{
  "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": 45.2546501159668,
      "action_relation_loss": 0.05340769365429878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.009354536950420954,
      "flow_matching_loss": 0.9186811864376068,
      "grad_norm": 7.205082416534424,
      "learning_rate": 6.000000000000001e-07,
      "loss": 1.042,
      "relation_loss": 0.12333939075469971,
      "selected_pairs": 26709.2,
      "state_query_norm": 45.25465393066406,
      "state_relation_loss": 0.06993169784545898,
      "step": 10,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.12333939075469971,
      "weighted_relation_loss": 0.12333939075469971
    },
    {
      "action_query_norm": 45.254662704467776,
      "action_relation_loss": 0.05570003427565098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.018709073900841908,
      "flow_matching_loss": 0.9052734434604645,
      "grad_norm": 6.959439754486084,
      "learning_rate": 1.2666666666666667e-06,
      "loss": 1.0282,
      "relation_loss": 0.12291600704193115,
      "selected_pairs": 27177.8,
      "state_query_norm": 45.25465393066406,
      "state_relation_loss": 0.06721597239375114,
      "step": 20,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.12291600704193115,
      "weighted_relation_loss": 0.12291600704193115
    },
    {
      "action_query_norm": 45.254676818847656,
      "action_relation_loss": 0.0516024362295866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.02806361085126286,
      "flow_matching_loss": 0.7601522088050843,
      "grad_norm": 4.142604351043701,
      "learning_rate": 1.9333333333333336e-06,
      "loss": 0.8761,
      "relation_loss": 0.11594739630818367,
      "selected_pairs": 26755.1,
      "state_query_norm": 45.25465393066406,
      "state_relation_loss": 0.06434495970606804,
      "step": 30,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.11594739630818367,
      "weighted_relation_loss": 0.11594739630818367
    },
    {
      "action_query_norm": 45.25470123291016,
      "action_relation_loss": 0.05144812539219856,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.037418147801683815,
      "flow_matching_loss": 0.5641243159770966,
      "grad_norm": 1.395561933517456,
      "learning_rate": 2.6e-06,
      "loss": 0.6771,
      "relation_loss": 0.11293624192476273,
      "selected_pairs": 27012.3,
      "state_query_norm": 45.25465393066406,
      "state_relation_loss": 0.06148811653256416,
      "step": 40,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.11293624192476273,
      "weighted_relation_loss": 0.11293624192476273
    },
    {
      "action_query_norm": 45.254717636108396,
      "action_relation_loss": 0.04782609865069389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.04677268475210477,
      "flow_matching_loss": 0.43103312849998476,
      "grad_norm": 1.229891061782837,
      "learning_rate": 3.2666666666666666e-06,
      "loss": 0.5342,
      "relation_loss": 0.10315800979733467,
      "selected_pairs": 26396.6,
      "state_query_norm": 45.25463523864746,
      "state_relation_loss": 0.055331910401582717,
      "step": 50,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.10315800979733467,
      "weighted_relation_loss": 0.10315800979733467
    },
    {
      "action_query_norm": 45.25471992492676,
      "action_relation_loss": 0.04565924257040024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.05612722170252572,
      "flow_matching_loss": 0.35517717003822324,
      "grad_norm": 1.0416944026947021,
      "learning_rate": 3.9333333333333335e-06,
      "loss": 0.4528,
      "relation_loss": 0.09760460034012794,
      "selected_pairs": 27364.8,
      "state_query_norm": 45.25461120605469,
      "state_relation_loss": 0.05194535627961159,
      "step": 60,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.09760460034012794,
      "weighted_relation_loss": 0.09760460034012794
    },
    {
      "action_query_norm": 45.254721069335936,
      "action_relation_loss": 0.043479495123028754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.06548175865294668,
      "flow_matching_loss": 0.30528165996074674,
      "grad_norm": 0.9752121567726135,
      "learning_rate": 4.6e-06,
      "loss": 0.3971,
      "relation_loss": 0.09177463501691818,
      "selected_pairs": 26454.3,
      "state_query_norm": 45.2545581817627,
      "state_relation_loss": 0.04829514026641846,
      "step": 70,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.09177463501691818,
      "weighted_relation_loss": 0.09177463501691818
    },
    {
      "action_query_norm": 45.25474319458008,
      "action_relation_loss": 0.042540339007973674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.07483629560336763,
      "flow_matching_loss": 0.270125550031662,
      "grad_norm": 0.7042507529258728,
      "learning_rate": 5.266666666666667e-06,
      "loss": 0.356,
      "relation_loss": 0.08583849817514419,
      "selected_pairs": 26614.2,
      "state_query_norm": 45.25447235107422,
      "state_relation_loss": 0.04329815916717052,
      "step": 80,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.08583849817514419,
      "weighted_relation_loss": 0.08583849817514419
    },
    {
      "action_query_norm": 45.254787826538085,
      "action_relation_loss": 0.03784787580370903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.08419083255378859,
      "flow_matching_loss": 0.25218638479709626,
      "grad_norm": 0.632448673248291,
      "learning_rate": 5.933333333333334e-06,
      "loss": 0.3289,
      "relation_loss": 0.07672819793224335,
      "selected_pairs": 26897.3,
      "state_query_norm": 45.25433464050293,
      "state_relation_loss": 0.03888032175600529,
      "step": 90,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.07672819793224335,
      "weighted_relation_loss": 0.07672819793224335
    },
    {
      "action_query_norm": 45.25485534667969,
      "action_relation_loss": 0.036040821298956874,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.09354536950420954,
      "flow_matching_loss": 0.22028050124645232,
      "grad_norm": 0.47354429960250854,
      "learning_rate": 6.6e-06,
      "loss": 0.2901,
      "relation_loss": 0.06979024335741997,
      "selected_pairs": 26853.0,
      "state_query_norm": 45.25419120788574,
      "state_relation_loss": 0.0337494220584631,
      "step": 100,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.06979024335741997,
      "weighted_relation_loss": 0.06979024335741997
    },
    {
      "action_query_norm": 45.254927444458005,
      "action_relation_loss": 0.0339972123503685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1028999064546305,
      "flow_matching_loss": 0.20416683554649354,
      "grad_norm": 0.45853376388549805,
      "learning_rate": 7.266666666666668e-06,
      "loss": 0.2673,
      "relation_loss": 0.06312858127057552,
      "selected_pairs": 26207.9,
      "state_query_norm": 45.254081344604494,
      "state_relation_loss": 0.02913136836141348,
      "step": 110,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.06312858127057552,
      "weighted_relation_loss": 0.06312858127057552
    },
    {
      "action_query_norm": 45.25500640869141,
      "action_relation_loss": 0.03382673989981413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.11225444340505145,
      "flow_matching_loss": 0.19342466592788696,
      "grad_norm": 0.6365541815757751,
      "learning_rate": 7.933333333333334e-06,
      "loss": 0.2513,
      "relation_loss": 0.057843136787414554,
      "selected_pairs": 26649.7,
      "state_query_norm": 45.25405616760254,
      "state_relation_loss": 0.024016397073864937,
      "step": 120,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.057843136787414554,
      "weighted_relation_loss": 0.057843136787414554
    },
    {
      "action_query_norm": 45.255110549926755,
      "action_relation_loss": 0.03353781457990408,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1216089803554724,
      "flow_matching_loss": 0.18607258498668672,
      "grad_norm": 0.7602685689926147,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.2404,
      "relation_loss": 0.054357655718922614,
      "selected_pairs": 27342.7,
      "state_query_norm": 45.2541446685791,
      "state_relation_loss": 0.02081984132528305,
      "step": 130,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.054357655718922614,
      "weighted_relation_loss": 0.054357655718922614
    },
    {
      "action_query_norm": 45.25524826049805,
      "action_relation_loss": 0.031831499002873895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.13096351730589337,
      "flow_matching_loss": 0.1784786656498909,
      "grad_norm": 0.6745913624763489,
      "learning_rate": 9.266666666666667e-06,
      "loss": 0.2285,
      "relation_loss": 0.05005505084991455,
      "selected_pairs": 27174.7,
      "state_query_norm": 45.254323959350586,
      "state_relation_loss": 0.018223552405834197,
      "step": 140,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.05005505084991455,
      "weighted_relation_loss": 0.05005505084991455
    },
    {
      "action_query_norm": 45.25536575317383,
      "action_relation_loss": 0.030255170539021492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.1403180542563143,
      "flow_matching_loss": 0.16992716044187545,
      "grad_norm": 0.5677897930145264,
      "learning_rate": 9.933333333333334e-06,
      "loss": 0.2167,
      "relation_loss": 0.046773720905184744,
      "selected_pairs": 26826.3,
      "state_query_norm": 45.25457077026367,
      "state_relation_loss": 0.01651855055242777,
      "step": 150,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.046773720905184744,
      "weighted_relation_loss": 0.046773720905184744
    },
    {
      "action_query_norm": 45.2554573059082,
      "action_relation_loss": 0.031104502268135548,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.14967259120673526,
      "flow_matching_loss": 0.15539080798625945,
      "grad_norm": 0.4092671871185303,
      "learning_rate": 1.06e-05,
      "loss": 0.2021,
      "relation_loss": 0.04667664766311645,
      "selected_pairs": 26622.1,
      "state_query_norm": 45.25482864379883,
      "state_relation_loss": 0.01557214604690671,
      "step": 160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.04667664766311645,
      "weighted_relation_loss": 0.04667664766311645
    },
    {
      "action_query_norm": 45.255516815185544,
      "action_relation_loss": 0.028075487725436686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.15902712815715622,
      "flow_matching_loss": 0.14875063300132751,
      "grad_norm": 0.45903268456459045,
      "learning_rate": 1.1266666666666667e-05,
      "loss": 0.1898,
      "relation_loss": 0.04109590724110603,
      "selected_pairs": 27272.6,
      "state_query_norm": 45.25500144958496,
      "state_relation_loss": 0.01302041970193386,
      "step": 170,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.04109590724110603,
      "weighted_relation_loss": 0.04109590724110603
    },
    {
      "action_query_norm": 45.255560302734374,
      "action_relation_loss": 0.029191733710467814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.16838166510757718,
      "flow_matching_loss": 0.14126872271299362,
      "grad_norm": 0.526023268699646,
      "learning_rate": 1.1933333333333333e-05,
      "loss": 0.182,
      "relation_loss": 0.0407618947327137,
      "selected_pairs": 27079.9,
      "state_query_norm": 45.255113983154295,
      "state_relation_loss": 0.011570161022245885,
      "step": 180,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0407618947327137,
      "weighted_relation_loss": 0.0407618947327137
    },
    {
      "action_query_norm": 45.25559539794922,
      "action_relation_loss": 0.02744417544454336,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.17773620205799812,
      "flow_matching_loss": 0.14210095405578613,
      "grad_norm": 0.5831652879714966,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.1806,
      "relation_loss": 0.038500251993536946,
      "selected_pairs": 26875.0,
      "state_query_norm": 45.25523452758789,
      "state_relation_loss": 0.011056076176464558,
      "step": 190,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.038500251993536946,
      "weighted_relation_loss": 0.038500251993536946
    },
    {
      "action_query_norm": 45.25559616088867,
      "action_relation_loss": 0.026968119479715825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.18709073900841908,
      "flow_matching_loss": 0.13882389068603515,
      "grad_norm": 0.371195524930954,
      "learning_rate": 1.3266666666666666e-05,
      "loss": 0.1763,
      "relation_loss": 0.03748415745794773,
      "selected_pairs": 26918.3,
      "state_query_norm": 45.25522956848145,
      "state_relation_loss": 0.010516037791967392,
      "step": 200,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.03748415745794773,
      "weighted_relation_loss": 0.03748415745794773
    },
    {
      "action_query_norm": 45.2555435180664,
      "action_relation_loss": 0.026001301035284995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.19644527595884004,
      "flow_matching_loss": 0.13043080344796182,
      "grad_norm": 0.6865203976631165,
      "learning_rate": 1.3933333333333334e-05,
      "loss": 0.1658,
      "relation_loss": 0.035338271781802176,
      "selected_pairs": 27081.4,
      "state_query_norm": 45.25526657104492,
      "state_relation_loss": 0.009336971119046212,
      "step": 210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.035338271781802176,
      "weighted_relation_loss": 0.035338271781802176
    },
    {
      "action_query_norm": 45.255522918701175,
      "action_relation_loss": 0.02681019399315119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.205799812909261,
      "flow_matching_loss": 0.13263167440891266,
      "grad_norm": 0.6951103210449219,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.1683,
      "relation_loss": 0.03565327525138855,
      "selected_pairs": 26688.9,
      "state_query_norm": 45.25487518310547,
      "state_relation_loss": 0.008843081071972847,
      "step": 220,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.03565327525138855,
      "weighted_relation_loss": 0.03565327525138855
    },
    {
      "action_query_norm": 45.255591201782224,
      "action_relation_loss": 0.024350964277982712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.21515434985968196,
      "flow_matching_loss": 0.1280208446085453,
      "grad_norm": 0.6017796397209167,
      "learning_rate": 1.5266666666666667e-05,
      "loss": 0.161,
      "relation_loss": 0.03298072591423988,
      "selected_pairs": 26947.2,
      "state_query_norm": 45.25430526733398,
      "state_relation_loss": 0.008629762008786202,
      "step": 230,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.03298072591423988,
      "weighted_relation_loss": 0.03298072591423988
    },
    {
      "action_query_norm": 45.25572128295899,
      "action_relation_loss": 0.023629947938024996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2245088868101029,
      "flow_matching_loss": 0.1257811650633812,
      "grad_norm": 0.5031418800354004,
      "learning_rate": 1.5933333333333332e-05,
      "loss": 0.1574,
      "relation_loss": 0.03163376022130251,
      "selected_pairs": 26659.7,
      "state_query_norm": 45.25411872863769,
      "state_relation_loss": 0.008003811538219451,
      "step": 240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.03163376022130251,
      "weighted_relation_loss": 0.03163376022130251
    },
    {
      "action_query_norm": 45.255760192871094,
      "action_relation_loss": 0.02301287539303303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.23386342376052385,
      "flow_matching_loss": 0.12486137077212334,
      "grad_norm": 0.4587470591068268,
      "learning_rate": 1.66e-05,
      "loss": 0.1556,
      "relation_loss": 0.030767418816685675,
      "selected_pairs": 26918.7,
      "state_query_norm": 45.253844833374025,
      "state_relation_loss": 0.007754543237388134,
      "step": 250,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.030767418816685675,
      "weighted_relation_loss": 0.030767418816685675
    },
    {
      "action_query_norm": 45.25579261779785,
      "action_relation_loss": 0.02234683632850647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2432179607109448,
      "flow_matching_loss": 0.12138062417507171,
      "grad_norm": 0.5096615552902222,
      "learning_rate": 1.726666666666667e-05,
      "loss": 0.151,
      "relation_loss": 0.029604020155966283,
      "selected_pairs": 26196.7,
      "state_query_norm": 45.25330696105957,
      "state_relation_loss": 0.0072571839671581985,
      "step": 260,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.029604020155966283,
      "weighted_relation_loss": 0.029604020155966283
    },
    {
      "action_query_norm": 45.25577621459961,
      "action_relation_loss": 0.02142379302531481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.25257249766136575,
      "flow_matching_loss": 0.11405032873153687,
      "grad_norm": 0.4877815544605255,
      "learning_rate": 1.7933333333333337e-05,
      "loss": 0.1427,
      "relation_loss": 0.028630470857024193,
      "selected_pairs": 27476.1,
      "state_query_norm": 45.25326156616211,
      "state_relation_loss": 0.007206677878275514,
      "step": 270,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.028630470857024193,
      "weighted_relation_loss": 0.028630470857024193
    },
    {
      "action_query_norm": 45.255548095703126,
      "action_relation_loss": 0.021677988581359386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.26192703461178674,
      "flow_matching_loss": 0.10981547310948372,
      "grad_norm": 0.548502504825592,
      "learning_rate": 1.86e-05,
      "loss": 0.1384,
      "relation_loss": 0.02857779935002327,
      "selected_pairs": 27263.9,
      "state_query_norm": 45.25312805175781,
      "state_relation_loss": 0.006899810675531626,
      "step": 280,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.02857779935002327,
      "weighted_relation_loss": 0.02857779935002327
    },
    {
      "action_query_norm": 45.25547904968262,
      "action_relation_loss": 0.021714557707309724,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.27128157156220767,
      "flow_matching_loss": 0.11467624083161354,
      "grad_norm": 0.69101881980896,
      "learning_rate": 1.926666666666667e-05,
      "loss": 0.1431,
      "relation_loss": 0.02842134181410074,
      "selected_pairs": 25961.5,
      "state_query_norm": 45.25268707275391,
      "state_relation_loss": 0.006706783920526505,
      "step": 290,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.02842134181410074,
      "weighted_relation_loss": 0.02842134181410074
    },
    {
      "action_query_norm": 45.255553817749025,
      "action_relation_loss": 0.02073005251586437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2806361085126286,
      "flow_matching_loss": 0.11313454061746597,
      "grad_norm": 0.5375874638557434,
      "learning_rate": 1.9933333333333334e-05,
      "loss": 0.1405,
      "relation_loss": 0.02732971478253603,
      "selected_pairs": 26620.2,
      "state_query_norm": 45.25251159667969,
      "state_relation_loss": 0.006599662266671658,
      "step": 300,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.02732971478253603,
      "weighted_relation_loss": 0.02732971478253603
    },
    {
      "action_query_norm": 45.25547866821289,
      "action_relation_loss": 0.020303963869810104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2899906454630496,
      "flow_matching_loss": 0.11110922023653984,
      "grad_norm": 0.5202985405921936,
      "learning_rate": 2.06e-05,
      "loss": 0.1375,
      "relation_loss": 0.02636074647307396,
      "selected_pairs": 26466.7,
      "state_query_norm": 45.25210762023926,
      "state_relation_loss": 0.006056782510131598,
      "step": 310,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.02636074647307396,
      "weighted_relation_loss": 0.02636074647307396
    },
    {
      "action_query_norm": 45.255184173583984,
      "action_relation_loss": 0.021514091454446316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.2993451824134705,
      "flow_matching_loss": 0.11011009737849235,
      "grad_norm": 0.41160497069358826,
      "learning_rate": 2.1266666666666667e-05,
      "loss": 0.1376,
      "relation_loss": 0.02749122753739357,
      "selected_pairs": 27105.2,
      "state_query_norm": 45.25157852172852,
      "state_relation_loss": 0.005977136269211769,
      "step": 320,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.02749122753739357,
      "weighted_relation_loss": 0.02749122753739357
    },
    {
      "action_query_norm": 45.25495071411133,
      "action_relation_loss": 0.02035189848393202,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3086997193638915,
      "flow_matching_loss": 0.10816551074385643,
      "grad_norm": 0.5841379165649414,
      "learning_rate": 2.1933333333333332e-05,
      "loss": 0.1344,
      "relation_loss": 0.02623775154352188,
      "selected_pairs": 26965.0,
      "state_query_norm": 45.25091819763183,
      "state_relation_loss": 0.0058858531527221205,
      "step": 330,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.02623775154352188,
      "weighted_relation_loss": 0.02623775154352188
    },
    {
      "action_query_norm": 45.25460929870606,
      "action_relation_loss": 0.02017104532569647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.31805425631431244,
      "flow_matching_loss": 0.10670405328273773,
      "grad_norm": 0.515902578830719,
      "learning_rate": 2.26e-05,
      "loss": 0.1325,
      "relation_loss": 0.025788164883852004,
      "selected_pairs": 26916.4,
      "state_query_norm": 45.25023727416992,
      "state_relation_loss": 0.00561711941845715,
      "step": 340,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.025788164883852004,
      "weighted_relation_loss": 0.025788164883852004
    },
    {
      "action_query_norm": 45.25480003356934,
      "action_relation_loss": 0.01915260571986437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3274087932647334,
      "flow_matching_loss": 0.10675496757030487,
      "grad_norm": 0.4401108920574188,
      "learning_rate": 2.326666666666667e-05,
      "loss": 0.1317,
      "relation_loss": 0.024936549738049506,
      "selected_pairs": 26783.9,
      "state_query_norm": 45.2500919342041,
      "state_relation_loss": 0.005783943878486753,
      "step": 350,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.024936549738049506,
      "weighted_relation_loss": 0.024936549738049506
    },
    {
      "action_query_norm": 45.25484008789063,
      "action_relation_loss": 0.019782018288969992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.33676333021515437,
      "flow_matching_loss": 0.10831172987818719,
      "grad_norm": 0.542881429195404,
      "learning_rate": 2.3933333333333337e-05,
      "loss": 0.1335,
      "relation_loss": 0.025180627778172492,
      "selected_pairs": 26576.7,
      "state_query_norm": 45.24973754882812,
      "state_relation_loss": 0.0053986092563718556,
      "step": 360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.025180627778172492,
      "weighted_relation_loss": 0.025180627778172492
    },
    {
      "action_query_norm": 45.25457763671875,
      "action_relation_loss": 0.019114269688725472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3461178671655753,
      "flow_matching_loss": 0.10300420448184014,
      "grad_norm": 0.45650333166122437,
      "learning_rate": 2.46e-05,
      "loss": 0.1275,
      "relation_loss": 0.024526365287601947,
      "selected_pairs": 26965.0,
      "state_query_norm": 45.24900436401367,
      "state_relation_loss": 0.005412095692008734,
      "step": 370,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.024526365287601947,
      "weighted_relation_loss": 0.024526365287601947
    },
    {
      "action_query_norm": 45.25443000793457,
      "action_relation_loss": 0.019005272723734378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.35547240411599623,
      "flow_matching_loss": 0.10097340345382691,
      "grad_norm": 0.3425696790218353,
      "learning_rate": 2.5266666666666666e-05,
      "loss": 0.125,
      "relation_loss": 0.024070602655410767,
      "selected_pairs": 26673.6,
      "state_query_norm": 45.24843521118164,
      "state_relation_loss": 0.005065329559147358,
      "step": 380,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.024070602655410767,
      "weighted_relation_loss": 0.024070602655410767
    },
    {
      "action_query_norm": 45.25430793762207,
      "action_relation_loss": 0.019058318808674812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3648269410664172,
      "flow_matching_loss": 0.104776631295681,
      "grad_norm": 0.46932733058929443,
      "learning_rate": 2.5933333333333338e-05,
      "loss": 0.1287,
      "relation_loss": 0.023947899229824543,
      "selected_pairs": 26691.8,
      "state_query_norm": 45.24837188720703,
      "state_relation_loss": 0.004889580234885216,
      "step": 390,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.023947899229824543,
      "weighted_relation_loss": 0.023947899229824543
    },
    {
      "action_query_norm": 45.254465103149414,
      "action_relation_loss": 0.01953979767858982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.37418147801683815,
      "flow_matching_loss": 0.10465312004089355,
      "grad_norm": 0.4319829046726227,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.1289,
      "relation_loss": 0.024292890354990958,
      "selected_pairs": 26829.5,
      "state_query_norm": 45.24752311706543,
      "state_relation_loss": 0.004753092909231782,
      "step": 400,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.024292890354990958,
      "weighted_relation_loss": 0.024292890354990958
    },
    {
      "action_query_norm": 45.254034423828124,
      "action_relation_loss": 0.018327061645686628,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.38353601496725914,
      "flow_matching_loss": 0.10044711232185363,
      "grad_norm": 0.4834316074848175,
      "learning_rate": 2.7266666666666668e-05,
      "loss": 0.1236,
      "relation_loss": 0.02313486225903034,
      "selected_pairs": 27249.4,
      "state_query_norm": 45.246554946899415,
      "state_relation_loss": 0.004807800520211458,
      "step": 410,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.02313486225903034,
      "weighted_relation_loss": 0.02313486225903034
    },
    {
      "action_query_norm": 45.25379829406738,
      "action_relation_loss": 0.01827105674892664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.3928905519176801,
      "flow_matching_loss": 0.10057056471705436,
      "grad_norm": 0.444985955953598,
      "learning_rate": 2.7933333333333332e-05,
      "loss": 0.1237,
      "relation_loss": 0.023164691403508186,
      "selected_pairs": 26803.8,
      "state_query_norm": 45.24683456420898,
      "state_relation_loss": 0.004893634654581547,
      "step": 420,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.023164691403508186,
      "weighted_relation_loss": 0.023164691403508186
    },
    {
      "action_query_norm": 45.25333824157715,
      "action_relation_loss": 0.0191911306232214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.402245088868101,
      "flow_matching_loss": 0.10469778552651406,
      "grad_norm": 0.3127943277359009,
      "learning_rate": 2.86e-05,
      "loss": 0.1287,
      "relation_loss": 0.023989731818437575,
      "selected_pairs": 26985.8,
      "state_query_norm": 45.24662704467774,
      "state_relation_loss": 0.004798601008951664,
      "step": 430,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.023989731818437575,
      "weighted_relation_loss": 0.023989731818437575
    },
    {
      "action_query_norm": 45.25268669128418,
      "action_relation_loss": 0.018193045072257517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.411599625818522,
      "flow_matching_loss": 0.10053487420082093,
      "grad_norm": 0.35809969902038574,
      "learning_rate": 2.926666666666667e-05,
      "loss": 0.1233,
      "relation_loss": 0.022813934087753295,
      "selected_pairs": 26354.5,
      "state_query_norm": 45.24534797668457,
      "state_relation_loss": 0.004620889201760292,
      "step": 440,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.022813934087753295,
      "weighted_relation_loss": 0.022813934087753295
    },
    {
      "action_query_norm": 45.25270500183105,
      "action_relation_loss": 0.01742879245430231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.42095416276894293,
      "flow_matching_loss": 0.09962858185172081,
      "grad_norm": 0.433498740196228,
      "learning_rate": 2.9933333333333337e-05,
      "loss": 0.1218,
      "relation_loss": 0.02220011204481125,
      "selected_pairs": 27014.6,
      "state_query_norm": 45.24424514770508,
      "state_relation_loss": 0.004771319404244423,
      "step": 450,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.02220011204481125,
      "weighted_relation_loss": 0.02220011204481125
    },
    {
      "action_query_norm": 45.25264282226563,
      "action_relation_loss": 0.01843823753297329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4303086997193639,
      "flow_matching_loss": 0.09762632846832275,
      "grad_norm": 0.47331658005714417,
      "learning_rate": 3.06e-05,
      "loss": 0.1208,
      "relation_loss": 0.023155465349555017,
      "selected_pairs": 26617.4,
      "state_query_norm": 45.24451789855957,
      "state_relation_loss": 0.004717227723449469,
      "step": 460,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.023155465349555017,
      "weighted_relation_loss": 0.023155465349555017
    },
    {
      "action_query_norm": 45.25234107971191,
      "action_relation_loss": 0.017821504920721053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.43966323666978485,
      "flow_matching_loss": 0.09421916529536248,
      "grad_norm": 0.2915874421596527,
      "learning_rate": 3.126666666666666e-05,
      "loss": 0.1163,
      "relation_loss": 0.02208873853087425,
      "selected_pairs": 25988.2,
      "state_query_norm": 45.24344825744629,
      "state_relation_loss": 0.00426723356358707,
      "step": 470,
      "valid_action_samples": 228.4,
      "weighted_auxiliary_loss": 0.02208873853087425,
      "weighted_relation_loss": 0.02208873853087425
    },
    {
      "action_query_norm": 45.25185775756836,
      "action_relation_loss": 0.017762031499296428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4490177736202058,
      "flow_matching_loss": 0.09674270451068878,
      "grad_norm": 0.3707540035247803,
      "learning_rate": 3.1933333333333335e-05,
      "loss": 0.1186,
      "relation_loss": 0.021868069469928742,
      "selected_pairs": 26416.2,
      "state_query_norm": 45.24256439208985,
      "state_relation_loss": 0.004106037924066186,
      "step": 480,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.021868069469928742,
      "weighted_relation_loss": 0.021868069469928742
    },
    {
      "action_query_norm": 45.25173454284668,
