm": 618.6935607910157,
      "action_relation_loss": 0.0008386890403926373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.912067352666043,
      "flow_matching_loss": 0.014674908109009266,
      "grad_norm": 0.08977382630109787,
      "learning_rate": 1.5997557587998868e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010989815928041935,
      "selected_pairs": 26480.1,
      "state_query_norm": 1091.2582275390625,
      "state_relation_loss": 0.00026029254513559863,
      "step": 27700,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0010989815928041935,
      "weighted_relation_loss": 0.0010989815928041935
    },
    {
      "action_query_norm": 615.96689453125,
      "action_relation_loss": 0.0008323932765051722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.921421889616465,
      "flow_matching_loss": 0.01406297693029046,
      "grad_norm": 0.08081335574388504,
      "learning_rate": 1.5859549198534606e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010979764745570719,
      "selected_pairs": 27104.8,
      "state_query_norm": 1115.0630126953124,
      "state_relation_loss": 0.0002655832111486234,
      "step": 27710,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0010979764745570719,
      "weighted_relation_loss": 0.0010979764745570719
    },
    {
      "action_query_norm": 622.4857788085938,
      "action_relation_loss": 0.0008113352057989686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.930776426566887,
      "flow_matching_loss": 0.016318674292415382,
      "grad_norm": 0.08239955455064774,
      "learning_rate": 1.572212908536963e-06,
      "loss": 0.0174,
      "relation_loss": 0.001076224457938224,
      "selected_pairs": 26603.9,
      "state_query_norm": 1111.0654541015624,
      "state_relation_loss": 0.00026488925650483,
      "step": 27720,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001076224457938224,
      "weighted_relation_loss": 0.001076224457938224
    },
    {
      "action_query_norm": 613.2239562988282,
      "action_relation_loss": 0.0008354709832929074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.940130963517305,
      "flow_matching_loss": 0.016388624254614115,
      "grad_norm": 0.08935034275054932,
      "learning_rate": 1.5585297415482202e-06,
      "loss": 0.0175,
      "relation_loss": 0.0011052093119360506,
      "selected_pairs": 27251.4,
      "state_query_norm": 1102.048681640625,
      "state_relation_loss": 0.00026973832282237706,
      "step": 27730,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011052093119360506,
      "weighted_relation_loss": 0.0011052093119360506
    },
    {
      "action_query_norm": 619.1877563476562,
      "action_relation_loss": 0.0008423006860539317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.949485500467727,
      "flow_matching_loss": 0.014931980427354575,
      "grad_norm": 0.09367313235998154,
      "learning_rate": 1.5449054355135717e-06,
      "loss": 0.016,
      "relation_loss": 0.0011109643848612905,
      "selected_pairs": 27147.6,
      "state_query_norm": 1123.9060791015625,
      "state_relation_loss": 0.0002686636973521672,
      "step": 27740,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011109643848612905,
      "weighted_relation_loss": 0.0011109643848612905
    },
    {
      "action_query_norm": 620.3861022949219,
      "action_relation_loss": 0.0008311159850563854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.95884003741815,
      "flow_matching_loss": 0.013889508787542582,
      "grad_norm": 0.09201688319444656,
      "learning_rate": 1.5313400069878415e-06,
      "loss": 0.015,
      "relation_loss": 0.0010979400132782758,
      "selected_pairs": 26940.9,
      "state_query_norm": 1112.5767578125,
      "state_relation_loss": 0.00026682401803554965,
      "step": 27750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010979400132782758,
      "weighted_relation_loss": 0.0010979400132782758
    },
    {
      "action_query_norm": 617.2739929199219,
      "action_relation_loss": 0.0008281812944915146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.968194574368567,
      "flow_matching_loss": 0.014554133266210556,
      "grad_norm": 0.09201418608427048,
      "learning_rate": 1.5178334724542887e-06,
      "loss": 0.0156,
      "relation_loss": 0.0010865498741623014,
      "selected_pairs": 26800.9,
      "state_query_norm": 1117.5578491210938,
      "state_relation_loss": 0.0002583685694844462,
      "step": 27760,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010865498741623014,
      "weighted_relation_loss": 0.0010865498741623014
    },
    {
      "action_query_norm": 619.3222045898438,
      "action_relation_loss": 0.0008566891076043249,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.97754911131899,
      "flow_matching_loss": 0.014654433261603117,
      "grad_norm": 0.08747142553329468,
      "learning_rate": 1.504385848324641e-06,
      "loss": 0.0158,
      "relation_loss": 0.001120999432168901,
      "selected_pairs": 26500.4,
      "state_query_norm": 1114.4439086914062,
      "state_relation_loss": 0.0002643103172886185,
      "step": 27770,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.001120999432168901,
      "weighted_relation_loss": 0.001120999432168901
    },
    {
      "action_query_norm": 613.1039916992188,
      "action_relation_loss": 0.0008600504253990949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.98690364826941,
      "flow_matching_loss": 0.01521929819136858,
      "grad_norm": 0.09356461465358734,
      "learning_rate": 1.490997150939033e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011240999680012464,
      "selected_pairs": 26393.8,
      "state_query_norm": 1109.107568359375,
      "state_relation_loss": 0.0002640495411469601,
      "step": 27780,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0011240999680012464,
      "weighted_relation_loss": 0.0011240999680012464
    },
    {
      "action_query_norm": 621.3279602050782,
      "action_relation_loss": 0.0008486334001645446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.996258185219833,
      "flow_matching_loss": 0.015714877005666494,
      "grad_norm": 0.12061353027820587,
      "learning_rate": 1.4776673965659793e-06,
      "loss": 0.0168,
      "relation_loss": 0.0011081115226261318,
      "selected_pairs": 26226.8,
      "state_query_norm": 1112.5331420898438,
      "state_relation_loss": 0.00025947811373043804,
      "step": 27790,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0011081115226261318,
      "weighted_relation_loss": 0.0011081115226261318
    },
    {
      "action_query_norm": 620.8456787109375,
      "action_relation_loss": 0.0007854826690163463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.00561272217025,
      "flow_matching_loss": 0.0146408268250525,
      "grad_norm": 0.08626297861337662,
      "learning_rate": 1.4643966014023957e-06,
      "loss": 0.0157,
      "relation_loss": 0.0010426613793242724,
      "selected_pairs": 23525.5,
      "state_query_norm": 1107.4719360351562,
      "state_relation_loss": 0.00025717871612869204,
      "step": 27800,
      "valid_action_samples": 210.5,
      "weighted_auxiliary_loss": 0.0010426613793242724,
      "weighted_relation_loss": 0.0010426613793242724
    },
    {
      "action_query_norm": 619.2338989257812,
      "action_relation_loss": 0.0008105967193841934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.014967259120674,
      "flow_matching_loss": 0.014750478230416775,
      "grad_norm": 0.07006634026765823,
      "learning_rate": 1.4511847815735503e-06,
      "loss": 0.0158,
      "relation_loss": 0.0010907263378612698,
      "selected_pairs": 26433.4,
      "state_query_norm": 1102.6039184570313,
      "state_relation_loss": 0.0002801296184770763,
      "step": 27810,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0010907263378612698,
      "weighted_relation_loss": 0.0010907263378612698
    },
    {
      "action_query_norm": 619.6925598144531,
      "action_relation_loss": 0.0008122584433294833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.024321796071096,
      "flow_matching_loss": 0.013351305294781924,
      "grad_norm": 0.06851544231176376,
      "learning_rate": 1.438031953133051e-06,
      "loss": 0.0144,
      "relation_loss": 0.0010799459414556623,
      "selected_pairs": 26899.3,
      "state_query_norm": 1103.5611450195313,
      "state_relation_loss": 0.0002676874937606044,
      "step": 27820,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010799459414556623,
      "weighted_relation_loss": 0.0010799459414556623
    },
    {
      "action_query_norm": 617.2293395996094,
      "action_relation_loss": 0.0008519443450495601,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.033676333021514,
      "flow_matching_loss": 0.01375562846660614,
      "grad_norm": 0.11388770490884781,
      "learning_rate": 1.4249381320628197e-06,
      "loss": 0.0149,
      "relation_loss": 0.0011171958525665105,
      "selected_pairs": 26736.4,
      "state_query_norm": 1118.3329345703125,
      "state_relation_loss": 0.00026525150169618426,
      "step": 27830,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011171958525665105,
      "weighted_relation_loss": 0.0011171958525665105
    },
    {
      "action_query_norm": 616.4401245117188,
      "action_relation_loss": 0.0008297039719764143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.043030869971936,
      "flow_matching_loss": 0.015015192702412606,
      "grad_norm": 0.10342344641685486,
      "learning_rate": 1.4119033342730902e-06,
      "loss": 0.0161,
      "relation_loss": 0.0010872974991798401,
      "selected_pairs": 26393.2,
      "state_query_norm": 1112.1565307617188,
      "state_relation_loss": 0.00025759353011380883,
      "step": 27840,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0010872974991798401,
      "weighted_relation_loss": 0.0010872974991798401
    },
    {
      "action_query_norm": 623.1960266113281,
      "action_relation_loss": 0.0008253420237451792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.052385406922358,
      "flow_matching_loss": 0.014388137869536877,
      "grad_norm": 0.07613803446292877,
      "learning_rate": 1.3989275756023656e-06,
      "loss": 0.0155,
      "relation_loss": 0.0010865693213418127,
      "selected_pairs": 26730.6,
      "state_query_norm": 1114.862939453125,
      "state_relation_loss": 0.00026122729323105884,
      "step": 27850,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010865693213418127,
      "weighted_relation_loss": 0.0010865693213418127
    },
    {
      "action_query_norm": 618.2064758300781,
      "action_relation_loss": 0.0007936235924717039,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.061739943872777,
      "flow_matching_loss": 0.014128872659057379,
      "grad_norm": 0.08635448664426804,
      "learning_rate": 1.386010871817417e-06,
      "loss": 0.0152,
      "relation_loss": 0.001058744709007442,
      "selected_pairs": 26758.3,
      "state_query_norm": 1101.524365234375,
      "state_relation_loss": 0.000265121110714972,
      "step": 27860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001058744709007442,
      "weighted_relation_loss": 0.001058744709007442
    },
    {
      "action_query_norm": 611.5615112304688,
      "action_relation_loss": 0.0008154962328262627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.0710944808232,
      "flow_matching_loss": 0.01365568144246936,
      "grad_norm": 0.08396444469690323,
      "learning_rate": 1.3731532386132616e-06,
      "loss": 0.0147,
      "relation_loss": 0.0010733727714978159,
      "selected_pairs": 27431.0,
      "state_query_norm": 1106.5683471679688,
      "state_relation_loss": 0.00025787654594751076,
      "step": 27870,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0010733727714978159,
      "weighted_relation_loss": 0.0010733727714978159
    },
    {
      "action_query_norm": 621.7972351074219,
      "action_relation_loss": 0.0008309653669130058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.08044901777362,
      "flow_matching_loss": 0.016042363829910755,
      "grad_norm": 0.0851588174700737,
      "learning_rate": 1.360354691613136e-06,
      "loss": 0.0171,
      "relation_loss": 0.0010972333024255932,
      "selected_pairs": 26596.1,
      "state_query_norm": 1115.282470703125,
      "state_relation_loss": 0.00026626793696777896,
      "step": 27880,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0010972333024255932,
      "weighted_relation_loss": 0.0010972333024255932
    },
    {
      "action_query_norm": 614.7834899902343,
      "action_relation_loss": 0.0008413741714321076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.089803554724043,
      "flow_matching_loss": 0.013620984368026256,
      "grad_norm": 0.11241939663887024,
      "learning_rate": 1.3476152463684776e-06,
      "loss": 0.0147,
      "relation_loss": 0.0011109239188954235,
      "selected_pairs": 27646.8,
      "state_query_norm": 1100.2457153320313,
      "state_relation_loss": 0.000269549764925614,
      "step": 27890,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0011109239188954235,
      "weighted_relation_loss": 0.0011109239188954235
    },
    {
      "action_query_norm": 613.7582641601563,
      "action_relation_loss": 0.0009073574794456363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.09915809167446,
      "flow_matching_loss": 0.015007703099399805,
      "grad_norm": 0.10848725587129593,
      "learning_rate": 1.3349349183589155e-06,
      "loss": 0.0162,
      "relation_loss": 0.0011657133349217475,
      "selected_pairs": 25914.5,
      "state_query_norm": 1106.7827514648438,
      "state_relation_loss": 0.000258355877303984,
      "step": 27900,
      "valid_action_samples": 228.1,
      "weighted_auxiliary_loss": 0.0011657133349217475,
      "weighted_relation_loss": 0.0011657133349217475
    },
    {
      "action_query_norm": 618.5631225585937,
      "action_relation_loss": 0.000823121314169839,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.108512628624883,
      "flow_matching_loss": 0.015253903437405824,
      "grad_norm": 0.07809542119503021,
      "learning_rate": 1.3223137229922356e-06,
      "loss": 0.0163,
      "relation_loss": 0.001080730208195746,
      "selected_pairs": 26860.2,
      "state_query_norm": 1116.6366577148438,
      "state_relation_loss": 0.00025760889693629,
      "step": 27910,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001080730208195746,
      "weighted_relation_loss": 0.001080730208195746
    },
    {
      "action_query_norm": 618.2179748535157,
      "action_relation_loss": 0.0008505380363203585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.117867165575305,
      "flow_matching_loss": 0.015196362975984812,
      "grad_norm": 0.08710870891809464,
      "learning_rate": 1.3097516756043981e-06,
      "loss": 0.0163,
      "relation_loss": 0.0011128497426398098,
      "selected_pairs": 27218.2,
      "state_query_norm": 1111.348974609375,
      "state_relation_loss": 0.00026231170777464287,
      "step": 27920,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011128497426398098,
      "weighted_relation_loss": 0.0011128497426398098
    },
    {
      "action_query_norm": 619.3208068847656,
      "action_relation_loss": 0.000820442201802507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.127221702525723,
      "flow_matching_loss": 0.014076055586338043,
      "grad_norm": 0.08774235099554062,
      "learning_rate": 1.29724879145946e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010915322112850846,
      "selected_pairs": 27204.3,
      "state_query_norm": 1101.5037475585937,
      "state_relation_loss": 0.0002710900123929605,
      "step": 27930,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0010915322112850846,
      "weighted_relation_loss": 0.0010915322112850846
    },
    {
      "action_query_norm": 615.320703125,
      "action_relation_loss": 0.0008286404307000339,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.136576239476145,
      "flow_matching_loss": 0.014143818709999322,
      "grad_norm": 0.09139693528413773,
      "learning_rate": 1.284805085749613e-06,
      "loss": 0.0152,
      "relation_loss": 0.0010821962379850448,
      "selected_pairs": 26619.8,
      "state_query_norm": 1111.6318359375,
      "state_relation_loss": 0.00025355581601615995,
      "step": 27940,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010821962379850448,
      "weighted_relation_loss": 0.0010821962379850448
    },
    {
      "action_query_norm": 620.3093322753906,
      "action_relation_loss": 0.0008340247615706176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.145930776426567,
      "flow_matching_loss": 0.014425010699778796,
      "grad_norm": 0.08122832328081131,
      "learning_rate": 1.2724205735951288e-06,
      "loss": 0.0155,
      "relation_loss": 0.0010862575145438313,
      "selected_pairs": 26919.6,
      "state_query_norm": 1097.5284912109375,
      "state_relation_loss": 0.0002522327413316816,
      "step": 27950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010862575145438313,
      "weighted_relation_loss": 0.0010862575145438313
    },
    {
      "action_query_norm": 609.6593688964844,
      "action_relation_loss": 0.0008407517976593226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.15528531337699,
      "flow_matching_loss": 0.014037551637738944,
      "grad_norm": 0.07002955675125122,
      "learning_rate": 1.2600952700443591e-06,
      "loss": 0.0151,
      "relation_loss": 0.0010997202945873142,
      "selected_pairs": 26902.0,
      "state_query_norm": 1121.3525634765624,
      "state_relation_loss": 0.0002589684954728,
      "step": 27960,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010997202945873142,
      "weighted_relation_loss": 0.0010997202945873142
    },
    {
      "action_query_norm": 623.705224609375,
      "action_relation_loss": 0.00080946123925969,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.164639850327408,
      "flow_matching_loss": 0.013487123604863881,
      "grad_norm": 0.10788053274154663,
      "learning_rate": 1.247829190073707e-06,
      "loss": 0.0146,
      "relation_loss": 0.001068321568891406,
      "selected_pairs": 26850.0,
      "state_query_norm": 1111.9467895507812,
      "state_relation_loss": 0.00025886033836286515,
      "step": 27970,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001068321568891406,
      "weighted_relation_loss": 0.001068321568891406
    },
    {
      "action_query_norm": 622.8070251464844,
      "action_relation_loss": 0.0008149009721819311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.17399438727783,
      "flow_matching_loss": 0.013869991898536682,
      "grad_norm": 0.09047145396471024,
      "learning_rate": 1.2356223485876173e-06,
      "loss": 0.0149,
      "relation_loss": 0.001074370334390551,
      "selected_pairs": 26825.6,
      "state_query_norm": 1103.0904418945313,
      "state_relation_loss": 0.00025946936511900276,
      "step": 27980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001074370334390551,