      "action_relation_loss": 0.017901567183434962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4583723105706268,
      "flow_matching_loss": 0.09588465467095375,
      "grad_norm": 0.3622055649757385,
      "learning_rate": 3.26e-05,
      "loss": 0.1177,
      "relation_loss": 0.02186008859425783,
      "selected_pairs": 26572.2,
      "state_query_norm": 45.24174461364746,
      "state_relation_loss": 0.003958521387539804,
      "step": 490,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.02186008859425783,
      "weighted_relation_loss": 0.02186008859425783
    },
    {
      "action_query_norm": 45.251773834228516,
      "action_relation_loss": 0.018087354488670826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4677268475210477,
      "flow_matching_loss": 0.09942597150802612,
      "grad_norm": 0.31317150592803955,
      "learning_rate": 3.326666666666667e-05,
      "loss": 0.1216,
      "relation_loss": 0.02213777173310518,
      "selected_pairs": 26546.6,
      "state_query_norm": 45.24165840148926,
      "state_relation_loss": 0.004050417453981936,
      "step": 500,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.02213777173310518,
      "weighted_relation_loss": 0.02213777173310518
    },
    {
      "action_query_norm": 45.25163116455078,
      "action_relation_loss": 0.01801095912232995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.47708138447146864,
      "flow_matching_loss": 0.09476407915353775,
      "grad_norm": 0.407890647649765,
      "learning_rate": 3.3933333333333336e-05,
      "loss": 0.1168,
      "relation_loss": 0.022015074640512465,
      "selected_pairs": 27058.2,
      "state_query_norm": 45.24082908630371,
      "state_relation_loss": 0.004004115425050259,
      "step": 510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.022015074640512465,
      "weighted_relation_loss": 0.022015074640512465
    },
    {
      "action_query_norm": 45.25171051025391,
      "action_relation_loss": 0.017730210535228254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.4864359214218896,
      "flow_matching_loss": 0.09404235258698464,
      "grad_norm": 0.5317059755325317,
      "learning_rate": 3.46e-05,
      "loss": 0.1159,
      "relation_loss": 0.02190112415701151,
      "selected_pairs": 27478.1,
      "state_query_norm": 45.24010963439942,
      "state_relation_loss": 0.0041709136683493854,
      "step": 520,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.02190112415701151,
      "weighted_relation_loss": 0.02190112415701151
    },
    {
      "action_query_norm": 45.251681900024415,
      "action_relation_loss": 0.01796914041042328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.49579045837231056,
      "flow_matching_loss": 0.09497443065047265,
      "grad_norm": 0.6325930953025818,
      "learning_rate": 3.526666666666667e-05,
      "loss": 0.1168,
      "relation_loss": 0.021788379922509194,
      "selected_pairs": 26298.5,
      "state_query_norm": 45.23976783752441,
      "state_relation_loss": 0.003819239512085915,
      "step": 530,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.021788379922509194,
      "weighted_relation_loss": 0.021788379922509194
    },
    {
      "action_query_norm": 45.2519889831543,
      "action_relation_loss": 0.01733931777998805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5051449953227315,
      "flow_matching_loss": 0.09619136527180672,
      "grad_norm": 0.38411423563957214,
      "learning_rate": 3.593333333333334e-05,
      "loss": 0.117,
      "relation_loss": 0.020785257034003735,
      "selected_pairs": 27060.5,
      "state_query_norm": 45.238642883300784,
      "state_relation_loss": 0.003445939370431006,
      "step": 540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.020785257034003735,
      "weighted_relation_loss": 0.020785257034003735
    },
    {
      "action_query_norm": 45.25195198059082,
      "action_relation_loss": 0.017672315798699857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5144995322731525,
      "flow_matching_loss": 0.097017540037632,
      "grad_norm": 0.3788808584213257,
      "learning_rate": 3.66e-05,
      "loss": 0.118,
      "relation_loss": 0.021017683297395708,
      "selected_pairs": 26225.1,
      "state_query_norm": 45.23784561157227,
      "state_relation_loss": 0.003345367475412786,
      "step": 550,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.021017683297395708,
      "weighted_relation_loss": 0.021017683297395708
    },
    {
      "action_query_norm": 45.251404190063475,
      "action_relation_loss": 0.01779889827594161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5238540692235735,
      "flow_matching_loss": 0.09466631710529327,
      "grad_norm": 0.4263041317462921,
      "learning_rate": 3.726666666666667e-05,
      "loss": 0.116,
      "relation_loss": 0.021351526491343975,
      "selected_pairs": 26645.2,
      "state_query_norm": 45.23787879943848,
      "state_relation_loss": 0.003552628168836236,
      "step": 560,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.021351526491343975,
      "weighted_relation_loss": 0.021351526491343975
    },
    {
      "action_query_norm": 45.250722885131836,
      "action_relation_loss": 0.01712980940937996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5332086061739943,
      "flow_matching_loss": 0.09114197865128518,
      "grad_norm": 0.4041295647621155,
      "learning_rate": 3.793333333333334e-05,
      "loss": 0.1117,
      "relation_loss": 0.0205789927393198,
      "selected_pairs": 26852.9,
      "state_query_norm": 45.23727264404297,
      "state_relation_loss": 0.0034491831436753275,
      "step": 570,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0205789927393198,
      "weighted_relation_loss": 0.0205789927393198
    },
    {
      "action_query_norm": 45.25048942565918,
      "action_relation_loss": 0.016816816199570893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5425631431244153,
      "flow_matching_loss": 0.09527507275342942,
      "grad_norm": 0.3408348560333252,
      "learning_rate": 3.86e-05,
      "loss": 0.1152,
      "relation_loss": 0.019963843375444414,
      "selected_pairs": 27201.6,
      "state_query_norm": 45.2351734161377,
      "state_relation_loss": 0.0031470270128920673,
      "step": 580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.019963843375444414,
      "weighted_relation_loss": 0.019963843375444414
    },
    {
      "action_query_norm": 45.250572204589844,
      "action_relation_loss": 0.017166643962264062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5519176800748363,
      "flow_matching_loss": 0.09499009475111961,
      "grad_norm": 0.2877660393714905,
      "learning_rate": 3.926666666666667e-05,
      "loss": 0.1156,
      "relation_loss": 0.020594651438295843,
      "selected_pairs": 26848.9,
      "state_query_norm": 45.23490562438965,
      "state_relation_loss": 0.003428007196635008,
      "step": 590,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.020594651438295843,
      "weighted_relation_loss": 0.020594651438295843
    },
    {
      "action_query_norm": 45.25027236938477,
      "action_relation_loss": 0.016886362805962563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5612722170252572,
      "flow_matching_loss": 0.09074569717049599,
      "grad_norm": 0.3071886897087097,
      "learning_rate": 3.993333333333333e-05,
      "loss": 0.1107,
      "relation_loss": 0.019962507672607897,
      "selected_pairs": 26365.9,
      "state_query_norm": 45.234135437011716,
      "state_relation_loss": 0.003076145029626787,
      "step": 600,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.019962507672607897,
      "weighted_relation_loss": 0.019962507672607897
    },
    {
      "action_query_norm": 45.25028038024902,
      "action_relation_loss": 0.017667298205196857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5706267539756782,
      "flow_matching_loss": 0.09103069603443145,
      "grad_norm": 0.35050368309020996,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.1119,
      "relation_loss": 0.020822177082300185,
      "selected_pairs": 26413.6,
      "state_query_norm": 45.232138061523436,
      "state_relation_loss": 0.0031548789935186504,
      "step": 610,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.020822177082300185,
      "weighted_relation_loss": 0.020822177082300185
    },
    {
      "action_query_norm": 45.250262069702146,
      "action_relation_loss": 0.01713280277326703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5799812909260992,
      "flow_matching_loss": 0.08794319555163384,
      "grad_norm": 0.3916497826576233,
      "learning_rate": 4.126666666666667e-05,
      "loss": 0.1084,
      "relation_loss": 0.02041163928806782,
      "selected_pairs": 27270.6,
      "state_query_norm": 45.232055282592775,
      "state_relation_loss": 0.0032788363518193365,
      "step": 620,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.02041163928806782,
      "weighted_relation_loss": 0.02041163928806782
    },
    {
      "action_query_norm": 45.25017776489258,
      "action_relation_loss": 0.01694265678524971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.5893358278765201,
      "flow_matching_loss": 0.09358008205890656,
      "grad_norm": 0.34026333689689636,
      "learning_rate": 4.1933333333333334e-05,
      "loss": 0.1136,
      "relation_loss": 0.019977390579879285,
      "selected_pairs": 26306.5,
      "state_query_norm": 45.23116264343262,
      "state_relation_loss": 0.0030347339110448957,
      "step": 630,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.019977390579879285,
      "weighted_relation_loss": 0.019977390579879285
    },
    {
      "action_query_norm": 45.24959831237793,
      "action_relation_loss": 0.017197964992374182,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.598690364826941,
      "flow_matching_loss": 0.09378017261624336,
      "grad_norm": 0.49405932426452637,
      "learning_rate": 4.26e-05,
      "loss": 0.1139,
      "relation_loss": 0.02010855097323656,
      "selected_pairs": 27222.9,
      "state_query_norm": 45.22997131347656,
      "state_relation_loss": 0.0029105860507115723,
      "step": 640,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.02010855097323656,
      "weighted_relation_loss": 0.02010855097323656
    },
    {
      "action_query_norm": 45.249197769165036,
      "action_relation_loss": 0.015258426684886217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.608044901777362,
      "flow_matching_loss": 0.08911287039518356,
      "grad_norm": 0.2663465738296509,
      "learning_rate": 4.3266666666666664e-05,
      "loss": 0.1073,
      "relation_loss": 0.018189277313649655,
      "selected_pairs": 26754.4,
      "state_query_norm": 45.22938499450684,
      "state_relation_loss": 0.002930850465781987,
      "step": 650,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.018189277313649655,
      "weighted_relation_loss": 0.018189277313649655
    },
    {
      "action_query_norm": 45.24873428344726,
      "action_relation_loss": 0.016313574835658075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.617399438727783,
      "flow_matching_loss": 0.09387655183672905,
      "grad_norm": 0.3367742896080017,
      "learning_rate": 4.3933333333333335e-05,
      "loss": 0.1132,
      "relation_loss": 0.01931520029902458,
      "selected_pairs": 26672.1,
      "state_query_norm": 45.229081344604495,
      "state_relation_loss": 0.003001625370234251,
      "step": 660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.01931520029902458,
      "weighted_relation_loss": 0.01931520029902458
    },
    {
      "action_query_norm": 45.24836349487305,
      "action_relation_loss": 0.0180350361391902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6267539756782039,
      "flow_matching_loss": 0.09230391010642051,
      "grad_norm": 0.4267338812351227,
      "learning_rate": 4.46e-05,
      "loss": 0.1134,
      "relation_loss": 0.021061863750219345,
      "selected_pairs": 27316.5,
      "state_query_norm": 45.22667198181152,
      "state_relation_loss": 0.0030268274480476976,
      "step": 670,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.021061863750219345,
      "weighted_relation_loss": 0.021061863750219345
    },
    {
      "action_query_norm": 45.24818801879883,
      "action_relation_loss": 0.01623802585527301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6361085126286249,
      "flow_matching_loss": 0.09394967555999756,
      "grad_norm": 0.6805102229118347,
      "learning_rate": 4.526666666666667e-05,
      "loss": 0.1131,
      "relation_loss": 0.019125356152653695,
      "selected_pairs": 26925.6,
      "state_query_norm": 45.22524299621582,
      "state_relation_loss": 0.002887330553494394,
      "step": 680,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.019125356152653695,
      "weighted_relation_loss": 0.019125356152653695
    },
    {
      "action_query_norm": 45.24791831970215,
      "action_relation_loss": 0.015486459992825986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6454630495790459,
      "flow_matching_loss": 0.09191064089536667,
      "grad_norm": 0.4052201807498932,
      "learning_rate": 4.5933333333333336e-05,
      "loss": 0.1105,
      "relation_loss": 0.018547202460467815,
      "selected_pairs": 27382.9,
      "state_query_norm": 45.225800323486325,
      "state_relation_loss": 0.003060742327943444,
      "step": 690,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.018547202460467815,
      "weighted_relation_loss": 0.018547202460467815
    },
    {
      "action_query_norm": 45.24760398864746,
      "action_relation_loss": 0.01566706392914057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6548175865294668,
      "flow_matching_loss": 0.09154412299394607,
      "grad_norm": 0.38509419560432434,
      "learning_rate": 4.660000000000001e-05,
      "loss": 0.1103,
      "relation_loss": 0.01875111274421215,
      "selected_pairs": 26574.2,
      "state_query_norm": 45.22491264343262,
      "state_relation_loss": 0.0030840485356748103,
      "step": 700,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.01875111274421215,
      "weighted_relation_loss": 0.01875111274421215
    },
    {
      "action_query_norm": 45.247004699707034,
      "action_relation_loss": 0.01596632841974497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6641721234798877,
      "flow_matching_loss": 0.09085827246308327,
      "grad_norm": 0.4964647889137268,
      "learning_rate": 4.726666666666667e-05,
      "loss": 0.1099,
      "relation_loss": 0.01903618164360523,
      "selected_pairs": 26826.8,
      "state_query_norm": 45.22362174987793,
      "state_relation_loss": 0.0030698533868417145,
      "step": 710,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.01903618164360523,
      "weighted_relation_loss": 0.01903618164360523
    },
    {
      "action_query_norm": 45.24629135131836,
      "action_relation_loss": 0.016742807906121017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6735266604303087,
      "flow_matching_loss": 0.09124287962913513,
      "grad_norm": 0.39358142018318176,
      "learning_rate": 4.793333333333334e-05,
      "loss": 0.1109,
      "relation_loss": 0.01965504474937916,
      "selected_pairs": 27413.5,
      "state_query_norm": 45.22109146118164,
      "state_relation_loss": 0.0029122369363904,
      "step": 720,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.01965504474937916,
      "weighted_relation_loss": 0.01965504474937916
    },
    {
      "action_query_norm": 45.24610595703125,
      "action_relation_loss": 0.016162890195846557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6828811973807296,
      "flow_matching_loss": 0.08869066536426544,
      "grad_norm": 0.4380916655063629,
      "learning_rate": 4.86e-05,
      "loss": 0.1078,
      "relation_loss": 0.01914167180657387,
      "selected_pairs": 26964.0,
      "state_query_norm": 45.21980361938476,
      "state_relation_loss": 0.0029787818202748895,
      "step": 730,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01914167180657387,
      "weighted_relation_loss": 0.01914167180657387
    },
    {
      "action_query_norm": 45.24496765136719,
      "action_relation_loss": 0.016128311958163978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.6922357343311506,
      "flow_matching_loss": 0.09056197628378868,
      "grad_norm": 0.2849609851837158,
      "learning_rate": 4.926666666666667e-05,
      "loss": 0.1095,
      "relation_loss": 0.018942568078637124,
      "selected_pairs": 27245.2,
      "state_query_norm": 45.221130752563475,
      "state_relation_loss": 0.0028142562368884683,
      "step": 740,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.018942568078637124,
      "weighted_relation_loss": 0.018942568078637124
    },
    {
      "action_query_norm": 45.244716262817384,
      "action_relation_loss": 0.015890187118202448,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7015902712815716,
      "flow_matching_loss": 0.08567168861627579,
      "grad_norm": 0.3443996012210846,
      "learning_rate": 4.993333333333334e-05,
      "loss": 0.1042,
      "relation_loss": 0.018538859486579896,
      "selected_pairs": 27059.8,
      "state_query_norm": 45.218732833862305,
      "state_relation_loss": 0.002648672112263739,
      "step": 750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.018538859486579896,
      "weighted_relation_loss": 0.018538859486579896
    },
    {
      "action_query_norm": 45.24415664672851,
      "action_relation_loss": 0.015977895725518464,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7109448082319925,
      "flow_matching_loss": 0.08723880127072334,
      "grad_norm": 0.3569867014884949,
      "learning_rate": 5.0600000000000003e-05,
      "loss": 0.106,
      "relation_loss": 0.018718411587178706,
      "selected_pairs": 26211.2,
      "state_query_norm": 45.21986808776855,
      "state_relation_loss": 0.002740515721961856,
      "step": 760,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.018718411587178706,
      "weighted_relation_loss": 0.018718411587178706
    },
    {
      "action_query_norm": 45.243505859375,
      "action_relation_loss": 0.01573842205107212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7202993451824135,
      "flow_matching_loss": 0.08480599075555802,
      "grad_norm": 0.5078528523445129,
      "learning_rate": 5.1266666666666675e-05,
      "loss": 0.1032,
      "relation_loss": 0.018369952403008937,
      "selected_pairs": 27034.8,
      "state_query_norm": 45.21811180114746,
      "state_relation_loss": 0.0026315304916352034,
      "step": 770,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.018369952403008937,
      "weighted_relation_loss": 0.018369952403008937
    },
    {
      "action_query_norm": 45.242591857910156,
      "action_relation_loss": 0.016202396992594004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7296538821328344,
      "flow_matching_loss": 0.08909786269068717,
      "grad_norm": 0.4206501245498657,
      "learning_rate": 5.193333333333333e-05,
      "loss": 0.1078,
      "relation_loss": 0.0187416385859251,
      "selected_pairs": 27082.2,
      "state_query_norm": 45.21708946228027,
      "state_relation_loss": 0.00253924154676497,
      "step": 780,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0187416385859251,
      "weighted_relation_loss": 0.0187416385859251
    },
    {
      "action_query_norm": 45.241742324829104,
      "action_relation_loss": 0.015532403159886599,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7390084190832554,
      "flow_matching_loss": 0.08745712265372277,
      "grad_norm": 0.39531105756759644,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 0.1055,
      "relation_loss": 0.018016451597213747,
      "selected_pairs": 27058.4,
      "state_query_norm": 45.21644248962402,
      "state_relation_loss": 0.0024840480880811812,
      "step": 790,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.018016451597213747,
      "weighted_relation_loss": 0.018016451597213747
    },
    {
      "action_query_norm": 45.240783309936525,
      "action_relation_loss": 0.015620024874806404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7483629560336763,
      "flow_matching_loss": 0.0867896094918251,
      "grad_norm": 0.35916441679000854,
      "learning_rate": 5.326666666666666e-05,
      "loss": 0.1049,
      "relation_loss": 0.018067002482712268,
      "selected_pairs": 27107.8,
      "state_query_norm": 45.215201950073244,
      "state_relation_loss": 0.0024469776777550577,
      "step": 800,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.018067002482712268,
      "weighted_relation_loss": 0.018067002482712268
    },
    {
      "action_query_norm": 45.24063301086426,
      "action_relation_loss": 0.015118071250617504,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7577174929840973,
      "flow_matching_loss": 0.08553403839468957,
      "grad_norm": 0.23882533609867096,
      "learning_rate": 5.3933333333333334e-05,
      "loss": 0.1031,
      "relation_loss": 0.017534817941486834,
      "selected_pairs": 27045.8,
      "state_query_norm": 45.21460304260254,
      "state_relation_loss": 0.00241674673743546,
      "step": 810,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.017534817941486834,
      "weighted_relation_loss": 0.017534817941486834
    },
    {
      "action_query_norm": 45.24035720825195,
      "action_relation_loss": 0.015308588463813066,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7670720299345183,
      "flow_matching_loss": 0.08752544894814492,
      "grad_norm": 0.29247793555259705,
      "learning_rate": 5.4600000000000006e-05,
      "loss": 0.1052,
      "relation_loss": 0.017678111046552657,
      "selected_pairs": 27290.8,
      "state_query_norm": 45.21291007995605,
      "state_relation_loss": 0.0023695227690041067,
      "step": 820,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.017678111046552657,
      "weighted_relation_loss": 0.017678111046552657
    },
    {
      "action_query_norm": 45.23991432189941,
      "action_relation_loss": 0.01499620834365487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7764265668849392,
      "flow_matching_loss": 0.09040054008364677,
      "grad_norm": 0.32674238085746765,
      "learning_rate": 5.5266666666666664e-05,
      "loss": 0.1078,
      "relation_loss": 0.017417001351714135,
      "selected_pairs": 26780.2,
      "state_query_norm": 45.21053199768066,
      "state_relation_loss": 0.002420793194323778,
      "step": 830,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.017417001351714135,
      "weighted_relation_loss": 0.017417001351714135
    },
    {
      "action_query_norm": 45.23933868408203,
      "action_relation_loss": 0.015127075836062431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7857811038353602,
      "flow_matching_loss": 0.08672384694218635,
      "grad_norm": 0.30303922295570374,
      "learning_rate": 5.5933333333333335e-05,
      "loss": 0.1043,
      "relation_loss": 0.017526969127357007,
      "selected_pairs": 27173.0,
      "state_query_norm": 45.20981712341309,
      "state_relation_loss": 0.002399893361143768,
      "step": 840,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.017526969127357007,
      "weighted_relation_loss": 0.017526969127357007
    },
    {
      "action_query_norm": 45.23963050842285,
      "action_relation_loss": 0.015310434345155954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.7951356407857811,
      "flow_matching_loss": 0.08847342580556869,
      "grad_norm": 0.3614964783191681,
      "learning_rate": 5.66e-05,
      "loss": 0.1061,
      "relation_loss": 0.01766414912417531,
      "selected_pairs": 26987.0,
      "state_query_norm": 45.21018295288086,
      "state_relation_loss": 0.0023537147557362915,
      "step": 850,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.01766414912417531,
      "weighted_relation_loss": 0.01766414912417531
    },
    {
      "action_query_norm": 45.23938636779785,
      "action_relation_loss": 0.01588479345664382,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.804490177736202,
      "flow_matching_loss": 0.08407943472266197,
      "grad_norm": 0.2910688519477844,
      "learning_rate": 5.726666666666667e-05,
      "loss": 0.1024,
      "relation_loss": 0.018297987058758736,
      "selected_pairs": 26989.0,
      "state_query_norm": 45.208286666870116,
      "state_relation_loss": 0.002413193741813302,
      "step": 860,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.018297987058758736,
      "weighted_relation_loss": 0.018297987058758736
    },
    {
      "action_query_norm": 45.23843879699707,
      "action_relation_loss": 0.015218568220734596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.813844714686623,
      "flow_matching_loss": 0.08810366243124008,
      "grad_norm": 0.4301912486553192,
      "learning_rate": 5.7933333333333337e-05,
      "loss": 0.1056,
      "relation_loss": 0.01749140340834856,
      "selected_pairs": 26689.3,
      "state_query_norm": 45.207759857177734,
      "state_relation_loss": 0.0022728352108970285,
      "step": 870,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.01749140340834856,
      "weighted_relation_loss": 0.01749140340834856
    },
    {
      "action_query_norm": 45.238271713256836,
      "action_relation_loss": 0.015814262349158524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.823199251637044,
      "flow_matching_loss": 0.08422252461314202,
      "grad_norm": 0.27496811747550964,
      "learning_rate": 5.86e-05,
      "loss": 0.1024,
      "relation_loss": 0.018181711621582507,
      "selected_pairs": 26760.4,
      "state_query_norm": 45.2078140258789,
      "state_relation_loss": 0.002367448969744146,
      "step": 880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.018181711621582507,
      "weighted_relation_loss": 0.018181711621582507
    },
    {
      "action_query_norm": 45.23675003051758,
      "action_relation_loss": 0.015504166297614575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8325537885874649,
      "flow_matching_loss": 0.08507129102945328,
      "grad_norm": 0.3716159164905548,
      "learning_rate": 5.926666666666667e-05,
      "loss": 0.1028,
      "relation_loss": 0.017682649567723273,
      "selected_pairs": 26894.0,
      "state_query_norm": 45.20508079528808,
      "state_relation_loss": 0.0021784833166748284,
      "step": 890,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.017682649567723273,
      "weighted_relation_loss": 0.017682649567723273
    },
    {
      "action_query_norm": 45.23460083007812,
      "action_relation_loss": 0.015866328869014977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8419083255378859,
      "flow_matching_loss": 0.08609712645411491,
      "grad_norm": 0.320712149143219,
      "learning_rate": 5.9933333333333345e-05,
      "loss": 0.1041,
      "relation_loss": 0.018046128377318383,
      "selected_pairs": 27289.2,
      "state_query_norm": 45.2049072265625,
      "state_relation_loss": 0.0021797996247187256,
      "step": 900,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.018046128377318383,
      "weighted_relation_loss": 0.018046128377318383
    },
    {
      "action_query_norm": 45.234492492675784,
      "action_relation_loss": 0.015384470019489527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8512628624883068,
      "flow_matching_loss": 0.08440793976187706,
      "grad_norm": 0.23439469933509827,
      "learning_rate": 6.06e-05,
      "loss": 0.1019,
      "relation_loss": 0.0174779218621552,
      "selected_pairs": 26898.3,
      "state_query_norm": 45.20393524169922,
      "state_relation_loss": 0.002093452087137848,
      "step": 910,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0174779218621552,
      "weighted_relation_loss": 0.0174779218621552
    },
    {
      "action_query_norm": 45.234799194335935,
      "action_relation_loss": 0.014631143305450678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8606173994387278,
      "flow_matching_loss": 0.08623598143458366,
      "grad_norm": 0.3781754970550537,
      "learning_rate": 6.126666666666667e-05,
      "loss": 0.103,
      "relation_loss": 0.01681299405172467,
      "selected_pairs": 26660.5,
      "state_query_norm": 45.19964790344238,
      "state_relation_loss": 0.002181850909255445,
      "step": 920,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.01681299405172467,
      "weighted_relation_loss": 0.01681299405172467
    },
    {
      "action_query_norm": 45.23465423583984,
      "action_relation_loss": 0.01499113067984581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8699719363891487,
      "flow_matching_loss": 0.08453229218721389,
      "grad_norm": 0.29635170102119446,
      "learning_rate": 6.193333333333333e-05,
      "loss": 0.1017,
      "relation_loss": 0.017154527828097343,
      "selected_pairs": 27086.8,
      "state_query_norm": 45.19918327331543,
      "state_relation_loss": 0.0021633970784023402,
      "step": 930,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.017154527828097343,
      "weighted_relation_loss": 0.017154527828097343
    },
    {
      "action_query_norm": 45.23439979553223,
      "action_relation_loss": 0.015027274098247289,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8793264733395697,
      "flow_matching_loss": 0.08420921266078948,
      "grad_norm": 0.3336855173110962,
      "learning_rate": 6.26e-05,
      "loss": 0.1014,
      "relation_loss": 0.017169402353465556,
      "selected_pairs": 27623.5,
      "state_query_norm": 45.198817825317384,
      "state_relation_loss": 0.0021421282086521386,
      "step": 940,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.017169402353465556,
      "weighted_relation_loss": 0.017169402353465556
    },
    {
      "action_query_norm": 45.23335227966309,
      "action_relation_loss": 0.014903416763991118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8886810102899907,
      "flow_matching_loss": 0.08306904882192612,
      "grad_norm": 0.220134437084198,
      "learning_rate": 6.326666666666667e-05,
      "loss": 0.1002,
      "relation_loss": 0.017091102432459593,
      "selected_pairs": 27269.9,
      "state_query_norm": 45.197472381591794,
      "state_relation_loss": 0.002187685645185411,
      "step": 950,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.017091102432459593,
      "weighted_relation_loss": 0.017091102432459593
    },
    {
      "action_query_norm": 45.23272972106933,
      "action_relation_loss": 0.014331561606377363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.8980355472404116,
      "flow_matching_loss": 0.0825296126306057,
      "grad_norm": 0.34893694519996643,
      "learning_rate": 6.393333333333333e-05,
      "loss": 0.0989,
      "relation_loss": 0.016413542535156013,
      "selected_pairs": 26485.8,
      "state_query_norm": 45.195552825927734,
      "state_relation_loss": 0.0020819810917600988,
      "step": 960,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.016413542535156013,
      "weighted_relation_loss": 0.016413542535156013
    },
    {
      "action_query_norm": 45.23115501403809,