      "weighted_relation_loss": 0.001074370334390551
    },
    {
      "action_query_norm": 619.5995971679688,
      "action_relation_loss": 0.0008335238322615624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.18334892422825,
      "flow_matching_loss": 0.013843294512480498,
      "grad_norm": 0.0793631374835968,
      "learning_rate": 1.2234747604185526e-06,
      "loss": 0.0149,
      "relation_loss": 0.0010896477149799466,
      "selected_pairs": 26924.3,
      "state_query_norm": 1106.8597534179687,
      "state_relation_loss": 0.00025612388999434187,
      "step": 27990,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010896477149799466,
      "weighted_relation_loss": 0.0010896477149799466
    },
    {
      "action_query_norm": 622.1828918457031,
      "action_relation_loss": 0.0008151115966029465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.19270346117867,
      "flow_matching_loss": 0.014442124031484127,
      "grad_norm": 0.08752003312110901,
      "learning_rate": 1.2113864403269836e-06,
      "loss": 0.0155,
      "relation_loss": 0.001069086603820324,
      "selected_pairs": 26739.9,
      "state_query_norm": 1116.3753540039063,
      "state_relation_loss": 0.0002539750057621859,
      "step": 28000,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001069086603820324,
      "weighted_relation_loss": 0.001069086603820324
    },
    {
      "action_query_norm": 617.3007568359375,
      "action_relation_loss": 0.0008187926257960498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.202057998129092,
      "flow_matching_loss": 0.015377395506948232,
      "grad_norm": 0.08392056822776794,
      "learning_rate": 1.1993574030013555e-06,
      "loss": 0.0164,
      "relation_loss": 0.0010714900447055698,
      "selected_pairs": 26754.3,
      "state_query_norm": 1110.792919921875,
      "state_relation_loss": 0.00025269742036471143,
      "step": 28010,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010714900447055698,
      "weighted_relation_loss": 0.0010714900447055698
    },
    {
      "action_query_norm": 615.7800598144531,
      "action_relation_loss": 0.0008441648853477091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.211412535079514,
      "flow_matching_loss": 0.01464685732498765,
      "grad_norm": 0.10225705802440643,
      "learning_rate": 1.1873876630580816e-06,
      "loss": 0.0158,
      "relation_loss": 0.0011101441108621658,
      "selected_pairs": 26880.9,
      "state_query_norm": 1103.3902709960937,
      "state_relation_loss": 0.00026597923133522274,
      "step": 28020,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011101441108621658,
      "weighted_relation_loss": 0.0011101441108621658
    },
    {
      "action_query_norm": 617.0663208007812,
      "action_relation_loss": 0.0008630492608062923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.220767072029936,
      "flow_matching_loss": 0.01559997834265232,
      "grad_norm": 0.08903438597917557,
      "learning_rate": 1.1754772350415278e-06,
      "loss": 0.0167,
      "relation_loss": 0.0011346590938046575,
      "selected_pairs": 27549.8,
      "state_query_norm": 1081.165771484375,
      "state_relation_loss": 0.0002716098417295143,
      "step": 28030,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0011346590938046575,
      "weighted_relation_loss": 0.0011346590938046575
    },
    {
      "action_query_norm": 618.6531188964843,
      "action_relation_loss": 0.0008108685491606593,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.230121608980355,
      "flow_matching_loss": 0.01490279771387577,
      "grad_norm": 0.07778288424015045,
      "learning_rate": 1.163626133423984e-06,
      "loss": 0.016,
      "relation_loss": 0.001084599772002548,
      "selected_pairs": 26825.9,
      "state_query_norm": 1117.684619140625,
      "state_relation_loss": 0.00027373122284188864,
      "step": 28040,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001084599772002548,
      "weighted_relation_loss": 0.001084599772002548
    },
    {
      "action_query_norm": 619.8964660644531,
      "action_relation_loss": 0.0008433902054093778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.239476145930777,
      "flow_matching_loss": 0.01418263502418995,
      "grad_norm": 0.10866653919219971,
      "learning_rate": 1.1518343726056591e-06,
      "loss": 0.0153,
      "relation_loss": 0.0011013361508958042,
      "selected_pairs": 26512.2,
      "state_query_norm": 1103.6179565429688,
      "state_relation_loss": 0.0002579459411208518,
      "step": 28050,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011013361508958042,
      "weighted_relation_loss": 0.0011013361508958042
    },
    {
      "action_query_norm": 621.0442504882812,
      "action_relation_loss": 0.0008106855791993439,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.2488306828812,
      "flow_matching_loss": 0.013888614904135465,
      "grad_norm": 0.07659295946359634,
      "learning_rate": 1.1401019669146474e-06,
      "loss": 0.015,
      "relation_loss": 0.0010826391167938708,
      "selected_pairs": 27363.8,
      "state_query_norm": 1115.9642456054687,
      "state_relation_loss": 0.00027195355214644225,
      "step": 28060,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0010826391167938708,
      "weighted_relation_loss": 0.0010826391167938708
    },
    {
      "action_query_norm": 624.6216064453125,
      "action_relation_loss": 0.0008444628969300538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.258185219831617,
      "flow_matching_loss": 0.01389394998550415,
      "grad_norm": 0.07310321927070618,
      "learning_rate": 1.128428930606934e-06,
      "loss": 0.015,
      "relation_loss": 0.0010973321041092277,
      "selected_pairs": 27109.1,
      "state_query_norm": 1119.0524536132812,
      "state_relation_loss": 0.0002528692013584077,
      "step": 28070,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0010973321041092277,
      "weighted_relation_loss": 0.0010973321041092277
    },
    {
      "action_query_norm": 620.0874267578125,
      "action_relation_loss": 0.0008188629581127316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.26753975678204,
      "flow_matching_loss": 0.013604473229497671,
      "grad_norm": 0.07628781348466873,
      "learning_rate": 1.1168152778663621e-06,
      "loss": 0.0147,
      "relation_loss": 0.001086292660329491,
      "selected_pairs": 27081.4,
      "state_query_norm": 1101.8138549804687,
      "state_relation_loss": 0.00026742970512714235,
      "step": 28080,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001086292660329491,
      "weighted_relation_loss": 0.001086292660329491
    },
    {
      "action_query_norm": 617.9654174804688,
      "action_relation_loss": 0.0008068605093285441,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.27689429373246,
      "flow_matching_loss": 0.014441251847893,
      "grad_norm": 0.10681509971618652,
      "learning_rate": 1.10526102280461e-06,
      "loss": 0.0155,
      "relation_loss": 0.001058465102687478,
      "selected_pairs": 27477.3,
      "state_query_norm": 1107.0323364257813,
      "state_relation_loss": 0.0002516045919037424,
      "step": 28090,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.001058465102687478,
      "weighted_relation_loss": 0.001058465102687478
    },
    {
      "action_query_norm": 620.0762512207032,
      "action_relation_loss": 0.0008115423494018615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.28624883068288,
      "flow_matching_loss": 0.013388293981552123,
      "grad_norm": 0.06770249456167221,
      "learning_rate": 1.0937661794611864e-06,
      "loss": 0.0145,
      "relation_loss": 0.001076372223906219,
      "selected_pairs": 27060.9,
      "state_query_norm": 1098.5898193359376,
      "state_relation_loss": 0.0002648298803251237,
      "step": 28100,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001076372223906219,
      "weighted_relation_loss": 0.001076372223906219
    },
    {
      "action_query_norm": 613.2794982910157,
      "action_relation_loss": 0.0008196057984605432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.2956033676333,
      "flow_matching_loss": 0.013621477689594031,
      "grad_norm": 0.0733429342508316,
      "learning_rate": 1.082330761803413e-06,
      "loss": 0.0147,
      "relation_loss": 0.0010864209732972085,
      "selected_pairs": 26294.4,
      "state_query_norm": 1098.0584228515625,
      "state_relation_loss": 0.0002668151777470484,
      "step": 28110,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0010864209732972085,
      "weighted_relation_loss": 0.0010864209732972085
    },
    {
      "action_query_norm": 620.594287109375,
      "action_relation_loss": 0.0008037900202907622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.304957904583723,
      "flow_matching_loss": 0.013219838775694371,
      "grad_norm": 0.07944127917289734,
      "learning_rate": 1.0709547837263966e-06,
      "loss": 0.0143,
      "relation_loss": 0.0010709712048992514,
      "selected_pairs": 26966.2,
      "state_query_norm": 1107.282177734375,
      "state_relation_loss": 0.00026718118460848926,
      "step": 28120,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010709712048992514,
      "weighted_relation_loss": 0.0010709712048992514
    },
    {
      "action_query_norm": 620.8877258300781,
      "action_relation_loss": 0.0008652376418467611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.314312441534145,
      "flow_matching_loss": 0.013354127202183009,
      "grad_norm": 0.08821423351764679,
      "learning_rate": 1.059638259053025e-06,
      "loss": 0.0145,
      "relation_loss": 0.0011346507933922112,
      "selected_pairs": 27038.1,
      "state_query_norm": 1120.3664306640626,
      "state_relation_loss": 0.0002694131719181314,
      "step": 28130,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011346507933922112,
      "weighted_relation_loss": 0.0011346507933922112
    },
    {
      "action_query_norm": 622.1157104492188,
      "action_relation_loss": 0.0008117220713756979,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.323666978484564,
      "flow_matching_loss": 0.013964398577809334,
      "grad_norm": 0.0865887999534607,
      "learning_rate": 1.0483812015339434e-06,
      "loss": 0.015,
      "relation_loss": 0.0010666439950000495,
      "selected_pairs": 26833.2,
      "state_query_norm": 1104.4935302734375,
      "state_relation_loss": 0.0002549219105276279,
      "step": 28140,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010666439950000495,
      "weighted_relation_loss": 0.0010666439950000495
    },
    {
      "action_query_norm": 616.1634155273438,
      "action_relation_loss": 0.0008386651636101306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.333021515434986,
      "flow_matching_loss": 0.013318336848169566,
      "grad_norm": 0.08140810579061508,
      "learning_rate": 1.037183624847543e-06,
      "loss": 0.0144,
      "relation_loss": 0.0011000246624462307,
      "selected_pairs": 27104.8,
      "state_query_norm": 1118.6147583007812,
      "state_relation_loss": 0.00026135949592571707,
      "step": 28150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011000246624462307,
      "weighted_relation_loss": 0.0011000246624462307
    },
    {
      "action_query_norm": 622.3636474609375,
      "action_relation_loss": 0.0008162366168107838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.342376052385408,
      "flow_matching_loss": 0.014590496383607387,
      "grad_norm": 0.0833163633942604,
      "learning_rate": 1.0260455425999238e-06,
      "loss": 0.0157,
      "relation_loss": 0.0010941875050775706,
      "selected_pairs": 26944.4,
      "state_query_norm": 1108.4346923828125,
      "state_relation_loss": 0.00027795089117717,
      "step": 28160,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010941875050775706,
      "weighted_relation_loss": 0.0010941875050775706
    },
    {
      "action_query_norm": 615.4891052246094,
      "action_relation_loss": 0.0008651573793031275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.351730589335826,
      "flow_matching_loss": 0.014136195089668036,
      "grad_norm": 0.08709968626499176,
      "learning_rate": 1.0149669683249208e-06,
      "loss": 0.0153,
      "relation_loss": 0.0011252596857957542,
      "selected_pairs": 27103.4,
      "state_query_norm": 1106.0159545898437,
      "state_relation_loss": 0.0002601023035822436,
      "step": 28170,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0011252596857957542,
      "weighted_relation_loss": 0.0011252596857957542
    },
    {
      "action_query_norm": 613.9982360839844,
      "action_relation_loss": 0.0008562220900785178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.36108512628625,
      "flow_matching_loss": 0.012859028298407793,
      "grad_norm": 0.0652908906340599,
      "learning_rate": 1.003947915484027e-06,
      "loss": 0.014,
      "relation_loss": 0.001122667419258505,
      "selected_pairs": 26802.4,
      "state_query_norm": 1099.1128173828124,
      "state_relation_loss": 0.0002664453480974771,
      "step": 28180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001122667419258505,
      "weighted_relation_loss": 0.001122667419258505
    },
    {
      "action_query_norm": 618.5929382324218,
      "action_relation_loss": 0.000802327023120597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.37043966323667,
      "flow_matching_loss": 0.014044733997434378,
      "grad_norm": 0.0828499123454094,
      "learning_rate": 9.92988397466449e-07,
      "loss": 0.0151,
      "relation_loss": 0.0010722290608100593,
      "selected_pairs": 27061.0,
      "state_query_norm": 1098.101171875,
      "state_relation_loss": 0.0002699020347790793,
      "step": 28190,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010722290608100593,
      "weighted_relation_loss": 0.0010722290608100593
    },
    {
      "action_query_norm": 613.9885375976562,
      "action_relation_loss": 0.0008372439362574369,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.379794200187092,
      "flow_matching_loss": 0.013912097737193107,
      "grad_norm": 0.09400152415037155,
      "learning_rate": 9.820884275890286e-07,
      "loss": 0.015,
      "relation_loss": 0.0010965099558234216,
      "selected_pairs": 26301.3,
      "state_query_norm": 1110.6424682617187,
      "state_relation_loss": 0.00025926603557309137,
      "step": 28200,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0010965099558234216,
      "weighted_relation_loss": 0.0010965099558234216
    },
    {
      "action_query_norm": 620.8074401855469,
      "action_relation_loss": 0.000830989790847525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.38914873713751,
      "flow_matching_loss": 0.015683515463024378,
      "grad_norm": 0.09185531735420227,
      "learning_rate": 9.712480190962546e-07,
      "loss": 0.0168,
      "relation_loss": 0.001097182370722294,
      "selected_pairs": 26941.7,
      "state_query_norm": 1102.4155395507812,
      "state_relation_loss": 0.0002661925871507265,
      "step": 28210,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001097182370722294,
      "weighted_relation_loss": 0.001097182370722294
    },
    {
      "action_query_norm": 616.8121276855469,
      "action_relation_loss": 0.0009073663852177561,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.398503274087933,
      "flow_matching_loss": 0.014251540135592222,
      "grad_norm": 0.08807871490716934,
      "learning_rate": 9.604671851602464e-07,
      "loss": 0.0154,
      "relation_loss": 0.0011636223527602852,
      "selected_pairs": 26439.2,
      "state_query_norm": 1098.9312255859375,
      "state_relation_loss": 0.00025625596899772065,
      "step": 28220,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011636223527602852,
      "weighted_relation_loss": 0.0011636223527602852
    },
    {
      "action_query_norm": 624.0466369628906,
      "action_relation_loss": 0.000805339787621051,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.407857811038355,
      "flow_matching_loss": 0.013121209386736155,
      "grad_norm": 0.08540171384811401,
      "learning_rate": 9.497459388807306e-07,
      "loss": 0.0142,
      "relation_loss": 0.0010718947160057724,
      "selected_pairs": 26755.4,
      "state_query_norm": 1116.2791015625,
      "state_relation_loss": 0.00026655491237761455,
      "step": 28230,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010718947160057724,
      "weighted_relation_loss": 0.0010718947160057724
    },
    {
      "action_query_norm": 622.359130859375,
      "action_relation_loss": 0.0008199107367545366,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.417212347988773,
      "flow_matching_loss": 0.015012998227030038,
      "grad_norm": 0.08766328543424606,
      "learning_rate": 9.390842932850364e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010858259978704155,
      "selected_pairs": 26965.5,
      "state_query_norm": 1112.6409912109375,
      "state_relation_loss": 0.00026591524947434665,
      "step": 28240,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010858259978704155,
      "weighted_relation_loss": 0.0010858259978704155
    },
    {
      "action_query_norm": 622.0974426269531,
      "action_relation_loss": 0.0008611588214989752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.426566884939195,
      "flow_matching_loss": 0.012995590548962354,
      "grad_norm": 0.09605240821838379,
      "learning_rate": 9.28482261328073e-07,
      "loss": 0.0141,
      "relation_loss": 0.0011186421732418239,
      "selected_pairs": 26763.9,
      "state_query_norm": 1102.5314208984375,
      "state_relation_loss": 0.0002574833531980403,
      "step": 28250,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011186421732418239,
      "weighted_relation_loss": 0.0011186421732418239
    },
    {
      "action_query_norm": 618.5344116210938,
      "action_relation_loss": 0.0008117280318401754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.435921421889617,
      "flow_matching_loss": 0.014607080817222595,
      "grad_norm": 0.06997402757406235,
      "learning_rate": 9.179398558923025e-07,
      "loss": 0.0157,
      "relation_loss": 0.0010729207249823957,
      "selected_pairs": 26596.3,
      "state_query_norm": 1109.673095703125,
      "state_relation_loss": 0.00026119269314222037,
      "step": 28260,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0010729207249823957,
      "weighted_relation_loss": 0.0010729207249823957
    },
    {
      "action_query_norm": 615.9343688964843,
      "action_relation_loss": 0.0008273438143078238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.44527595884004,
      "flow_matching_loss": 0.014763747993856668,
      "grad_norm": 0.11539750546216965,
      "learning_rate": 9.074570897877388e-07,
      "loss": 0.0159,
      "relation_loss": 0.001093745301477611,
      "selected_pairs": 26827.6,
      "state_query_norm": 1105.1882080078126,
      "state_relation_loss": 0.0002664015191840008,
      "step": 28270,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001093745301477611,
      "weighted_relation_loss": 0.001093745301477611
    },
    {
      "action_query_norm": 619.6181884765625,
      "action_relation_loss": 0.0008446455583907663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.454630495790457,
      "flow_matching_loss": 0.013810763880610467,
      "grad_norm": 0.07545831054449081,
      "learning_rate": 8.970339757519375e-07,
      "loss": 0.0149,
      "relation_loss": 0.001113811496179551,
      "selected_pairs": 26453.5,
      "state_query_norm": 1105.2320434570313,
      "state_relation_loss": 0.00026916594069916757,
      "step": 28280,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001113811496179551,
      "weighted_relation_loss": 0.001113811496179551
    },
    {
      "action_query_norm": 618.5907592773438,
      "action_relation_loss": 0.0008337522449437529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.46398503274088,
      "flow_matching_loss": 0.013362241256982089,
      "grad_norm": 0.07521916925907135,
      "learning_rate": 8.866705264499619e-07,
      "loss": 0.0145,
      "relation_loss": 0.0010998738696798682,
      "selected_pairs": 27339.7,
      "state_query_norm": 1101.6943603515624,
      "state_relation_loss": 0.00026612161746015773,
      "step": 28290,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010998738696798682,
      "weighted_relation_loss": 0.0010998738696798682
    },
    {
      "action_query_norm": 621.5918823242188,
      "action_relation_loss": 0.0008251936931628733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.4733395696913,
      "flow_matching_loss": 0.014179187174886466,
      "grad_norm": 0.10893958806991577,
      "learning_rate": 8.763667544743836e-07,
      "loss": 0.0153,
      "relation_loss": 0.0010894511360675097,
      "selected_pairs": 26894.4,
      "state_query_norm": 1112.864111328125,
      "state_relation_loss": 0.0002642574574565515,
      "step": 28300,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010894511360675097,
      "weighted_relation_loss": 0.0010894511360675097
    },
    {
      "action_query_norm": 617.3046569824219,
      "action_relation_loss": 0.0008158734708558768,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.48269410664172,
      "flow_matching_loss": 0.0168392782099545,
      "grad_norm": 0.0851941704750061,
      "learning_rate": 8.661226723452543e-07,
      "loss": 0.0179,
      "relation_loss": 0.0010872928658500313,
      "selected_pairs": 27079.4,
      "state_query_norm": 1111.6618896484374,
      "state_relation_loss": 0.00027141938917338846,
      "step": 28310,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010872928658500313,
      "weighted_relation_loss": 0.0010872928658500313
    },
    {
      "action_query_norm": 615.459130859375,
      "action_relation_loss": 0.0008197052229661494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.492048643592142,
      "flow_matching_loss": 0.013992969039827585,
      "grad_norm": 0.11236134171485901,
      "learning_rate": 8.559382925101001e-07,
      "loss": 0.0151,
      "relation_loss": 0.0010839900933206081,
      "selected_pairs": 26578.0,
      "state_query_norm": 1124.3814086914062,
      "state_relation_loss": 0.0002642848732648417,
      "step": 28320,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0010839900933206081,
      "weighted_relation_loss": 0.0010839900933206081
    },
    {
      "action_query_norm": 616.0666931152343,
      "action_relation_loss": 0.0008266468590591103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.501403180542564,
      "flow_matching_loss": 0.013127172831445932,
      "grad_norm": 0.059721093624830246,
      "learning_rate": 8.458136273438943e-07,
      "loss": 0.0142,
      "relation_loss": 0.0010920630185864866,
      "selected_pairs": 27505.6,
      "state_query_norm": 1095.6564453125,
      "state_relation_loss": 0.0002654161464306526,
      "step": 28330,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0010920630185864866,
      "weighted_relation_loss": 0.0010920630185864866
    },
    {
      "action_query_norm": 617.8614379882813,
      "action_relation_loss": 0.0008500534400809556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.510757717492986,
      "flow_matching_loss": 0.014322496391832829,
      "grad_norm": 0.09896283596754074,
      "learning_rate": 8.357486891490795e-07,
      "loss": 0.0154,
      "relation_loss": 0.0011114864377304912,
      "selected_pairs": 26677.5,
      "state_query_norm": 1111.9490844726563,
      "state_relation_loss": 0.0002614329830976203,
      "step": 28340,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011114864377304912,
      "weighted_relation_loss": 0.0011114864377304912
    },
    {
      "action_query_norm": 612.5882568359375,
      "action_relation_loss": 0.0009094934328459203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.520112254443404,
      "flow_matching_loss": 0.014066546130925417,
      "grad_norm": 0.09422695636749268,
      "learning_rate": 8.257434901554895e-07,
      "loss": 0.0152,
      "relation_loss": 0.0011715466855093837,
      "selected_pairs": 26203.8,
      "state_query_norm": 1104.6031494140625,
      "state_relation_loss": 0.0002620532453875057,
      "step": 28350,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0011715466855093837,
      "weighted_relation_loss": 0.0011715466855093837
    },
    {
      "action_query_norm": 627.4719970703125,
      "action_relation_loss": 0.0008198141353204846,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.529466791393826,
      "flow_matching_loss": 0.014676261413842439,
      "grad_norm": 0.0829429179430008,
      "learning_rate": 8.157980425203937e-07,
      "loss": 0.0158,
      "relation_loss": 0.0010820347350090742,
      "selected_pairs": 27811.4,
      "state_query_norm": 1109.0951171875,
      "state_relation_loss": 0.0002622205967782065,
      "step": 28360,
      "valid_action_samples": 236.3,
      "weighted_auxiliary_loss": 0.0010820347350090742,
      "weighted_relation_loss": 0.0010820347350090742
    },
    {
      "action_query_norm": 618.6598876953125,
      "action_relation_loss": 0.000810435542371124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.53882132834425,
      "flow_matching_loss": 0.014389482699334622,
      "grad_norm": 0.08632789552211761,
      "learning_rate": 8.059123583284367e-07,
      "loss": 0.0155,
      "relation_loss": 0.0010829498991370202,
      "selected_pairs": 26949.1,
      "state_query_norm": 1101.7872314453125,
      "state_relation_loss": 0.0002725143582210876,
      "step": 28370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010829498991370202,
      "weighted_relation_loss": 0.0010829498991370202
    },
    {
      "action_query_norm": 617.9216552734375,
      "action_relation_loss": 0.000850708014331758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.548175865294667,
      "flow_matching_loss": 0.015906106401234864,
      "grad_norm": 0.07823330163955688,
      "learning_rate": 7.960864495916654e-07,
      "loss": 0.017,
      "relation_loss": 0.0011154854786582292,
      "selected_pairs": 26763.2,
      "state_query_norm": 1112.692236328125,
      "state_relation_loss": 0.0002647774730576202,
      "step": 28380,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011154854786582292,
      "weighted_relation_loss": 0.0011154854786582292
    },
    {
      "action_query_norm": 617.2816528320312,
      "action_relation_loss": 0.0008448107400909066,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.55753040224509,
      "flow_matching_loss": 0.013106690347194671,
      "grad_norm": 0.1064184159040451,
      "learning_rate": 7.863203282494847e-07,
      "loss": 0.0142,
      "relation_loss": 0.0011028802953660488,
      "selected_pairs": 26762.7,
      "state_query_norm": 1116.857080078125,
      "state_relation_loss": 0.00025806954654399304,
      "step": 28390,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011028802953660488,
      "weighted_relation_loss": 0.0011028802953660488
    },
    {
      "action_query_norm": 619.2946594238281,
      "action_relation_loss": 0.0008290864585433156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.56688493919551,
      "flow_matching_loss": 0.015949054341763257,
      "grad_norm": 0.1366545408964157,
      "learning_rate": 7.766140061686522e-07,
      "loss": 0.017,
      "relation_loss": 0.0010939106345176698,
      "selected_pairs": 27036.6,
      "state_query_norm": 1109.694091796875,
      "state_relation_loss": 0.0002648241730639711,
      "step": 28400,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010939106345176698,
      "weighted_relation_loss": 0.0010939106345176698
    },
    {
      "action_query_norm": 619.3067077636719,
      "action_relation_loss": 0.0008163229154888541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.57623947614593,
      "flow_matching_loss": 0.015783474035561084,
      "grad_norm": 0.08453916758298874,
      "learning_rate": 7.669674951432615e-07,
      "loss": 0.0169,
      "relation_loss": 0.0010810065199621023,
      "selected_pairs": 26948.5,
      "state_query_norm": 1116.4892944335938,
      "state_relation_loss": 0.00026468361902516336,
      "step": 28410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010810065199621023,
      "weighted_relation_loss": 0.0010810065199621023
    },
    {
      "action_query_norm": 617.476123046875,
      "action_relation_loss": 0.0007878191536292434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.58559401309635,
      "flow_matching_loss": 0.013372647017240525,
      "grad_norm": 0.06089961156249046,
      "learning_rate": 7.573808068947363e-07,
      "loss": 0.0144,
      "relation_loss": 0.0010557029978372156,
      "selected_pairs": 26869.9,
      "state_query_norm": 1096.961376953125,
      "state_relation_loss": 0.0002678838340216316,
      "step": 28420,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010557029978372156,
      "weighted_relation_loss": 0.0010557029978372156
    },
    {
      "action_query_norm": 618.7237548828125,
      "action_relation_loss": 0.0008265340293291956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.594948550046773,
      "flow_matching_loss": 0.015598202683031559,
      "grad_norm": 0.1107734963297844,
      "learning_rate": 7.478539530718087e-07,
      "loss": 0.0167,
      "relation_loss": 0.0010845833458006382,
      "selected_pairs": 26874.9,
      "state_query_norm": 1107.5837890625,
      "state_relation_loss": 0.00025804934266489,
      "step": 28430,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010845833458006382,
      "weighted_relation_loss": 0.0010845833458006382
    },
    {
      "action_query_norm": 622.1080810546875,
      "action_relation_loss": 0.000828181451652199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.604303086997195,
      "flow_matching_loss": 0.015852685552090406,
      "grad_norm": 0.1030033603310585,
      "learning_rate": 7.383869452504965e-07,
      "loss": 0.017,
      "relation_loss": 0.0011044753482565284,
      "selected_pairs": 26556.0,
      "state_query_norm": 1106.7157958984376,
      "state_relation_loss": 0.00027629389660432935,
      "step": 28440,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011044753482565284,
      "weighted_relation_loss": 0.0011044753482565284
    },
    {
      "action_query_norm": 618.662548828125,
      "action_relation_loss": 0.0007767039001919329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.613657623947613,
      "flow_matching_loss": 0.013684718497097492,
      "grad_norm": 0.1230694055557251,
      "learning_rate": 7.289797949341204e-07,
      "loss": 0.0147,
      "relation_loss": 0.0010441180318593978,
      "selected_pairs": 26809.8,
      "state_query_norm": 1109.6511474609374,
      "state_relation_loss": 0.0002674141403986141,
      "step": 28450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010441180318593978,
      "weighted_relation_loss": 0.0010441180318593978
    },
    {
      "action_query_norm": 617.5353271484375,
      "action_relation_loss": 0.0008341630280483514,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.623012160898035,
      "flow_matching_loss": 0.013648666720837355,
      "grad_norm": 0.08612935990095139,
      "learning_rate": 7.196325135532423e-07,
      "loss": 0.0147,
      "relation_loss": 0.0011006580549292266,
      "selected_pairs": 27050.9,
      "state_query_norm": 1112.1209106445312,
      "state_relation_loss": 0.0002664950210601091,
      "step": 28460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011006580549292266,
      "weighted_relation_loss": 0.0011006580549292266
    },
    {
      "action_query_norm": 612.835302734375,
      "action_relation_loss": 0.0008395149430725724,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.632366697848457,
      "flow_matching_loss": 0.01553082261234522,
      "grad_norm": 0.0775952860713005,
      "learning_rate": 7.10345112465699e-07,
      "loss": 0.0166,
      "relation_loss": 0.0011137191904708743,
      "selected_pairs": 26548.4,
      "state_query_norm": 1110.6589233398438,
      "state_relation_loss": 0.00027420427359174936,
      "step": 28470,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011137191904708743,
      "weighted_relation_loss": 0.0011137191904708743
    },
    {
      "action_query_norm": 618.2530395507813,
      "action_relation_loss": 0.0008601569337770343,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.641721234798876,
      "flow_matching_loss": 0.01427926616743207,
      "grad_norm": 0.07824429124593735,
      "learning_rate": 7.011176029565525e-07,
      "loss": 0.0154,
      "relation_loss": 0.00112394712632522,
      "selected_pairs": 26919.3,
      "state_query_norm": 1100.020947265625,
      "state_relation_loss": 0.0002637901867274195,
      "step": 28480,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00112394712632522,
      "weighted_relation_loss": 0.00112394712632522
    },
    {
      "action_query_norm": 614.1108215332031,
      "action_relation_loss": 0.0008228792226873338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.651075771749298,
      "flow_matching_loss": 0.012336931936442853,
      "grad_norm": 0.08301328122615814,
      "learning_rate": 6.919499962381004e-07,
      "loss": 0.0134,
      "relation_loss": 0.0010852134088054298,
      "selected_pairs": 26856.5,
      "state_query_norm": 1093.5078491210938,
      "state_relation_loss": 0.00026233419339405373,
      "step": 28490,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010852134088054298,
      "weighted_relation_loss": 0.0010852134088054298
    },
    {
      "action_query_norm": 616.4872802734375,
      "action_relation_loss": 0.0008250944083556533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.66043030869972,
      "flow_matching_loss": 0.015338450018316508,
      "grad_norm": 0.07672113925218582,
      "learning_rate": 6.828423034498488e-07,
      "loss": 0.0164,
      "relation_loss": 0.0010897157713770866,
      "selected_pairs": 26417.8,
      "state_query_norm": 1102.2801635742187,
      "state_relation_loss": 0.00026462135720066725,
      "step": 28500,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0010897157713770866,
      "weighted_relation_loss": 0.0010897157713770866
    },
    {
      "action_query_norm": 616.1624694824219,
      "action_relation_loss": 0.0008203535398934036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.669784845650142,
      "flow_matching_loss": 0.0145433209836483,
      "grad_norm": 0.07238220423460007,
      "learning_rate": 6.737945356585007e-07,
      "loss": 0.0156,
      "relation_loss": 0.0010833108215592801,
      "selected_pairs": 26731.4,
      "state_query_norm": 1118.3996826171874,
      "state_relation_loss": 0.00026295728457625955,
      "step": 28510,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010833108215592801,
      "weighted_relation_loss": 0.0010833108215592801
    },
    {
      "action_query_norm": 611.6121765136719,
      "action_relation_loss": 0.0008235624642111361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.67913938260056,
      "flow_matching_loss": 0.014364887215197086,
      "grad_norm": 0.06389238685369492,
      "learning_rate": 6.648067038579508e-07,
      "loss": 0.0155,
      "relation_loss": 0.001088994333986193,
      "selected_pairs": 26295.4,
      "state_query_norm": 1107.4714111328126,
      "state_relation_loss": 0.0002654318552231416,
      "step": 28520,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.001088994333986193,
      "weighted_relation_loss": 0.001088994333986193
    },
    {
      "action_query_norm": 616.4162475585938,
      "action_relation_loss": 0.0008059137791860849,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.688493919550982,
      "flow_matching_loss": 0.014759942609816789,
      "grad_norm": 0.10329744964838028,
      "learning_rate": 6.558788189692578e-07,
      "loss": 0.0158,
      "relation_loss": 0.0010707279783673585,
      "selected_pairs": 26465.6,
      "state_query_norm": 1104.3674438476562,
      "state_relation_loss": 0.00026481419918127357,
      "step": 28530,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010707279783673585,
      "weighted_relation_loss": 0.0010707279783673585
    },
    {
      "action_query_norm": 621.4523803710938,
      "action_relation_loss": 0.0008522428455762565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.697848456501404,
      "flow_matching_loss": 0.016099434159696103,
      "grad_norm": 0.08130329847335815,
      "learning_rate": 6.470108918406492e-07,
      "loss": 0.0172,
      "relation_loss": 0.0011249892530031502,
      "selected_pairs": 26969.8,
      "state_query_norm": 1102.1342163085938,
      "state_relation_loss": 0.0002727464190684259,
      "step": 28540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011249892530031502,
      "weighted_relation_loss": 0.0011249892530031502
    },
    {
      "action_query_norm": 620.3367553710938,
      "action_relation_loss": 0.0008348011353518814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.707202993451823,
      "flow_matching_loss": 0.014946019183844328,
      "grad_norm": 0.0918872207403183,
      "learning_rate": 6.382029332474893e-07,
      "loss": 0.016,
      "relation_loss": 0.0011037596850655973,
      "selected_pairs": 26464.8,
      "state_query_norm": 1102.9371948242188,
      "state_relation_loss": 0.0002689585468033329,
      "step": 28550,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0011037596850655973,
      "weighted_relation_loss": 0.0011037596850655973
    },
    {
      "action_query_norm": 617.6098510742188,
      "action_relation_loss": 0.0008605857845395803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.716557530402245,
      "flow_matching_loss": 0.014579310547560453,
      "grad_norm": 0.10828136652708054,
      "learning_rate": 6.294549538922778e-07,
      "loss": 0.0157,
      "relation_loss": 0.001129447459243238,
      "selected_pairs": 27220.1,
      "state_query_norm": 1102.7366088867188,
      "state_relation_loss": 0.0002688616848899983,
      "step": 28560,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.001129447459243238,