      "action_relation_loss": 0.01480875350534916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9073900841908326,
      "flow_matching_loss": 0.08234061375260353,
      "grad_norm": 0.3802785575389862,
      "learning_rate": 6.460000000000001e-05,
      "loss": 0.0994,
      "relation_loss": 0.017082030512392522,
      "selected_pairs": 26531.5,
      "state_query_norm": 45.19747619628906,
      "state_relation_loss": 0.0022732768207788467,
      "step": 970,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.017082030512392522,
      "weighted_relation_loss": 0.017082030512392522
    },
    {
      "action_query_norm": 45.23049774169922,
      "action_relation_loss": 0.014633988495916128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9167446211412535,
      "flow_matching_loss": 0.08729752078652382,
      "grad_norm": 0.5503228902816772,
      "learning_rate": 6.526666666666666e-05,
      "loss": 0.1042,
      "relation_loss": 0.01686289580538869,
      "selected_pairs": 26920.1,
      "state_query_norm": 45.194083786010744,
      "state_relation_loss": 0.0022289074724540113,
      "step": 980,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.01686289580538869,
      "weighted_relation_loss": 0.01686289580538869
    },
    {
      "action_query_norm": 45.22949447631836,
      "action_relation_loss": 0.014132872223854065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9260991580916744,
      "flow_matching_loss": 0.08388752937316894,
      "grad_norm": 0.3402090072631836,
      "learning_rate": 6.593333333333334e-05,
      "loss": 0.1003,
      "relation_loss": 0.01640469627454877,
      "selected_pairs": 27012.6,
      "state_query_norm": 45.19390335083008,
      "state_relation_loss": 0.0022718242835253477,
      "step": 990,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.01640469627454877,
      "weighted_relation_loss": 0.01640469627454877
    },
    {
      "action_query_norm": 45.230389785766604,
      "action_relation_loss": 0.01493560429662466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9354536950420954,
      "flow_matching_loss": 0.08831989467144012,
      "grad_norm": 0.3671901822090149,
      "learning_rate": 6.66e-05,
      "loss": 0.1055,
      "relation_loss": 0.017204423528164627,
      "selected_pairs": 26686.1,
      "state_query_norm": 45.192769622802736,
      "state_relation_loss": 0.002268819254823029,
      "step": 1000,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.017204423528164627,
      "weighted_relation_loss": 0.017204423528164627
    },
    {
      "action_query_norm": 45.22917251586914,
      "action_relation_loss": 0.014839021395891905,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9448082319925164,
      "flow_matching_loss": 0.08576714247465134,
      "grad_norm": 0.45678356289863586,
      "learning_rate": 6.726666666666667e-05,
      "loss": 0.1028,
      "relation_loss": 0.01707411641255021,
      "selected_pairs": 26782.0,
      "state_query_norm": 45.19110221862793,
      "state_relation_loss": 0.0022350951679982246,
      "step": 1010,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.01707411641255021,
      "weighted_relation_loss": 0.01707411641255021
    },
    {
      "action_query_norm": 45.22872657775879,
      "action_relation_loss": 0.014426990784704686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9541627689429373,
      "flow_matching_loss": 0.08302350342273712,
      "grad_norm": 0.4533548057079315,
      "learning_rate": 6.793333333333334e-05,
      "loss": 0.0996,
      "relation_loss": 0.016581074707210063,
      "selected_pairs": 27155.1,
      "state_query_norm": 45.18942642211914,
      "state_relation_loss": 0.0021540841553360224,
      "step": 1020,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.016581074707210063,
      "weighted_relation_loss": 0.016581074707210063
    },
    {
      "action_query_norm": 45.228383255004886,
      "action_relation_loss": 0.0155910006724298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9635173058933583,
      "flow_matching_loss": 0.08730457499623298,
      "grad_norm": 0.4829663336277008,
      "learning_rate": 6.860000000000001e-05,
      "loss": 0.1049,
      "relation_loss": 0.017616871185600756,
      "selected_pairs": 26726.0,
      "state_query_norm": 45.189924240112305,
      "state_relation_loss": 0.0020258706528693436,
      "step": 1030,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.017616871185600756,
      "weighted_relation_loss": 0.017616871185600756
    },
    {
      "action_query_norm": 45.227640533447264,
      "action_relation_loss": 0.014838677924126386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9728718428437793,
      "flow_matching_loss": 0.08427319452166557,
      "grad_norm": 0.2915290296077728,
      "learning_rate": 6.926666666666667e-05,
      "loss": 0.1012,
      "relation_loss": 0.016878592874854804,
      "selected_pairs": 27193.9,
      "state_query_norm": 45.18847160339355,
      "state_relation_loss": 0.0020399152534082534,
      "step": 1040,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.016878592874854804,
      "weighted_relation_loss": 0.016878592874854804
    },
    {
      "action_query_norm": 45.22617607116699,
      "action_relation_loss": 0.014104980882257223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9822263797942001,
      "flow_matching_loss": 0.08361369147896766,
      "grad_norm": 0.2607865631580353,
      "learning_rate": 6.993333333333334e-05,
      "loss": 0.0997,
      "relation_loss": 0.016070484463125468,
      "selected_pairs": 26000.8,
      "state_query_norm": 45.18704452514648,
      "state_relation_loss": 0.001965503697283566,
      "step": 1050,
      "valid_action_samples": 228.5,
      "weighted_auxiliary_loss": 0.016070484463125468,
      "weighted_relation_loss": 0.016070484463125468
    },
    {
      "action_query_norm": 45.22651748657226,
      "action_relation_loss": 0.015012943092733622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 0.9915809167446211,
      "flow_matching_loss": 0.08240759819746017,
      "grad_norm": 0.2312202751636505,
      "learning_rate": 7.06e-05,
      "loss": 0.0994,
      "relation_loss": 0.016964469384402037,
      "selected_pairs": 27615.0,
      "state_query_norm": 45.184682083129886,
      "state_relation_loss": 0.0019515262451022864,
      "step": 1060,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.016964469384402037,
      "weighted_relation_loss": 0.016964469384402037
    },
    {
      "action_query_norm": 45.2234691619873,
      "action_relation_loss": 0.014037975762039422,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.000935453695042,
      "flow_matching_loss": 0.08254989981651306,
      "grad_norm": 0.33497855067253113,
      "learning_rate": 7.126666666666667e-05,
      "loss": 0.0984,
      "relation_loss": 0.015835512802004813,
      "selected_pairs": 24331.3,
      "state_query_norm": 45.18278617858887,
      "state_relation_loss": 0.0017975370399653911,
      "step": 1070,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.015835512802004813,
      "weighted_relation_loss": 0.015835512802004813
    },
    {
      "action_query_norm": 45.2240177154541,
      "action_relation_loss": 0.014649110846221447,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.010289990645463,
      "flow_matching_loss": 0.08616936206817627,
      "grad_norm": 0.3519594371318817,
      "learning_rate": 7.193333333333334e-05,
      "loss": 0.1028,
      "relation_loss": 0.01659090770408511,
      "selected_pairs": 26854.3,
      "state_query_norm": 45.18325996398926,
      "state_relation_loss": 0.0019417969859205186,
      "step": 1080,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.01659090770408511,
      "weighted_relation_loss": 0.01659090770408511
    },
    {
      "action_query_norm": 45.22231216430664,
      "action_relation_loss": 0.014475313480943441,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.019644527595884,
      "flow_matching_loss": 0.0815578393638134,
      "grad_norm": 0.27067112922668457,
      "learning_rate": 7.26e-05,
      "loss": 0.0979,
      "relation_loss": 0.016361000668257474,
      "selected_pairs": 26685.2,
      "state_query_norm": 45.18095817565918,
      "state_relation_loss": 0.001885687408503145,
      "step": 1090,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.016361000668257474,
      "weighted_relation_loss": 0.016361000668257474
    },
    {
      "action_query_norm": 45.22121810913086,
      "action_relation_loss": 0.015227117668837309,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.028999064546305,
      "flow_matching_loss": 0.08826726526021958,
      "grad_norm": 0.2901850640773773,
      "learning_rate": 7.326666666666667e-05,
      "loss": 0.1055,
      "relation_loss": 0.01721990704536438,
      "selected_pairs": 27292.0,
      "state_query_norm": 45.179677200317386,
      "state_relation_loss": 0.001992789166979492,
      "step": 1100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.01721990704536438,
      "weighted_relation_loss": 0.01721990704536438
    },
    {
      "action_query_norm": 45.22046585083008,
      "action_relation_loss": 0.015114766359329224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0383536014967258,
      "flow_matching_loss": 0.08347492814064025,
      "grad_norm": 0.26806434988975525,
      "learning_rate": 7.393333333333333e-05,
      "loss": 0.1006,
      "relation_loss": 0.01710454961284995,
      "selected_pairs": 26446.3,
      "state_query_norm": 45.177028274536134,
      "state_relation_loss": 0.001989783311728388,
      "step": 1110,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.01710454961284995,
      "weighted_relation_loss": 0.01710454961284995
    },
    {
      "action_query_norm": 45.21931610107422,
      "action_relation_loss": 0.014988487865775824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.047708138447147,
      "flow_matching_loss": 0.08499652445316315,
      "grad_norm": 0.23490740358829498,
      "learning_rate": 7.46e-05,
      "loss": 0.102,
      "relation_loss": 0.01699934471398592,
      "selected_pairs": 27480.5,
      "state_query_norm": 45.17562828063965,
      "state_relation_loss": 0.0020108567085117103,
      "step": 1120,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.01699934471398592,
      "weighted_relation_loss": 0.01699934471398592
    },
    {
      "action_query_norm": 45.219177627563475,
      "action_relation_loss": 0.014586679823696613,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0570626753975678,
      "flow_matching_loss": 0.08238864317536354,
      "grad_norm": 0.21791821718215942,
      "learning_rate": 7.526666666666668e-05,
      "loss": 0.0989,
      "relation_loss": 0.016487162187695505,
      "selected_pairs": 27033.8,
      "state_query_norm": 45.17518539428711,
      "state_relation_loss": 0.0019004823639988898,
      "step": 1130,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.016487162187695505,
      "weighted_relation_loss": 0.016487162187695505
    },
    {
      "action_query_norm": 45.2169692993164,
      "action_relation_loss": 0.015072960685938597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0664172123479887,
      "flow_matching_loss": 0.0836298830807209,
      "grad_norm": 0.2738470137119293,
      "learning_rate": 7.593333333333334e-05,
      "loss": 0.1005,
      "relation_loss": 0.016919235046952964,
      "selected_pairs": 26409.6,
      "state_query_norm": 45.17271270751953,
      "state_relation_loss": 0.001846274279523641,
      "step": 1140,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.016919235046952964,
      "weighted_relation_loss": 0.016919235046952964
    },
    {
      "action_query_norm": 45.21809730529785,
      "action_relation_loss": 0.015354607813060283,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0757717492984098,
      "flow_matching_loss": 0.08309889659285545,
      "grad_norm": 0.27231383323669434,
      "learning_rate": 7.66e-05,
      "loss": 0.1004,
      "relation_loss": 0.017258261051028966,
      "selected_pairs": 27662.5,
      "state_query_norm": 45.17346611022949,
      "state_relation_loss": 0.0019036530633457006,
      "step": 1150,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.017258261051028966,
      "weighted_relation_loss": 0.017258261051028966
    },
    {
      "action_query_norm": 45.215098190307614,
      "action_relation_loss": 0.014587251655757427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0851262862488307,
      "flow_matching_loss": 0.08260500505566597,
      "grad_norm": 0.32315593957901,
      "learning_rate": 7.726666666666667e-05,
      "loss": 0.0991,
      "relation_loss": 0.016512981988489626,
      "selected_pairs": 26627.6,
      "state_query_norm": 45.17118453979492,
      "state_relation_loss": 0.0019257305306382478,
      "step": 1160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.016512981988489626,
      "weighted_relation_loss": 0.016512981988489626
    },
    {
      "action_query_norm": 45.21553192138672,
      "action_relation_loss": 0.013891409523785115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.0944808231992516,
      "flow_matching_loss": 0.0817290723323822,
      "grad_norm": 0.23647873103618622,
      "learning_rate": 7.793333333333333e-05,
      "loss": 0.0974,
      "relation_loss": 0.01568777682259679,
      "selected_pairs": 26926.1,
      "state_query_norm": 45.17119026184082,
      "state_relation_loss": 0.001796367298811674,
      "step": 1170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.01568777682259679,
      "weighted_relation_loss": 0.01568777682259679
    },
    {
      "action_query_norm": 45.214822387695314,
      "action_relation_loss": 0.014579750876873732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1038353601496727,
      "flow_matching_loss": 0.08295455798506737,
      "grad_norm": 0.3010614812374115,
      "learning_rate": 7.860000000000001e-05,
      "loss": 0.0993,
      "relation_loss": 0.016386524960398673,
      "selected_pairs": 26729.7,
      "state_query_norm": 45.169747543334964,
      "state_relation_loss": 0.001806774188298732,
      "step": 1180,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.016386524960398673,
      "weighted_relation_loss": 0.016386524960398673
    },
    {
      "action_query_norm": 45.21381607055664,
      "action_relation_loss": 0.015328202955424785,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1131898971000935,
      "flow_matching_loss": 0.08338217437267303,
      "grad_norm": 0.19001102447509766,
      "learning_rate": 7.926666666666666e-05,
      "loss": 0.1005,
      "relation_loss": 0.01714021535590291,
      "selected_pairs": 27038.9,
      "state_query_norm": 45.16888275146484,
      "state_relation_loss": 0.0018120122840628028,
      "step": 1190,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.01714021535590291,
      "weighted_relation_loss": 0.01714021535590291
    },
    {
      "action_query_norm": 45.214202117919925,
      "action_relation_loss": 0.014249502588063479,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1225444340505144,
      "flow_matching_loss": 0.08267278969287872,
      "grad_norm": 0.22649399936199188,
      "learning_rate": 7.993333333333334e-05,
      "loss": 0.0988,
      "relation_loss": 0.016092130169272422,
      "selected_pairs": 26204.3,
      "state_query_norm": 45.16953544616699,
      "state_relation_loss": 0.0018426275462843477,
      "step": 1200,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.016092130169272422,
      "weighted_relation_loss": 0.016092130169272422
    },
    {
      "action_query_norm": 45.21206512451172,
      "action_relation_loss": 0.013970756065100431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1318989710009355,
      "flow_matching_loss": 0.08149964734911919,
      "grad_norm": 0.31179195642471313,
      "learning_rate": 8.060000000000001e-05,
      "loss": 0.0972,
      "relation_loss": 0.015660001616925,
      "selected_pairs": 26853.9,
      "state_query_norm": 45.16533508300781,
      "state_relation_loss": 0.0016892454586923122,
      "step": 1210,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.015660001616925,
      "weighted_relation_loss": 0.015660001616925
    },
    {
      "action_query_norm": 45.21050033569336,
      "action_relation_loss": 0.014624392054975033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1412535079513564,
      "flow_matching_loss": 0.08138922527432442,
      "grad_norm": 0.2230483591556549,
      "learning_rate": 8.126666666666667e-05,
      "loss": 0.0977,
      "relation_loss": 0.016345786582678556,
      "selected_pairs": 26970.1,
      "state_query_norm": 45.16340713500976,
      "state_relation_loss": 0.0017213947954587638,
      "step": 1220,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.016345786582678556,
      "weighted_relation_loss": 0.016345786582678556
    },
    {
      "action_query_norm": 45.20899772644043,
      "action_relation_loss": 0.015318758878856897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1506080449017775,
      "flow_matching_loss": 0.08230296298861503,
      "grad_norm": 0.2706463932991028,
      "learning_rate": 8.193333333333334e-05,
      "loss": 0.0994,
      "relation_loss": 0.017127034813165666,
      "selected_pairs": 26903.1,
      "state_query_norm": 45.16119956970215,
      "state_relation_loss": 0.00180827610893175,
      "step": 1230,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.017127034813165666,
      "weighted_relation_loss": 0.017127034813165666
    },
    {
      "action_query_norm": 45.208530044555665,
      "action_relation_loss": 0.014008806832134724,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1599625818521984,
      "flow_matching_loss": 0.08454965725541115,
      "grad_norm": 0.36809539794921875,
      "learning_rate": 8.26e-05,
      "loss": 0.1003,
      "relation_loss": 0.01579523105174303,
      "selected_pairs": 26621.1,
      "state_query_norm": 45.160593032836914,
      "state_relation_loss": 0.0017864242778159677,
      "step": 1240,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.01579523105174303,
      "weighted_relation_loss": 0.01579523105174303
    },
    {
      "action_query_norm": 45.20847625732422,
      "action_relation_loss": 0.01473131999373436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1693171188026192,
      "flow_matching_loss": 0.0833884447813034,
      "grad_norm": 0.37335875630378723,
      "learning_rate": 8.326666666666667e-05,
      "loss": 0.0998,
      "relation_loss": 0.01639096299186349,
      "selected_pairs": 27023.8,
      "state_query_norm": 45.15850601196289,
      "state_relation_loss": 0.0016596429515630007,
      "step": 1250,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.01639096299186349,
      "weighted_relation_loss": 0.01639096299186349
    },
    {
      "action_query_norm": 45.20839500427246,
      "action_relation_loss": 0.014220582507550716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1786716557530403,
      "flow_matching_loss": 0.08675330281257629,
      "grad_norm": 0.29121559858322144,
      "learning_rate": 8.393333333333335e-05,
      "loss": 0.1025,
      "relation_loss": 0.015768306329846382,
      "selected_pairs": 27037.5,
      "state_query_norm": 45.156398391723634,
      "state_relation_loss": 0.001547723705880344,
      "step": 1260,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.015768306329846382,
      "weighted_relation_loss": 0.015768306329846382
    },
    {
      "action_query_norm": 45.207482528686526,
      "action_relation_loss": 0.014384204801172019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.1880261927034612,
      "flow_matching_loss": 0.08019518926739692,
      "grad_norm": 0.3290378451347351,
      "learning_rate": 8.46e-05,
      "loss": 0.0962,
      "relation_loss": 0.01599266827106476,
      "selected_pairs": 27059.5,
      "state_query_norm": 45.15679092407227,
      "state_relation_loss": 0.0016084634582512081,
      "step": 1270,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.01599266827106476,
      "weighted_relation_loss": 0.01599266827106476
    },
    {
      "action_query_norm": 45.20674362182617,
      "action_relation_loss": 0.014112206827849149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.197380729653882,
      "flow_matching_loss": 0.08449653163552284,
      "grad_norm": 0.2834104597568512,
      "learning_rate": 8.526666666666667e-05,
      "loss": 0.1002,
      "relation_loss": 0.015734068490564822,
      "selected_pairs": 26659.5,
      "state_query_norm": 45.154484558105466,
      "state_relation_loss": 0.0016218617791309952,
      "step": 1280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.015734068490564822,
      "weighted_relation_loss": 0.015734068490564822
    },
    {
      "action_query_norm": 45.20387077331543,
      "action_relation_loss": 0.015283450298011304,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2067352666043032,
      "flow_matching_loss": 0.08340805098414421,
      "grad_norm": 0.2149086594581604,
      "learning_rate": 8.593333333333333e-05,
      "loss": 0.1004,
      "relation_loss": 0.016968264989554883,
      "selected_pairs": 27108.8,
      "state_query_norm": 45.15318298339844,
      "state_relation_loss": 0.0016848147031851112,
      "step": 1290,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.016968264989554883,
      "weighted_relation_loss": 0.016968264989554883
    },
    {
      "action_query_norm": 45.202537536621094,
      "action_relation_loss": 0.014677897468209267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.216089803554724,
      "flow_matching_loss": 0.08014190569519997,
      "grad_norm": 0.22621633112430573,
      "learning_rate": 8.66e-05,
      "loss": 0.0964,
      "relation_loss": 0.016268413979560137,
      "selected_pairs": 26697.4,
      "state_query_norm": 45.15415802001953,
      "state_relation_loss": 0.001590516825672239,
      "step": 1300,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.016268413979560137,
      "weighted_relation_loss": 0.016268413979560137
    },
    {
      "action_query_norm": 45.20025405883789,
      "action_relation_loss": 0.014328118320554495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.225444340505145,
      "flow_matching_loss": 0.08267568424344063,
      "grad_norm": 0.33700987696647644,
      "learning_rate": 8.726666666666667e-05,
      "loss": 0.0988,
      "relation_loss": 0.016095809731632472,
      "selected_pairs": 26462.9,
      "state_query_norm": 45.15214500427246,
      "state_relation_loss": 0.0017676912946626544,
      "step": 1310,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.016095809731632472,
      "weighted_relation_loss": 0.016095809731632472
    },
    {
      "action_query_norm": 45.19883003234863,
      "action_relation_loss": 0.013792127929627895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.234798877455566,
      "flow_matching_loss": 0.08171737864613533,
      "grad_norm": 0.309990257024765,
      "learning_rate": 8.793333333333333e-05,
      "loss": 0.0973,
      "relation_loss": 0.015591917186975479,
      "selected_pairs": 27136.2,
      "state_query_norm": 45.15162620544434,
      "state_relation_loss": 0.0017997894086875022,
      "step": 1320,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.015591917186975479,
      "weighted_relation_loss": 0.015591917186975479
    },
    {
      "action_query_norm": 45.19645309448242,
      "action_relation_loss": 0.014437450189143419,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.244153414405987,
      "flow_matching_loss": 0.08090632632374764,
      "grad_norm": 0.284868448972702,
      "learning_rate": 8.86e-05,
      "loss": 0.097,
      "relation_loss": 0.01612152876332402,
      "selected_pairs": 26976.3,
      "state_query_norm": 45.15064010620117,
      "state_relation_loss": 0.0016840784694068134,
      "step": 1330,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01612152876332402,
      "weighted_relation_loss": 0.01612152876332402
    },
    {
      "action_query_norm": 45.19515762329102,
      "action_relation_loss": 0.01416391134262085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2535079513564078,
      "flow_matching_loss": 0.07885362729430198,
      "grad_norm": 0.3651602268218994,
      "learning_rate": 8.926666666666668e-05,
      "loss": 0.0947,
      "relation_loss": 0.015804447792470457,
      "selected_pairs": 27592.7,
      "state_query_norm": 45.14743003845215,
      "state_relation_loss": 0.0016405363916419447,
      "step": 1340,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.015804447792470457,
      "weighted_relation_loss": 0.015804447792470457
    },
    {
      "action_query_norm": 45.193759536743165,
      "action_relation_loss": 0.014512355998158455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.262862488306829,
      "flow_matching_loss": 0.07609353214502335,
      "grad_norm": 0.3006051182746887,
      "learning_rate": 8.993333333333334e-05,
      "loss": 0.0922,
      "relation_loss": 0.01612571608275175,
      "selected_pairs": 26967.6,
      "state_query_norm": 45.146054077148435,
      "state_relation_loss": 0.0016133601195178926,
      "step": 1350,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01612571608275175,
      "weighted_relation_loss": 0.01612571608275175
    },
    {
      "action_query_norm": 45.19460220336914,
      "action_relation_loss": 0.013786600064486265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2722170252572498,
      "flow_matching_loss": 0.07901373133063316,
      "grad_norm": 0.35259366035461426,
      "learning_rate": 9.06e-05,
      "loss": 0.0944,
      "relation_loss": 0.015375040099024772,
      "selected_pairs": 26574.5,
      "state_query_norm": 45.14633369445801,
      "state_relation_loss": 0.0015884400112554432,
      "step": 1360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.015375040099024772,
      "weighted_relation_loss": 0.015375040099024772
    },
    {
      "action_query_norm": 45.19319381713867,
      "action_relation_loss": 0.01439945176243782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2815715622076707,
      "flow_matching_loss": 0.07919137626886368,
      "grad_norm": 0.30996012687683105,
      "learning_rate": 9.126666666666667e-05,
      "loss": 0.0951,
      "relation_loss": 0.015862166415899992,
      "selected_pairs": 26501.1,
      "state_query_norm": 45.14339714050293,
      "state_relation_loss": 0.001462714362423867,
      "step": 1370,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.015862166415899992,
      "weighted_relation_loss": 0.015862166415899992
    },
    {
      "action_query_norm": 45.191520309448244,
      "action_relation_loss": 0.013789815921336412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.2909260991580918,
      "flow_matching_loss": 0.08282613083720207,
      "grad_norm": 0.3383494019508362,
      "learning_rate": 9.193333333333334e-05,
      "loss": 0.0982,
      "relation_loss": 0.015348293911665678,
      "selected_pairs": 27263.0,
      "state_query_norm": 45.144250106811526,
      "state_relation_loss": 0.0015584778157062829,
      "step": 1380,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.015348293911665678,
      "weighted_relation_loss": 0.015348293911665678
    },
    {
      "action_query_norm": 45.1888916015625,
      "action_relation_loss": 0.014497818984091282,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3002806361085126,
      "flow_matching_loss": 0.08360137268900872,
      "grad_norm": 0.29604214429855347,
      "learning_rate": 9.260000000000001e-05,
      "loss": 0.0997,
      "relation_loss": 0.016127450950443746,
      "selected_pairs": 26693.4,
      "state_query_norm": 45.14247093200684,
      "state_relation_loss": 0.0016296319314278661,
      "step": 1390,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.016127450950443746,
      "weighted_relation_loss": 0.016127450950443746
    },
    {
      "action_query_norm": 45.18539619445801,
      "action_relation_loss": 0.014250156376510859,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3096351730589335,
      "flow_matching_loss": 0.08307350724935532,
      "grad_norm": 0.3245616555213928,
      "learning_rate": 9.326666666666667e-05,
      "loss": 0.099,
      "relation_loss": 0.015930748730897903,
      "selected_pairs": 26919.3,
      "state_query_norm": 45.140755844116214,
      "state_relation_loss": 0.0016805926687084138,
      "step": 1400,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.015930748730897903,
      "weighted_relation_loss": 0.015930748730897903
    },
    {
      "action_query_norm": 45.183494567871094,
      "action_relation_loss": 0.013678541127592325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3189897100093546,
      "flow_matching_loss": 0.08001318201422691,
      "grad_norm": 0.28769299387931824,
      "learning_rate": 9.393333333333334e-05,
      "loss": 0.0951,
      "relation_loss": 0.015136257745325565,
      "selected_pairs": 27339.2,
      "state_query_norm": 45.14040603637695,
      "state_relation_loss": 0.0014577163732610644,
      "step": 1410,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.015136257745325565,
      "weighted_relation_loss": 0.015136257745325565
    },
    {
      "action_query_norm": 45.18200454711914,
      "action_relation_loss": 0.014372128248214721,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3283442469597755,
      "flow_matching_loss": 0.08131279423832893,
      "grad_norm": 0.29507169127464294,
      "learning_rate": 9.46e-05,
      "loss": 0.0972,
      "relation_loss": 0.015893350541591644,
      "selected_pairs": 26665.1,
      "state_query_norm": 45.138827896118165,
      "state_relation_loss": 0.001521222572773695,
      "step": 1420,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.015893350541591644,
      "weighted_relation_loss": 0.015893350541591644
    },
    {
      "action_query_norm": 45.1808162689209,
      "action_relation_loss": 0.014131669048219919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3376987839101964,
      "flow_matching_loss": 0.0800245687365532,
      "grad_norm": 0.2618167996406555,
      "learning_rate": 9.526666666666667e-05,
      "loss": 0.0957,
      "relation_loss": 0.015654105693101883,
      "selected_pairs": 26660.6,
      "state_query_norm": 45.13630676269531,
      "state_relation_loss": 0.0015224367496557533,
      "step": 1430,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.015654105693101883,
      "weighted_relation_loss": 0.015654105693101883
    },
    {
      "action_query_norm": 45.17859535217285,
      "action_relation_loss": 0.013537310343235732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3470533208606175,
      "flow_matching_loss": 0.08199491798877716,
      "grad_norm": 0.294070839881897,
      "learning_rate": 9.593333333333334e-05,
      "loss": 0.097,
      "relation_loss": 0.014999831281602382,
      "selected_pairs": 26951.2,
      "state_query_norm": 45.134024429321286,
      "state_relation_loss": 0.0014625207404606044,
      "step": 1440,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.014999831281602382,
      "weighted_relation_loss": 0.014999831281602382
    },
    {
      "action_query_norm": 45.17676620483398,