      "weighted_relation_loss": 0.001129447459243238
    },
    {
      "action_query_norm": 614.0748596191406,
      "action_relation_loss": 0.0008059953281190247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.725912067352667,
      "flow_matching_loss": 0.014316375087946653,
      "grad_norm": 0.08503563702106476,
      "learning_rate": 6.207669644046344e-07,
      "loss": 0.0154,
      "relation_loss": 0.0010695551813114435,
      "selected_pairs": 26270.6,
      "state_query_norm": 1095.343115234375,
      "state_relation_loss": 0.00026355985028203577,
      "step": 28570,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0010695551813114435,
      "weighted_relation_loss": 0.0010695551813114435
    },
    {
      "action_query_norm": 618.6307189941406,
      "action_relation_loss": 0.000810440577333793,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.73526660430309,
      "flow_matching_loss": 0.014295777771621943,
      "grad_norm": 0.07822573184967041,
      "learning_rate": 6.121389753412865e-07,
      "loss": 0.0154,
      "relation_loss": 0.0010796724702231586,
      "selected_pairs": 26530.8,
      "state_query_norm": 1096.5077758789062,
      "state_relation_loss": 0.000269231888523791,
      "step": 28580,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010796724702231586,
      "weighted_relation_loss": 0.0010796724702231586
    },
    {
      "action_query_norm": 621.8798095703125,
      "action_relation_loss": 0.0008392812160309404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.744621141253507,
      "flow_matching_loss": 0.014480405300855637,
      "grad_norm": 0.09097335487604141,
      "learning_rate": 6.035709971860592e-07,
      "loss": 0.0156,
      "relation_loss": 0.0011026583844795824,
      "selected_pairs": 27202.7,
      "state_query_norm": 1097.8921875,
      "state_relation_loss": 0.0002633771495311521,
      "step": 28590,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011026583844795824,
      "weighted_relation_loss": 0.0011026583844795824
    },
    {
      "action_query_norm": 608.0009399414063,
      "action_relation_loss": 0.0008868408622220159,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.75397567820393,
      "flow_matching_loss": 0.014739084336906672,
      "grad_norm": 0.07693696767091751,
      "learning_rate": 5.950630403498469e-07,
      "loss": 0.0159,
      "relation_loss": 0.0011543866130523384,
      "selected_pairs": 26524.3,
      "state_query_norm": 1097.3982421875,
      "state_relation_loss": 0.00026754575228551405,
      "step": 28600,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0011543866130523384,
      "weighted_relation_loss": 0.0011543866130523384
    },
    {
      "action_query_norm": 626.5984375,
      "action_relation_loss": 0.0008255575259681792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.76333021515435,
      "flow_matching_loss": 0.013927893806248903,
      "grad_norm": 0.07438506186008453,
      "learning_rate": 5.866151151706189e-07,
      "loss": 0.015,
      "relation_loss": 0.001080324314534664,
      "selected_pairs": 26920.2,
      "state_query_norm": 1111.5316650390625,
      "state_relation_loss": 0.0002547667842009105,
      "step": 28610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001080324314534664,
      "weighted_relation_loss": 0.001080324314534664
    },
    {
      "action_query_norm": 623.2137512207031,
      "action_relation_loss": 0.000827722012763843,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.77268475210477,
      "flow_matching_loss": 0.014157631248235703,
      "grad_norm": 0.0659097209572792,
      "learning_rate": 5.782272319134086e-07,
      "loss": 0.0152,
      "relation_loss": 0.0010893163620494307,
      "selected_pairs": 26525.9,
      "state_query_norm": 1115.9872924804688,
      "state_relation_loss": 0.0002615943289129063,
      "step": 28620,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010893163620494307,
      "weighted_relation_loss": 0.0010893163620494307
    },
    {
      "action_query_norm": 626.3394470214844,
      "action_relation_loss": 0.0008106000605039299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.78203928905519,
      "flow_matching_loss": 0.015130092669278384,
      "grad_norm": 0.08767393231391907,
      "learning_rate": 5.698994007702796e-07,
      "loss": 0.0162,
      "relation_loss": 0.0010723719955421985,
      "selected_pairs": 27614.9,
      "state_query_norm": 1106.389990234375,
      "state_relation_loss": 0.0002617719292175025,
      "step": 28630,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0010723719955421985,
      "weighted_relation_loss": 0.0010723719955421985
    },
    {
      "action_query_norm": 619.133251953125,
      "action_relation_loss": 0.0008061963657382875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.791393826005613,
      "flow_matching_loss": 0.01582372421398759,
      "grad_norm": 0.07161958515644073,
      "learning_rate": 5.616316318603321e-07,
      "loss": 0.0169,
      "relation_loss": 0.0010738072858657687,
      "selected_pairs": 26989.4,
      "state_query_norm": 1103.470703125,
      "state_relation_loss": 0.00026761090266518294,
      "step": 28640,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010738072858657687,
      "weighted_relation_loss": 0.0010738072858657687
    },
    {
      "action_query_norm": 616.3559326171875,
      "action_relation_loss": 0.0008106386289000511,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.800748362956035,
      "flow_matching_loss": 0.015399123262614011,
      "grad_norm": 0.09723711013793945,
      "learning_rate": 5.534239352296799e-07,
      "loss": 0.0165,
      "relation_loss": 0.0010660822736099362,
      "selected_pairs": 27045.7,
      "state_query_norm": 1118.6032348632812,
      "state_relation_loss": 0.00025544365926180037,
      "step": 28650,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010660822736099362,
      "weighted_relation_loss": 0.0010660822736099362
    },
    {
      "action_query_norm": 623.0846374511718,
      "action_relation_loss": 0.0008106729132123291,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.810102899906454,
      "flow_matching_loss": 0.01476492015644908,
      "grad_norm": 0.09157650172710419,
      "learning_rate": 5.452763208514621e-07,
      "loss": 0.0158,
      "relation_loss": 0.0010732527007348835,
      "selected_pairs": 27084.3,
      "state_query_norm": 1108.7381713867187,
      "state_relation_loss": 0.0002625797758810222,
      "step": 28660,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010732527007348835,
      "weighted_relation_loss": 0.0010732527007348835
    },
    {
      "action_query_norm": 613.6093688964844,
      "action_relation_loss": 0.0008199026400689035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.819457436856876,
      "flow_matching_loss": 0.01464729718863964,
      "grad_norm": 0.1004929319024086,
      "learning_rate": 5.371887986257873e-07,
      "loss": 0.0157,
      "relation_loss": 0.0010910490178503095,
      "selected_pairs": 26415.9,
      "state_query_norm": 1122.51513671875,
      "state_relation_loss": 0.00027114635740872474,
      "step": 28670,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0010910490178503095,
      "weighted_relation_loss": 0.0010910490178503095
    },
    {
      "action_query_norm": 620.6125122070313,
      "action_relation_loss": 0.0008182505902368575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.828811973807298,
      "flow_matching_loss": 0.014152885880321264,
      "grad_norm": 0.10618368536233902,
      "learning_rate": 5.291613783797611e-07,
      "loss": 0.0152,
      "relation_loss": 0.001082359196152538,
      "selected_pairs": 26874.0,
      "state_query_norm": 1097.9964477539063,
      "state_relation_loss": 0.00026410860155010594,
      "step": 28680,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001082359196152538,
      "weighted_relation_loss": 0.001082359196152538
    },
    {
      "action_query_norm": 619.439208984375,
      "action_relation_loss": 0.0008159543678630144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.838166510757716,
      "flow_matching_loss": 0.014666783902794123,
      "grad_norm": 0.09066314995288849,
      "learning_rate": 5.211940698674534e-07,
      "loss": 0.0157,
      "relation_loss": 0.0010813991888426245,
      "selected_pairs": 27107.9,
      "state_query_norm": 1100.9600341796875,
      "state_relation_loss": 0.0002654448093380779,
      "step": 28690,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0010813991888426245,
      "weighted_relation_loss": 0.0010813991888426245
    },
    {
      "action_query_norm": 623.6515258789062,
      "action_relation_loss": 0.0008330124197527766,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.84752104770814,
      "flow_matching_loss": 0.014637120440602303,
      "grad_norm": 0.061533164232969284,
      "learning_rate": 5.132868827698978e-07,
      "loss": 0.0157,
      "relation_loss": 0.0010931640747003258,
      "selected_pairs": 27614.3,
      "state_query_norm": 1124.5108276367187,
      "state_relation_loss": 0.0002601516403956339,
      "step": 28700,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0010931640747003258,
      "weighted_relation_loss": 0.0010931640747003258
    },
    {
      "action_query_norm": 616.4401916503906,
      "action_relation_loss": 0.0008171646273694933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.85687558465856,
      "flow_matching_loss": 0.015273954160511494,
      "grad_norm": 0.1169666275382042,
      "learning_rate": 5.054398266950755e-07,
      "loss": 0.0164,
      "relation_loss": 0.0010785506223328412,
      "selected_pairs": 26969.0,
      "state_query_norm": 1098.6066284179688,
      "state_relation_loss": 0.00026138600369449703,
      "step": 28710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010785506223328412,
      "weighted_relation_loss": 0.0010785506223328412
    },
    {
      "action_query_norm": 618.1739562988281,
      "action_relation_loss": 0.0008271270140539855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.86623012160898,
      "flow_matching_loss": 0.01577477380633354,
      "grad_norm": 0.08899658173322678,
      "learning_rate": 4.976529111778872e-07,
      "loss": 0.0169,
      "relation_loss": 0.0011018039658665657,
      "selected_pairs": 27363.5,
      "state_query_norm": 1103.0520751953125,
      "state_relation_loss": 0.00027467695763334633,
      "step": 28720,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0011018039658665657,
      "weighted_relation_loss": 0.0011018039658665657
    },
    {
      "action_query_norm": 619.36142578125,
      "action_relation_loss": 0.0007763643108773977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.8755846585594,
      "flow_matching_loss": 0.013825156725943089,
      "grad_norm": 0.07821638882160187,
      "learning_rate": 4.899261456801862e-07,
      "loss": 0.0149,
      "relation_loss": 0.0010310079145710916,
      "selected_pairs": 27175.2,
      "state_query_norm": 1104.2426391601562,
      "state_relation_loss": 0.0002546436051488854,
      "step": 28730,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0010310079145710916,
      "weighted_relation_loss": 0.0010310079145710916
    },
    {
      "action_query_norm": 615.2838500976562,
      "action_relation_loss": 0.0008445453830063343,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.884939195509823,
      "flow_matching_loss": 0.014079720340669155,
      "grad_norm": 0.06881042569875717,
      "learning_rate": 4.822595395907126e-07,
      "loss": 0.0152,
      "relation_loss": 0.0011124867247417568,
      "selected_pairs": 26594.9,
      "state_query_norm": 1112.8271606445312,
      "state_relation_loss": 0.00026794133009389043,
      "step": 28740,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011124867247417568,
      "weighted_relation_loss": 0.0011124867247417568
    },
    {
      "action_query_norm": 620.1319152832032,
      "action_relation_loss": 0.0008225373865570873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.894293732460245,
      "flow_matching_loss": 0.013811381720006465,
      "grad_norm": 0.06943446397781372,
      "learning_rate": 4.7465310222510886e-07,
      "loss": 0.0149,
      "relation_loss": 0.0010878965200390666,
      "selected_pairs": 27334.8,
      "state_query_norm": 1096.91728515625,
      "state_relation_loss": 0.00026535911893006415,
      "step": 28750,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010878965200390666,
      "weighted_relation_loss": 0.0010878965200390666
    },
    {
      "action_query_norm": 620.74912109375,
      "action_relation_loss": 0.0008473758818581701,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.903648269410663,
      "flow_matching_loss": 0.014718778710812331,
      "grad_norm": 0.08298972249031067,
      "learning_rate": 4.6710684282593175e-07,
      "loss": 0.0158,
      "relation_loss": 0.0011135770822875203,
      "selected_pairs": 26942.0,
      "state_query_norm": 1121.8023559570313,
      "state_relation_loss": 0.0002662012106156908,
      "step": 28760,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011135770822875203,
      "weighted_relation_loss": 0.0011135770822875203
    },
    {
      "action_query_norm": 612.0140991210938,
      "action_relation_loss": 0.0008347228402271866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.913002806361085,
      "flow_matching_loss": 0.013930467329919339,
      "grad_norm": 0.08981639891862869,
      "learning_rate": 4.5962077056257993e-07,
      "loss": 0.015,
      "relation_loss": 0.0010998154757544398,
      "selected_pairs": 26663.3,
      "state_query_norm": 1096.1667236328126,
      "state_relation_loss": 0.00026509263407206165,
      "step": 28770,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010998154757544398,
      "weighted_relation_loss": 0.0010998154757544398
    },
    {
      "action_query_norm": 619.3376220703125,
      "action_relation_loss": 0.0008226563455536962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.922357343311507,
      "flow_matching_loss": 0.012815982569009066,
      "grad_norm": 0.0656752660870552,
      "learning_rate": 4.521948945313492e-07,
      "loss": 0.0139,
      "relation_loss": 0.001081656152382493,
      "selected_pairs": 26896.0,
      "state_query_norm": 1111.21552734375,
      "state_relation_loss": 0.0002589998039184138,
      "step": 28780,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001081656152382493,
      "weighted_relation_loss": 0.001081656152382493
    },
    {
      "action_query_norm": 616.0321105957031,
      "action_relation_loss": 0.00078041524393484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.931711880261926,
      "flow_matching_loss": 0.014996115397661925,
      "grad_norm": 0.08396289497613907,
      "learning_rate": 4.4482922375537195e-07,
      "loss": 0.016,
      "relation_loss": 0.0010409408481791616,
      "selected_pairs": 25845.2,
      "state_query_norm": 1113.2130249023437,
      "state_relation_loss": 0.00026052559551317247,
      "step": 28790,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.0010409408481791616,
      "weighted_relation_loss": 0.0010409408481791616
    },
    {
      "action_query_norm": 620.7712341308594,
      "action_relation_loss": 0.0008175526279956102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.941066417212348,
      "flow_matching_loss": 0.01434104535728693,
      "grad_norm": 0.08926745504140854,
      "learning_rate": 4.375237671846333e-07,
      "loss": 0.0154,
      "relation_loss": 0.0010891524725593626,
      "selected_pairs": 26368.6,
      "state_query_norm": 1107.9133178710938,
      "state_relation_loss": 0.0002715998562052846,
      "step": 28800,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0010891524725593626,
      "weighted_relation_loss": 0.0010891524725593626
    },
    {
      "action_query_norm": 611.2299255371094,
      "action_relation_loss": 0.0008582983980886638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.95042095416277,
      "flow_matching_loss": 0.013902620133012532,
      "grad_norm": 0.11563637852668762,
      "learning_rate": 4.302785336959547e-07,
      "loss": 0.015,
      "relation_loss": 0.0011219717212952673,
      "selected_pairs": 26246.3,
      "state_query_norm": 1112.7409545898438,
      "state_relation_loss": 0.00026367332466179503,
      "step": 28810,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0011219717212952673,
      "weighted_relation_loss": 0.0011219717212952673
    },
    {
      "action_query_norm": 615.9085571289063,
      "action_relation_loss": 0.0008403468877077102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.95977549111319,
      "flow_matching_loss": 0.014736461732536555,
      "grad_norm": 0.07680217176675797,
      "learning_rate": 4.2309353209297744e-07,
      "loss": 0.0158,
      "relation_loss": 0.0011110582039691509,
      "selected_pairs": 26515.1,
      "state_query_norm": 1107.2068969726563,
      "state_relation_loss": 0.00027071130753029136,
      "step": 28820,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011110582039691509,
      "weighted_relation_loss": 0.0011110582039691509
    },
    {
      "action_query_norm": 619.6317443847656,
      "action_relation_loss": 0.0008554792904760689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.96913002806361,
      "flow_matching_loss": 0.01337565528228879,
      "grad_norm": 0.07870550453662872,
      "learning_rate": 4.159687711061566e-07,
      "loss": 0.0145,
      "relation_loss": 0.0011203889385797083,
      "selected_pairs": 27667.7,
      "state_query_norm": 1116.5863525390625,
      "state_relation_loss": 0.0002649096539244056,
      "step": 28830,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0011203889385797083,
      "weighted_relation_loss": 0.0011203889385797083
    },
    {
      "action_query_norm": 617.4419128417969,
      "action_relation_loss": 0.0008219901646953076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.978484565014032,
      "flow_matching_loss": 0.015606017969548702,
      "grad_norm": 0.07806658744812012,
      "learning_rate": 4.0890425939275055e-07,
      "loss": 0.0167,
      "relation_loss": 0.0010853758256416767,
      "selected_pairs": 27366.7,
      "state_query_norm": 1106.9544677734375,
      "state_relation_loss": 0.0002633856565807946,
      "step": 28840,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0010853758256416767,
      "weighted_relation_loss": 0.0010853758256416767
    },
    {
      "action_query_norm": 620.1434020996094,
      "action_relation_loss": 0.0008298138855025172,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.987839101964454,
      "flow_matching_loss": 0.013839254155755043,
      "grad_norm": 0.07743548601865768,
      "learning_rate": 4.0190000553679827e-07,
      "loss": 0.0149,
      "relation_loss": 0.0011005267500877381,
      "selected_pairs": 27371.7,
      "state_query_norm": 1100.8752563476562,
      "state_relation_loss": 0.00027071286021964627,
      "step": 28850,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0011005267500877381,
      "weighted_relation_loss": 0.0011005267500877381
    },
    {
      "action_query_norm": 617.47392578125,
      "action_relation_loss": 0.0008112425159197301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.997193638914872,
      "flow_matching_loss": 0.014973171427845955,
      "grad_norm": 0.07963912934064865,
      "learning_rate": 3.9495601804913627e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010776913492009043,
      "selected_pairs": 26738.6,
      "state_query_norm": 1112.2427124023438,
      "state_relation_loss": 0.0002664488318259828,
      "step": 28860,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010776913492009043,
      "weighted_relation_loss": 0.0010776913492009043
    },
    {
      "action_query_norm": 617.3505615234375,
      "action_relation_loss": 0.0008854821033310145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.006548175865294,
      "flow_matching_loss": 0.01420775605365634,
      "grad_norm": 0.07761629670858383,
      "learning_rate": 3.880723053673652e-07,
      "loss": 0.0154,
      "relation_loss": 0.0011486822855658828,
      "selected_pairs": 24389.7,
      "state_query_norm": 1109.9459594726563,
      "state_relation_loss": 0.00026320017204852777,
      "step": 28870,
      "valid_action_samples": 214.5,
      "weighted_auxiliary_loss": 0.0011486822855658828,
      "weighted_relation_loss": 0.0011486822855658828
    },
    {
      "action_query_norm": 612.2885192871094,
      "action_relation_loss": 0.0008087342896033079,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.015902712815716,
      "flow_matching_loss": 0.014688661508262157,
      "grad_norm": 0.08364301919937134,
      "learning_rate": 3.812488758558386e-07,
      "loss": 0.0158,
      "relation_loss": 0.001075868180487305,
      "selected_pairs": 26616.7,
      "state_query_norm": 1104.2475463867188,
      "state_relation_loss": 0.00026713389088399707,
      "step": 28880,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001075868180487305,
      "weighted_relation_loss": 0.001075868180487305
    },
    {
      "action_query_norm": 614.8928466796875,
      "action_relation_loss": 0.0008224855642765761,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.02525724976614,
      "flow_matching_loss": 0.014319937489926816,
      "grad_norm": 0.08820104598999023,
      "learning_rate": 3.744857378056743e-07,
      "loss": 0.0154,
      "relation_loss": 0.0010850967490114271,
      "selected_pairs": 25756.0,
      "state_query_norm": 1126.09970703125,
      "state_relation_loss": 0.0002626111963763833,
      "step": 28890,
      "valid_action_samples": 227.4,
      "weighted_auxiliary_loss": 0.0010850967490114271,
      "weighted_relation_loss": 0.0010850967490114271
    },
    {
      "action_query_norm": 610.2525146484375,
      "action_relation_loss": 0.0008673621690832078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.034611786716557,
      "flow_matching_loss": 0.014587721973657607,
      "grad_norm": 0.08977656066417694,
      "learning_rate": 3.677828994347154e-07,
      "loss": 0.0157,
      "relation_loss": 0.0011289572576060891,
      "selected_pairs": 27298.1,
      "state_query_norm": 1111.0894409179687,
      "state_relation_loss": 0.00026159508270211517,
      "step": 28900,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011289572576060891,
      "weighted_relation_loss": 0.0011289572576060891
    },
    {
      "action_query_norm": 615.4986877441406,
      "action_relation_loss": 0.0008362086024135351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.04396632366698,
      "flow_matching_loss": 0.013785500079393387,
      "grad_norm": 0.06500627845525742,
      "learning_rate": 3.61140368887547e-07,
      "loss": 0.0149,
      "relation_loss": 0.0011003727675415575,
      "selected_pairs": 26803.4,
      "state_query_norm": 1104.2534545898438,
      "state_relation_loss": 0.00026416416367283093,
      "step": 28910,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011003727675415575,
      "weighted_relation_loss": 0.0011003727675415575
    },
    {
      "action_query_norm": 613.3655822753906,
      "action_relation_loss": 0.0008430798479821533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.0533208606174,
      "flow_matching_loss": 0.014340039994567632,
      "grad_norm": 0.07537747919559479,
      "learning_rate": 3.545581542354681e-07,
      "loss": 0.0154,
      "relation_loss": 0.001105284201912582,
      "selected_pairs": 26695.0,
      "state_query_norm": 1099.872412109375,
      "state_relation_loss": 0.00026220434810966255,
      "step": 28920,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.001105284201912582,
      "weighted_relation_loss": 0.001105284201912582
    },
    {
      "action_query_norm": 618.4900817871094,
      "action_relation_loss": 0.000804916670313105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.06267539756782,
      "flow_matching_loss": 0.014448327012360096,
      "grad_norm": 0.08996396511793137,
      "learning_rate": 3.480362634764922e-07,
      "loss": 0.0155,
      "relation_loss": 0.001071962434798479,
      "selected_pairs": 26777.4,
      "state_query_norm": 1119.3388671875,
      "state_relation_loss": 0.00026704577321652324,
      "step": 28930,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001071962434798479,
      "weighted_relation_loss": 0.001071962434798479
    },
    {
      "action_query_norm": 620.5155029296875,
      "action_relation_loss": 0.0008190110675059259,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.07202993451824,
      "flow_matching_loss": 0.013714949041604996,
      "grad_norm": 0.08632910996675491,
      "learning_rate": 3.4157470453533014e-07,
      "loss": 0.0148,
      "relation_loss": 0.001083095499780029,
      "selected_pairs": 26833.5,
      "state_query_norm": 1101.6081909179688,
      "state_relation_loss": 0.00026408441772218794,
      "step": 28940,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001083095499780029,
      "weighted_relation_loss": 0.001083095499780029
    },
    {
      "action_query_norm": 619.7192932128906,
      "action_relation_loss": 0.0008193698013201356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.081384471468663,
      "flow_matching_loss": 0.014705788809806108,
      "grad_norm": 0.10435350984334946,
      "learning_rate": 3.3517348526339034e-07,
      "loss": 0.0158,
      "relation_loss": 0.0010817270842380822,
      "selected_pairs": 26879.9,
      "state_query_norm": 1119.3663208007813,
      "state_relation_loss": 0.0002623573032906279,
      "step": 28950,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010817270842380822,
      "weighted_relation_loss": 0.0010817270842380822
    },
    {
      "action_query_norm": 613.232763671875,
      "action_relation_loss": 0.0008361082815099508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.09073900841908,
      "flow_matching_loss": 0.015114408358931542,
      "grad_norm": 0.07667295634746552,
      "learning_rate": 3.288326134387454e-07,
      "loss": 0.0162,
      "relation_loss": 0.0010952071752399206,
      "selected_pairs": 26700.1,
      "state_query_norm": 1092.8106506347656,
      "state_relation_loss": 0.0002590988951851614,
      "step": 28960,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010952071752399206,
      "weighted_relation_loss": 0.0010952071752399206
    },
    {
      "action_query_norm": 622.033251953125,
      "action_relation_loss": 0.0008302894479129463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.100093545369504,
      "flow_matching_loss": 0.015543559566140175,
      "grad_norm": 0.08129340410232544,
      "learning_rate": 3.225520967661655e-07,
      "loss": 0.0166,
      "relation_loss": 0.001089729554951191,
      "selected_pairs": 25933.8,
      "state_query_norm": 1102.224609375,
      "state_relation_loss": 0.00025944009976228697,
      "step": 28970,
      "valid_action_samples": 228.2,
      "weighted_auxiliary_loss": 0.001089729554951191,
      "weighted_relation_loss": 0.001089729554951191
    },
    {
      "action_query_norm": 621.6001647949219,
      "action_relation_loss": 0.0008267819415777921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.109448082319926,
      "flow_matching_loss": 0.013179596792906523,
      "grad_norm": 0.08671446889638901,
      "learning_rate": 3.163319428770628e-07,
      "loss": 0.0143,
      "relation_loss": 0.0010901764151640236,
      "selected_pairs": 26861.3,
      "state_query_norm": 1110.4447387695313,
      "state_relation_loss": 0.00026339449104852977,
      "step": 28980,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010901764151640236,
      "weighted_relation_loss": 0.0010901764151640236
    },
    {
      "action_query_norm": 615.3464599609375,
      "action_relation_loss": 0.0008245397708378732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.118802619270348,
      "flow_matching_loss": 0.015308899339288473,
      "grad_norm": 0.08331071585416794,
      "learning_rate": 3.1017215932951374e-07,
      "loss": 0.0164,
      "relation_loss": 0.0010902032838203013,
      "selected_pairs": 25896.5,
      "state_query_norm": 1107.657470703125,
      "state_relation_loss": 0.0002656635275343433,
      "step": 28990,
      "valid_action_samples": 228.0,
      "weighted_auxiliary_loss": 0.0010902032838203013,
      "weighted_relation_loss": 0.0010902032838203013
    },
    {
      "action_query_norm": 625.3902221679688,
      "action_relation_loss": 0.0008124719082843512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.128157156220766,
      "flow_matching_loss": 0.015261872950941325,
      "grad_norm": 0.09340006113052368,
      "learning_rate": 3.040727536082366e-07,
      "loss": 0.0163,
      "relation_loss": 0.0010831673978827893,
      "selected_pairs": 27011.8,
      "state_query_norm": 1098.9400756835937,
      "state_relation_loss": 0.0002706954983295873,
      "step": 29000,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010831673978827893,
      "weighted_relation_loss": 0.0010831673978827893
    },
    {
      "action_query_norm": 622.1182312011719,
      "action_relation_loss": 0.0008164186379872263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.137511693171188,
      "flow_matching_loss": 0.012795083411037921,
      "grad_norm": 0.0793418288230896,
      "learning_rate": 2.980337331245864e-07,
      "loss": 0.0139,
      "relation_loss": 0.0010794351692311466,
      "selected_pairs": 26690.9,
      "state_query_norm": 1099.8771850585938,
      "state_relation_loss": 0.00026301651960238817,
      "step": 29010,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010794351692311466,
      "weighted_relation_loss": 0.0010794351692311466
    },
    {
      "action_query_norm": 619.0156860351562,
      "action_relation_loss": 0.000820298120379448,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.14686623012161,
      "flow_matching_loss": 0.014654656127095222,
      "grad_norm": 0.08222619444131851,
      "learning_rate": 2.9205510521653213e-07,
      "loss": 0.0157,
      "relation_loss": 0.001090788992587477,
      "selected_pairs": 26415.2,
      "state_query_norm": 1112.6939697265625,
      "state_relation_loss": 0.0002704908809391782,
      "step": 29020,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001090788992587477,
      "weighted_relation_loss": 0.001090788992587477
    },
    {
      "action_query_norm": 624.8096069335937,
      "action_relation_loss": 0.0008598399756010622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.15622076707203,
      "flow_matching_loss": 0.014097781293094159,
      "grad_norm": 0.08185016363859177,
      "learning_rate": 2.86136877148685e-07,
      "loss": 0.0152,
      "relation_loss": 0.001127914770040661,
      "selected_pairs": 27251.0,
      "state_query_norm": 1115.2285278320312,
      "state_relation_loss": 0.00026807479152921585,
      "step": 29030,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001127914770040661,
      "weighted_relation_loss": 0.001127914770040661
    },
    {
      "action_query_norm": 623.0313171386719,
      "action_relation_loss": 0.0008212654443923384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.16557530402245,
      "flow_matching_loss": 0.014276695530861616,
      "grad_norm": 0.11026635020971298,
      "learning_rate": 2.8027905611223704e-07,
      "loss": 0.0154,
      "relation_loss": 0.00108997889328748,
      "selected_pairs": 27244.7,
      "state_query_norm": 1118.7176879882813,
      "state_relation_loss": 0.00026871346635743977,
      "step": 29040,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.00108997889328748,
      "weighted_relation_loss": 0.00108997889328748
    },
    {
      "action_query_norm": 620.6155090332031,
      "action_relation_loss": 0.0007991950085852295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.174929840972872,
      "flow_matching_loss": 0.01390859941020608,
      "grad_norm": 0.066830575466156,
      "learning_rate": 2.7448164922500573e-07,
      "loss": 0.015,
      "relation_loss": 0.00106534983497113,
      "selected_pairs": 26881.1,
      "state_query_norm": 1104.66240234375,
      "state_relation_loss": 0.00026615482638590036,
      "step": 29050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00106534983497113,
      "weighted_relation_loss": 0.00106534983497113
    },
    {
      "action_query_norm": 620.7587036132812,
      "action_relation_loss": 0.00080186165869236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.184284377923294,
      "flow_matching_loss": 0.014815541729331017,
      "grad_norm": 0.09705644100904465,
      "learning_rate": 2.687446635313784e-07,
      "loss": 0.0159,
      "relation_loss": 0.001066222880035639,
      "selected_pairs": 26736.4,
      "state_query_norm": 1092.2039672851563,
      "state_relation_loss": 0.0002643612358951941,
      "step": 29060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001066222880035639,
      "weighted_relation_loss": 0.001066222880035639
    },
    {
      "action_query_norm": 622.8098449707031,
      "action_relation_loss": 0.0007924005331005901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.193638914873713,
      "flow_matching_loss": 0.0145086913369596,
      "grad_norm": 0.0852561891078949,
      "learning_rate": 2.630681060023343e-07,
      "loss": 0.0156,
      "relation_loss": 0.0010492499452084303,
      "selected_pairs": 26973.5,
      "state_query_norm": 1091.0394775390625,
      "state_relation_loss": 0.0002568494106526487,
      "step": 29070,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010492499452084303,
      "weighted_relation_loss": 0.0010492499452084303
    },
    {
      "action_query_norm": 627.5529724121094,
      "action_relation_loss": 0.000811877561500296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.202993451824135,
      "flow_matching_loss": 0.015002332255244255,
      "grad_norm": 0.06183132529258728,
      "learning_rate": 2.5745198353542833e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010801701049786062,
      "selected_pairs": 27291.9,
      "state_query_norm": 1097.5504150390625,
      "state_relation_loss": 0.0002682925303815864,
      "step": 29080,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0010801701049786062,
      "weighted_relation_loss": 0.0010801701049786062
    },
    {
      "action_query_norm": 609.4898681640625,
      "action_relation_loss": 0.0008311832440085709,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.212347988774557,
      "flow_matching_loss": 0.012252556439489127,
      "grad_norm": 0.07127208262681961,
      "learning_rate": 2.518963029547794e-07,
      "loss": 0.0133,
      "relation_loss": 0.0010937742074020207,
      "selected_pairs": 26821.9,
      "state_query_norm": 1102.8678833007812,
      "state_relation_loss": 0.0002625909692142159,
      "step": 29090,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010937742074020207,
      "weighted_relation_loss": 0.0010937742074020207
    },
    {
      "action_query_norm": 623.3421875,
      "action_relation_loss": 0.0008083920634817332,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.221702525724975,
      "flow_matching_loss": 0.012614264898002148,
      "grad_norm": 0.07945854961872101,
      "learning_rate": 2.464010710110598e-07,
      "loss": 0.0137,
      "relation_loss": 0.0010707585490308702,
      "selected_pairs": 26685.8,
      "state_query_norm": 1112.99599609375,
      "state_relation_loss": 0.0002623664782731794,
      "step": 29100,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010707585490308702,
      "weighted_relation_loss": 0.0010707585490308702
    },
    {
      "action_query_norm": 613.3040588378906,
      "action_relation_loss": 0.0008310308156069368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.231057062675397,
      "flow_matching_loss": 0.013888490106910466,
      "grad_norm": 0.09466632455587387,
      "learning_rate": 2.4096629438150054e-07,
      "loss": 0.015,
      "relation_loss": 0.001104744104668498,
      "selected_pairs": 27080.2,
      "state_query_norm": 1114.2689331054687,
      "state_relation_loss": 0.00027371327742002903,
      "step": 29110,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001104744104668498,
      "weighted_relation_loss": 0.001104744104668498
    },
    {
      "action_query_norm": 620.2772521972656,
      "action_relation_loss": 0.0008335697290021926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.24041159962582,
      "flow_matching_loss": 0.015048054698854684,
      "grad_norm": 0.08173083513975143,
      "learning_rate": 2.3559197966985802e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010926137096248568,
      "selected_pairs": 26675.5,
      "state_query_norm": 1112.150732421875,
      "state_relation_loss": 0.00025904396461555733,
      "step": 29120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010926137096248568,
      "weighted_relation_loss": 0.0010926137096248568
    },
    {
      "action_query_norm": 623.2168395996093,
      "action_relation_loss": 0.0008158324984833598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.24976613657624,
      "flow_matching_loss": 0.01342915277928114,
      "grad_norm": 0.10176478326320648,
      "learning_rate": 2.3027813340644188e-07,
      "loss": 0.0145,
      "relation_loss": 0.0010884804534725845,
      "selected_pairs": 27268.6,
      "state_query_norm": 1110.384228515625,
      "state_relation_loss": 0.00027264796080999075,
      "step": 29130,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0010884804534725845,
      "weighted_relation_loss": 0.0010884804534725845
    },
    {
      "action_query_norm": 621.6209411621094,
      "action_relation_loss": 0.0008215644105803221,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.25912067352666,
      "flow_matching_loss": 0.01559103224426508,
      "grad_norm": 0.06691280752420425,
      "learning_rate": 2.2502476204807055e-07,
      "loss": 0.0167,
      "relation_loss": 0.001088057749439031,
      "selected_pairs": 27477.1,
      "state_query_norm": 1109.4573974609375,
      "state_relation_loss": 0.0002664933446794748,