      "action_relation_loss": 0.012680767010897398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3564078578110383,
      "flow_matching_loss": 0.07925659120082855,
      "grad_norm": 0.30843019485473633,
      "learning_rate": 9.66e-05,
      "loss": 0.0934,
      "relation_loss": 0.014123235084116458,
      "selected_pairs": 26723.7,
      "state_query_norm": 45.13328247070312,
      "state_relation_loss": 0.0014424681314267219,
      "step": 1450,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.014123235084116458,
      "weighted_relation_loss": 0.014123235084116458
    },
    {
      "action_query_norm": 45.174625396728516,
      "action_relation_loss": 0.013562359008938074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3657623947614592,
      "flow_matching_loss": 0.08055609688162804,
      "grad_norm": 0.26303839683532715,
      "learning_rate": 9.726666666666667e-05,
      "loss": 0.0955,
      "relation_loss": 0.014974349644035101,
      "selected_pairs": 27150.1,
      "state_query_norm": 45.13132934570312,
      "state_relation_loss": 0.0014119905885308982,
      "step": 1460,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.014974349644035101,
      "weighted_relation_loss": 0.014974349644035101
    },
    {
      "action_query_norm": 45.17434196472168,
      "action_relation_loss": 0.012601378839462996,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3751169317118803,
      "flow_matching_loss": 0.07552218288183213,
      "grad_norm": 0.26848700642585754,
      "learning_rate": 9.793333333333333e-05,
      "loss": 0.0895,
      "relation_loss": 0.0140051051042974,
      "selected_pairs": 27109.4,
      "state_query_norm": 45.13034324645996,
      "state_relation_loss": 0.0014037261018529535,
      "step": 1470,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0140051051042974,
      "weighted_relation_loss": 0.0140051051042974
    },
    {
      "action_query_norm": 45.17098846435547,
      "action_relation_loss": 0.01392301581799984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.3844714686623012,
      "flow_matching_loss": 0.07826128676533699,
      "grad_norm": 0.2779375910758972,
      "learning_rate": 9.86e-05,
      "loss": 0.0937,
      "relation_loss": 0.015391257405281068,
      "selected_pairs": 26978.5,
      "state_query_norm": 45.1290901184082,
      "state_relation_loss": 0.0014682415756396949,
      "step": 1480,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.015391257405281068,
      "weighted_relation_loss": 0.015391257405281068
    },
    {
      "action_query_norm": 45.17010383605957,
      "action_relation_loss": 0.01385676497593522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.393826005612722,
      "flow_matching_loss": 0.08114760220050812,
      "grad_norm": 0.3615885078907013,
      "learning_rate": 9.926666666666668e-05,
      "loss": 0.0964,
      "relation_loss": 0.015268620755523443,
      "selected_pairs": 26717.2,
      "state_query_norm": 45.12526626586914,
      "state_relation_loss": 0.0014118559076450766,
      "step": 1490,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.015268620755523443,
      "weighted_relation_loss": 0.015268620755523443
    },
    {
      "action_query_norm": 45.16568222045898,
      "action_relation_loss": 0.013639620412141085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4031805425631432,
      "flow_matching_loss": 0.0768810547888279,
      "grad_norm": 0.2790745198726654,
      "learning_rate": 9.993333333333334e-05,
      "loss": 0.0919,
      "relation_loss": 0.015010160580277442,
      "selected_pairs": 26481.1,
      "state_query_norm": 45.12571640014649,
      "state_relation_loss": 0.001370540214702487,
      "step": 1500,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.015010160580277442,
      "weighted_relation_loss": 0.015010160580277442
    },
    {
      "action_query_norm": 45.16421966552734,
      "action_relation_loss": 0.014082630537450314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.412535079513564,
      "flow_matching_loss": 0.08068454414606094,
      "grad_norm": 0.3235214948654175,
      "learning_rate": 9.999997539434007e-05,
      "loss": 0.0963,
      "relation_loss": 0.015589841827750205,
      "selected_pairs": 26966.8,
      "state_query_norm": 45.122764587402344,
      "state_relation_loss": 0.0015072113834321498,
      "step": 1510,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.015589841827750205,
      "weighted_relation_loss": 0.015589841827750205
    },
    {
      "action_query_norm": 45.16282501220703,
      "action_relation_loss": 0.012965264543890954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.421889616463985,
      "flow_matching_loss": 0.08086610287427902,
      "grad_norm": 0.3297918438911438,
      "learning_rate": 9.999989033776898e-05,
      "loss": 0.0953,
      "relation_loss": 0.014392683375626802,
      "selected_pairs": 27545.6,
      "state_query_norm": 45.12194404602051,
      "state_relation_loss": 0.0014274187153205276,
      "step": 1520,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.014392683375626802,
      "weighted_relation_loss": 0.014392683375626802
    },
    {
      "action_query_norm": 45.1609561920166,
      "action_relation_loss": 0.013010305538773536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.431244153414406,
      "flow_matching_loss": 0.07985863760113716,
      "grad_norm": 0.3296429216861725,
      "learning_rate": 9.999974452661641e-05,
      "loss": 0.0942,
      "relation_loss": 0.014311800897121429,
      "selected_pairs": 26709.9,
      "state_query_norm": 45.120549011230466,
      "state_relation_loss": 0.0013014954864047467,
      "step": 1530,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.014311800897121429,
      "weighted_relation_loss": 0.014311800897121429
    },
    {
      "action_query_norm": 45.15901832580566,
      "action_relation_loss": 0.013040730450302362,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.440598690364827,
      "flow_matching_loss": 0.07873924225568771,
      "grad_norm": 0.27162453532218933,
      "learning_rate": 9.999953796105959e-05,
      "loss": 0.0931,
      "relation_loss": 0.014409908000379801,
      "selected_pairs": 26944.6,
      "state_query_norm": 45.11763458251953,
      "state_relation_loss": 0.001369177724700421,
      "step": 1540,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.014409908000379801,
      "weighted_relation_loss": 0.014409908000379801
    },
    {
      "action_query_norm": 45.15506172180176,
      "action_relation_loss": 0.013627804722636938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4499532273152478,
      "flow_matching_loss": 0.07624591290950775,
      "grad_norm": 0.2713761031627655,
      "learning_rate": 9.999927064134949e-05,
      "loss": 0.0913,
      "relation_loss": 0.01500996258109808,
      "selected_pairs": 26316.8,
      "state_query_norm": 45.117062377929685,
      "state_relation_loss": 0.0013821579166688025,
      "step": 1550,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.01500996258109808,
      "weighted_relation_loss": 0.01500996258109808
    },
    {
      "action_query_norm": 45.15515632629395,
      "action_relation_loss": 0.013106690160930157,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4593077642656689,
      "flow_matching_loss": 0.08058197274804116,
      "grad_norm": 0.22069574892520905,
      "learning_rate": 9.999894256781095e-05,
      "loss": 0.0951,
      "relation_loss": 0.01451193643733859,
      "selected_pairs": 26826.8,
      "state_query_norm": 45.11571578979492,
      "state_relation_loss": 0.0014052462414838375,
      "step": 1560,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.01451193643733859,
      "weighted_relation_loss": 0.01451193643733859
    },
    {
      "action_query_norm": 45.15345726013184,
      "action_relation_loss": 0.01393719231709838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4686623012160898,
      "flow_matching_loss": 0.07781763076782226,
      "grad_norm": 0.2964685261249542,
      "learning_rate": 9.99985537408426e-05,
      "loss": 0.0932,
      "relation_loss": 0.015380370151251554,
      "selected_pairs": 27079.3,
      "state_query_norm": 45.11461334228515,
      "state_relation_loss": 0.0014431779040023685,
      "step": 1570,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.015380370151251554,
      "weighted_relation_loss": 0.015380370151251554
    },
    {
      "action_query_norm": 45.15404930114746,
      "action_relation_loss": 0.013488473277539015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4780168381665106,
      "flow_matching_loss": 0.07704597860574722,
      "grad_norm": 0.2506782114505768,
      "learning_rate": 9.999810416091688e-05,
      "loss": 0.092,
      "relation_loss": 0.014936034381389619,
      "selected_pairs": 27011.6,
      "state_query_norm": 45.11415786743164,
      "state_relation_loss": 0.001447560975793749,
      "step": 1580,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.014936034381389619,
      "weighted_relation_loss": 0.014936034381389619
    },
    {
      "action_query_norm": 45.15279846191406,
      "action_relation_loss": 0.012673931661993265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4873713751169317,
      "flow_matching_loss": 0.0780644841492176,
      "grad_norm": 0.26288944482803345,
      "learning_rate": 9.99975938285801e-05,
      "loss": 0.0922,
      "relation_loss": 0.014087096508592366,
      "selected_pairs": 26389.4,
      "state_query_norm": 45.11135330200195,
      "state_relation_loss": 0.001413164916448295,
      "step": 1590,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.014087096508592366,
      "weighted_relation_loss": 0.014087096508592366
    },
    {
      "action_query_norm": 45.14969329833984,
      "action_relation_loss": 0.013370811566710472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.4967259120673526,
      "flow_matching_loss": 0.07906955480575562,
      "grad_norm": 0.17635399103164673,
      "learning_rate": 9.999702274445236e-05,
      "loss": 0.0939,
      "relation_loss": 0.014784155413508415,
      "selected_pairs": 27599.6,
      "state_query_norm": 45.11210174560547,
      "state_relation_loss": 0.0014133438118733466,
      "step": 1600,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.014784155413508415,
      "weighted_relation_loss": 0.014784155413508415
    },
    {
      "action_query_norm": 45.147256469726564,
      "action_relation_loss": 0.012824203167110682,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5060804490177735,
      "flow_matching_loss": 0.07835074439644814,
      "grad_norm": 0.24739566445350647,
      "learning_rate": 9.999639090922756e-05,
      "loss": 0.0927,
      "relation_loss": 0.014332873281091452,
      "selected_pairs": 26897.9,
      "state_query_norm": 45.11183586120605,
      "state_relation_loss": 0.0015086701489053667,
      "step": 1610,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.014332873281091452,
      "weighted_relation_loss": 0.014332873281091452
    },
    {
      "action_query_norm": 45.14577140808105,
      "action_relation_loss": 0.01306261271238327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5154349859681946,
      "flow_matching_loss": 0.07474321350455285,
      "grad_norm": 0.33673539757728577,
      "learning_rate": 9.999569832367346e-05,
      "loss": 0.0892,
      "relation_loss": 0.014498801808804274,
      "selected_pairs": 27243.8,
      "state_query_norm": 45.111124420166014,
      "state_relation_loss": 0.0014361890382133425,
      "step": 1620,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.014498801808804274,
      "weighted_relation_loss": 0.014498801808804274
    },
    {
      "action_query_norm": 45.14118995666504,
      "action_relation_loss": 0.012939379643648862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5247895229186157,
      "flow_matching_loss": 0.0779972493648529,
      "grad_norm": 0.2792033851146698,
      "learning_rate": 9.999494498863162e-05,
      "loss": 0.0923,
      "relation_loss": 0.014313377905637026,
      "selected_pairs": 26732.0,
      "state_query_norm": 45.10628967285156,
      "state_relation_loss": 0.001373998261988163,
      "step": 1630,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.014313377905637026,
      "weighted_relation_loss": 0.014313377905637026
    },
    {
      "action_query_norm": 45.138352966308595,
      "action_relation_loss": 0.01294904127717018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5341440598690363,
      "flow_matching_loss": 0.07719481140375137,
      "grad_norm": 0.1813494712114334,
      "learning_rate": 9.99941309050174e-05,
      "loss": 0.0915,
      "relation_loss": 0.01427627969533205,
      "selected_pairs": 27085.0,
      "state_query_norm": 45.10582809448242,
      "state_relation_loss": 0.001327238359954208,
      "step": 1640,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.01427627969533205,
      "weighted_relation_loss": 0.01427627969533205
    },
    {
      "action_query_norm": 45.1360897064209,
      "action_relation_loss": 0.013404911197721959,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5434985968194574,
      "flow_matching_loss": 0.07889341339468955,
      "grad_norm": 0.2393966019153595,
      "learning_rate": 9.999325607381999e-05,
      "loss": 0.0937,
      "relation_loss": 0.014787635579705239,
      "selected_pairs": 26320.7,
      "state_query_norm": 45.1059383392334,
      "state_relation_loss": 0.0013827243936248123,
      "step": 1650,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.014787635579705239,
      "weighted_relation_loss": 0.014787635579705239
    },
    {
      "action_query_norm": 45.13316345214844,
      "action_relation_loss": 0.012999864388257265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5528531337698785,
      "flow_matching_loss": 0.08013203740119934,
      "grad_norm": 0.18401454389095306,
      "learning_rate": 9.999232049610238e-05,
      "loss": 0.0944,
      "relation_loss": 0.014285468868911267,
      "selected_pairs": 26759.7,
      "state_query_norm": 45.10299797058106,
      "state_relation_loss": 0.001285604585427791,
      "step": 1660,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.014285468868911267,
      "weighted_relation_loss": 0.014285468868911267
    },
    {
      "action_query_norm": 45.13312339782715,
      "action_relation_loss": 0.013231632579118014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5622076707202992,
      "flow_matching_loss": 0.07626880332827568,
      "grad_norm": 0.3069908320903778,
      "learning_rate": 9.999132417300142e-05,
      "loss": 0.0908,
      "relation_loss": 0.014538425579667091,
      "selected_pairs": 26814.8,
      "state_query_norm": 45.103316116333005,
      "state_relation_loss": 0.0013067928608506917,
      "step": 1670,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.014538425579667091,
      "weighted_relation_loss": 0.014538425579667091
    },
    {
      "action_query_norm": 45.133405685424805,
      "action_relation_loss": 0.013377804402261972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5715622076707203,
      "flow_matching_loss": 0.07869788557291031,
      "grad_norm": 0.2997399568557739,
      "learning_rate": 9.99902671057277e-05,
      "loss": 0.0933,
      "relation_loss": 0.014588013477623462,
      "selected_pairs": 26572.3,
      "state_query_norm": 45.101942443847655,
      "state_relation_loss": 0.0012102089240215719,
      "step": 1680,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.014588013477623462,
      "weighted_relation_loss": 0.014588013477623462
    },
    {
      "action_query_norm": 45.1296443939209,
      "action_relation_loss": 0.012550909258425235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5809167446211414,
      "flow_matching_loss": 0.08011947497725487,
      "grad_norm": 0.33694618940353394,
      "learning_rate": 9.998914929556569e-05,
      "loss": 0.094,
      "relation_loss": 0.013915895391255617,
      "selected_pairs": 26998.2,
      "state_query_norm": 45.10038681030274,
      "state_relation_loss": 0.001364986365661025,
      "step": 1690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.013915895391255617,
      "weighted_relation_loss": 0.013915895391255617
    },
    {
      "action_query_norm": 45.12739028930664,
      "action_relation_loss": 0.012455405481159688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.590271281571562,
      "flow_matching_loss": 0.08179178833961487,
      "grad_norm": 0.3128647804260254,
      "learning_rate": 9.998797074387361e-05,
      "loss": 0.0956,
      "relation_loss": 0.013832388073205948,
      "selected_pairs": 26685.9,
      "state_query_norm": 45.09721488952637,
      "state_relation_loss": 0.0013769825920462609,
      "step": 1700,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.013832388073205948,
      "weighted_relation_loss": 0.013832388073205948
    },
    {
      "action_query_norm": 45.12403411865235,
      "action_relation_loss": 0.013091714307665824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.5996258185219832,
      "flow_matching_loss": 0.07750123292207718,
      "grad_norm": 0.22233474254608154,
      "learning_rate": 9.99867314520835e-05,
      "loss": 0.0919,
      "relation_loss": 0.014416566956788302,
      "selected_pairs": 26668.3,
      "state_query_norm": 45.09856414794922,
      "state_relation_loss": 0.0013248527189716696,
      "step": 1710,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.014416566956788302,
      "weighted_relation_loss": 0.014416566956788302
    },
    {
      "action_query_norm": 45.12444496154785,
      "action_relation_loss": 0.012959856353700162,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6089803554724043,
      "flow_matching_loss": 0.07710445150732995,
      "grad_norm": 0.33044520020484924,
      "learning_rate": 9.998543142170126e-05,
      "loss": 0.0914,
      "relation_loss": 0.014286814630031586,
      "selected_pairs": 26461.7,
      "state_query_norm": 45.09683456420898,
      "state_relation_loss": 0.001326958485879004,
      "step": 1720,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.014286814630031586,
      "weighted_relation_loss": 0.014286814630031586
    },
    {
      "action_query_norm": 45.12019577026367,
      "action_relation_loss": 0.0121173738501966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.618334892422825,
      "flow_matching_loss": 0.08075598254799843,
      "grad_norm": 0.3607083559036255,
      "learning_rate": 9.99840706543065e-05,
      "loss": 0.0943,
      "relation_loss": 0.013549461495131254,
      "selected_pairs": 27149.7,
      "state_query_norm": 45.094726181030275,
      "state_relation_loss": 0.0014320876449346541,
      "step": 1730,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.013549461495131254,
      "weighted_relation_loss": 0.013549461495131254
    },
    {
      "action_query_norm": 45.11767158508301,
      "action_relation_loss": 0.013261358253657818,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.627689429373246,
      "flow_matching_loss": 0.07755099460482598,
      "grad_norm": 0.34884732961654663,
      "learning_rate": 9.998264915155274e-05,
      "loss": 0.0922,
      "relation_loss": 0.014606043044477702,
      "selected_pairs": 26826.4,
      "state_query_norm": 45.09305648803711,
      "state_relation_loss": 0.001344684651121497,
      "step": 1740,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.014606043044477702,
      "weighted_relation_loss": 0.014606043044477702
    },
    {
      "action_query_norm": 45.11832504272461,
      "action_relation_loss": 0.01235485915094614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.637043966323667,
      "flow_matching_loss": 0.07759790048003197,
      "grad_norm": 0.29285728931427,
      "learning_rate": 9.998116691516718e-05,
      "loss": 0.0913,
      "relation_loss": 0.013699735142290592,
      "selected_pairs": 26640.6,
      "state_query_norm": 45.09407615661621,
      "state_relation_loss": 0.0013448760146275163,
      "step": 1750,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.013699735142290592,
      "weighted_relation_loss": 0.013699735142290592
    },
    {
      "action_query_norm": 45.116682815551755,
      "action_relation_loss": 0.01427143281325698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.646398503274088,
      "flow_matching_loss": 0.07958639711141587,
      "grad_norm": 0.2527390420436859,
      "learning_rate": 9.997962394695091e-05,
      "loss": 0.0952,
      "relation_loss": 0.015608522016555071,
      "selected_pairs": 26644.2,
      "state_query_norm": 45.0941463470459,
      "state_relation_loss": 0.0013370893662795424,
      "step": 1760,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.015608522016555071,
      "weighted_relation_loss": 0.015608522016555071
    },
    {
      "action_query_norm": 45.11215019226074,
      "action_relation_loss": 0.012603800371289253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6557530402245089,
      "flow_matching_loss": 0.07552361860871315,
      "grad_norm": 0.19573526084423065,
      "learning_rate": 9.997802024877875e-05,
      "loss": 0.0894,
      "relation_loss": 0.013926211092621089,
      "selected_pairs": 26640.7,
      "state_query_norm": 45.093156814575195,
      "state_relation_loss": 0.0013224107795394957,
      "step": 1770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.013926211092621089,
      "weighted_relation_loss": 0.013926211092621089
    },
    {
      "action_query_norm": 45.10874290466309,
      "action_relation_loss": 0.01291544958949089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.66510757717493,
      "flow_matching_loss": 0.077745421230793,
      "grad_norm": 0.20510295033454895,
      "learning_rate": 9.99763558225994e-05,
      "loss": 0.0919,
      "relation_loss": 0.014183186832815409,
      "selected_pairs": 26638.9,
      "state_query_norm": 45.09255676269531,
      "state_relation_loss": 0.001267737289890647,
      "step": 1780,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.014183186832815409,
      "weighted_relation_loss": 0.014183186832815409
    },
    {
      "action_query_norm": 45.106249237060545,
      "action_relation_loss": 0.013311308156698943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6744621141253508,
      "flow_matching_loss": 0.07515786364674568,
      "grad_norm": 0.2473999261856079,
      "learning_rate": 9.997463067043526e-05,
      "loss": 0.0898,
      "relation_loss": 0.014606872946023941,
      "selected_pairs": 27342.1,
      "state_query_norm": 45.08914756774902,
      "state_relation_loss": 0.001295564731117338,
      "step": 1790,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.014606872946023941,
      "weighted_relation_loss": 0.014606872946023941
    },
    {
      "action_query_norm": 45.105537796020506,
      "action_relation_loss": 0.012086125183850526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6838166510757717,
      "flow_matching_loss": 0.07706638798117638,
      "grad_norm": 0.20011916756629944,
      "learning_rate": 9.997284479438253e-05,
      "loss": 0.0904,
      "relation_loss": 0.013313659373670817,
      "selected_pairs": 27130.7,
      "state_query_norm": 45.08720932006836,
      "state_relation_loss": 0.0012275339919142424,
      "step": 1800,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.013313659373670817,
      "weighted_relation_loss": 0.013313659373670817
    },
    {
      "action_query_norm": 45.10229072570801,
      "action_relation_loss": 0.012749570701271296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.6931711880261928,
      "flow_matching_loss": 0.07696511894464493,
      "grad_norm": 0.20272298157215118,
      "learning_rate": 9.997099819661127e-05,
      "loss": 0.091,
      "relation_loss": 0.01400076700374484,
      "selected_pairs": 26480.5,
      "state_query_norm": 45.087640380859376,
      "state_relation_loss": 0.0012511960696429015,
      "step": 1810,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.01400076700374484,
      "weighted_relation_loss": 0.01400076700374484
    },
    {
      "action_query_norm": 45.10147361755371,
      "action_relation_loss": 0.012800135463476182,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7025257249766137,
      "flow_matching_loss": 0.07754967510700225,
      "grad_norm": 0.24661393463611603,
      "learning_rate": 9.996909087936524e-05,
      "loss": 0.0916,
      "relation_loss": 0.014024465717375279,
      "selected_pairs": 27339.8,
      "state_query_norm": 45.08820762634277,
      "state_relation_loss": 0.0012243301840499043,
      "step": 1820,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.014024465717375279,
      "weighted_relation_loss": 0.014024465717375279
    },
    {
      "action_query_norm": 45.099792861938475,
      "action_relation_loss": 0.012928474321961403,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7118802619270346,
      "flow_matching_loss": 0.08120860680937766,
      "grad_norm": 0.2345314472913742,
      "learning_rate": 9.996712284496201e-05,
      "loss": 0.0954,
      "relation_loss": 0.014161999244242907,
      "selected_pairs": 27126.5,
      "state_query_norm": 45.08702201843262,
      "state_relation_loss": 0.0012335250503383577,
      "step": 1830,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.014161999244242907,
      "weighted_relation_loss": 0.014161999244242907
    },
    {
      "action_query_norm": 45.09629821777344,
      "action_relation_loss": 0.012559243571013212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7212347988774557,
      "flow_matching_loss": 0.07613077759742737,
      "grad_norm": 0.23459717631340027,
      "learning_rate": 9.996509409579293e-05,
      "loss": 0.0901,
      "relation_loss": 0.01394827226176858,
      "selected_pairs": 26743.6,
      "state_query_norm": 45.085578536987306,
      "state_relation_loss": 0.0013890285743400454,
      "step": 1840,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.01394827226176858,
      "weighted_relation_loss": 0.01394827226176858
    },
    {
      "action_query_norm": 45.0952335357666,
      "action_relation_loss": 0.012934996653348208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7305893358278766,
      "flow_matching_loss": 0.07966068014502525,
      "grad_norm": 0.16581469774246216,
      "learning_rate": 9.996300463432312e-05,
      "loss": 0.0939,
      "relation_loss": 0.014284304715692997,
      "selected_pairs": 27202.9,
      "state_query_norm": 45.08088417053223,
      "state_relation_loss": 0.0013493082369677722,
      "step": 1850,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.014284304715692997,
      "weighted_relation_loss": 0.014284304715692997
    },
    {
      "action_query_norm": 45.0927116394043,
      "action_relation_loss": 0.012622499000281095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7399438727782974,
      "flow_matching_loss": 0.07257256135344506,
      "grad_norm": 0.20768730342388153,
      "learning_rate": 9.996085446309148e-05,
      "loss": 0.0865,
      "relation_loss": 0.013916655536741018,
      "selected_pairs": 26477.1,
      "state_query_norm": 45.079199600219724,
      "state_relation_loss": 0.0012941565015353262,
      "step": 1860,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.013916655536741018,
      "weighted_relation_loss": 0.013916655536741018
    },
    {
      "action_query_norm": 45.09132690429688,
      "action_relation_loss": 0.012105782981961966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7492984097287185,
      "flow_matching_loss": 0.07523631528019906,
      "grad_norm": 0.23234036564826965,
      "learning_rate": 9.995864358471066e-05,
      "loss": 0.0885,
      "relation_loss": 0.013305955659598112,
      "selected_pairs": 26878.5,
      "state_query_norm": 45.07599983215332,
      "state_relation_loss": 0.0012001728289760649,
      "step": 1870,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.013305955659598112,
      "weighted_relation_loss": 0.013305955659598112
    },
    {
      "action_query_norm": 45.091962051391604,
      "action_relation_loss": 0.012362024839967489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7586529466791394,
      "flow_matching_loss": 0.07423983886837959,
      "grad_norm": 0.2503344714641571,
      "learning_rate": 9.99563720018671e-05,
      "loss": 0.0877,
      "relation_loss": 0.013498826231807471,
      "selected_pairs": 27152.7,
      "state_query_norm": 45.07840690612793,
      "state_relation_loss": 0.0011368014151230455,
      "step": 1880,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.013498826231807471,
      "weighted_relation_loss": 0.013498826231807471
    },
    {
      "action_query_norm": 45.091064071655275,
      "action_relation_loss": 0.013041036017239094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7680074836295603,
      "flow_matching_loss": 0.0766624502837658,
      "grad_norm": 0.24156789481639862,
      "learning_rate": 9.995403971732098e-05,
      "loss": 0.0908,
      "relation_loss": 0.014125224947929383,
      "selected_pairs": 27152.4,
      "state_query_norm": 45.07663345336914,
      "state_relation_loss": 0.0010841890121810137,
      "step": 1890,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.014125224947929383,
      "weighted_relation_loss": 0.014125224947929383
    },
    {
      "action_query_norm": 45.08653221130371,
      "action_relation_loss": 0.013208841532468795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7773620205799814,
      "flow_matching_loss": 0.07263269647955894,
      "grad_norm": 0.2583112120628357,
      "learning_rate": 9.995164673390625e-05,
      "loss": 0.0869,
      "relation_loss": 0.014304338581860065,
      "selected_pairs": 27033.6,
      "state_query_norm": 45.07470855712891,
      "state_relation_loss": 0.0010954971192404627,
      "step": 1900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.014304338581860065,
      "weighted_relation_loss": 0.014304338581860065
    },
    {
      "action_query_norm": 45.085619354248045,
      "action_relation_loss": 0.013136611878871917,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7867165575304023,
      "flow_matching_loss": 0.07635155394673347,
      "grad_norm": 0.29145535826683044,
      "learning_rate": 9.994919305453059e-05,
      "loss": 0.0906,
      "relation_loss": 0.014233471639454365,
      "selected_pairs": 26486.7,
      "state_query_norm": 45.07321968078613,
      "state_relation_loss": 0.0010968596965540201,
      "step": 1910,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.014233471639454365,
      "weighted_relation_loss": 0.014233471639454365
    },
    {
      "action_query_norm": 45.08140106201172,
      "action_relation_loss": 0.011739598214626312,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.7960710944808231,
      "flow_matching_loss": 0.07878564968705178,
      "grad_norm": 0.29375478625297546,
      "learning_rate": 9.994667868217548e-05,
      "loss": 0.0917,
      "relation_loss": 0.012869262788444757,
      "selected_pairs": 27858.6,
      "state_query_norm": 45.07329216003418,
      "state_relation_loss": 0.001129664748441428,
      "step": 1920,