      "step": 29140,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.001088057749439031,
      "weighted_relation_loss": 0.001088057749439031
    },
    {
      "action_query_norm": 609.4668212890625,
      "action_relation_loss": 0.000843870680546388,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.26847521047708,
      "flow_matching_loss": 0.013715119566768408,
      "grad_norm": 0.11248984932899475,
      "learning_rate": 2.1983187197808786e-07,
      "loss": 0.0148,
      "relation_loss": 0.0011003251536749303,
      "selected_pairs": 26501.6,
      "state_query_norm": 1111.1592895507813,
      "state_relation_loss": 0.00025645447458373384,
      "step": 29150,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011003251536749303,
      "weighted_relation_loss": 0.0011003251536749303
    },
    {
      "action_query_norm": 617.80185546875,
      "action_relation_loss": 0.0008116240845993161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.277829747427504,
      "flow_matching_loss": 0.013694295287132263,
      "grad_norm": 0.06381017714738846,
      "learning_rate": 2.1469946950634644e-07,
      "loss": 0.0148,
      "relation_loss": 0.001084488013293594,
      "selected_pairs": 26661.5,
      "state_query_norm": 1103.4771118164062,
      "state_relation_loss": 0.0002728639461565763,
      "step": 29160,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001084488013293594,
      "weighted_relation_loss": 0.001084488013293594
    },
    {
      "action_query_norm": 618.6238647460938,
      "action_relation_loss": 0.0008525003853719681,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.287184284377922,
      "flow_matching_loss": 0.015551540348678828,
      "grad_norm": 0.0641450583934784,
      "learning_rate": 2.096275608691911e-07,
      "loss": 0.0167,
      "relation_loss": 0.0011231350828893483,
      "selected_pairs": 26831.2,
      "state_query_norm": 1111.9019409179687,
      "state_relation_loss": 0.00027063470042776316,
      "step": 29170,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011231350828893483,
      "weighted_relation_loss": 0.0011231350828893483
    },
    {
      "action_query_norm": 618.7046752929688,
      "action_relation_loss": 0.0008346445683855563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.296538821328344,
      "flow_matching_loss": 0.013543965015560388,
      "grad_norm": 0.11766500771045685,
      "learning_rate": 2.046161522294754e-07,
      "loss": 0.0146,
      "relation_loss": 0.0010982882231473923,
      "selected_pairs": 26462.4,
      "state_query_norm": 1096.8893310546875,
      "state_relation_loss": 0.0002636436460306868,
      "step": 29180,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010982882231473923,
      "weighted_relation_loss": 0.0010982882231473923
    },
    {
      "action_query_norm": 620.3298950195312,
      "action_relation_loss": 0.0008418906363658607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.305893358278766,
      "flow_matching_loss": 0.012963528372347355,
      "grad_norm": 0.07412684708833694,
      "learning_rate": 1.9966524967653944e-07,
      "loss": 0.0141,
      "relation_loss": 0.0011093039123807103,
      "selected_pairs": 27080.7,
      "state_query_norm": 1102.5712646484376,
      "state_relation_loss": 0.00026741327747004107,
      "step": 29190,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011093039123807103,
      "weighted_relation_loss": 0.0011093039123807103
    },
    {
      "action_query_norm": 619.6281066894531,
      "action_relation_loss": 0.000837829295778647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.315247895229184,
      "flow_matching_loss": 0.014462325442582368,
      "grad_norm": 0.0742013156414032,
      "learning_rate": 1.9477485922618222e-07,
      "loss": 0.0156,
      "relation_loss": 0.0011071501299738883,
      "selected_pairs": 26930.0,
      "state_query_norm": 1104.352197265625,
      "state_relation_loss": 0.00026932083419524134,
      "step": 29200,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011071501299738883,
      "weighted_relation_loss": 0.0011071501299738883
    },
    {
      "action_query_norm": 620.3347045898438,
      "action_relation_loss": 0.0008070947485975921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.324602432179606,
      "flow_matching_loss": 0.014382623508572579,
      "grad_norm": 0.08750039339065552,
      "learning_rate": 1.899449868207004e-07,
      "loss": 0.0155,
      "relation_loss": 0.0010769380605779588,
      "selected_pairs": 27219.7,
      "state_query_norm": 1103.843798828125,
      "state_relation_loss": 0.00026984333235304805,
      "step": 29210,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0010769380605779588,
      "weighted_relation_loss": 0.0010769380605779588
    },
    {
      "action_query_norm": 617.5798095703125,
      "action_relation_loss": 0.0008293530670925974,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.33395696913003,
      "flow_matching_loss": 0.013774258736521005,
      "grad_norm": 0.0848989188671112,
      "learning_rate": 1.851756383288439e-07,
      "loss": 0.0149,
      "relation_loss": 0.001081688777776435,
      "selected_pairs": 27020.5,
      "state_query_norm": 1092.67373046875,
      "state_relation_loss": 0.0002523357150494121,
      "step": 29220,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001081688777776435,
      "weighted_relation_loss": 0.001081688777776435
    },
    {
      "action_query_norm": 619.1358642578125,
      "action_relation_loss": 0.0008270624442957341,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.34331150608045,
      "flow_matching_loss": 0.013588584121316671,
      "grad_norm": 0.10060538351535797,
      "learning_rate": 1.8046681954581035e-07,
      "loss": 0.0147,
      "relation_loss": 0.0010899085085839034,
      "selected_pairs": 26618.3,
      "state_query_norm": 1108.4354248046875,
      "state_relation_loss": 0.00026284604828106237,
      "step": 29230,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010899085085839034,
      "weighted_relation_loss": 0.0010899085085839034
    },
    {
      "action_query_norm": 620.5741638183594,
      "action_relation_loss": 0.0008135732205118984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.35266604303087,
      "flow_matching_loss": 0.014245041366666555,
      "grad_norm": 0.09787802398204803,
      "learning_rate": 1.7581853619327294e-07,
      "loss": 0.0153,
      "relation_loss": 0.0010830648941919207,
      "selected_pairs": 26754.3,
      "state_query_norm": 1104.2188110351562,
      "state_relation_loss": 0.000269491683866363,
      "step": 29240,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010830648941919207,
      "weighted_relation_loss": 0.0010830648941919207
    },
    {
      "action_query_norm": 613.6631958007813,
      "action_relation_loss": 0.0008161477860994637,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.36202057998129,
      "flow_matching_loss": 0.013345963787287474,
      "grad_norm": 0.11739394068717957,
      "learning_rate": 1.7123079391932472e-07,
      "loss": 0.0144,
      "relation_loss": 0.0010705367545597254,
      "selected_pairs": 27107.3,
      "state_query_norm": 1111.80986328125,
      "state_relation_loss": 0.00025438897137064487,
      "step": 29250,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0010705367545597254,
      "weighted_relation_loss": 0.0010705367545597254
    },
    {
      "action_query_norm": 621.1248657226563,
      "action_relation_loss": 0.0008055167738348246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.371375116931713,
      "flow_matching_loss": 0.013430062774568796,
      "grad_norm": 0.07469248026609421,
      "learning_rate": 1.6670359829850657e-07,
      "loss": 0.0145,
      "relation_loss": 0.0010570088401436807,
      "selected_pairs": 26942.3,
      "state_query_norm": 1110.8321166992187,
      "state_relation_loss": 0.0002514920503017493,
      "step": 29260,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010570088401436807,
      "weighted_relation_loss": 0.0010570088401436807
    },
    {
      "action_query_norm": 618.9515930175781,
      "action_relation_loss": 0.0008144091523718088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.38072965388213,
      "flow_matching_loss": 0.013620627578347921,
      "grad_norm": 0.09839612245559692,
      "learning_rate": 1.6223695483179037e-07,
      "loss": 0.0147,
      "relation_loss": 0.0010787004313897342,
      "selected_pairs": 26880.1,
      "state_query_norm": 1101.3276733398438,
      "state_relation_loss": 0.00026429129065945743,
      "step": 29270,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010787004313897342,
      "weighted_relation_loss": 0.0010787004313897342
    },
    {
      "action_query_norm": 611.6279418945312,
      "action_relation_loss": 0.0008105781278572976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.390084190832553,
      "flow_matching_loss": 0.015069131460040807,
      "grad_norm": 0.09093067795038223,
      "learning_rate": 1.5783086894656795e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010645030881278217,
      "selected_pairs": 26976.1,
      "state_query_norm": 1110.4181030273437,
      "state_relation_loss": 0.0002539249646360986,
      "step": 29280,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010645030881278217,
      "weighted_relation_loss": 0.0010645030881278217
    },
    {
      "action_query_norm": 625.1032409667969,
      "action_relation_loss": 0.0007914013753179461,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.399438727782975,
      "flow_matching_loss": 0.014971911814063788,
      "grad_norm": 0.08148931711912155,
      "learning_rate": 1.5348534599665121e-07,
      "loss": 0.016,
      "relation_loss": 0.0010525844059884547,
      "selected_pairs": 26027.0,
      "state_query_norm": 1111.767138671875,
      "state_relation_loss": 0.0002611830408568494,
      "step": 29290,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.0010525844059884547,
      "weighted_relation_loss": 0.0010525844059884547
    },
    {
      "action_query_norm": 618.393115234375,
      "action_relation_loss": 0.0008564787392970174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.408793264733397,
      "flow_matching_loss": 0.013794309459626675,
      "grad_norm": 0.0969688668847084,
      "learning_rate": 1.4920039126225527e-07,
      "loss": 0.0149,
      "relation_loss": 0.0011166754527948796,
      "selected_pairs": 27033.8,
      "state_query_norm": 1105.0089233398437,
      "state_relation_loss": 0.00026019671495305376,
      "step": 29300,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011166754527948796,
      "weighted_relation_loss": 0.0011166754527948796
    },
    {
      "action_query_norm": 618.0460510253906,
      "action_relation_loss": 0.0008424962870776653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.418147801683816,
      "flow_matching_loss": 0.014568406529724598,
      "grad_norm": 0.07486551254987717,
      "learning_rate": 1.449760099500097e-07,
      "loss": 0.0157,
      "relation_loss": 0.0011073414352722466,
      "selected_pairs": 26390.8,
      "state_query_norm": 1104.3385986328126,
      "state_relation_loss": 0.0002648451380082406,
      "step": 29310,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0011073414352722466,
      "weighted_relation_loss": 0.0011073414352722466
    },
    {
      "action_query_norm": 614.6770690917969,
      "action_relation_loss": 0.0008121960330754518,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.427502338634238,
      "flow_matching_loss": 0.015112743061035872,
      "grad_norm": 0.08121062070131302,
      "learning_rate": 1.4081220719293075e-07,
      "loss": 0.0162,
      "relation_loss": 0.0010749901179224254,
      "selected_pairs": 27198.1,
      "state_query_norm": 1104.3335083007812,
      "state_relation_loss": 0.00026279408775735645,
      "step": 29320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0010749901179224254,
      "weighted_relation_loss": 0.0010749901179224254
    },
    {
      "action_query_norm": 621.0972290039062,
      "action_relation_loss": 0.0008213466091547162,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.43685687558466,
      "flow_matching_loss": 0.012723187264055014,
      "grad_norm": 0.06527264416217804,
      "learning_rate": 1.3670898805043797e-07,
      "loss": 0.0138,
      "relation_loss": 0.0010830624494701623,
      "selected_pairs": 26898.8,
      "state_query_norm": 1105.5985961914062,
      "state_relation_loss": 0.00026171583886025473,
      "step": 29330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010830624494701623,
      "weighted_relation_loss": 0.0010830624494701623
    },
    {
      "action_query_norm": 612.0870727539062,
      "action_relation_loss": 0.0008171691210009157,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.446211412535078,
      "flow_matching_loss": 0.013936523720622063,
      "grad_norm": 0.06281623989343643,
      "learning_rate": 1.326663575083209e-07,
      "loss": 0.015,
      "relation_loss": 0.0010724955704063178,
      "selected_pairs": 26940.5,
      "state_query_norm": 1114.6319213867187,
      "state_relation_loss": 0.00025532644212944433,
      "step": 29340,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010724955704063178,
      "weighted_relation_loss": 0.0010724955704063178
    },
    {
      "action_query_norm": 619.5797790527344,
      "action_relation_loss": 0.0008208808198105544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.4555659494855,
      "flow_matching_loss": 0.015991835761815308,
      "grad_norm": 0.1071300134062767,
      "learning_rate": 1.2868432047876688e-07,
      "loss": 0.0171,
      "relation_loss": 0.0010895279468968511,
      "selected_pairs": 26902.1,
      "state_query_norm": 1109.172998046875,
      "state_relation_loss": 0.00026864715619012713,
      "step": 29350,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0010895279468968511,
      "weighted_relation_loss": 0.0010895279468968511
    },
    {
      "action_query_norm": 622.1376831054688,
      "action_relation_loss": 0.0008118179044686258,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.464920486435922,
      "flow_matching_loss": 0.015034086909145117,
      "grad_norm": 0.1231742724776268,
      "learning_rate": 1.2476288180032213e-07,
      "loss": 0.0161,
      "relation_loss": 0.0010793319554068149,
      "selected_pairs": 27131.4,
      "state_query_norm": 1102.29697265625,
      "state_relation_loss": 0.00026751404511742293,
      "step": 29360,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010793319554068149,
      "weighted_relation_loss": 0.0010793319554068149
    },
    {
      "action_query_norm": 620.58330078125,
      "action_relation_loss": 0.0008123310806695371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.474275023386344,
      "flow_matching_loss": 0.014298356138169765,
      "grad_norm": 0.0851050615310669,
      "learning_rate": 1.209020462379029e-07,
      "loss": 0.0154,
      "relation_loss": 0.0010748634347692132,
      "selected_pairs": 26826.4,
      "state_query_norm": 1102.0586547851562,
      "state_relation_loss": 0.00026253235264448447,
      "step": 29370,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010748634347692132,
      "weighted_relation_loss": 0.0010748634347692132
    },
    {
      "action_query_norm": 620.694677734375,
      "action_relation_loss": 0.0008177638927008956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.483629560336762,
      "flow_matching_loss": 0.014218033663928509,
      "grad_norm": 0.09919741749763489,
      "learning_rate": 1.1710181848278435e-07,
      "loss": 0.0153,
      "relation_loss": 0.001089072262402624,
      "selected_pairs": 27408.3,
      "state_query_norm": 1094.8015991210937,
      "state_relation_loss": 0.000271308355149813,
      "step": 29380,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001089072262402624,
      "weighted_relation_loss": 0.001089072262402624
    },
    {
      "action_query_norm": 616.6705139160156,
      "action_relation_loss": 0.0008133910596370697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.492984097287184,
      "flow_matching_loss": 0.013259899523109197,
      "grad_norm": 0.06538171321153641,
      "learning_rate": 1.133622031526116e-07,
      "loss": 0.0143,
      "relation_loss": 0.0010852244507987053,
      "selected_pairs": 26802.8,
      "state_query_norm": 1105.4291259765625,
      "state_relation_loss": 0.00027183339261682704,
      "step": 29390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010852244507987053,
      "weighted_relation_loss": 0.0010852244507987053
    },
    {
      "action_query_norm": 624.1847290039062,
      "action_relation_loss": 0.0008149902569130063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.502338634237606,
      "flow_matching_loss": 0.013570186123251915,
      "grad_norm": 0.12470772117376328,
      "learning_rate": 1.0968320479136095e-07,
      "loss": 0.0146,
      "relation_loss": 0.0010797182330861688,
      "selected_pairs": 26826.5,
      "state_query_norm": 1100.7078979492187,
      "state_relation_loss": 0.0002647279718075879,
      "step": 29400,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010797182330861688,
      "weighted_relation_loss": 0.0010797182330861688
    },
    {
      "action_query_norm": 619.760498046875,
      "action_relation_loss": 0.0008673629490658641,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.511693171188025,
      "flow_matching_loss": 0.014288400951772928,
      "grad_norm": 0.07491739094257355,
      "learning_rate": 1.0606482786936767e-07,
      "loss": 0.0154,
      "relation_loss": 0.0011262394371442496,
      "selected_pairs": 26880.6,
      "state_query_norm": 1116.0439819335938,
      "state_relation_loss": 0.0002588764997199178,
      "step": 29410,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011262394371442496,
      "weighted_relation_loss": 0.0011262394371442496
    },
    {
      "action_query_norm": 617.5392883300781,
      "action_relation_loss": 0.0008300233806949109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.521047708138447,
      "flow_matching_loss": 0.01580086499452591,
      "grad_norm": 0.06381753087043762,
      "learning_rate": 1.0250707678329808e-07,
      "loss": 0.0169,
      "relation_loss": 0.0010948337847366929,
      "selected_pairs": 26209.2,
      "state_query_norm": 1101.5437866210937,
      "state_relation_loss": 0.0002648104084073566,
      "step": 29420,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0010948337847366929,
      "weighted_relation_loss": 0.0010948337847366929
    },
    {
      "action_query_norm": 616.9441589355469,
      "action_relation_loss": 0.0008203615376260132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.53040224508887,
      "flow_matching_loss": 0.014698297064751387,
      "grad_norm": 0.09268291294574738,
      "learning_rate": 9.900995585615525e-08,
      "loss": 0.0158,
      "relation_loss": 0.0010893513564951718,
      "selected_pairs": 26810.7,
      "state_query_norm": 1115.525390625,