      "valid_action_samples": 236.5,
      "weighted_auxiliary_loss": 0.012869262788444757,
      "weighted_relation_loss": 0.012869262788444757
    },
    {
      "action_query_norm": 45.07601432800293,
      "action_relation_loss": 0.013129170145839452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8054256314312442,
      "flow_matching_loss": 0.07699382826685905,
      "grad_norm": 0.22404152154922485,
      "learning_rate": 9.99441036198961e-05,
      "loss": 0.0913,
      "relation_loss": 0.01428686324506998,
      "selected_pairs": 27152.7,
      "state_query_norm": 45.06900329589844,
      "state_relation_loss": 0.0011576930293813348,
      "step": 1930,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.01428686324506998,
      "weighted_relation_loss": 0.01428686324506998
    },
    {
      "action_query_norm": 45.07505989074707,
      "action_relation_loss": 0.013096177577972412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8147801683816651,
      "flow_matching_loss": 0.07600754648447036,
      "grad_norm": 0.3163326382637024,
      "learning_rate": 9.99414678708214e-05,
      "loss": 0.0902,
      "relation_loss": 0.014209241140633821,
      "selected_pairs": 26075.9,
      "state_query_norm": 45.06916847229004,
      "state_relation_loss": 0.0011130635335575788,
      "step": 1940,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.014209241140633821,
      "weighted_relation_loss": 0.014209241140633821
    },
    {
      "action_query_norm": 45.06918220520019,
      "action_relation_loss": 0.012018426228314638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.824134705332086,
      "flow_matching_loss": 0.0739514522254467,
      "grad_norm": 0.30062437057495117,
      "learning_rate": 9.993877143815407e-05,
      "loss": 0.0871,
      "relation_loss": 0.013109857123345137,
      "selected_pairs": 26204.5,
      "state_query_norm": 45.068647384643555,
      "state_relation_loss": 0.0010914309823419899,
      "step": 1950,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.013109857123345137,
      "weighted_relation_loss": 0.013109857123345137
    },
    {
      "action_query_norm": 45.065527725219724,
      "action_relation_loss": 0.01302949795499444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.833489242282507,
      "flow_matching_loss": 0.07340925037860871,
      "grad_norm": 0.2491753101348877,
      "learning_rate": 9.993601432517053e-05,
      "loss": 0.0875,
      "relation_loss": 0.014077496156096459,
      "selected_pairs": 26369.8,
      "state_query_norm": 45.06706581115723,
      "state_relation_loss": 0.0010479982476681471,
      "step": 1960,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.014077496156096459,
      "weighted_relation_loss": 0.014077496156096459
    },
    {
      "action_query_norm": 45.063592529296876,
      "action_relation_loss": 0.012758486624807119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.842843779232928,
      "flow_matching_loss": 0.07305736094713211,
      "grad_norm": 0.22951185703277588,
      "learning_rate": 9.993319653522091e-05,
      "loss": 0.087,
      "relation_loss": 0.013898764736950397,
      "selected_pairs": 26216.1,
      "state_query_norm": 45.0622974395752,
      "state_relation_loss": 0.0011402780015487223,
      "step": 1970,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.013898764736950397,
      "weighted_relation_loss": 0.013898764736950397
    },
    {
      "action_query_norm": 45.06616134643555,
      "action_relation_loss": 0.012507059331983327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8521983161833488,
      "flow_matching_loss": 0.07520819902420044,
      "grad_norm": 0.22311833500862122,
      "learning_rate": 9.993031807172911e-05,
      "loss": 0.0888,
      "relation_loss": 0.01360583482310176,
      "selected_pairs": 26673.4,
      "state_query_norm": 45.065264511108396,
      "state_relation_loss": 0.001098775491118431,
      "step": 1980,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.01360583482310176,
      "weighted_relation_loss": 0.01360583482310176
    },
    {
      "action_query_norm": 45.060917663574216,
      "action_relation_loss": 0.012058415170758962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.86155285313377,
      "flow_matching_loss": 0.07519553527235985,
      "grad_norm": 0.16246120631694794,
      "learning_rate": 9.992737893819273e-05,
      "loss": 0.0883,
      "relation_loss": 0.013111897464841604,
      "selected_pairs": 26311.4,
      "state_query_norm": 45.06254920959473,
      "state_relation_loss": 0.0010534822125919164,
      "step": 1990,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.013111897464841604,
      "weighted_relation_loss": 0.013111897464841604
    },
    {
      "action_query_norm": 45.05859603881836,
      "action_relation_loss": 0.012424052134156227,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8709073900841908,
      "flow_matching_loss": 0.07474018633365631,
      "grad_norm": 0.2720249593257904,
      "learning_rate": 9.992437913818312e-05,
      "loss": 0.0883,
      "relation_loss": 0.013556331396102905,
      "selected_pairs": 26806.5,
      "state_query_norm": 45.05905113220215,
      "state_relation_loss": 0.0011322793201543392,
      "step": 2000,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.013556331396102905,
      "weighted_relation_loss": 0.013556331396102905
    },
    {
      "action_query_norm": 45.052519989013675,
      "action_relation_loss": 0.012499096430838108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8802619270346117,
      "flow_matching_loss": 0.07457466199994087,
      "grad_norm": 0.2561357915401459,
      "learning_rate": 9.992131867534526e-05,
      "loss": 0.0882,
      "relation_loss": 0.013621976412832738,
      "selected_pairs": 27383.8,
      "state_query_norm": 45.05854263305664,
      "state_relation_loss": 0.0011228799005039037,
      "step": 2010,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.013621976412832738,
      "weighted_relation_loss": 0.013621976412832738
    },
    {
      "action_query_norm": 45.05081405639648,
      "action_relation_loss": 0.012465731799602508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8896164639850328,
      "flow_matching_loss": 0.07618649303913116,
      "grad_norm": 0.20400717854499817,
      "learning_rate": 9.991819755339796e-05,
      "loss": 0.0897,
      "relation_loss": 0.013511249795556068,
      "selected_pairs": 26844.6,
      "state_query_norm": 45.05574493408203,
      "state_relation_loss": 0.00104551812983118,
      "step": 2020,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.013511249795556068,
      "weighted_relation_loss": 0.013511249795556068
    },
    {
      "action_query_norm": 45.04685592651367,
      "action_relation_loss": 0.012215229123830796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.8989710009354537,
      "flow_matching_loss": 0.0746017873287201,
      "grad_norm": 0.21732695400714874,
      "learning_rate": 9.991501577613365e-05,
      "loss": 0.0879,
      "relation_loss": 0.013249435741454364,
      "selected_pairs": 26619.7,
      "state_query_norm": 45.055724716186525,
      "state_relation_loss": 0.0010342066234443337,
      "step": 2030,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.013249435741454364,
      "weighted_relation_loss": 0.013249435741454364
    },
    {
      "action_query_norm": 45.04971351623535,
      "action_relation_loss": 0.011708487290889025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9083255378858746,
      "flow_matching_loss": 0.07550494223833085,
      "grad_norm": 0.24039649963378906,
      "learning_rate": 9.99117733474185e-05,
      "loss": 0.0883,
      "relation_loss": 0.012768849264830351,
      "selected_pairs": 26871.5,
      "state_query_norm": 45.05273284912109,
      "state_relation_loss": 0.0010603619855828583,
      "step": 2040,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.012768849264830351,
      "weighted_relation_loss": 0.012768849264830351
    },
    {
      "action_query_norm": 45.04511642456055,
      "action_relation_loss": 0.012044950295239687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9176800748362957,
      "flow_matching_loss": 0.07663976401090622,
      "grad_norm": 0.23461730778217316,
      "learning_rate": 9.990847027119234e-05,
      "loss": 0.0897,
      "relation_loss": 0.013105950132012368,
      "selected_pairs": 26341.8,
      "state_query_norm": 45.05161819458008,
      "state_relation_loss": 0.0010609997203573584,
      "step": 2050,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.013105950132012368,
      "weighted_relation_loss": 0.013105950132012368
    },
    {
      "action_query_norm": 45.04253387451172,
      "action_relation_loss": 0.01166589865460992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9270346117867165,
      "flow_matching_loss": 0.07154318951070308,
      "grad_norm": 0.2314489334821701,
      "learning_rate": 9.990510655146877e-05,
      "loss": 0.0843,
      "relation_loss": 0.012711598258465528,
      "selected_pairs": 26916.8,
      "state_query_norm": 45.051456451416016,
      "state_relation_loss": 0.0010456994466949254,
      "step": 2060,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.012711598258465528,
      "weighted_relation_loss": 0.012711598258465528
    },
    {
      "action_query_norm": 45.039765167236325,
      "action_relation_loss": 0.011266223154962063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9363891487371374,
      "flow_matching_loss": 0.0743184246122837,
      "grad_norm": 0.25447869300842285,
      "learning_rate": 9.990168219233496e-05,
      "loss": 0.0867,
      "relation_loss": 0.012345515098422766,
      "selected_pairs": 27174.8,
      "state_query_norm": 45.04839973449707,
      "state_relation_loss": 0.0010792921006213874,
      "step": 2070,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.012345515098422766,
      "weighted_relation_loss": 0.012345515098422766
    },
    {
      "action_query_norm": 45.038000869750974,
      "action_relation_loss": 0.012089511938393116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9457436856875585,
      "flow_matching_loss": 0.07355519235134125,
      "grad_norm": 0.18083330988883972,
      "learning_rate": 9.989819719795188e-05,
      "loss": 0.0867,
      "relation_loss": 0.013169898837804794,
      "selected_pairs": 27056.8,
      "state_query_norm": 45.04972801208496,
      "state_relation_loss": 0.0010803869809024036,
      "step": 2080,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.013169898837804794,
      "weighted_relation_loss": 0.013169898837804794
    },
    {
      "action_query_norm": 45.036560440063475,
      "action_relation_loss": 0.012050179112702609,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9550982226379794,
      "flow_matching_loss": 0.07032805457711219,
      "grad_norm": 0.22176845371723175,
      "learning_rate": 9.989465157255412e-05,
      "loss": 0.0835,
      "relation_loss": 0.013131276424974202,
      "selected_pairs": 26812.1,
      "state_query_norm": 45.04936485290527,
      "state_relation_loss": 0.0010810972831677645,
      "step": 2090,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.013131276424974202,
      "weighted_relation_loss": 0.013131276424974202
    },
    {
      "action_query_norm": 45.03514060974121,
      "action_relation_loss": 0.012839552201330661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9644527595884003,
      "flow_matching_loss": 0.07466611862182618,
      "grad_norm": 0.27989640831947327,
      "learning_rate": 9.989104532044994e-05,
      "loss": 0.0885,
      "relation_loss": 0.013881358411163091,
      "selected_pairs": 26047.8,
      "state_query_norm": 45.04973030090332,
      "state_relation_loss": 0.0010418061050586403,
      "step": 2100,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.013881358411163091,
      "weighted_relation_loss": 0.013881358411163091
    },
    {
      "action_query_norm": 45.033098220825195,
      "action_relation_loss": 0.012179682403802872,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9738072965388214,
      "flow_matching_loss": 0.07217254564166069,
      "grad_norm": 0.281342089176178,
      "learning_rate": 9.988737844602128e-05,
      "loss": 0.0854,
      "relation_loss": 0.013221915252506733,
      "selected_pairs": 27162.7,
      "state_query_norm": 45.04469451904297,
      "state_relation_loss": 0.001042232825420797,
      "step": 2110,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.013221915252506733,
      "weighted_relation_loss": 0.013221915252506733
    },
    {
      "action_query_norm": 45.02958374023437,
      "action_relation_loss": 0.011647993978112937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9831618334892422,
      "flow_matching_loss": 0.07347209602594376,
      "grad_norm": 0.20461000502109528,
      "learning_rate": 9.988365095372372e-05,
      "loss": 0.0861,
      "relation_loss": 0.012593142036348582,
      "selected_pairs": 27179.4,
      "state_query_norm": 45.04429817199707,
      "state_relation_loss": 0.0009451480756979436,
      "step": 2120,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.012593142036348582,
      "weighted_relation_loss": 0.012593142036348582
    },
    {
      "action_query_norm": 45.02027015686035,
      "action_relation_loss": 0.012358248699456453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 1.9925163704396631,
      "flow_matching_loss": 0.0732746172696352,
      "grad_norm": 0.1711350679397583,
      "learning_rate": 9.987986284808654e-05,
      "loss": 0.0866,
      "relation_loss": 0.013358073588460685,
      "selected_pairs": 26916.4,
      "state_query_norm": 45.04419860839844,
      "state_relation_loss": 0.000999824987957254,
      "step": 2130,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.013358073588460685,
      "weighted_relation_loss": 0.013358073588460685
    },
    {
      "action_query_norm": 45.019721603393556,
      "action_relation_loss": 0.011374303419142962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.001870907390084,
      "flow_matching_loss": 0.07314998544752598,
      "grad_norm": 0.19197607040405273,
      "learning_rate": 9.987601413371264e-05,
      "loss": 0.0855,
      "relation_loss": 0.012331152986735105,
      "selected_pairs": 24556.9,
      "state_query_norm": 45.03945121765137,
      "state_relation_loss": 0.0009568495152052492,
      "step": 2140,
      "valid_action_samples": 215.2,
      "weighted_auxiliary_loss": 0.012331152986735105,
      "weighted_relation_loss": 0.012331152986735105
    },
    {
      "action_query_norm": 45.01828193664551,
      "action_relation_loss": 0.012000446300953626,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.0112254443405053,
      "flow_matching_loss": 0.07883658334612846,
      "grad_norm": 0.2544567286968231,
      "learning_rate": 9.987210481527855e-05,
      "loss": 0.0918,
      "relation_loss": 0.013011598307639361,
      "selected_pairs": 26400.0,
      "state_query_norm": 45.038364028930665,
      "state_relation_loss": 0.0010111519251950086,
      "step": 2150,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.013011598307639361,
      "weighted_relation_loss": 0.013011598307639361
    },
    {
      "action_query_norm": 45.019202041625974,
      "action_relation_loss": 0.01200074478983879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.020579981290926,
      "flow_matching_loss": 0.0753841146826744,
      "grad_norm": 0.300721138715744,
      "learning_rate": 9.98681348975345e-05,
      "loss": 0.0883,
      "relation_loss": 0.01296262526884675,
      "selected_pairs": 26648.1,
      "state_query_norm": 45.03812103271484,
      "state_relation_loss": 0.0009618803567718714,
      "step": 2160,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.01296262526884675,
      "weighted_relation_loss": 0.01296262526884675
    },
    {
      "action_query_norm": 45.01735916137695,
      "action_relation_loss": 0.012589353043586016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.029934518241347,
      "flow_matching_loss": 0.0747525416314602,
      "grad_norm": 0.19548317790031433,
      "learning_rate": 9.986410438530427e-05,
      "loss": 0.0883,
      "relation_loss": 0.013593303970992565,
      "selected_pairs": 27128.4,
      "state_query_norm": 45.03539886474609,
      "state_relation_loss": 0.0010039508575573563,
      "step": 2170,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.013593303970992565,
      "weighted_relation_loss": 0.013593303970992565
    },
    {
      "action_query_norm": 45.02068824768067,
      "action_relation_loss": 0.011827005911618471,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.039289055191768,
      "flow_matching_loss": 0.07132745906710625,
      "grad_norm": 0.2319287359714508,
      "learning_rate": 9.986001328348534e-05,
      "loss": 0.0841,
      "relation_loss": 0.012788462173193692,
      "selected_pairs": 27291.6,
      "state_query_norm": 45.03465538024902,
      "state_relation_loss": 0.0009614562615752221,
      "step": 2180,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.012788462173193692,
      "weighted_relation_loss": 0.012788462173193692
    },
    {
      "action_query_norm": 45.01231842041015,
      "action_relation_loss": 0.011974423099309206,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.048643592142189,
      "flow_matching_loss": 0.07008712403476239,
      "grad_norm": 0.24836835265159607,
      "learning_rate": 9.985586159704878e-05,
      "loss": 0.0829,
      "relation_loss": 0.012852758169174194,
      "selected_pairs": 26761.1,
      "state_query_norm": 45.03410873413086,
      "state_relation_loss": 0.0008783351688180119,
      "step": 2190,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.012852758169174194,
      "weighted_relation_loss": 0.012852758169174194
    },
    {
      "action_query_norm": 45.01022605895996,
      "action_relation_loss": 0.011566811427474021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.05799812909261,
      "flow_matching_loss": 0.07357986718416214,
      "grad_norm": 0.1801110804080963,
      "learning_rate": 9.985164933103929e-05,
      "loss": 0.0861,
      "relation_loss": 0.01249657105654478,
      "selected_pairs": 26971.8,
      "state_query_norm": 45.03311004638672,
      "state_relation_loss": 0.0009297596348915249,
      "step": 2200,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.01249657105654478,
      "weighted_relation_loss": 0.01249657105654478
    },
    {
      "action_query_norm": 45.00823059082031,
      "action_relation_loss": 0.011741234362125397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.067352666043031,
      "flow_matching_loss": 0.07229062393307686,
      "grad_norm": 0.2581319510936737,
      "learning_rate": 9.984737649057513e-05,
      "loss": 0.085,
      "relation_loss": 0.012734964862465859,
      "selected_pairs": 26928.6,
      "state_query_norm": 45.0316879272461,
      "state_relation_loss": 0.0009937305527273565,
      "step": 2210,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.012734964862465859,
      "weighted_relation_loss": 0.012734964862465859
    },
    {
      "action_query_norm": 45.00345039367676,
      "action_relation_loss": 0.012371599581092596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.0767072029934517,
      "flow_matching_loss": 0.07417081072926521,
      "grad_norm": 0.1729535013437271,
      "learning_rate": 9.984304308084827e-05,
      "loss": 0.0875,
      "relation_loss": 0.013321974780410528,
      "selected_pairs": 26822.9,
      "state_query_norm": 45.031700134277344,
      "state_relation_loss": 0.0009503752109594644,
      "step": 2220,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.013321974780410528,
      "weighted_relation_loss": 0.013321974780410528
    },
    {
      "action_query_norm": 44.999171829223634,
      "action_relation_loss": 0.012140569090843201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.086061739943873,
      "flow_matching_loss": 0.07512362822890281,
      "grad_norm": 0.2285943180322647,
      "learning_rate": 9.983864910712416e-05,
      "loss": 0.0883,
      "relation_loss": 0.013134855963289737,
      "selected_pairs": 27156.8,
      "state_query_norm": 45.02971839904785,
      "state_relation_loss": 0.0009942868317011743,
      "step": 2230,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.013134855963289737,
      "weighted_relation_loss": 0.013134855963289737
    },
    {
      "action_query_norm": 44.99949264526367,
      "action_relation_loss": 0.011781747918576002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.095416276894294,
      "flow_matching_loss": 0.07453542351722717,
      "grad_norm": 0.1941005438566208,
      "learning_rate": 9.98341945747419e-05,
      "loss": 0.0873,
      "relation_loss": 0.012797153927385807,
      "selected_pairs": 27174.6,
      "state_query_norm": 45.02790832519531,
      "state_relation_loss": 0.001015405950602144,
      "step": 2240,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.012797153927385807,
      "weighted_relation_loss": 0.012797153927385807
    },
    {
      "action_query_norm": 44.99829292297363,
      "action_relation_loss": 0.01172624034807086,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1047708138447145,
      "flow_matching_loss": 0.07643972113728523,
      "grad_norm": 0.16580119729042053,
      "learning_rate": 9.98296794891142e-05,
      "loss": 0.0892,
      "relation_loss": 0.012758167646825313,
      "selected_pairs": 27131.5,
      "state_query_norm": 45.02574119567871,
      "state_relation_loss": 0.0010319272172637285,
      "step": 2250,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.012758167646825313,
      "weighted_relation_loss": 0.012758167646825313
    },
    {
      "action_query_norm": 44.9968017578125,
      "action_relation_loss": 0.012116635963320733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1141253507951356,
      "flow_matching_loss": 0.0773981012403965,
      "grad_norm": 0.234228253364563,
      "learning_rate": 9.982510385572725e-05,
      "loss": 0.0906,
      "relation_loss": 0.013153400365263223,
      "selected_pairs": 26824.4,
      "state_query_norm": 45.02608451843262,
      "state_relation_loss": 0.001036764495074749,
      "step": 2260,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.013153400365263223,
      "weighted_relation_loss": 0.013153400365263223
    },
    {
      "action_query_norm": 44.99062385559082,
      "action_relation_loss": 0.011898133158683776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1234798877455567,
      "flow_matching_loss": 0.07678029537200928,
      "grad_norm": 0.2878287434577942,
      "learning_rate": 9.982046768014094e-05,
      "loss": 0.0897,
      "relation_loss": 0.012956959567964077,
      "selected_pairs": 26792.6,
      "state_query_norm": 45.02564582824707,
      "state_relation_loss": 0.0010588265140540898,
      "step": 2270,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.012956959567964077,
      "weighted_relation_loss": 0.012956959567964077
    },
    {
      "action_query_norm": 44.99194145202637,
      "action_relation_loss": 0.011441702954471111,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1328344246959774,
      "flow_matching_loss": 0.07152113541960717,
      "grad_norm": 0.22964707016944885,
      "learning_rate": 9.981577096798863e-05,
      "loss": 0.0839,
      "relation_loss": 0.012399575393646955,
      "selected_pairs": 26942.0,
      "state_query_norm": 45.023658752441406,
      "state_relation_loss": 0.0009578725090250373,
      "step": 2280,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.012399575393646955,
      "weighted_relation_loss": 0.012399575393646955
    },
    {
      "action_query_norm": 44.99151725769043,
      "action_relation_loss": 0.011639506090432405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1421889616463985,
      "flow_matching_loss": 0.07141118645668029,
      "grad_norm": 0.22504231333732605,
      "learning_rate": 9.981101372497727e-05,
      "loss": 0.084,
      "relation_loss": 0.012604463752359152,
      "selected_pairs": 26789.0,
      "state_query_norm": 45.02409439086914,
      "state_relation_loss": 0.0009649576735682785,
      "step": 2290,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.012604463752359152,
      "weighted_relation_loss": 0.012604463752359152
    },
    {
      "action_query_norm": 44.98466186523437,
      "action_relation_loss": 0.011153768096119166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1515434985968196,
      "flow_matching_loss": 0.07070473358035087,
      "grad_norm": 0.1688978374004364,
      "learning_rate": 9.980619595688737e-05,
      "loss": 0.0828,
      "relation_loss": 0.012133438978344202,
      "selected_pairs": 26996.1,
      "state_query_norm": 45.02128448486328,
      "state_relation_loss": 0.0009796709113288672,
      "step": 2300,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.012133438978344202,
      "weighted_relation_loss": 0.012133438978344202
    },
    {
      "action_query_norm": 44.9855541229248,
      "action_relation_loss": 0.011230128165334464,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1608980355472402,
      "flow_matching_loss": 0.07211866676807403,
      "grad_norm": 0.2949434518814087,
      "learning_rate": 9.980131766957295e-05,
      "loss": 0.0843,
      "relation_loss": 0.012181219551712275,
      "selected_pairs": 26411.7,
      "state_query_norm": 45.0219783782959,
      "state_relation_loss": 0.0009510913339909166,
      "step": 2310,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.012181219551712275,
      "weighted_relation_loss": 0.012181219551712275
    },
    {
      "action_query_norm": 44.98130416870117,
      "action_relation_loss": 0.011799391824752092,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1702525724976613,
      "flow_matching_loss": 0.0747261993587017,
      "grad_norm": 0.2606983780860901,
      "learning_rate": 9.979637886896163e-05,
      "loss": 0.0875,
      "relation_loss": 0.012818972114473581,
      "selected_pairs": 26457.8,
      "state_query_norm": 45.02121696472168,
      "state_relation_loss": 0.0010195802664384247,
      "step": 2320,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.012818972114473581,
      "weighted_relation_loss": 0.012818972114473581
    },
    {
      "action_query_norm": 44.98053817749023,
      "action_relation_loss": 0.01175501262769103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.1796071094480824,
      "flow_matching_loss": 0.07280634343624115,
      "grad_norm": 0.2281946837902069,
      "learning_rate": 9.979137956105447e-05,
      "loss": 0.0855,
      "relation_loss": 0.012658810056746007,
      "selected_pairs": 26272.8,
      "state_query_norm": 45.01672134399414,
      "state_relation_loss": 0.0009037974756211043,
      "step": 2330,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.012658810056746007,
      "weighted_relation_loss": 0.012658810056746007
    },
    {
      "action_query_norm": 44.98101272583008,
      "action_relation_loss": 0.01201686281710863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.188961646398503,
      "flow_matching_loss": 0.07394763305783272,
      "grad_norm": 0.21026766300201416,
      "learning_rate": 9.978631975192613e-05,
      "loss": 0.0869,
      "relation_loss": 0.012940130848437547,
      "selected_pairs": 27174.2,
      "state_query_norm": 45.014878845214845,
      "state_relation_loss": 0.0009232680546119809,
      "step": 2340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.012940130848437547,
      "weighted_relation_loss": 0.012940130848437547
    },
    {
      "action_query_norm": 44.98126335144043,
      "action_relation_loss": 0.011828952468931675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.198316183348924,
      "flow_matching_loss": 0.06982225067913532,
      "grad_norm": 0.1711663007736206,
      "learning_rate": 9.978119944772475e-05,
      "loss": 0.0826,
      "relation_loss": 0.012737864442169667,
      "selected_pairs": 27102.6,
      "state_query_norm": 45.01271781921387,
      "state_relation_loss": 0.0009089120430871844,
      "step": 2350,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.012737864442169667,
      "weighted_relation_loss": 0.012737864442169667
    },
    {
      "action_query_norm": 44.97556991577149,
      "action_relation_loss": 0.01227588513866067,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2076707202993453,
      "flow_matching_loss": 0.07364902049303054,
      "grad_norm": 0.2672383487224579,
      "learning_rate": 9.977601865467197e-05,
      "loss": 0.0869,
      "relation_loss": 0.013217470049858094,
      "selected_pairs": 26735.8,
      "state_query_norm": 45.01035919189453,
      "state_relation_loss": 0.0009415849752258509,
      "step": 2360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.013217470049858094,
      "weighted_relation_loss": 0.013217470049858094
    },
    {
      "action_query_norm": 44.973345565795896,
      "action_relation_loss": 0.012219636794179678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.217025257249766,
      "flow_matching_loss": 0.07418671324849128,
      "grad_norm": 0.18414361774921417,
      "learning_rate": 9.977077737906297e-05,
      "loss": 0.0873,
      "relation_loss": 0.013136938959360123,
      "selected_pairs": 27174.6,
      "state_query_norm": 45.00948028564453,
      "state_relation_loss": 0.0009173022292088718,
      "step": 2370,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.013136938959360123,
      "weighted_relation_loss": 0.013136938959360123
    },
    {
      "action_query_norm": 44.97079963684082,
      "action_relation_loss": 0.011145231872797012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.226379794200187,
      "flow_matching_loss": 0.0723235361278057,
      "grad_norm": 0.17498570680618286,
      "learning_rate": 9.976547562726636e-05,
      "loss": 0.0844,
      "relation_loss": 0.012106426618993282,
      "selected_pairs": 26759.0,
      "state_query_norm": 45.00740013122559,
      "state_relation_loss": 0.000961194938281551,
      "step": 2380,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.012106426618993282,
      "weighted_relation_loss": 0.012106426618993282
    },
    {
      "action_query_norm": 44.962451171875,
      "action_relation_loss": 0.011580782849341631,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.235734331150608,
      "flow_matching_loss": 0.06974669471383095,
      "grad_norm": 0.1911175400018692,
      "learning_rate": 9.976011340572429e-05,
      "loss": 0.0823,
      "relation_loss": 0.012528859172016382,
      "selected_pairs": 27153.0,
      "state_query_norm": 45.00579681396484,
      "state_relation_loss": 0.0009480763517785817,
      "step": 2390,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.012528859172016382,
      "weighted_relation_loss": 0.012528859172016382
    },
    {
      "action_query_norm": 44.96398582458496,
      "action_relation_loss": 0.011357927508652211,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.245088868101029,
      "flow_matching_loss": 0.07332026362419128,
      "grad_norm": 0.19388031959533691,
      "learning_rate": 9.975469072095237e-05,
      "loss": 0.0856,
      "relation_loss": 0.012260048184543848,
      "selected_pairs": 26781.3,
      "state_query_norm": 45.00832748413086,
      "state_relation_loss": 0.0009021207340992987,
      "step": 2400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.012260048184543848,
      "weighted_relation_loss": 0.012260048184543848
    },
    {
      "action_query_norm": 44.960931396484376,
      "action_relation_loss": 0.010873091407120228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.25444340505145,
      "flow_matching_loss": 0.07163400948047638,
      "grad_norm": 0.21477831900119781,
      "learning_rate": 9.974920757953965e-05,
      "loss": 0.0834,
      "relation_loss": 0.011809319350868463,
      "selected_pairs": 26433.6,
      "state_query_norm": 45.005330657958986,
      "state_relation_loss": 0.000936227961210534,
      "step": 2410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.011809319350868463,
      "weighted_relation_loss": 0.011809319350868463
    },
    {
      "action_query_norm": 44.95571327209473,
      "action_relation_loss": 0.010911241732537746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.263797942001871,
      "flow_matching_loss": 0.0703727077692747,
      "grad_norm": 0.22327497601509094,
      "learning_rate": 9.97436639881487e-05,
      "loss": 0.0822,
      "relation_loss": 0.011842901539057493,
      "selected_pairs": 27059.9,
      "state_query_norm": 45.004351425170896,
      "state_relation_loss": 0.0009316597192082554,
      "step": 2420,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.011842901539057493,
      "weighted_relation_loss": 0.011842901539057493
    },
    {
      "action_query_norm": 44.95166931152344,
      "action_relation_loss": 0.011332139000296592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2731524789522917,
      "flow_matching_loss": 0.07167470529675483,
      "grad_norm": 0.1772642731666565,
      "learning_rate": 9.973805995351545e-05,
      "loss": 0.084,
      "relation_loss": 0.012305320426821708,
      "selected_pairs": 26532.1,
      "state_query_norm": 45.00357246398926,
      "state_relation_loss": 0.0009731813392136246,
      "step": 2430,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.012305320426821708,
      "weighted_relation_loss": 0.012305320426821708
    },
    {
      "action_query_norm": 44.95399971008301,
      "action_relation_loss": 0.011517503764480352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.2825070159027128,
      "flow_matching_loss": 0.0735580176115036,
      "grad_norm": 0.1587660163640976,
      "learning_rate": 9.973239548244939e-05,
      "loss": 0.086,
      "relation_loss": 0.01243015481159091,
      "selected_pairs": 26501.9,
      "state_query_norm": 45.000976181030275,
      "state_relation_loss": 0.0009126509947236627,
      "step": 2440,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.01243015481159091,
      "weighted_relation_loss": 0.01243015481159091
    },
    {
      "action_query_norm": 44.95223121643066,
      "action_relation_loss": 0.011245640739798547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.291861552853134,
      "flow_matching_loss": 0.07228179946541786,
      "grad_norm": 0.20324696600437164,
      "learning_rate": 9.972667058183333e-05,
      "loss": 0.0844,
      "relation_loss": 0.012100636772811414,
      "selected_pairs": 26342.6,
      "state_query_norm": 45.001809692382814,
      "state_relation_loss": 0.0008549961319658905,
      "step": 2450,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.012100636772811414,
      "weighted_relation_loss": 0.012100636772811414
    },
    {
      "action_query_norm": 44.953812789916995,
      "action_relation_loss": 0.011822010390460491,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.301216089803555,
      "flow_matching_loss": 0.07030661068856717,
      "grad_norm": 0.26009029150009155,
      "learning_rate": 9.972088525862362e-05,
      "loss": 0.083,
      "relation_loss": 0.012705072946846484,
      "selected_pairs": 26660.3,
      "state_query_norm": 44.99949226379395,
      "state_relation_loss": 0.0008830626145936548,
      "step": 2460,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.012705072946846484,
      "weighted_relation_loss": 0.012705072946846484
    },
    {
      "action_query_norm": 44.94292373657227,
      "action_relation_loss": 0.011228303425014019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3105706267539756,
      "flow_matching_loss": 0.07254029288887978,
      "grad_norm": 0.26120442152023315,
      "learning_rate": 9.971503951984995e-05,
      "loss": 0.0847,
      "relation_loss": 0.01215129904448986,
      "selected_pairs": 26997.4,
      "state_query_norm": 44.999430084228514,
      "state_relation_loss": 0.0009229956718627364,
      "step": 2470,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.01215129904448986,
      "weighted_relation_loss": 0.01215129904448986
    },
    {
      "action_query_norm": 44.9452709197998,
      "action_relation_loss": 0.01141803916543722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3199251637043967,
      "flow_matching_loss": 0.07343965955078602,
      "grad_norm": 0.29065564274787903,
      "learning_rate": 9.970913337261543e-05,
      "loss": 0.0858,
      "relation_loss": 0.012392653338611125,
      "selected_pairs": 26327.1,
      "state_query_norm": 44.99717254638672,
      "state_relation_loss": 0.0009746140392962843,
      "step": 2480,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.012392653338611125,
      "weighted_relation_loss": 0.012392653338611125
    },
    {
      "action_query_norm": 44.944190979003906,
      "action_relation_loss": 0.011799865588545799,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3292797006548174,
      "flow_matching_loss": 0.07076315209269524,
      "grad_norm": 0.1833970546722412,
      "learning_rate": 9.97031668240966e-05,
      "loss": 0.0835,
      "relation_loss": 0.012736952025443315,
      "selected_pairs": 26544.0,
      "state_query_norm": 44.99543380737305,
      "state_relation_loss": 0.0009370863961521536,
      "step": 2490,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.012736952025443315,
      "weighted_relation_loss": 0.012736952025443315
    },
    {
      "action_query_norm": 44.93693046569824,
      "action_relation_loss": 0.01159356776624918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3386342376052385,
      "flow_matching_loss": 0.0702321782708168,
      "grad_norm": 0.19508452713489532,
      "learning_rate": 9.969713988154339e-05,
      "loss": 0.0827,
      "relation_loss": 0.012512745149433614,
      "selected_pairs": 26850.5,
      "state_query_norm": 44.996313858032224,
      "state_relation_loss": 0.0009191774646751582,
      "step": 2500,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.012512745149433614,
      "weighted_relation_loss": 0.012512745149433614
    },
    {
      "action_query_norm": 44.94007453918457,
      "action_relation_loss": 0.011700630653649569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3479887745556596,
      "flow_matching_loss": 0.06765017062425613,
      "grad_norm": 0.17681048810482025,
      "learning_rate": 9.969105255227906e-05,
      "loss": 0.0803,
      "relation_loss": 0.012620887719094753,
      "selected_pairs": 27292.8,
      "state_query_norm": 44.997670364379886,
      "state_relation_loss": 0.0009202571935020387,
      "step": 2510,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.012620887719094753,
      "weighted_relation_loss": 0.012620887719094753
    },
    {
      "action_query_norm": 44.93537559509277,
      "action_relation_loss": 0.01125227902084589,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3573433115060807,
      "flow_matching_loss": 0.067777856066823,
      "grad_norm": 0.22448644042015076,
      "learning_rate": 9.968490484370035e-05,
      "loss": 0.0799,
      "relation_loss": 0.012142121978104114,
      "selected_pairs": 26249.9,
      "state_query_norm": 44.99356193542481,
      "state_relation_loss": 0.0008898429747205228,
      "step": 2520,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.012142121978104114,
      "weighted_relation_loss": 0.012142121978104114
    },
    {
      "action_query_norm": 44.93324699401855,
      "action_relation_loss": 0.011870872974395753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3666978484565013,
      "flow_matching_loss": 0.07206048890948295,
      "grad_norm": 0.1903230845928192,
      "learning_rate": 9.967869676327726e-05,
      "loss": 0.0848,
      "relation_loss": 0.012776136305183173,
      "selected_pairs": 27034.6,
      "state_query_norm": 44.99193077087402,
      "state_relation_loss": 0.0009052633831743151,
      "step": 2530,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.012776136305183173,
      "weighted_relation_loss": 0.012776136305183173
    },
    {
      "action_query_norm": 44.934376907348636,
      "action_relation_loss": 0.011298277694731951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.3760523854069224,
      "flow_matching_loss": 0.06584054678678512,
      "grad_norm": 0.24720782041549683,
      "learning_rate": 9.96724283185532e-05,
      "loss": 0.078,
      "relation_loss": 0.012150601483881474,
      "selected_pairs": 27012.2,
      "state_query_norm": 44.991343307495114,
      "state_relation_loss": 0.0008523238473571837,
      "step": 2540,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.012150601483881474,
      "weighted_relation_loss": 0.012150601483881474
    },
    {
      "action_query_norm": 44.93136291503906,
      "action_relation_loss": 0.011588378809392452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.385406922357343,
      "flow_matching_loss": 0.06984102576971055,
      "grad_norm": 0.23437097668647766,
      "learning_rate": 9.966609951714494e-05,
      "loss": 0.0823,
      "relation_loss": 0.012439709529280663,
      "selected_pairs": 27196.4,
      "state_query_norm": 44.98951873779297,
      "state_relation_loss": 0.0008513306092936546,
      "step": 2550,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.012439709529280663,
      "weighted_relation_loss": 0.012439709529280663
    },
    {
      "action_query_norm": 44.926479721069335,
      "action_relation_loss": 0.011293249297887087,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.394761459307764,
      "flow_matching_loss": 0.06762780398130416,
      "grad_norm": 0.13686680793762207,
      "learning_rate": 9.965971036674255e-05,
      "loss": 0.0798,
      "relation_loss": 0.012190628051757812,
      "selected_pairs": 27084.3,
      "state_query_norm": 44.98636474609375,
      "state_relation_loss": 0.0008973788993898779,
      "step": 2560,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.012190628051757812,
      "weighted_relation_loss": 0.012190628051757812
    },
    {
      "action_query_norm": 44.92741813659668,
      "action_relation_loss": 0.011631500627845525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4041159962581853,
      "flow_matching_loss": 0.06968053728342057,
      "grad_norm": 0.17620690166950226,
      "learning_rate": 9.965326087510947e-05,
      "loss": 0.0822,
      "relation_loss": 0.012508855201303959,
      "selected_pairs": 26484.4,
      "state_query_norm": 44.987907791137694,
      "state_relation_loss": 0.000877354567637667,
      "step": 2570,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.012508855201303959,
      "weighted_relation_loss": 0.012508855201303959
    },
    {
      "action_query_norm": 44.92230033874512,
      "action_relation_loss": 0.010961262974888086,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4134705332086064,
      "flow_matching_loss": 0.0708658903837204,
      "grad_norm": 0.20926114916801453,
      "learning_rate": 9.964675105008244e-05,
      "loss": 0.0827,
      "relation_loss": 0.01179443635046482,
      "selected_pairs": 27292.2,
      "state_query_norm": 44.98499221801758,
      "state_relation_loss": 0.0008331734745297581,
      "step": 2580,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.01179443635046482,
      "weighted_relation_loss": 0.01179443635046482
    },
    {
      "action_query_norm": 44.91548156738281,
      "action_relation_loss": 0.011754566989839076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.422825070159027,
      "flow_matching_loss": 0.0692298136651516,
      "grad_norm": 0.17989197373390198,
      "learning_rate": 9.964018089957147e-05,
      "loss": 0.0818,
      "relation_loss": 0.012610552366822958,
      "selected_pairs": 26940.8,
      "state_query_norm": 44.982377243041995,
      "state_relation_loss": 0.0008559853071346879,
      "step": 2590,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.012610552366822958,
      "weighted_relation_loss": 0.012610552366822958
    },
    {
      "action_query_norm": 44.92113494873047,
      "action_relation_loss": 0.010783608071506023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.432179607109448,
      "flow_matching_loss": 0.07199509516358375,
      "grad_norm": 0.18305398523807526,
      "learning_rate": 9.963355043155996e-05,
      "loss": 0.0836,
      "relation_loss": 0.011648677382618189,
      "selected_pairs": 27510.0,
      "state_query_norm": 44.98551635742187,
      "state_relation_loss": 0.0008650692412629724,
      "step": 2600,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.011648677382618189,
      "weighted_relation_loss": 0.011648677382618189
    },
    {
      "action_query_norm": 44.90741119384766,
      "action_relation_loss": 0.011434193514287472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.441534144059869,
      "flow_matching_loss": 0.06908653303980827,
      "grad_norm": 0.16626454889774323,
      "learning_rate": 9.962685965410455e-05,
      "loss": 0.0814,
      "relation_loss": 0.012322483584284782,
      "selected_pairs": 27034.6,
      "state_query_norm": 44.98034934997558,
      "state_relation_loss": 0.000888290093280375,
      "step": 2610,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.012322483584284782,
      "weighted_relation_loss": 0.012322483584284782
    },
    {
      "action_query_norm": 44.902574157714845,
      "action_relation_loss": 0.011360792070627212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.45088868101029,
      "flow_matching_loss": 0.0708805114030838,
      "grad_norm": 0.15640434622764587,
      "learning_rate": 9.962010857533514e-05,
      "loss": 0.0832,
      "relation_loss": 0.01230196300894022,
      "selected_pairs": 27198.1,
      "state_query_norm": 44.9776252746582,
      "state_relation_loss": 0.0009411710023414343,
      "step": 2620,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.01230196300894022,
      "weighted_relation_loss": 0.01230196300894022
    },
    {
      "action_query_norm": 44.900861740112305,
      "action_relation_loss": 0.01171939205378294,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.460243217960711,
      "flow_matching_loss": 0.07314113304018974,
      "grad_norm": 0.1942974179983139,
      "learning_rate": 9.961329720345493e-05,
      "loss": 0.0857,
      "relation_loss": 0.012602482084184885,
      "selected_pairs": 26640.0,
      "state_query_norm": 44.978908920288085,
      "state_relation_loss": 0.000883089954731986,
      "step": 2630,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.012602482084184885,
      "weighted_relation_loss": 0.012602482084184885
    },
    {
      "action_query_norm": 44.90378952026367,
      "action_relation_loss": 0.011211552377790212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.469597754911132,
      "flow_matching_loss": 0.07085808254778385,
      "grad_norm": 0.20636595785617828,
      "learning_rate": 9.96064255467404e-05,
      "loss": 0.083,
      "relation_loss": 0.012117481976747512,
      "selected_pairs": 27201.6,
      "state_query_norm": 44.97616882324219,
      "state_relation_loss": 0.00090592946507968,
      "step": 2640,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.012117481976747512,
      "weighted_relation_loss": 0.012117481976747512
    },
    {
      "action_query_norm": 44.8996768951416,
      "action_relation_loss": 0.010829063039273024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.4789522918615527,
      "flow_matching_loss": 0.06969154551625252,
      "grad_norm": 0.19912803173065186,
      "learning_rate": 9.959949361354126e-05,
      "loss": 0.0815,
      "relation_loss": 0.011782785039395094,
      "selected_pairs": 27217.7,
      "state_query_norm": 44.974478530883786,
      "state_relation_loss": 0.0009537218138575554,
      "step": 2650,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.011782785039395094,
      "weighted_relation_loss": 0.011782785039395094
    },
    {
      "action_query_norm": 44.900917053222656,
      "action_relation_loss": 0.011969286203384399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.488306828811974,
      "flow_matching_loss": 0.07150096967816352,
      "grad_norm": 0.2286519557237625,
      "learning_rate": 9.959250141228045e-05,
      "loss": 0.0844,
      "relation_loss": 0.01292281886562705,
      "selected_pairs": 26270.9,
      "state_query_norm": 44.974670028686525,
      "state_relation_loss": 0.0009535327611956745,
      "step": 2660,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.01292281886562705,
      "weighted_relation_loss": 0.01292281886562705
    },
    {
      "action_query_norm": 44.89425811767578,
      "action_relation_loss": 0.010913683380931617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.497661365762395,
      "flow_matching_loss": 0.07095120698213578,
      "grad_norm": 0.18568243086338043,
      "learning_rate": 9.958544895145414e-05,
      "loss": 0.0828,
      "relation_loss": 0.011809489969164134,
      "selected_pairs": 26920.5,
      "state_query_norm": 44.97176895141602,
      "state_relation_loss": 0.0008958065882325172,
      "step": 2670,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.011809489969164134,
      "weighted_relation_loss": 0.011809489969164134
    },
    {
      "action_query_norm": 44.88583679199219,
      "action_relation_loss": 0.010539034195244313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5070159027128156,
      "flow_matching_loss": 0.06859979927539825,
      "grad_norm": 0.13823267817497253,
      "learning_rate": 9.957833623963177e-05,
      "loss": 0.08,
      "relation_loss": 0.011384271830320359,
      "selected_pairs": 27243.8,
      "state_query_norm": 44.974524307250974,
      "state_relation_loss": 0.0008452374953776598,
      "step": 2680,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.011384271830320359,
      "weighted_relation_loss": 0.011384271830320359
    },
    {
      "action_query_norm": 44.88512840270996,
      "action_relation_loss": 0.010831086430698633,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5163704396632367,
      "flow_matching_loss": 0.07020974829792977,
      "grad_norm": 0.17995524406433105,
      "learning_rate": 9.957116328545593e-05,
      "loss": 0.0819,
      "relation_loss": 0.01167496144771576,
      "selected_pairs": 27427.9,
      "state_query_norm": 44.97302780151367,
      "state_relation_loss": 0.0008438750577624887,
      "step": 2690,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.01167496144771576,
      "weighted_relation_loss": 0.01167496144771576
    },
    {
      "action_query_norm": 44.88229255676269,
      "action_relation_loss": 0.010937930271029472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.525724976613658,
      "flow_matching_loss": 0.06888384521007537,
      "grad_norm": 0.16782256960868835,
      "learning_rate": 9.956393009764244e-05,
      "loss": 0.0806,
      "relation_loss": 0.011744681652635336,
      "selected_pairs": 26687.0,
      "state_query_norm": 44.97225151062012,
      "state_relation_loss": 0.0008067513641435653,
      "step": 2700,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.011744681652635336,
      "weighted_relation_loss": 0.011744681652635336
    },
    {
      "action_query_norm": 44.881914901733396,
      "action_relation_loss": 0.010694170463830232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5350795135640785,
      "flow_matching_loss": 0.06908611506223679,
      "grad_norm": 0.1641542613506317,
      "learning_rate": 9.955663668498032e-05,
      "loss": 0.0806,
      "relation_loss": 0.011536771431565285,
      "selected_pairs": 27298.0,
      "state_query_norm": 44.971953582763675,
      "state_relation_loss": 0.0008426010783296079,
      "step": 2710,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.011536771431565285,
      "weighted_relation_loss": 0.011536771431565285
    },
    {
      "action_query_norm": 44.876526260375975,
      "action_relation_loss": 0.010792282316833735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5444340505144996,
      "flow_matching_loss": 0.06976198554039001,
      "grad_norm": 0.16459016501903534,
      "learning_rate": 9.954928305633173e-05,
      "loss": 0.0814,
      "relation_loss": 0.011611390952020883,
      "selected_pairs": 26898.4,
      "state_query_norm": 44.968649291992186,
      "state_relation_loss": 0.0008191084489226341,
      "step": 2720,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.011611390952020883,
      "weighted_relation_loss": 0.011611390952020883
    },
    {
      "action_query_norm": 44.87554016113281,
      "action_relation_loss": 0.010765784047544002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.55378858746492,
      "flow_matching_loss": 0.06671322323381901,
      "grad_norm": 0.15973792970180511,
      "learning_rate": 9.954186922063204e-05,
      "loss": 0.0784,
      "relation_loss": 0.011637720931321382,
      "selected_pairs": 26890.7,
      "state_query_norm": 44.96558876037598,
      "state_relation_loss": 0.0008719369070604444,
      "step": 2730,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.011637720931321382,
      "weighted_relation_loss": 0.011637720931321382
    },
    {
      "action_query_norm": 44.87071304321289,
      "action_relation_loss": 0.010663795005530118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5631431244153413,
      "flow_matching_loss": 0.06901969201862812,
      "grad_norm": 0.19632209837436676,
      "learning_rate": 9.953439518688974e-05,
      "loss": 0.0806,
      "relation_loss": 0.011530768685042858,
      "selected_pairs": 26826.6,
      "state_query_norm": 44.964715576171876,
      "state_relation_loss": 0.0008669735980220139,
      "step": 2740,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.011530768685042858,
      "weighted_relation_loss": 0.011530768685042858
    },
    {
      "action_query_norm": 44.86960716247559,
      "action_relation_loss": 0.011124265659600497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5724976613657624,
      "flow_matching_loss": 0.07026622332632541,
      "grad_norm": 0.1766485720872879,
      "learning_rate": 9.952686096418652e-05,
      "loss": 0.0822,
      "relation_loss": 0.011967555992305279,
      "selected_pairs": 26439.8,
      "state_query_norm": 44.96647911071777,
      "state_relation_loss": 0.0008432902512140572,
      "step": 2750,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.011967555992305279,
      "weighted_relation_loss": 0.011967555992305279
    },
    {
      "action_query_norm": 44.86138000488281,
      "action_relation_loss": 0.010508381482213735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.5818521983161835,
      "flow_matching_loss": 0.06946047469973564,
      "grad_norm": 0.23311707377433777,
      "learning_rate": 9.951926656167714e-05,
      "loss": 0.0808,
      "relation_loss": 0.011375718843191862,
      "selected_pairs": 26209.9,
      "state_query_norm": 44.96399078369141,
      "state_relation_loss": 0.0008673373784404248,
      "step": 2760,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.011375718843191862,
      "weighted_relation_loss": 0.011375718843191862
    },
    {
      "action_query_norm": 44.85812263488769,
      "action_relation_loss": 0.011215066444128752,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.591206735266604,
      "flow_matching_loss": 0.06911600828170776,
      "grad_norm": 0.23351728916168213,
      "learning_rate": 9.951161198858953e-05,
      "loss": 0.0812,
      "relation_loss": 0.01208655470982194,
      "selected_pairs": 27171.9,
      "state_query_norm": 44.96313629150391,
      "state_relation_loss": 0.0008714881958439946,
      "step": 2770,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.01208655470982194,
      "weighted_relation_loss": 0.01208655470982194
    },
    {
      "action_query_norm": 44.853433227539064,
      "action_relation_loss": 0.011086639948189259,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6005612722170253,
      "flow_matching_loss": 0.06976166963577271,
      "grad_norm": 0.21872395277023315,
      "learning_rate": 9.950389725422471e-05,
      "loss": 0.0817,
      "relation_loss": 0.011953829415142536,
      "selected_pairs": 26412.3,
      "state_query_norm": 44.95890846252441,
      "state_relation_loss": 0.0008671894727740436,
      "step": 2780,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.011953829415142536,
      "weighted_relation_loss": 0.011953829415142536
    },
    {
      "action_query_norm": 44.8518569946289,
      "action_relation_loss": 0.010420882422477008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6099158091674464,
      "flow_matching_loss": 0.06812901087105275,
      "grad_norm": 0.2007896602153778,
      "learning_rate": 9.949612236795682e-05,
      "loss": 0.0794,
      "relation_loss": 0.011305247340351344,
      "selected_pairs": 26475.2,
      "state_query_norm": 44.95934906005859,
      "state_relation_loss": 0.0008843650226481259,
      "step": 2790,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.011305247340351344,
      "weighted_relation_loss": 0.011305247340351344
    },
    {
      "action_query_norm": 44.85164566040039,
      "action_relation_loss": 0.010827072709798814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.619270346117867,
      "flow_matching_loss": 0.07124928124248982,
      "grad_norm": 0.20615363121032715,
      "learning_rate": 9.948828733923305e-05,
      "loss": 0.0829,
      "relation_loss": 0.011611092090606689,
      "selected_pairs": 26981.5,
      "state_query_norm": 44.9602165222168,
      "state_relation_loss": 0.0007840193633455783,
      "step": 2800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.011611092090606689,
      "weighted_relation_loss": 0.011611092090606689
    },
    {
      "action_query_norm": 44.84670753479004,
      "action_relation_loss": 0.011290707532316447,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.628624883068288,
      "flow_matching_loss": 0.07054553776979447,
      "grad_norm": 0.16455614566802979,
      "learning_rate": 9.948039217757374e-05,
      "loss": 0.0827,
      "relation_loss": 0.01213512709364295,
      "selected_pairs": 26777.2,
      "state_query_norm": 44.95792503356934,
      "state_relation_loss": 0.0008444195031188428,
      "step": 2810,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.01213512709364295,
      "weighted_relation_loss": 0.01213512709364295
    },
    {
      "action_query_norm": 44.841691207885745,
      "action_relation_loss": 0.010269022826105357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.637979420018709,
      "flow_matching_loss": 0.06695613525807857,
      "grad_norm": 0.16363735496997833,
      "learning_rate": 9.947243689257225e-05,
      "loss": 0.078,
      "relation_loss": 0.011084669921547174,
      "selected_pairs": 26851.5,
      "state_query_norm": 44.956494140625,
      "state_relation_loss": 0.0008156471478287131,
      "step": 2820,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.011084669921547174,
      "weighted_relation_loss": 0.011084669921547174
    },
    {
      "action_query_norm": 44.843467712402344,
      "action_relation_loss": 0.010568186081945896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.64733395696913,
      "flow_matching_loss": 0.06588243916630745,
      "grad_norm": 0.18050263822078705,
      "learning_rate": 9.946442149389497e-05,
      "loss": 0.0773,
      "relation_loss": 0.011399821285158396,
      "selected_pairs": 27103.5,
      "state_query_norm": 44.95458450317383,
      "state_relation_loss": 0.0008316352672409266,
      "step": 2830,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.011399821285158396,
      "weighted_relation_loss": 0.011399821285158396
    },
    {
      "action_query_norm": 44.84859657287598,
      "action_relation_loss": 0.010248535219579935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.656688493919551,
      "flow_matching_loss": 0.06862531527876854,
      "grad_norm": 0.24039530754089355,
      "learning_rate": 9.945634599128139e-05,
      "loss": 0.0797,
      "relation_loss": 0.011072110198438168,
      "selected_pairs": 26966.8,
      "state_query_norm": 44.95568199157715,
      "state_relation_loss": 0.0008235750021412969,
      "step": 2840,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.011072110198438168,
      "weighted_relation_loss": 0.011072110198438168
    },
    {
      "action_query_norm": 44.838349151611325,
      "action_relation_loss": 0.010782852116972209,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.666043030869972,
      "flow_matching_loss": 0.07096589580178261,
      "grad_norm": 0.15532541275024414,
      "learning_rate": 9.944821039454402e-05,
      "loss": 0.0826,
      "relation_loss": 0.011617148108780385,
      "selected_pairs": 26851.4,
      "state_query_norm": 44.951530838012694,
      "state_relation_loss": 0.0008342960150912404,
      "step": 2850,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.011617148108780385,
      "weighted_relation_loss": 0.011617148108780385
    },
    {
      "action_query_norm": 44.838859176635744,
      "action_relation_loss": 0.01086726551875472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.6753975678203927,
      "flow_matching_loss": 0.06798711605370045,
      "grad_norm": 0.21066005527973175,
      "learning_rate": 9.944001471356835e-05,
      "loss": 0.0797,
      "relation_loss": 0.01170002119615674,
      "selected_pairs": 26412.8,
      "state_query_norm": 44.95114440917969,
      "state_relation_loss": 0.000832755834562704,
      "step": 2860,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.01170002119615674,
      "weighted_relation_loss": 0.01170002119615674
    },
    {
      "action_query_norm": 44.84717063903808,
      "action_relation_loss": 0.01105652330443263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.684752104770814,
      "flow_matching_loss": 0.0676095649600029,
      "grad_norm": 0.20773549377918243,
      "learning_rate": 9.94317589583129e-05,
      "loss": 0.0795,
      "relation_loss": 0.011856691911816597,
      "selected_pairs": 26855.5,
      "state_query_norm": 44.94594192504883,
      "state_relation_loss": 0.0008001686539500952,
      "step": 2870,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.011856691911816597,
      "weighted_relation_loss": 0.011856691911816597
    },
    {