      "state_relation_loss": 0.0002689898028620519,
      "step": 29430,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010893513564951718,
      "weighted_relation_loss": 0.0010893513564951718
    },
    {
      "action_query_norm": 619.061865234375,
      "action_relation_loss": 0.0008082732558250428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.53975678203929,
      "flow_matching_loss": 0.014592828135937453,
      "grad_norm": 0.09977005422115326,
      "learning_rate": 9.55734693372734e-08,
      "loss": 0.0157,
      "relation_loss": 0.0010745253646746277,
      "selected_pairs": 26461.8,
      "state_query_norm": 1097.2245849609376,
      "state_relation_loss": 0.0002662521175807342,
      "step": 29440,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010745253646746277,
      "weighted_relation_loss": 0.0010745253646746277
    },
    {
      "action_query_norm": 616.1823120117188,
      "action_relation_loss": 0.0008325383940245956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.54911131898971,
      "flow_matching_loss": 0.013804428651928902,
      "grad_norm": 0.09110639244318008,
      "learning_rate": 9.219762140231236e-08,
      "loss": 0.0149,
      "relation_loss": 0.0010932183009572326,
      "selected_pairs": 26781.2,
      "state_query_norm": 1107.2663452148438,
      "state_relation_loss": 0.00026067990984302014,
      "step": 29450,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0010932183009572326,
      "weighted_relation_loss": 0.0010932183009572326
    },
    {
      "action_query_norm": 612.0083068847656,
      "action_relation_loss": 0.000827961292816326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.55846585594013,
      "flow_matching_loss": 0.014338503405451775,
      "grad_norm": 0.09204660356044769,
      "learning_rate": 8.888241615322978e-08,
      "loss": 0.0154,
      "relation_loss": 0.00109110256889835,
      "selected_pairs": 27061.8,
      "state_query_norm": 1106.010791015625,
      "state_relation_loss": 0.0002631412964547053,
      "step": 29460,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.00109110256889835,
      "weighted_relation_loss": 0.00109110256889835
    },
    {
      "action_query_norm": 614.7427490234375,
      "action_relation_loss": 0.0008544363314285875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.567820392890553,
      "flow_matching_loss": 0.015209079440683126,
      "grad_norm": 0.11117984354496002,
      "learning_rate": 8.562785761833114e-08,
      "loss": 0.0163,
      "relation_loss": 0.0011180884670466185,
      "selected_pairs": 26577.7,
      "state_query_norm": 1102.9464477539063,
      "state_relation_loss": 0.0002636521341628395,
      "step": 29470,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011180884670466185,
      "weighted_relation_loss": 0.0011180884670466185
    },
    {
      "action_query_norm": 618.0057495117187,
      "action_relation_loss": 0.0008118387486319989,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.57717492984097,
      "flow_matching_loss": 0.01247135316953063,
      "grad_norm": 0.13411551713943481,
      "learning_rate": 8.243394975219753e-08,
      "loss": 0.0136,
      "relation_loss": 0.0010791706386953591,
      "selected_pairs": 27336.7,
      "state_query_norm": 1111.3411987304687,
      "state_relation_loss": 0.00026733188424259423,
      "step": 29480,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010791706386953591,
      "weighted_relation_loss": 0.0010791706386953591
    },
    {
      "action_query_norm": 614.8445861816406,
      "action_relation_loss": 0.0008195718342904002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.586529466791394,
      "flow_matching_loss": 0.014729801379144192,
      "grad_norm": 0.08941654860973358,
      "learning_rate": 7.930069643573568e-08,
      "loss": 0.0158,
      "relation_loss": 0.0010914458427578211,
      "selected_pairs": 27339.0,
      "state_query_norm": 1116.823779296875,
      "state_relation_loss": 0.0002718740230193362,
      "step": 29490,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010914458427578211,
      "weighted_relation_loss": 0.0010914458427578211
    },
    {
      "action_query_norm": 615.1526428222656,
      "action_relation_loss": 0.0008439281838946045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.595884003741816,
      "flow_matching_loss": 0.014345854055136442,
      "grad_norm": 0.07794954627752304,
      "learning_rate": 7.622810147614456e-08,
      "loss": 0.0155,
      "relation_loss": 0.0011098466464318335,
      "selected_pairs": 26743.2,
      "state_query_norm": 1105.7518798828125,
      "state_relation_loss": 0.0002659184581716545,
      "step": 29500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011098466464318335,
      "weighted_relation_loss": 0.0011098466464318335
    },
    {
      "action_query_norm": 615.3922485351562,
      "action_relation_loss": 0.000832680327584967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.605238540692234,
      "flow_matching_loss": 0.014090316835790873,
      "grad_norm": 0.10123098641633987,
      "learning_rate": 7.321616860690995e-08,
      "loss": 0.0152,
      "relation_loss": 0.0010927489725872873,
      "selected_pairs": 27046.2,
      "state_query_norm": 1106.7277465820312,
      "state_relation_loss": 0.00026006865518866105,
      "step": 29510,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010927489725872873,
      "weighted_relation_loss": 0.0010927489725872873
    },
    {
      "action_query_norm": 610.2355163574218,
      "action_relation_loss": 0.0008171478868462146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.614593077642656,
      "flow_matching_loss": 0.015263091027736663,
      "grad_norm": 0.07665140181779861,
      "learning_rate": 7.026490148782095e-08,
      "loss": 0.0163,
      "relation_loss": 0.0010786415892653166,
      "selected_pairs": 26688.2,
      "state_query_norm": 1103.4974243164063,
      "state_relation_loss": 0.00026149371406063436,
      "step": 29520,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010786415892653166,
      "weighted_relation_loss": 0.0010786415892653166
    },
    {
      "action_query_norm": 620.3502319335937,
      "action_relation_loss": 0.0008492477179970592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.623947614593078,
      "flow_matching_loss": 0.014373745396733284,
      "grad_norm": 0.06901586055755615,
      "learning_rate": 6.737430370494236e-08,
      "loss": 0.0155,
      "relation_loss": 0.0011210808879695833,
      "selected_pairs": 27292.7,
      "state_query_norm": 1094.8719482421875,
      "state_relation_loss": 0.00027183315833099184,
      "step": 29530,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011210808879695833,
      "weighted_relation_loss": 0.0011210808879695833
    },
    {
      "action_query_norm": 624.06201171875,
      "action_relation_loss": 0.0008148861117661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.6333021515435,
      "flow_matching_loss": 0.014632190205156803,
      "grad_norm": 0.0873405709862709,
      "learning_rate": 6.454437877062569e-08,
      "loss": 0.0157,
      "relation_loss": 0.0010836008237674831,
      "selected_pairs": 26433.9,
      "state_query_norm": 1113.8999877929687,
      "state_relation_loss": 0.0002687147294636816,
      "step": 29540,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0010836008237674831,
      "weighted_relation_loss": 0.0010836008237674831
    },
    {
      "action_query_norm": 614.7067016601562,
      "action_relation_loss": 0.0008076844387687743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.64265668849392,
      "flow_matching_loss": 0.01386419553309679,
      "grad_norm": 0.09311797469854355,
      "learning_rate": 6.177513012349256e-08,
      "loss": 0.0149,
      "relation_loss": 0.0010739724384620786,
      "selected_pairs": 26987.9,
      "state_query_norm": 1122.9862548828125,
      "state_relation_loss": 0.000266287979320623,
      "step": 29550,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010739724384620786,
      "weighted_relation_loss": 0.0010739724384620786
    },
    {
      "action_query_norm": 614.5992980957031,
      "action_relation_loss": 0.0008472867310047149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.65201122544434,
      "flow_matching_loss": 0.013607475068420171,
      "grad_norm": 0.08983369171619415,
      "learning_rate": 5.9066561128445775e-08,
      "loss": 0.0147,
      "relation_loss": 0.0011046894127503038,
      "selected_pairs": 27199.1,
      "state_query_norm": 1092.6670288085938,
      "state_relation_loss": 0.0002574026744696312,
      "step": 29560,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011046894127503038,
      "weighted_relation_loss": 0.0011046894127503038
    },
    {
      "action_query_norm": 616.6831787109375,
      "action_relation_loss": 0.0008216781076043844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.661365762394762,
      "flow_matching_loss": 0.01447568777948618,
      "grad_norm": 0.09112448990345001,
      "learning_rate": 5.6418675076641556e-08,
      "loss": 0.0156,
      "relation_loss": 0.001099531736690551,
      "selected_pairs": 26895.2,
      "state_query_norm": 1110.9281860351562,
      "state_relation_loss": 0.0002778536261757836,
      "step": 29570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001099531736690551,
      "weighted_relation_loss": 0.001099531736690551
    },
    {
      "action_query_norm": 626.1865905761719,
      "action_relation_loss": 0.0008209114719647914,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.67072029934518,
      "flow_matching_loss": 0.014339076448231935,
      "grad_norm": 0.07947231829166412,
      "learning_rate": 5.383147518552845e-08,
      "loss": 0.0154,
      "relation_loss": 0.0010805422090925277,
      "selected_pairs": 27481.5,
      "state_query_norm": 1115.9005004882813,
      "state_relation_loss": 0.0002596307371277362,
      "step": 29580,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0010805422090925277,
      "weighted_relation_loss": 0.0010805422090925277
    },
    {
      "action_query_norm": 623.0331848144531,
      "action_relation_loss": 0.0007913161301985383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.680074836295603,
      "flow_matching_loss": 0.014883198495954274,
      "grad_norm": 0.0754825696349144,
      "learning_rate": 5.1304964598786196e-08,
      "loss": 0.0159,
      "relation_loss": 0.0010600243578664959,
      "selected_pairs": 26572.2,
      "state_query_norm": 1112.973193359375,
      "state_relation_loss": 0.00026870823639910666,
      "step": 29590,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0010600243578664959,
      "weighted_relation_loss": 0.0010600243578664959
    },
    {
      "action_query_norm": 615.5589538574219,
      "action_relation_loss": 0.000826229655649513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.689429373246025,
      "flow_matching_loss": 0.01399777764454484,
      "grad_norm": 0.06576447933912277,
      "learning_rate": 4.883914638636467e-08,
      "loss": 0.0151,
      "relation_loss": 0.0010902176145464182,
      "selected_pairs": 26477.0,
      "state_query_norm": 1098.7216796875,
      "state_relation_loss": 0.00026398796326247974,
      "step": 29600,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0010902176145464182,
      "weighted_relation_loss": 0.0010902176145464182
    },
    {
      "action_query_norm": 617.2796630859375,
      "action_relation_loss": 0.0008255545748397708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.698783910196447,
      "flow_matching_loss": 0.01363518601283431,
      "grad_norm": 0.08903788775205612,
      "learning_rate": 4.643402354446713e-08,
      "loss": 0.0147,
      "relation_loss": 0.0010990556911565363,
      "selected_pairs": 27080.7,
      "state_query_norm": 1095.3600952148438,
      "state_relation_loss": 0.00027350110904080795,
      "step": 29610,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010990556911565363,
      "weighted_relation_loss": 0.0010990556911565363
    },
    {
      "action_query_norm": 622.5187072753906,
      "action_relation_loss": 0.0008282716386020184,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.708138447146865,
      "flow_matching_loss": 0.014330529142171144,
      "grad_norm": 0.0978851169347763,
      "learning_rate": 4.4089598995544766e-08,
      "loss": 0.0154,
      "relation_loss": 0.0010867806267924607,
      "selected_pairs": 27015.6,
      "state_query_norm": 1113.41181640625,
      "state_relation_loss": 0.000258508967817761,
      "step": 29620,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010867806267924607,
      "weighted_relation_loss": 0.0010867806267924607
    },
    {
      "action_query_norm": 619.4325317382812,
      "action_relation_loss": 0.0008010987425222993,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.717492984097287,
      "flow_matching_loss": 0.013136500120162964,
      "grad_norm": 0.07416737824678421,
      "learning_rate": 4.180587558829663e-08,
      "loss": 0.0142,
      "relation_loss": 0.0010684580425731837,
      "selected_pairs": 26874.9,
      "state_query_norm": 1105.4189819335938,
      "state_relation_loss": 0.0002673592898645438,
      "step": 29630,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010684580425731837,
      "weighted_relation_loss": 0.0010684580425731837
    },
    {
      "action_query_norm": 615.1431091308593,
      "action_relation_loss": 0.0008272126142401248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.72684752104771,
      "flow_matching_loss": 0.015701802726835013,
      "grad_norm": 0.08470812439918518,
      "learning_rate": 3.958285609765855e-08,
      "loss": 0.0168,
      "relation_loss": 0.0010936936945654453,
      "selected_pairs": 26825.2,
      "state_query_norm": 1104.7060302734376,
      "state_relation_loss": 0.00026648105995263904,
      "step": 29640,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010936936945654453,
      "weighted_relation_loss": 0.0010936936945654453
    },
    {
      "action_query_norm": 617.1728210449219,
      "action_relation_loss": 0.0007868081680499017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.736202057998128,
      "flow_matching_loss": 0.0134380005300045,
      "grad_norm": 0.07827011495828629,
      "learning_rate": 3.74205432248087e-08,
      "loss": 0.0145,
      "relation_loss": 0.0010504246572963893,
      "selected_pairs": 26736.3,
      "state_query_norm": 1105.0510131835938,
      "state_relation_loss": 0.0002636165008880198,
      "step": 29650,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010504246572963893,
      "weighted_relation_loss": 0.0010504246572963893
    },
    {
      "action_query_norm": 609.934130859375,
      "action_relation_loss": 0.0008474525529891253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.74555659494855,
      "flow_matching_loss": 0.014292063098400832,
      "grad_norm": 0.0889718234539032,
      "learning_rate": 3.531893959716204e-08,
      "loss": 0.0154,
      "relation_loss": 0.0011099604074843229,
      "selected_pairs": 26763.2,
      "state_query_norm": 1114.1520629882812,
      "state_relation_loss": 0.000262507850129623,
      "step": 29660,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011099604074843229,
      "weighted_relation_loss": 0.0011099604074843229
    },
    {
      "action_query_norm": 621.7381774902344,
      "action_relation_loss": 0.0008065849775448442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.75491113189897,
      "flow_matching_loss": 0.015313843544572592,
      "grad_norm": 0.1085202619433403,
      "learning_rate": 3.327804776837029e-08,
      "loss": 0.0164,
      "relation_loss": 0.0010781251708976924,
      "selected_pairs": 26829.4,
      "state_query_norm": 1105.3947265625,
      "state_relation_loss": 0.0002715401817113161,
      "step": 29670,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010781251708976924,
      "weighted_relation_loss": 0.0010781251708976924
    },
    {
      "action_query_norm": 615.8177734375,
      "action_relation_loss": 0.0008304319868329913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.764265668849394,
      "flow_matching_loss": 0.013971754070371389,
      "grad_norm": 0.07828331738710403,
      "learning_rate": 3.129787021829977e-08,
      "loss": 0.0151,
      "relation_loss": 0.0011084748257417232,
      "selected_pairs": 27131.7,
      "state_query_norm": 1102.2412841796875,
      "state_relation_loss": 0.0002780428156256676,
      "step": 29680,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011084748257417232,
      "weighted_relation_loss": 0.0011084748257417232
    },
    {
      "action_query_norm": 614.7712097167969,
      "action_relation_loss": 0.0008485973929055035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.773620205799812,
      "flow_matching_loss": 0.013652219250798226,
      "grad_norm": 0.0654301643371582,
      "learning_rate": 2.9378409353059133e-08,
      "loss": 0.0148,
      "relation_loss": 0.0011212451267056168,
      "selected_pairs": 26986.1,
      "state_query_norm": 1097.459765625,
      "state_relation_loss": 0.00027264773088973013,
      "step": 29690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011212451267056168,
      "weighted_relation_loss": 0.0011212451267056168
    },
    {
      "action_query_norm": 616.1333801269532,
      "action_relation_loss": 0.0008371192845515907,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.782974742750234,
      "flow_matching_loss": 0.015173452440649271,
      "grad_norm": 0.08279412239789963,
      "learning_rate": 2.7519667504971593e-08,
      "loss": 0.0163,
      "relation_loss": 0.0011091798078268766,
      "selected_pairs": 27147.4,
      "state_query_norm": 1101.9783935546875,
      "state_relation_loss": 0.00027206052909605204,
      "step": 29700,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0011091798078268766,
      "weighted_relation_loss": 0.0011091798078268766
    },
    {
      "action_query_norm": 621.9782653808594,
      "action_relation_loss": 0.0007828007626812905,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.792329279700656,
      "flow_matching_loss": 0.014501205924898387,
      "grad_norm": 0.10326807200908661,
      "learning_rate": 2.572164693258605e-08,
      "loss": 0.0156,
      "relation_loss": 0.0010523351025767624,
      "selected_pairs": 26616.7,
      "state_query_norm": 1105.9201416015626,
      "state_relation_loss": 0.00026953434862662105,
      "step": 29710,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010523351025767624,
      "weighted_relation_loss": 0.0010523351025767624
    },
    {
      "action_query_norm": 616.7877197265625,
      "action_relation_loss": 0.0008348431438207626,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.801683816651074,
      "flow_matching_loss": 0.012781769130378961,
      "grad_norm": 0.07790075242519379,
      "learning_rate": 2.3984349820665996e-08,
      "loss": 0.0139,
      "relation_loss": 0.0011114082881249487,
      "selected_pairs": 26397.5,
      "state_query_norm": 1103.2232177734375,