      "action_query_norm": 44.83796730041504,
      "action_relation_loss": 0.010977223329246045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.694106641721235,
      "flow_matching_loss": 0.06973748281598091,
      "grad_norm": 0.1847505420446396,
      "learning_rate": 9.942344313880922e-05,
      "loss": 0.0815,
      "relation_loss": 0.01178236324340105,
      "selected_pairs": 26990.4,
      "state_query_norm": 44.94366226196289,
      "state_relation_loss": 0.0008051400654949248,
      "step": 2880,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.01178236324340105,
      "weighted_relation_loss": 0.01178236324340105
    },
    {
      "action_query_norm": 44.831475830078126,
      "action_relation_loss": 0.011199579201638698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.703461178671656,
      "flow_matching_loss": 0.07143157869577407,
      "grad_norm": 0.24344319105148315,
      "learning_rate": 9.941506726516179e-05,
      "loss": 0.0835,
      "relation_loss": 0.012036711163818836,
      "selected_pairs": 26854.9,
      "state_query_norm": 44.94412078857422,
      "state_relation_loss": 0.000837131985463202,
      "step": 2890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.012036711163818836,
      "weighted_relation_loss": 0.012036711163818836
    },
    {
      "action_query_norm": 44.82925910949707,
      "action_relation_loss": 0.010869255382567644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7128157156220767,
      "flow_matching_loss": 0.07044810876250267,
      "grad_norm": 0.15952980518341064,
      "learning_rate": 9.94066313475481e-05,
      "loss": 0.0821,
      "relation_loss": 0.01167505318298936,
      "selected_pairs": 27153.3,
      "state_query_norm": 44.94506759643555,
      "state_relation_loss": 0.0008057978004217148,
      "step": 2900,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.01167505318298936,
      "weighted_relation_loss": 0.01167505318298936
    },
    {
      "action_query_norm": 44.825050354003906,
      "action_relation_loss": 0.010934039577841758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.722170252572498,
      "flow_matching_loss": 0.07284700199961662,
      "grad_norm": 0.22401371598243713,
      "learning_rate": 9.939813539621857e-05,
      "loss": 0.0846,
      "relation_loss": 0.011764067783951759,
      "selected_pairs": 27087.3,
      "state_query_norm": 44.94102935791015,
      "state_relation_loss": 0.0008300281013362109,
      "step": 2910,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.011764067783951759,
      "weighted_relation_loss": 0.011764067783951759
    },
    {
      "action_query_norm": 44.82075805664063,
      "action_relation_loss": 0.01117693493142724,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7315247895229184,
      "flow_matching_loss": 0.0702387548983097,
      "grad_norm": 0.20993854105472565,
      "learning_rate": 9.93895794214966e-05,
      "loss": 0.0822,
      "relation_loss": 0.01199390646070242,
      "selected_pairs": 27243.4,
      "state_query_norm": 44.937760543823245,
      "state_relation_loss": 0.000816971471067518,
      "step": 2920,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.01199390646070242,
      "weighted_relation_loss": 0.01199390646070242
    },
    {
      "action_query_norm": 44.81670799255371,
      "action_relation_loss": 0.010262746829539537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7408793264733395,
      "flow_matching_loss": 0.07155658304691315,
      "grad_norm": 0.2832571566104889,
      "learning_rate": 9.938096343377852e-05,
      "loss": 0.0827,
      "relation_loss": 0.011094193905591965,
      "selected_pairs": 27533.1,
      "state_query_norm": 44.93633766174317,
      "state_relation_loss": 0.00083144725067541,
      "step": 2930,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.011094193905591965,
      "weighted_relation_loss": 0.011094193905591965
    },
    {
      "action_query_norm": 44.8175594329834,
      "action_relation_loss": 0.010805576294660568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7502338634237606,
      "flow_matching_loss": 0.06934522390365601,
      "grad_norm": 0.24162323772907257,
      "learning_rate": 9.937228744353353e-05,
      "loss": 0.081,
      "relation_loss": 0.011626306176185607,
      "selected_pairs": 27085.0,
      "state_query_norm": 44.93742790222168,
      "state_relation_loss": 0.0008207298640627414,
      "step": 2940,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.011626306176185607,
      "weighted_relation_loss": 0.011626306176185607
    },
    {
      "action_query_norm": 44.814518737792966,
      "action_relation_loss": 0.010945138242095708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7595884003741817,
      "flow_matching_loss": 0.0702494129538536,
      "grad_norm": 0.2525192201137543,
      "learning_rate": 9.936355146130379e-05,
      "loss": 0.082,
      "relation_loss": 0.011738917045295238,
      "selected_pairs": 26713.4,
      "state_query_norm": 44.9359634399414,
      "state_relation_loss": 0.0007937787275295704,
      "step": 2950,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.011738917045295238,
      "weighted_relation_loss": 0.011738917045295238
    },
    {
      "action_query_norm": 44.80935287475586,
      "action_relation_loss": 0.010914412047713994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7689429373246024,
      "flow_matching_loss": 0.07169725447893142,
      "grad_norm": 0.18923181295394897,
      "learning_rate": 9.935475549770435e-05,
      "loss": 0.0834,
      "relation_loss": 0.011751255951821804,
      "selected_pairs": 26845.7,
      "state_query_norm": 44.93378219604492,
      "state_relation_loss": 0.0008368438517209143,
      "step": 2960,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.011751255951821804,
      "weighted_relation_loss": 0.011751255951821804
    },
    {
      "action_query_norm": 44.803450393676755,
      "action_relation_loss": 0.011183993425220251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7782974742750235,
      "flow_matching_loss": 0.0714639626443386,
      "grad_norm": 0.26774707436561584,
      "learning_rate": 9.934589956342315e-05,
      "loss": 0.0835,
      "relation_loss": 0.011989343911409378,
      "selected_pairs": 26941.4,
      "state_query_norm": 44.9363525390625,
      "state_relation_loss": 0.0008053504570852965,
      "step": 2970,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.011989343911409378,
      "weighted_relation_loss": 0.011989343911409378
    },
    {
      "action_query_norm": 44.80520439147949,
      "action_relation_loss": 0.010579197574406863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.787652011225444,
      "flow_matching_loss": 0.06899393685162067,
      "grad_norm": 0.1935265064239502,
      "learning_rate": 9.933698366922093e-05,
      "loss": 0.0805,
      "relation_loss": 0.011456896737217902,
      "selected_pairs": 26626.7,
      "state_query_norm": 44.929209518432614,
      "state_relation_loss": 0.000877699168631807,
      "step": 2980,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.011456896737217902,
      "weighted_relation_loss": 0.011456896737217902
    },
    {
      "action_query_norm": 44.795627975463866,
      "action_relation_loss": 0.010955859813839198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.7970065481758652,
      "flow_matching_loss": 0.07176174074411393,
      "grad_norm": 0.20405517518520355,
      "learning_rate": 9.93280078259314e-05,
      "loss": 0.0835,
      "relation_loss": 0.011766880098730326,
      "selected_pairs": 27068.2,
      "state_query_norm": 44.92998275756836,
      "state_relation_loss": 0.0008110201626550407,
      "step": 2990,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.011766880098730326,
      "weighted_relation_loss": 0.011766880098730326
    },
    {
      "action_query_norm": 44.798838424682614,
      "action_relation_loss": 0.010949541628360749,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8063610851262863,
      "flow_matching_loss": 0.06918729841709137,
      "grad_norm": 0.18528912961483002,
      "learning_rate": 9.931897204446104e-05,
      "loss": 0.081,
      "relation_loss": 0.01180212115868926,
      "selected_pairs": 27126.1,
      "state_query_norm": 44.92725944519043,
      "state_relation_loss": 0.0008525796758476645,
      "step": 3000,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.01180212115868926,
      "weighted_relation_loss": 0.01180212115868926
    },
    {
      "action_query_norm": 44.790612030029294,
      "action_relation_loss": 0.010718497075140476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8157156220767074,
      "flow_matching_loss": 0.06675715446472168,
      "grad_norm": 0.21020419895648956,
      "learning_rate": 9.930987633578915e-05,
      "loss": 0.0783,
      "relation_loss": 0.011530837789177895,
      "selected_pairs": 26575.7,
      "state_query_norm": 44.92818641662598,
      "state_relation_loss": 0.000812340626725927,
      "step": 3010,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.011530837789177895,
      "weighted_relation_loss": 0.011530837789177895
    },
    {
      "action_query_norm": 44.79325828552246,
      "action_relation_loss": 0.011219866760075092,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.825070159027128,
      "flow_matching_loss": 0.06735524795949459,
      "grad_norm": 0.14865055680274963,
      "learning_rate": 9.93007207109679e-05,
      "loss": 0.0795,
      "relation_loss": 0.012112617958337069,
      "selected_pairs": 26312.2,
      "state_query_norm": 44.92591514587402,
      "state_relation_loss": 0.0008927511400543154,
      "step": 3020,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.012112617958337069,
      "weighted_relation_loss": 0.012112617958337069
    },
    {
      "action_query_norm": 44.7843017578125,
      "action_relation_loss": 0.011113011185079813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.834424695977549,
      "flow_matching_loss": 0.07012971341609955,
      "grad_norm": 0.1625695824623108,
      "learning_rate": 9.929150518112224e-05,
      "loss": 0.0821,
      "relation_loss": 0.011924877949059009,
      "selected_pairs": 26597.0,
      "state_query_norm": 44.924800872802734,
      "state_relation_loss": 0.0008118668978568166,
      "step": 3030,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.011924877949059009,
      "weighted_relation_loss": 0.011924877949059009
    },
    {
      "action_query_norm": 44.78249244689941,
      "action_relation_loss": 0.010685690678656101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.84377923292797,
      "flow_matching_loss": 0.0683840423822403,
      "grad_norm": 0.20705577731132507,
      "learning_rate": 9.928222975744991e-05,
      "loss": 0.0799,
      "relation_loss": 0.011510562058538199,
      "selected_pairs": 27404.9,
      "state_query_norm": 44.926943969726565,
      "state_relation_loss": 0.000824871368240565,
      "step": 3040,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.011510562058538199,
      "weighted_relation_loss": 0.011510562058538199
    },
    {
      "action_query_norm": 44.78505249023438,
      "action_relation_loss": 0.010500089079141618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.853133769878391,
      "flow_matching_loss": 0.06886198483407498,
      "grad_norm": 0.20910923182964325,
      "learning_rate": 9.92728944512214e-05,
      "loss": 0.0801,
      "relation_loss": 0.011251124553382397,
      "selected_pairs": 26829.9,
      "state_query_norm": 44.9239616394043,
      "state_relation_loss": 0.0007510354917030782,
      "step": 3050,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.011251124553382397,
      "weighted_relation_loss": 0.011251124553382397
    },
    {
      "action_query_norm": 44.78805503845215,
      "action_relation_loss": 0.010529043152928352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.862488306828812,
      "flow_matching_loss": 0.06990318074822426,
      "grad_norm": 0.17461737990379333,
      "learning_rate": 9.926349927378001e-05,
      "loss": 0.0812,
      "relation_loss": 0.011273283138871193,
      "selected_pairs": 26735.0,
      "state_query_norm": 44.92206344604492,
      "state_relation_loss": 0.0007442400266882033,
      "step": 3060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.011273283138871193,
      "weighted_relation_loss": 0.011273283138871193
    },
    {
      "action_query_norm": 44.796121978759764,
      "action_relation_loss": 0.010191685426980257,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.871842843779233,
      "flow_matching_loss": 0.0655982717871666,
      "grad_norm": 0.17745737731456757,
      "learning_rate": 9.925404423654174e-05,
      "loss": 0.0765,
      "relation_loss": 0.010907185729593039,
      "selected_pairs": 26500.2,
      "state_query_norm": 44.922811889648436,
      "state_relation_loss": 0.0007155003200750798,
      "step": 3070,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.010907185729593039,
      "weighted_relation_loss": 0.010907185729593039
    },
    {
      "action_query_norm": 44.7880672454834,
      "action_relation_loss": 0.010536475293338298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.881197380729654,
      "flow_matching_loss": 0.0689591147005558,
      "grad_norm": 0.19097046554088593,
      "learning_rate": 9.924452935099537e-05,
      "loss": 0.0802,
      "relation_loss": 0.011266200058162212,
      "selected_pairs": 26760.8,
      "state_query_norm": 44.919881439208986,
      "state_relation_loss": 0.0007297248055692762,
      "step": 3080,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.011266200058162212,
      "weighted_relation_loss": 0.011266200058162212
    },
    {
      "action_query_norm": 44.780521392822266,
      "action_relation_loss": 0.011007704492658376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.890551917680075,
      "flow_matching_loss": 0.0673981498926878,
      "grad_norm": 0.23073095083236694,
      "learning_rate": 9.92349546287024e-05,
      "loss": 0.0792,
      "relation_loss": 0.011764065641909838,
      "selected_pairs": 26761.6,
      "state_query_norm": 44.91982460021973,
      "state_relation_loss": 0.0007563611667137593,
      "step": 3090,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.011764065641909838,
      "weighted_relation_loss": 0.011764065641909838
    },
    {
      "action_query_norm": 44.78310432434082,
      "action_relation_loss": 0.010643423162400722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.8999064546304956,
      "flow_matching_loss": 0.06552321724593639,
      "grad_norm": 0.17821672558784485,
      "learning_rate": 9.9225320081297e-05,
      "loss": 0.0769,
      "relation_loss": 0.011382449790835381,
      "selected_pairs": 26596.9,
      "state_query_norm": 44.916962432861325,
      "state_relation_loss": 0.0007390266051515937,
      "step": 3100,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.011382449790835381,
      "weighted_relation_loss": 0.011382449790835381
    },
    {
      "action_query_norm": 44.78247489929199,
      "action_relation_loss": 0.011080241668969393,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9092609915809167,
      "flow_matching_loss": 0.06611399650573731,
      "grad_norm": 0.21347060799598694,
      "learning_rate": 9.921562572048606e-05,
      "loss": 0.078,
      "relation_loss": 0.011841343250125647,
      "selected_pairs": 26825.8,
      "state_query_norm": 44.9166015625,
      "state_relation_loss": 0.0007611016742885112,
      "step": 3110,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.011841343250125647,
      "weighted_relation_loss": 0.011841343250125647
    },
    {
      "action_query_norm": 44.77081985473633,
      "action_relation_loss": 0.01049388963729143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9186155285313378,
      "flow_matching_loss": 0.06694575287401676,
      "grad_norm": 0.19587436318397522,
      "learning_rate": 9.920587155804913e-05,
      "loss": 0.0782,
      "relation_loss": 0.011248006485402584,
      "selected_pairs": 26923.6,
      "state_query_norm": 44.91567726135254,
      "state_relation_loss": 0.0007541167899034918,
      "step": 3120,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.011248006485402584,
      "weighted_relation_loss": 0.011248006485402584
    },
    {
      "action_query_norm": 44.76050987243652,
      "action_relation_loss": 0.01098272753879428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.927970065481759,
      "flow_matching_loss": 0.06750271059572696,
      "grad_norm": 0.2219531238079071,
      "learning_rate": 9.919605760583845e-05,
      "loss": 0.0793,
      "relation_loss": 0.011819742154330015,
      "selected_pairs": 27149.6,
      "state_query_norm": 44.91420211791992,
      "state_relation_loss": 0.000837014481658116,
      "step": 3130,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.011819742154330015,
      "weighted_relation_loss": 0.011819742154330015
    },
    {
      "action_query_norm": 44.758191680908205,
      "action_relation_loss": 0.01113033266738057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9373246024321795,
      "flow_matching_loss": 0.06569714061915874,
      "grad_norm": 0.18832913041114807,
      "learning_rate": 9.91861838757789e-05,
      "loss": 0.0777,
      "relation_loss": 0.011983017250895501,
      "selected_pairs": 26799.0,
      "state_query_norm": 44.91426010131836,
      "state_relation_loss": 0.0008526845544110983,
      "step": 3140,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.011983017250895501,
      "weighted_relation_loss": 0.011983017250895501
    },
    {
      "action_query_norm": 44.75685768127441,
      "action_relation_loss": 0.010309570375829935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9466791393826006,
      "flow_matching_loss": 0.06676119193434715,
      "grad_norm": 0.2362222671508789,
      "learning_rate": 9.917625037986798e-05,
      "loss": 0.0779,
      "relation_loss": 0.011160065606236457,
      "selected_pairs": 26776.1,
      "state_query_norm": 44.9091609954834,
      "state_relation_loss": 0.0008504951605573296,
      "step": 3150,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.011160065606236457,
      "weighted_relation_loss": 0.011160065606236457
    },
    {
      "action_query_norm": 44.752859115600586,
      "action_relation_loss": 0.010456490144133569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9560336763330213,
      "flow_matching_loss": 0.06523156762123108,
      "grad_norm": 0.17667768895626068,
      "learning_rate": 9.916625713017583e-05,
      "loss": 0.0766,
      "relation_loss": 0.011339125037193299,
      "selected_pairs": 26776.4,
      "state_query_norm": 44.90840110778809,
      "state_relation_loss": 0.0008826348523143679,
      "step": 3160,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.011339125037193299,
      "weighted_relation_loss": 0.011339125037193299
    },
    {
      "action_query_norm": 44.75317726135254,
      "action_relation_loss": 0.010300293657928705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9653882132834424,
      "flow_matching_loss": 0.06829124204814434,
      "grad_norm": 0.1817275881767273,
      "learning_rate": 9.915620413884519e-05,
      "loss": 0.0794,
      "relation_loss": 0.011115715373307467,
      "selected_pairs": 26784.3,
      "state_query_norm": 44.90863037109375,
      "state_relation_loss": 0.0008154216047842056,
      "step": 3170,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.011115715373307467,
      "weighted_relation_loss": 0.011115715373307467
    },
    {
      "action_query_norm": 44.751654052734374,
      "action_relation_loss": 0.010305400006473064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9747427502338635,
      "flow_matching_loss": 0.06676280684769154,
      "grad_norm": 0.1938733458518982,
      "learning_rate": 9.914609141809139e-05,
      "loss": 0.0779,
      "relation_loss": 0.011145893670618533,
      "selected_pairs": 26921.9,
      "state_query_norm": 44.90995750427246,
      "state_relation_loss": 0.0008404935651924461,
      "step": 3180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.011145893670618533,
      "weighted_relation_loss": 0.011145893670618533
    },
    {
      "action_query_norm": 44.75080184936523,
      "action_relation_loss": 0.010920108761638403,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9840972871842846,
      "flow_matching_loss": 0.07046679183840751,
      "grad_norm": 0.16512738168239594,
      "learning_rate": 9.913591898020235e-05,
      "loss": 0.0822,
      "relation_loss": 0.011696037836372852,
      "selected_pairs": 27247.1,
      "state_query_norm": 44.9114372253418,
      "state_relation_loss": 0.0007759290398098529,
      "step": 3190,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.011696037836372852,
      "weighted_relation_loss": 0.011696037836372852
    },
    {
      "action_query_norm": 44.74716072082519,
      "action_relation_loss": 0.010786558315157891,
      "auxiliary_loss_weight": 1.0,
      "epoch": 2.9934518241347052,
      "flow_matching_loss": 0.07127896323800087,
      "grad_norm": 0.18120183050632477,
      "learning_rate": 9.912568683753853e-05,
      "loss": 0.0829,
      "relation_loss": 0.011575582250952721,
      "selected_pairs": 26684.8,
      "state_query_norm": 44.907948684692386,
      "state_relation_loss": 0.0007890239241532982,
      "step": 3200,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.011575582250952721,
      "weighted_relation_loss": 0.011575582250952721
    },
    {
      "action_query_norm": 44.75960121154785,
      "action_relation_loss": 0.010281292907893657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0028063610851263,
      "flow_matching_loss": 0.06829253509640694,
      "grad_norm": 0.17044982314109802,
      "learning_rate": 9.911539500253295e-05,
      "loss": 0.0793,
      "relation_loss": 0.011055780760943889,
      "selected_pairs": 23666.0,
      "state_query_norm": 44.91119613647461,
      "state_relation_loss": 0.0007744877599179745,
      "step": 3210,
      "valid_action_samples": 211.5,
      "weighted_auxiliary_loss": 0.011055780760943889,
      "weighted_relation_loss": 0.011055780760943889
    },
    {
      "action_query_norm": 44.74777107238769,
      "action_relation_loss": 0.010633463691920041,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0121608980355474,
      "flow_matching_loss": 0.06799619421362876,
      "grad_norm": 0.18050138652324677,
      "learning_rate": 9.910504348769118e-05,
      "loss": 0.0794,
      "relation_loss": 0.011417146679013968,
      "selected_pairs": 26881.7,
      "state_query_norm": 44.90371437072754,
      "state_relation_loss": 0.0007836830685846507,
      "step": 3220,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.011417146679013968,
      "weighted_relation_loss": 0.011417146679013968
    },
    {
      "action_query_norm": 44.75414352416992,
      "action_relation_loss": 0.01056278795003891,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.021515434985968,
      "flow_matching_loss": 0.06713378243148327,
      "grad_norm": 0.2002459466457367,
      "learning_rate": 9.909463230559127e-05,
      "loss": 0.0785,
      "relation_loss": 0.01137151475995779,
      "selected_pairs": 27205.0,
      "state_query_norm": 44.90467720031738,
      "state_relation_loss": 0.0008087267458904534,
      "step": 3230,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.01137151475995779,
      "weighted_relation_loss": 0.01137151475995779
    },
    {
      "action_query_norm": 44.74621391296387,
      "action_relation_loss": 0.010788393579423428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.030869971936389,
      "flow_matching_loss": 0.06792221665382385,
      "grad_norm": 0.22205327451229095,
      "learning_rate": 9.908416146888376e-05,
      "loss": 0.0794,
      "relation_loss": 0.011514197196811437,
      "selected_pairs": 27251.6,
      "state_query_norm": 44.90389213562012,
      "state_relation_loss": 0.0007258035358972848,
      "step": 3240,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.011514197196811437,
      "weighted_relation_loss": 0.011514197196811437
    },
    {
      "action_query_norm": 44.74203033447266,
      "action_relation_loss": 0.011101554799824952,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.0402245088868103,
      "flow_matching_loss": 0.06801382713019848,
      "grad_norm": 0.22530904412269592,
      "learning_rate": 9.907363099029175e-05,
      "loss": 0.0799,
      "relation_loss": 0.011886334884911776,
      "selected_pairs": 26344.2,
      "state_query_norm": 44.89755516052246,
      "state_relation_loss": 0.0007847799570299685,
      "step": 3250,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.011886334884911776,
      "weighted_relation_loss": 0.011886334884911776
    },
    {
      "action_query_norm": 44.74009323120117,
      "action_relation_loss": 0.010637327283620834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.049579045837231,
      "flow_matching_loss": 0.06729690954089165,
      "grad_norm": 0.1548159420490265,
      "learning_rate": 9.906304088261073e-05,
      "loss": 0.0787,
      "relation_loss": 0.011382497008889913,
      "selected_pairs": 27233.9,
      "state_query_norm": 44.898410415649415,
      "state_relation_loss": 0.0007451696204952895,
      "step": 3260,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.011382497008889913,
      "weighted_relation_loss": 0.011382497008889913
    },
    {
      "action_query_norm": 44.72823638916016,
      "action_relation_loss": 0.010698532592505217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.058933582787652,
      "flow_matching_loss": 0.0683873176574707,
      "grad_norm": 0.14787934720516205,
      "learning_rate": 9.905239115870872e-05,
      "loss": 0.0799,
      "relation_loss": 0.01146625131368637,
      "selected_pairs": 26854.1,
      "state_query_norm": 44.90051765441895,
      "state_relation_loss": 0.0007677186105865986,
      "step": 3270,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.01146625131368637,
      "weighted_relation_loss": 0.01146625131368637
    },
    {
      "action_query_norm": 44.72445449829102,
      "action_relation_loss": 0.010114847309887409,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.068288119738073,
      "flow_matching_loss": 0.06924454718828202,
      "grad_norm": 0.11764223873615265,
      "learning_rate": 9.90416818315261e-05,
      "loss": 0.0801,
      "relation_loss": 0.010901324916630983,
      "selected_pairs": 26877.9,
      "state_query_norm": 44.896012496948245,
      "state_relation_loss": 0.0007864776009228081,
      "step": 3280,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.010901324916630983,
      "weighted_relation_loss": 0.010901324916630983
    },
    {
      "action_query_norm": 44.71718139648438,
      "action_relation_loss": 0.009713182132691145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.077642656688494,
      "flow_matching_loss": 0.06639755591750145,
      "grad_norm": 0.1344873607158661,
      "learning_rate": 9.903091291407573e-05,
      "loss": 0.0768,
      "relation_loss": 0.010447814781218766,
      "selected_pairs": 27170.2,
      "state_query_norm": 44.898461532592776,
      "state_relation_loss": 0.0007346326834522188,
      "step": 3290,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010447814781218766,
      "weighted_relation_loss": 0.010447814781218766
    },
    {
      "action_query_norm": 44.715203857421876,
      "action_relation_loss": 0.010009503550827503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.086997193638915,
      "flow_matching_loss": 0.06759591028094292,
      "grad_norm": 0.12376567721366882,
      "learning_rate": 9.902008441944286e-05,
      "loss": 0.0784,
      "relation_loss": 0.010803824104368687,
      "selected_pairs": 26972.9,
      "state_query_norm": 44.893385314941405,
      "state_relation_loss": 0.0007943206292111427,
      "step": 3300,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.010803824104368687,
      "weighted_relation_loss": 0.010803824104368687
    },
    {
      "action_query_norm": 44.7097785949707,
      "action_relation_loss": 0.009681046754121781,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.096351730589336,
      "flow_matching_loss": 0.06366303861141205,
      "grad_norm": 0.1507566124200821,
      "learning_rate": 9.900919636078512e-05,
      "loss": 0.0741,
      "relation_loss": 0.010472579300403595,
      "selected_pairs": 26237.4,
      "state_query_norm": 44.888423919677734,
      "state_relation_loss": 0.000791532505536452,
      "step": 3310,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.010472579300403595,
      "weighted_relation_loss": 0.010472579300403595
    },
    {
      "action_query_norm": 44.71474266052246,
      "action_relation_loss": 0.010503351781517267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1057062675397566,
      "flow_matching_loss": 0.06459641270339489,
      "grad_norm": 0.1291285753250122,
      "learning_rate": 9.899824875133255e-05,
      "loss": 0.0758,
      "relation_loss": 0.011243292316794395,
      "selected_pairs": 26895.5,
      "state_query_norm": 44.89030532836914,
      "state_relation_loss": 0.0007399404596071691,
      "step": 3320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.011243292316794395,
      "weighted_relation_loss": 0.011243292316794395
    },
    {
      "action_query_norm": 44.70870590209961,
      "action_relation_loss": 0.010101682506501675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1150608044901777,
      "flow_matching_loss": 0.06543073281645775,
      "grad_norm": 0.12068121880292892,
      "learning_rate": 9.898724160438749e-05,
      "loss": 0.0762,
      "relation_loss": 0.010816503036767244,
      "selected_pairs": 26834.6,
      "state_query_norm": 44.88449325561523,
      "state_relation_loss": 0.0007148206117562949,
      "step": 3330,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.010816503036767244,
      "weighted_relation_loss": 0.010816503036767244
    },
    {
      "action_query_norm": 44.702806091308595,
      "action_relation_loss": 0.010167440585792064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.124415341440599,
      "flow_matching_loss": 0.06662421934306621,
      "grad_norm": 0.21718011796474457,
      "learning_rate": 9.89761749333247e-05,
      "loss": 0.0775,
      "relation_loss": 0.010863586515188216,
      "selected_pairs": 27175.2,
      "state_query_norm": 44.88448600769043,
      "state_relation_loss": 0.000696146028349176,
      "step": 3340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010863586515188216,
      "weighted_relation_loss": 0.010863586515188216
    },
    {
      "action_query_norm": 44.697146606445315,
      "action_relation_loss": 0.009683116991072892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1337698783910195,
      "flow_matching_loss": 0.06822118498384952,
      "grad_norm": 0.19276751577854156,
      "learning_rate": 9.896504875159122e-05,
      "loss": 0.0786,
      "relation_loss": 0.010389809682965278,
      "selected_pairs": 27089.8,
      "state_query_norm": 44.88348617553711,