      "state_relation_loss": 0.0002765651326626539,
      "step": 29720,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0011114082881249487,
      "weighted_relation_loss": 0.0011114082881249487
    },
    {
      "action_query_norm": 620.3628051757812,
      "action_relation_loss": 0.0008382412197533995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.811038353601496,
      "flow_matching_loss": 0.015105026960372924,
      "grad_norm": 0.0806896984577179,
      "learning_rate": 2.2307778280189485e-08,
      "loss": 0.0162,
      "relation_loss": 0.001101447409018874,
      "selected_pairs": 26730.5,
      "state_query_norm": 1113.2739868164062,
      "state_relation_loss": 0.0002632061907206662,
      "step": 29730,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001101447409018874,
      "weighted_relation_loss": 0.001101447409018874
    },
    {
      "action_query_norm": 616.8568359375,
      "action_relation_loss": 0.0008086017216555775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.82039289055192,
      "flow_matching_loss": 0.013926022127270699,
      "grad_norm": 0.0830826461315155,
      "learning_rate": 2.0691934348354704e-08,
      "loss": 0.015,
      "relation_loss": 0.0010682870983146132,
      "selected_pairs": 26941.9,
      "state_query_norm": 1117.5600830078124,
      "state_relation_loss": 0.0002596853868453763,
      "step": 29740,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010682870983146132,
      "weighted_relation_loss": 0.0010682870983146132
    },
    {
      "action_query_norm": 619.30380859375,
      "action_relation_loss": 0.0008102457271888852,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.82974742750234,
      "flow_matching_loss": 0.013420376181602477,
      "grad_norm": 0.10201393812894821,
      "learning_rate": 1.9136819988557763e-08,
      "loss": 0.0145,
      "relation_loss": 0.0010757075157016515,
      "selected_pairs": 27041.0,
      "state_query_norm": 1109.3293823242188,
      "state_relation_loss": 0.0002654617550433613,
      "step": 29750,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010757075157016515,
      "weighted_relation_loss": 0.0010757075157016515
    },
    {
      "action_query_norm": 616.3905456542968,
      "action_relation_loss": 0.0008619187283329666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.83910196445276,
      "flow_matching_loss": 0.014059675019234419,
      "grad_norm": 0.09244580566883087,
      "learning_rate": 1.7642437090414908e-08,
      "loss": 0.0152,
      "relation_loss": 0.0011291198432445526,
      "selected_pairs": 26848.6,
      "state_query_norm": 1112.11953125,
      "state_relation_loss": 0.00026720112800830974,
      "step": 29760,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0011291198432445526,
      "weighted_relation_loss": 0.0011291198432445526
    },
    {
      "action_query_norm": 619.7566589355469,
      "action_relation_loss": 0.0008596215164288878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.84845650140318,
      "flow_matching_loss": 0.013982641603797674,
      "grad_norm": 0.09297584742307663,
      "learning_rate": 1.6208787469734755e-08,
      "loss": 0.0151,
      "relation_loss": 0.0011352556641213596,
      "selected_pairs": 27035.4,
      "state_query_norm": 1103.363916015625,
      "state_relation_loss": 0.00027563416224438695,
      "step": 29770,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011352556641213596,
      "weighted_relation_loss": 0.0011352556641213596
    },
    {
      "action_query_norm": 619.815185546875,
      "action_relation_loss": 0.0008370308263693005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.857811038353603,
      "flow_matching_loss": 0.01298196380957961,
      "grad_norm": 0.07917752861976624,
      "learning_rate": 1.483587286854604e-08,
      "loss": 0.0141,
      "relation_loss": 0.0011079163872636856,
      "selected_pairs": 26507.4,
      "state_query_norm": 1117.1444213867187,
      "state_relation_loss": 0.00027088556671515105,
      "step": 29780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011079163872636856,
      "weighted_relation_loss": 0.0011079163872636856
    },
    {
      "action_query_norm": 623.3654846191406,
      "action_relation_loss": 0.0008477108203805983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.86716557530402,
      "flow_matching_loss": 0.014179089851677418,
      "grad_norm": 0.07705258578062057,
      "learning_rate": 1.3523694955064336e-08,
      "loss": 0.0153,
      "relation_loss": 0.0011163745424710215,
      "selected_pairs": 26895.5,
      "state_query_norm": 1122.220556640625,
      "state_relation_loss": 0.0002686637337319553,
      "step": 29790,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011163745424710215,
      "weighted_relation_loss": 0.0011163745424710215
    },
    {
      "action_query_norm": 621.631494140625,
      "action_relation_loss": 0.0008499476418364793,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.876520112254443,
      "flow_matching_loss": 0.013760673440992832,
      "grad_norm": 0.06696468591690063,
      "learning_rate": 1.2272255323708681e-08,
      "loss": 0.0149,
      "relation_loss": 0.0011090298881754278,
      "selected_pairs": 27479.5,
      "state_query_norm": 1099.3919555664063,
      "state_relation_loss": 0.0002590822579804808,
      "step": 29800,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0011090298881754278,
      "weighted_relation_loss": 0.0011090298881754278
    },
    {
      "action_query_norm": 620.7939270019531,
      "action_relation_loss": 0.000807645358145237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.885874649204865,
      "flow_matching_loss": 0.01527192872017622,
      "grad_norm": 0.07707979530096054,
      "learning_rate": 1.1081555495096042e-08,
      "loss": 0.0164,
      "relation_loss": 0.0010855937260203063,
      "selected_pairs": 27056.0,
      "state_query_norm": 1090.5203857421875,
      "state_relation_loss": 0.0002779483562335372,
      "step": 29810,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010855937260203063,
      "weighted_relation_loss": 0.0010855937260203063
    },
    {
      "action_query_norm": 619.1930908203125,
      "action_relation_loss": 0.0007960167655255646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.895229186155284,
      "flow_matching_loss": 0.01356249935925007,
      "grad_norm": 0.09096585214138031,
      "learning_rate": 9.951596916041306e-09,
      "loss": 0.0146,
      "relation_loss": 0.0010548573103733362,
      "selected_pairs": 26649.5,
      "state_query_norm": 1113.0029174804688,
      "state_relation_loss": 0.0002588405361166224,
      "step": 29820,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0010548573103733362,
      "weighted_relation_loss": 0.0010548573103733362
    },
    {
      "action_query_norm": 616.6369201660157,
      "action_relation_loss": 0.0008453472459223121,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.904583723105706,
      "flow_matching_loss": 0.013965188525617123,
      "grad_norm": 0.08056356757879257,
      "learning_rate": 8.882380959551739e-09,
      "loss": 0.0151,
      "relation_loss": 0.0011163500021211803,
      "selected_pairs": 26952.6,
      "state_query_norm": 1118.1204711914063,
      "state_relation_loss": 0.0002710027707507834,
      "step": 29830,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011163500021211803,
      "weighted_relation_loss": 0.0011163500021211803
    },
    {
      "action_query_norm": 618.125927734375,
      "action_relation_loss": 0.0007994654821231961,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.913938260056128,
      "flow_matching_loss": 0.013632963225245476,
      "grad_norm": 0.10389513522386551,
      "learning_rate": 7.873908924821428e-09,
      "loss": 0.0147,
      "relation_loss": 0.001053426373982802,
      "selected_pairs": 26874.9,
      "state_query_norm": 1104.583154296875,
      "state_relation_loss": 0.00025396088749403135,
      "step": 29840,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001053426373982802,
      "weighted_relation_loss": 0.001053426373982802
    },
    {
      "action_query_norm": 617.3119812011719,
      "action_relation_loss": 0.0008156041207257658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.92329279700655,
      "flow_matching_loss": 0.01475926786661148,
      "grad_norm": 0.06755360215902328,
      "learning_rate": 6.926182037242379e-09,
      "loss": 0.0158,
      "relation_loss": 0.0010719592333771288,
      "selected_pairs": 27057.3,
      "state_query_norm": 1117.337744140625,
      "state_relation_loss": 0.00025635512138251216,
      "step": 29850,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010719592333771288,
      "weighted_relation_loss": 0.0010719592333771288
    },
    {
      "action_query_norm": 621.4758178710938,
      "action_relation_loss": 0.000845705548999831,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.932647333956968,
      "flow_matching_loss": 0.014188439305871724,
      "grad_norm": 0.09278814494609833,
      "learning_rate": 6.039201448387876e-09,
      "loss": 0.0153,
      "relation_loss": 0.0011105747194960713,
      "selected_pairs": 27502.6,
      "state_query_norm": 1111.3801879882812,
      "state_relation_loss": 0.00026486915739951653,
      "step": 29860,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0011105747194960713,
      "weighted_relation_loss": 0.0011105747194960713
    },
    {
      "action_query_norm": 616.2384765625,
      "action_relation_loss": 0.0008521595154888928,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.94200187090739,
      "flow_matching_loss": 0.016150588169693947,
      "grad_norm": 0.07604807615280151,
      "learning_rate": 5.212968236029125e-09,
      "loss": 0.0173,
      "relation_loss": 0.0011162811191752554,
      "selected_pairs": 27643.9,
      "state_query_norm": 1107.8594360351562,
      "state_relation_loss": 0.0002641216095071286,
      "step": 29870,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0011162811191752554,
      "weighted_relation_loss": 0.0011162811191752554
    },
    {
      "action_query_norm": 628.392578125,
      "action_relation_loss": 0.000795466749696061,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.951356407857812,
      "flow_matching_loss": 0.01500521246343851,
      "grad_norm": 0.126658096909523,
      "learning_rate": 4.447483404118602e-09,
      "loss": 0.0161,
      "relation_loss": 0.0010518035560380667,
      "selected_pairs": 26883.5,
      "state_query_norm": 1112.0451416015626,
      "state_relation_loss": 0.00025633678887970746,
      "step": 29880,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010518035560380667,
      "weighted_relation_loss": 0.0010518035560380667
    },
    {
      "action_query_norm": 614.0675537109375,
      "action_relation_loss": 0.0008635733858682216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.96071094480823,
      "flow_matching_loss": 0.015769664943218232,
      "grad_norm": 0.09872947633266449,
      "learning_rate": 3.742747882784503e-09,
      "loss": 0.0169,
      "relation_loss": 0.0011278990423306823,
      "selected_pairs": 26183.9,
      "state_query_norm": 1107.9262084960938,
      "state_relation_loss": 0.00026432565646246077,
      "step": 29890,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0011278990423306823,
      "weighted_relation_loss": 0.0011278990423306823
    },
    {
      "action_query_norm": 620.2516540527344,
      "action_relation_loss": 0.0007959993788972497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.970065481758652,
      "flow_matching_loss": 0.013983285427093506,
      "grad_norm": 0.0794258639216423,
      "learning_rate": 3.0987625283473987e-09,
      "loss": 0.015,
      "relation_loss": 0.001061515515903011,
      "selected_pairs": 26968.3,
      "state_query_norm": 1105.0541137695313,
      "state_relation_loss": 0.00026551615446805953,
      "step": 29900,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001061515515903011,
      "weighted_relation_loss": 0.001061515515903011
    },
    {
      "action_query_norm": 620.4952514648437,
      "action_relation_loss": 0.0008447810541838408,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.979420018709074,
      "flow_matching_loss": 0.014603567961603402,
      "grad_norm": 0.08143659681081772,
      "learning_rate": 2.515528123320232e-09,
      "loss": 0.0157,
      "relation_loss": 0.0010969626600854099,
      "selected_pairs": 26639.2,
      "state_query_norm": 1111.2832275390624,
      "state_relation_loss": 0.0002521816189982928,
      "step": 29910,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0010969626600854099,
      "weighted_relation_loss": 0.0010969626600854099
    },
    {
      "action_query_norm": 618.8040466308594,
      "action_relation_loss": 0.0008171054767444729,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.988774555659496,
      "flow_matching_loss": 0.013478819746524096,
      "grad_norm": 0.0737033486366272,
      "learning_rate": 1.9930453763750133e-09,
      "loss": 0.0146,
      "relation_loss": 0.001078131760004908,
      "selected_pairs": 26734.2,
      "state_query_norm": 1112.250830078125,
      "state_relation_loss": 0.00026102628180524334,
      "step": 29920,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001078131760004908,
      "weighted_relation_loss": 0.001078131760004908
    },
    {
      "action_query_norm": 616.320849609375,
      "action_relation_loss": 0.0008508775499649346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.998129092609915,
      "flow_matching_loss": 0.014839960448443889,
      "grad_norm": 0.09575824439525604,
      "learning_rate": 1.5313149223872281e-09,
      "loss": 0.016,
      "relation_loss": 0.0011172773432917894,
      "selected_pairs": 26664.9,
      "state_query_norm": 1105.6996215820313,
      "state_relation_loss": 0.00026639979041647165,
      "step": 29930,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011172773432917894,
      "weighted_relation_loss": 0.0011172773432917894
    },
    {
      "action_query_norm": 619.80966796875,
      "action_relation_loss": 0.0008553729217965156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.007483629560337,
      "flow_matching_loss": 0.014511366188526154,
      "grad_norm": 0.08950380980968475,
      "learning_rate": 1.1303373224025305e-09,
      "loss": 0.0156,
      "relation_loss": 0.0011114425491541625,
      "selected_pairs": 24608.7,
      "state_query_norm": 1098.8938598632812,
      "state_relation_loss": 0.0002560696331784129,
      "step": 29940,
      "valid_action_samples": 215.5,
      "weighted_auxiliary_loss": 0.0011114425491541625,
      "weighted_relation_loss": 0.0011114425491541625
    },
    {
      "action_query_norm": 618.6730407714844,
      "action_relation_loss": 0.0007886443403549492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.01683816651076,
      "flow_matching_loss": 0.014219619892537593,
      "grad_norm": 0.08460482954978943,
      "learning_rate": 7.901130636367438e-10,
      "loss": 0.0153,
      "relation_loss": 0.0010501272510737181,
      "selected_pairs": 26558.4,
      "state_query_norm": 1108.6057373046874,
      "state_relation_loss": 0.0002614828903460875,
      "step": 29950,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010501272510737181,
      "weighted_relation_loss": 0.0010501272510737181
    },
    {
      "action_query_norm": 612.1407348632813,
      "action_relation_loss": 0.0008282601775135845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.026192703461177,
      "flow_matching_loss": 0.014957841392606497,
      "grad_norm": 0.07222571223974228,
      "learning_rate": 5.106425595036158e-10,
      "loss": 0.0161,
      "relation_loss": 0.0010948510142043232,
      "selected_pairs": 26233.1,
      "state_query_norm": 1129.1011840820313,
      "state_relation_loss": 0.00026659083960112184,
      "step": 29960,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0010948510142043232,
      "weighted_relation_loss": 0.0010948510142043232
    },
    {
      "action_query_norm": 616.2565246582031,
      "action_relation_loss": 0.0008228482271078974,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.0355472404116,
      "flow_matching_loss": 0.0133236950263381,
      "grad_norm": 0.07240737974643707,
      "learning_rate": 2.9192614958706286e-10,
      "loss": 0.0144,
      "relation_loss": 0.001095154054928571,
      "selected_pairs": 26875.3,
      "state_query_norm": 1119.4534301757812,
      "state_relation_loss": 0.00027230582491029055,
      "step": 29970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001095154054928571,
      "weighted_relation_loss": 0.001095154054928571
    },
    {
      "action_query_norm": 625.4807250976562,
      "action_relation_loss": 0.0008233656350057572,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.04490177736202,
      "flow_matching_loss": 0.014654789306223392,
      "grad_norm": 0.09076283127069473,
      "learning_rate": 1.3396409964117064e-10,
      "loss": 0.0157,
      "relation_loss": 0.0010906691197305918,
      "selected_pairs": 26808.6,
      "state_query_norm": 1118.3710815429688,
      "state_relation_loss": 0.0002673034861800261,
      "step": 29980,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010906691197305918,
      "weighted_relation_loss": 0.0010906691197305918
    },
    {
      "action_query_norm": 620.5613525390625,
      "action_relation_loss": 0.0008181019744370133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.054256314312443,
      "flow_matching_loss": 0.013550006877630949,
      "grad_norm": 0.08824998885393143,
      "learning_rate": 3.6756601606846574e-11,
      "loss": 0.0146,
      "relation_loss": 0.001086013426538557,
      "selected_pairs": 27151.3,
      "state_query_norm": 1094.9636840820312,
      "state_relation_loss": 0.0002679114550119266,
      "step": 29990,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001086013426538557,
      "weighted_relation_loss": 0.001086013426538557
    },
    {
      "action_query_norm": 620.515087890625,
      "action_relation_loss": 0.0008586762007325888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.06361085126286,
      "flow_matching_loss": 0.014997001178562641,
      "grad_norm": 0.10004635155200958,
      "learning_rate": 3.0377360626943076e-13,
      "loss": 0.0161,
      "relation_loss": 0.0011240038729738444,
      "selected_pairs": 27014.0,
      "state_query_norm": 1114.5035278320313,
      "state_relation_loss": 0.00026532766787568107,
      "step": 30000,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011240038729738444,
      "weighted_relation_loss": 0.0011240038729738444
    },
    {
      "epoch": 28.06361085126286,
      "step": 30000,
      "total_flos": 0.0,
      "train_loss": 0.04157031177828709,
      "train_runtime": 58947.3734,
      "train_samples_per_second": 130.286,
      "train_steps_per_second": 0.509
    }
  ],
  "logging_steps": 10,
  "max_steps": 30000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 29,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}