      "state_relation_loss": 0.0007066927093546838,
      "step": 3350,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.010389809682965278,
      "weighted_relation_loss": 0.010389809682965278
    },
    {
      "action_query_norm": 44.69978332519531,
      "action_relation_loss": 0.009544021729379893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1431244153414406,
      "flow_matching_loss": 0.0701854657381773,
      "grad_norm": 0.17859023809432983,
      "learning_rate": 9.89538630727064e-05,
      "loss": 0.0804,
      "relation_loss": 0.010238398239016533,
      "selected_pairs": 27230.4,
      "state_query_norm": 44.88203010559082,
      "state_relation_loss": 0.000694376410683617,
      "step": 3360,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.010238398239016533,
      "weighted_relation_loss": 0.010238398239016533
    },
    {
      "action_query_norm": 44.69366340637207,
      "action_relation_loss": 0.00971148768439889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1524789522918617,
      "flow_matching_loss": 0.06780361942946911,
      "grad_norm": 0.20863866806030273,
      "learning_rate": 9.894261791026189e-05,
      "loss": 0.0782,
      "relation_loss": 0.0103691965341568,
      "selected_pairs": 26273.9,
      "state_query_norm": 44.88173980712891,
      "state_relation_loss": 0.0006577088788617402,
      "step": 3370,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0103691965341568,
      "weighted_relation_loss": 0.0103691965341568
    },
    {
      "action_query_norm": 44.692385864257815,
      "action_relation_loss": 0.009563535451889038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1618334892422824,
      "flow_matching_loss": 0.06647207252681256,
      "grad_norm": 0.2114645391702652,
      "learning_rate": 9.893131327792165e-05,
      "loss": 0.0767,
      "relation_loss": 0.01025154795497656,
      "selected_pairs": 27385.8,
      "state_query_norm": 44.87884635925293,
      "state_relation_loss": 0.0006880124623421579,
      "step": 3380,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.01025154795497656,
      "weighted_relation_loss": 0.01025154795497656
    },
    {
      "action_query_norm": 44.68887100219727,
      "action_relation_loss": 0.009961420483887196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1711880261927035,
      "flow_matching_loss": 0.06593386493623257,
      "grad_norm": 0.2304060161113739,
      "learning_rate": 9.891994918942182e-05,
      "loss": 0.0766,
      "relation_loss": 0.010635135229676962,
      "selected_pairs": 27008.0,
      "state_query_norm": 44.88076477050781,
      "state_relation_loss": 0.0006737148738466203,
      "step": 3390,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.010635135229676962,
      "weighted_relation_loss": 0.010635135229676962
    },
    {
      "action_query_norm": 44.697439193725586,
      "action_relation_loss": 0.009923606272786855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1805425631431246,
      "flow_matching_loss": 0.06826701648533344,
      "grad_norm": 0.15884746611118317,
      "learning_rate": 9.890852565857092e-05,
      "loss": 0.0789,
      "relation_loss": 0.010638901218771934,
      "selected_pairs": 27017.3,
      "state_query_norm": 44.88149528503418,
      "state_relation_loss": 0.0007152950100135058,
      "step": 3400,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.010638901218771934,
      "weighted_relation_loss": 0.010638901218771934
    },
    {
      "action_query_norm": 44.68629188537598,
      "action_relation_loss": 0.009885875089094043,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.189897100093545,
      "flow_matching_loss": 0.06454330310225487,
      "grad_norm": 0.14010494947433472,
      "learning_rate": 9.889704269924954e-05,
      "loss": 0.0751,
      "relation_loss": 0.010603992827236652,
      "selected_pairs": 27201.0,
      "state_query_norm": 44.87942810058594,
      "state_relation_loss": 0.000718117575161159,
      "step": 3410,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.010603992827236652,
      "weighted_relation_loss": 0.010603992827236652
    },
    {
      "action_query_norm": 44.685990524291995,
      "action_relation_loss": 0.009760934673249722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.1992516370439663,
      "flow_matching_loss": 0.06573434211313725,
      "grad_norm": 0.12850676476955414,
      "learning_rate": 9.888550032541059e-05,
      "loss": 0.0762,
      "relation_loss": 0.01048999112099409,
      "selected_pairs": 26783.7,
      "state_query_norm": 44.87577514648437,
      "state_relation_loss": 0.0007290564943104982,
      "step": 3420,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.01048999112099409,
      "weighted_relation_loss": 0.01048999112099409
    },
    {
      "action_query_norm": 44.685905075073244,
      "action_relation_loss": 0.010387562680989504,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2086061739943874,
      "flow_matching_loss": 0.06688251905143261,
      "grad_norm": 0.2018786370754242,
      "learning_rate": 9.887389855107916e-05,
      "loss": 0.078,
      "relation_loss": 0.011091927252709865,
      "selected_pairs": 26970.5,
      "state_query_norm": 44.874458694458006,
      "state_relation_loss": 0.0007043646241072565,
      "step": 3430,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.011091927252709865,
      "weighted_relation_loss": 0.011091927252709865
    },
    {
      "action_query_norm": 44.683480453491214,
      "action_relation_loss": 0.009898201562464237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.217960710944808,
      "flow_matching_loss": 0.06698884591460227,
      "grad_norm": 0.16453728079795837,
      "learning_rate": 9.886223739035248e-05,
      "loss": 0.0776,
      "relation_loss": 0.010602188762277365,
      "selected_pairs": 26622.7,
      "state_query_norm": 44.87176170349121,
      "state_relation_loss": 0.0007039872463792563,
      "step": 3440,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.010602188762277365,
      "weighted_relation_loss": 0.010602188762277365
    },
    {
      "action_query_norm": 44.67947273254394,
      "action_relation_loss": 0.009814968146383763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.227315247895229,
      "flow_matching_loss": 0.06629270650446414,
      "grad_norm": 0.14052556455135345,
      "learning_rate": 9.885051685739997e-05,
      "loss": 0.0768,
      "relation_loss": 0.010478009842336178,
      "selected_pairs": 26367.4,
      "state_query_norm": 44.87014503479004,
      "state_relation_loss": 0.0006630416086409241,
      "step": 3450,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.010478009842336178,
      "weighted_relation_loss": 0.010478009842336178
    },
    {
      "action_query_norm": 44.676043319702146,
      "action_relation_loss": 0.01037879679352045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2366697848456503,
      "flow_matching_loss": 0.0643036350607872,
      "grad_norm": 0.14784207940101624,
      "learning_rate": 9.883873696646316e-05,
      "loss": 0.0754,
      "relation_loss": 0.011075034737586975,
      "selected_pairs": 27223.7,
      "state_query_norm": 44.86981735229492,
      "state_relation_loss": 0.0006962378858588636,
      "step": 3460,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.011075034737586975,
      "weighted_relation_loss": 0.011075034737586975
    },
    {
      "action_query_norm": 44.67076568603515,
      "action_relation_loss": 0.009969459753483534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.246024321796071,
      "flow_matching_loss": 0.0652785588055849,
      "grad_norm": 0.19167344272136688,
      "learning_rate": 9.882689773185575e-05,
      "loss": 0.0759,
      "relation_loss": 0.010636158287525177,
      "selected_pairs": 26931.8,
      "state_query_norm": 44.86830940246582,
      "state_relation_loss": 0.0006666983827017248,
      "step": 3470,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.010636158287525177,
      "weighted_relation_loss": 0.010636158287525177
    },
    {
      "action_query_norm": 44.67172203063965,
      "action_relation_loss": 0.010367987491190434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.255378858746492,
      "flow_matching_loss": 0.06609270498156547,
      "grad_norm": 0.1858828216791153,
      "learning_rate": 9.881499916796353e-05,
      "loss": 0.0772,
      "relation_loss": 0.011107753776013852,
      "selected_pairs": 27037.1,
      "state_query_norm": 44.868530654907225,
      "state_relation_loss": 0.0007397664187010378,
      "step": 3480,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.011107753776013852,
      "weighted_relation_loss": 0.011107753776013852
    },
    {
      "action_query_norm": 44.67383308410645,
      "action_relation_loss": 0.009445722680538893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.264733395696913,
      "flow_matching_loss": 0.0637446541339159,
      "grad_norm": 0.21234063804149628,
      "learning_rate": 9.880304128924434e-05,
      "loss": 0.0739,
      "relation_loss": 0.010172367189079523,
      "selected_pairs": 26569.1,
      "state_query_norm": 44.86443862915039,
      "state_relation_loss": 0.000726644363021478,
      "step": 3490,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.010172367189079523,
      "weighted_relation_loss": 0.010172367189079523
    },
    {
      "action_query_norm": 44.662471008300784,
      "action_relation_loss": 0.009848077036440372,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.2740879326473338,
      "flow_matching_loss": 0.06703648082911969,
      "grad_norm": 0.22225777804851532,
      "learning_rate": 9.879102411022817e-05,
      "loss": 0.0776,
      "relation_loss": 0.010560112167149782,
      "selected_pairs": 26096.5,
      "state_query_norm": 44.86320190429687,
      "state_relation_loss": 0.0007120352005586028,
      "step": 3500,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.010560112167149782,
      "weighted_relation_loss": 0.010560112167149782
    },
    {
      "action_query_norm": 44.65904197692871,
      "action_relation_loss": 0.009560687653720378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.283442469597755,
      "flow_matching_loss": 0.0634719729423523,
      "grad_norm": 0.18375718593597412,
      "learning_rate": 9.877894764551703e-05,
      "loss": 0.0737,
      "relation_loss": 0.010273277945816516,
      "selected_pairs": 27153.3,
      "state_query_norm": 44.86320571899414,
      "state_relation_loss": 0.000712590228067711,
      "step": 3510,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.010273277945816516,
      "weighted_relation_loss": 0.010273277945816516
    },
    {
      "action_query_norm": 44.65331268310547,
      "action_relation_loss": 0.009855825826525689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.292797006548176,
      "flow_matching_loss": 0.06484064869582654,
      "grad_norm": 0.15138499438762665,
      "learning_rate": 9.876681190978494e-05,
      "loss": 0.0754,
      "relation_loss": 0.010564466938376427,
      "selected_pairs": 26804.7,
      "state_query_norm": 44.86005897521973,
      "state_relation_loss": 0.0007086411642376333,
      "step": 3520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010564466938376427,
      "weighted_relation_loss": 0.010564466938376427
    },
    {
      "action_query_norm": 44.64619255065918,
      "action_relation_loss": 0.009718927647918462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3021515434985966,
      "flow_matching_loss": 0.06607462465763092,
      "grad_norm": 0.1513838917016983,
      "learning_rate": 9.875461691777797e-05,
      "loss": 0.0765,
      "relation_loss": 0.010410762391984463,
      "selected_pairs": 26123.5,
      "state_query_norm": 44.86170120239258,
      "state_relation_loss": 0.0006918347324244678,
      "step": 3530,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.010410762391984463,
      "weighted_relation_loss": 0.010410762391984463
    },
    {
      "action_query_norm": 44.653485488891604,
      "action_relation_loss": 0.010226206947118044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3115060804490177,
      "flow_matching_loss": 0.06633955053985119,
      "grad_norm": 0.20859280228614807,
      "learning_rate": 9.874236268431417e-05,
      "loss": 0.0773,
      "relation_loss": 0.010960685182362795,
      "selected_pairs": 26579.6,
      "state_query_norm": 44.85921669006348,
      "state_relation_loss": 0.0007344780722633004,
      "step": 3540,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.010960685182362795,
      "weighted_relation_loss": 0.010960685182362795
    },
    {
      "action_query_norm": 44.65253868103027,
      "action_relation_loss": 0.010048741195350885,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.320860617399439,
      "flow_matching_loss": 0.0636835340410471,
      "grad_norm": 0.15203352272510529,
      "learning_rate": 9.873004922428361e-05,
      "loss": 0.0744,
      "relation_loss": 0.010765333566814661,
      "selected_pairs": 26803.9,
      "state_query_norm": 44.853841400146486,
      "state_relation_loss": 0.0007165924413129688,
      "step": 3550,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010765333566814661,
      "weighted_relation_loss": 0.010765333566814661
    },
    {
      "action_query_norm": 44.65639228820801,
      "action_relation_loss": 0.009387954510748387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.33021515434986,
      "flow_matching_loss": 0.06340510658919811,
      "grad_norm": 0.18922346830368042,
      "learning_rate": 9.871767655264829e-05,
      "loss": 0.0735,
      "relation_loss": 0.010090416669845581,
      "selected_pairs": 26502.2,
      "state_query_norm": 44.85634536743164,
      "state_relation_loss": 0.0007024622114840895,
      "step": 3560,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.010090416669845581,
      "weighted_relation_loss": 0.010090416669845581
    },
    {
      "action_query_norm": 44.64644012451172,
      "action_relation_loss": 0.009297141153365374,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3395696913002806,
      "flow_matching_loss": 0.06510452553629875,
      "grad_norm": 0.11258786171674728,
      "learning_rate": 9.87052446844422e-05,
      "loss": 0.075,
      "relation_loss": 0.009936848282814026,
      "selected_pairs": 27664.0,
      "state_query_norm": 44.85329055786133,
      "state_relation_loss": 0.0006397072109393775,
      "step": 3570,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.009936848282814026,
      "weighted_relation_loss": 0.009936848282814026
    },
    {
      "action_query_norm": 44.648397827148436,
      "action_relation_loss": 0.009881804697215557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3489242282507017,
      "flow_matching_loss": 0.06874060407280921,
      "grad_norm": 0.1718769669532776,
      "learning_rate": 9.869275363477122e-05,
      "loss": 0.0793,
      "relation_loss": 0.01057386128231883,
      "selected_pairs": 27594.8,
      "state_query_norm": 44.85338096618652,
      "state_relation_loss": 0.0006920565967448056,
      "step": 3580,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.01057386128231883,
      "weighted_relation_loss": 0.01057386128231883
    },
    {
      "action_query_norm": 44.6504207611084,
      "action_relation_loss": 0.00963349426165223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3582787652011223,
      "flow_matching_loss": 0.06487002372741699,
      "grad_norm": 0.15329095721244812,
      "learning_rate": 9.868020341881312e-05,
      "loss": 0.0752,
      "relation_loss": 0.010309962369501591,
      "selected_pairs": 27530.4,
      "state_query_norm": 44.85410118103027,
      "state_relation_loss": 0.0006764681485947222,
      "step": 3590,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.010309962369501591,
      "weighted_relation_loss": 0.010309962369501591
    },
    {
      "action_query_norm": 44.64547996520996,
      "action_relation_loss": 0.009552357997745276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3676333021515434,
      "flow_matching_loss": 0.06768376119434834,
      "grad_norm": 0.26447370648384094,
      "learning_rate": 9.866759405181765e-05,
      "loss": 0.078,
      "relation_loss": 0.010267543233931064,
      "selected_pairs": 27323.8,
      "state_query_norm": 44.850640869140626,
      "state_relation_loss": 0.0007151851721573621,
      "step": 3600,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.010267543233931064,
      "weighted_relation_loss": 0.010267543233931064
    },
    {
      "action_query_norm": 44.641272354125974,
      "action_relation_loss": 0.00974039463326335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3769878391019645,
      "flow_matching_loss": 0.06493034660816192,
      "grad_norm": 0.2934746742248535,
      "learning_rate": 9.865492554910633e-05,
      "loss": 0.0753,
      "relation_loss": 0.010414069797843695,
      "selected_pairs": 26366.8,
      "state_query_norm": 44.848811340332034,
      "state_relation_loss": 0.0006736749724950642,
      "step": 3610,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.010414069797843695,
      "weighted_relation_loss": 0.010414069797843695
    },
    {
      "action_query_norm": 44.635270309448245,
      "action_relation_loss": 0.010272290650755167,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3863423760523856,
      "flow_matching_loss": 0.06510315202176571,
      "grad_norm": 0.24617545306682587,
      "learning_rate": 9.864219792607262e-05,
      "loss": 0.0761,
      "relation_loss": 0.010952951665967704,
      "selected_pairs": 26994.0,
      "state_query_norm": 44.846147537231445,
      "state_relation_loss": 0.0006806608871556818,
      "step": 3620,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.010952951665967704,
      "weighted_relation_loss": 0.010952951665967704
    },
    {
      "action_query_norm": 44.62169990539551,
      "action_relation_loss": 0.009836078155785799,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.3956969130028063,
      "flow_matching_loss": 0.06396746672689915,
      "grad_norm": 0.16145239770412445,
      "learning_rate": 9.862941119818177e-05,
      "loss": 0.0745,
      "relation_loss": 0.010523509420454501,
      "selected_pairs": 27164.5,
      "state_query_norm": 44.84297332763672,
      "state_relation_loss": 0.0006874310900457203,
      "step": 3630,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.010523509420454501,
      "weighted_relation_loss": 0.010523509420454501
    },
    {
      "action_query_norm": 44.62603607177734,
      "action_relation_loss": 0.00969639141112566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4050514499532274,
      "flow_matching_loss": 0.06676412224769593,
      "grad_norm": 0.13312683999538422,
      "learning_rate": 9.861656538097086e-05,
      "loss": 0.0771,
      "relation_loss": 0.01037228936329484,
      "selected_pairs": 27365.8,
      "state_query_norm": 44.84334449768066,
      "state_relation_loss": 0.0006758979114238172,
      "step": 3640,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.01037228936329484,
      "weighted_relation_loss": 0.01037228936329484
    },
    {
      "action_query_norm": 44.625048828125,
      "action_relation_loss": 0.010161132179200649,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.414405986903648,
      "flow_matching_loss": 0.06577147245407104,
      "grad_norm": 0.17047911882400513,
      "learning_rate": 9.860366049004877e-05,
      "loss": 0.0766,
      "relation_loss": 0.010846975911408662,
      "selected_pairs": 26963.9,
      "state_query_norm": 44.842593383789065,
      "state_relation_loss": 0.0006858436216134578,
      "step": 3650,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.010846975911408662,
      "weighted_relation_loss": 0.010846975911408662
    },
    {
      "action_query_norm": 44.621881103515626,
      "action_relation_loss": 0.009817736595869065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.423760523854069,
      "flow_matching_loss": 0.06535049974918365,
      "grad_norm": 0.21030010282993317,
      "learning_rate": 9.859069654109615e-05,
      "loss": 0.0759,
      "relation_loss": 0.01053200475871563,
      "selected_pairs": 26942.7,
      "state_query_norm": 44.842232131958006,
      "state_relation_loss": 0.0007142681861296296,
      "step": 3660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.01053200475871563,
      "weighted_relation_loss": 0.01053200475871563
    },
    {
      "action_query_norm": 44.62148742675781,
      "action_relation_loss": 0.009662124235183001,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4331150608044902,
      "flow_matching_loss": 0.06318131648004055,
      "grad_norm": 0.252419650554657,
      "learning_rate": 9.857767354986545e-05,
      "loss": 0.0736,
      "relation_loss": 0.010391457378864289,
      "selected_pairs": 26559.0,
      "state_query_norm": 44.84272994995117,
      "state_relation_loss": 0.0007293331145774573,
      "step": 3670,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.010391457378864289,
      "weighted_relation_loss": 0.010391457378864289
    },
    {
      "action_query_norm": 44.610504913330075,
      "action_relation_loss": 0.009751736372709274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4424695977549113,
      "flow_matching_loss": 0.06392262876033783,
      "grad_norm": 0.20328539609909058,
      "learning_rate": 9.856459153218078e-05,
      "loss": 0.0744,
      "relation_loss": 0.010429023019969463,
      "selected_pairs": 26618.7,
      "state_query_norm": 44.83686904907226,
      "state_relation_loss": 0.0006772864842787385,
      "step": 3680,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.010429023019969463,
      "weighted_relation_loss": 0.010429023019969463
    },
    {
      "action_query_norm": 44.60885391235352,
      "action_relation_loss": 0.010165373049676418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.451824134705332,
      "flow_matching_loss": 0.06469358503818512,
      "grad_norm": 0.19218483567237854,
      "learning_rate": 9.855145050393808e-05,
      "loss": 0.0756,
      "relation_loss": 0.010904673859477043,
      "selected_pairs": 26457.6,
      "state_query_norm": 44.83539009094238,
      "state_relation_loss": 0.0007393008330836892,
      "step": 3690,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.010904673859477043,
      "weighted_relation_loss": 0.010904673859477043
    },
    {
      "action_query_norm": 44.617475509643555,
      "action_relation_loss": 0.009980433993041516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.461178671655753,
      "flow_matching_loss": 0.06531939543783664,
      "grad_norm": 0.15757982432842255,
      "learning_rate": 9.85382504811049e-05,
      "loss": 0.076,
      "relation_loss": 0.010669376607984305,
      "selected_pairs": 26831.6,
      "state_query_norm": 44.836743927001955,
      "state_relation_loss": 0.0006889425334520638,
      "step": 3700,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.010669376607984305,
      "weighted_relation_loss": 0.010669376607984305
    },
    {
      "action_query_norm": 44.61569786071777,
      "action_relation_loss": 0.009783374797552823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.4705332086061738,
      "flow_matching_loss": 0.065739706158638,
      "grad_norm": 0.17862994968891144,
      "learning_rate": 9.852499147972054e-05,
      "loss": 0.0763,
      "relation_loss": 0.010540014784783124,
      "selected_pairs": 27253.7,
      "state_query_norm": 44.83313484191895,
      "state_relation_loss": 0.0007566400512587279,
      "step": 3710,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.010540014784783124,
      "weighted_relation_loss": 0.010540014784783124
    },
    {
      "action_query_norm": 44.61522445678711,
      "action_relation_loss": 0.009430779423564673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.479887745556595,
      "flow_matching_loss": 0.06465134955942631,
      "grad_norm": 0.1811298131942749,
      "learning_rate": 9.851167351589592e-05,
      "loss": 0.0748,
      "relation_loss": 0.01014057146385312,
      "selected_pairs": 26944.6,
      "state_query_norm": 44.833164596557616,
      "state_relation_loss": 0.0007097920286469161,
      "step": 3720,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.01014057146385312,
      "weighted_relation_loss": 0.01014057146385312
    },
    {
      "action_query_norm": 44.62185363769531,
      "action_relation_loss": 0.009846734534949064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.489242282507016,
      "flow_matching_loss": 0.06771280616521835,
      "grad_norm": 0.17623119056224823,
      "learning_rate": 9.849829660581363e-05,
      "loss": 0.0782,
      "relation_loss": 0.010477170627564193,
      "selected_pairs": 26756.8,
      "state_query_norm": 44.833734130859376,
      "state_relation_loss": 0.0006304362555965781,
      "step": 3730,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.010477170627564193,
      "weighted_relation_loss": 0.010477170627564193
    },
    {
      "action_query_norm": 44.61119270324707,
      "action_relation_loss": 0.009674530383199454,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.498596819457437,
      "flow_matching_loss": 0.06745526269078254,
      "grad_norm": 0.14262843132019043,
      "learning_rate": 9.848486076572787e-05,
      "loss": 0.0778,
      "relation_loss": 0.010344042535871267,
      "selected_pairs": 26853.2,
      "state_query_norm": 44.83416481018067,
      "state_relation_loss": 0.0006695121875964106,
      "step": 3740,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.010344042535871267,
      "weighted_relation_loss": 0.010344042535871267
    },
    {
      "action_query_norm": 44.59804573059082,
      "action_relation_loss": 0.009884994057938457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5079513564078577,
      "flow_matching_loss": 0.06423044726252555,
      "grad_norm": 0.19233529269695282,
      "learning_rate": 9.847136601196446e-05,
      "loss": 0.0747,
      "relation_loss": 0.01051328405737877,
      "selected_pairs": 27221.8,
      "state_query_norm": 44.829607009887695,
      "state_relation_loss": 0.0006282899295911193,
      "step": 3750,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.01051328405737877,
      "weighted_relation_loss": 0.01051328405737877
    },
    {
      "action_query_norm": 44.60128631591797,
      "action_relation_loss": 0.009681223053485155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.517305893358279,
      "flow_matching_loss": 0.0667033351957798,
      "grad_norm": 0.25113511085510254,
      "learning_rate": 9.845781236092078e-05,
      "loss": 0.077,
      "relation_loss": 0.010336226131767035,
      "selected_pairs": 26802.9,
      "state_query_norm": 44.82999229431152,
      "state_relation_loss": 0.0006550029793288559,
      "step": 3760,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010336226131767035,
      "weighted_relation_loss": 0.010336226131767035
    },
    {
      "action_query_norm": 44.5930419921875,
      "action_relation_loss": 0.009596714563667775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5266604303086995,
      "flow_matching_loss": 0.06703572943806649,
      "grad_norm": 0.19950515031814575,
      "learning_rate": 9.844419982906583e-05,
      "loss": 0.0773,
      "relation_loss": 0.010221755504608155,
      "selected_pairs": 26096.2,
      "state_query_norm": 44.82457275390625,
      "state_relation_loss": 0.0006250409875065088,
      "step": 3770,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.010221755504608155,
      "weighted_relation_loss": 0.010221755504608155
    },
    {
      "action_query_norm": 44.59294815063477,
      "action_relation_loss": 0.009309724951162935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5360149672591206,
      "flow_matching_loss": 0.0655294306576252,
      "grad_norm": 0.17853055894374847,
      "learning_rate": 9.843052843294008e-05,
      "loss": 0.0755,
      "relation_loss": 0.010017015039920807,
      "selected_pairs": 26648.6,
      "state_query_norm": 44.82015876770019,
      "state_relation_loss": 0.0007072900771163404,
      "step": 3780,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.010017015039920807,
      "weighted_relation_loss": 0.010017015039920807
    },
    {
      "action_query_norm": 44.585733795166014,
      "action_relation_loss": 0.009830484911799432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5453695042095417,
      "flow_matching_loss": 0.06476896218955516,
      "grad_norm": 0.25888752937316895,
      "learning_rate": 9.841679818915559e-05,
      "loss": 0.0753,
      "relation_loss": 0.010519937053322792,
      "selected_pairs": 26686.9,
      "state_query_norm": 44.82321472167969,
      "state_relation_loss": 0.0006894520309288054,
      "step": 3790,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.010519937053322792,
      "weighted_relation_loss": 0.010519937053322792
    },
    {
      "action_query_norm": 44.58385162353515,
      "action_relation_loss": 0.00963443135842681,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5547240411599628,
      "flow_matching_loss": 0.06428001411259174,
      "grad_norm": 0.22658155858516693,
      "learning_rate": 9.840300911439591e-05,
      "loss": 0.0746,
      "relation_loss": 0.010295066330581904,
      "selected_pairs": 26385.9,
      "state_query_norm": 44.824127197265625,
      "state_relation_loss": 0.0006606349488720298,
      "step": 3800,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.010295066330581904,
      "weighted_relation_loss": 0.010295066330581904
    },
    {
      "action_query_norm": 44.59057426452637,
      "action_relation_loss": 0.010288762021809816,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5640785781103834,
      "flow_matching_loss": 0.06820032447576523,
      "grad_norm": 0.17907188832759857,
      "learning_rate": 9.838916122541603e-05,
      "loss": 0.0792,
      "relation_loss": 0.011002376768738031,
      "selected_pairs": 27129.3,
      "state_query_norm": 44.82142333984375,
      "state_relation_loss": 0.0007136146305128932,
      "step": 3810,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.011002376768738031,
      "weighted_relation_loss": 0.011002376768738031
    },
    {
      "action_query_norm": 44.574025344848636,
      "action_relation_loss": 0.00981123885139823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5734331150608045,
      "flow_matching_loss": 0.06519823260605336,
      "grad_norm": 0.1990414708852768,
      "learning_rate": 9.837525453904246e-05,
      "loss": 0.0757,
      "relation_loss": 0.010520709306001663,
      "selected_pairs": 26584.5,
      "state_query_norm": 44.81837692260742,
      "state_relation_loss": 0.0007094704371411354,
      "step": 3820,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.010520709306001663,
      "weighted_relation_loss": 0.010520709306001663
    },
    {
      "action_query_norm": 44.57905731201172,
      "action_relation_loss": 0.00963675109669566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.582787652011225,
      "flow_matching_loss": 0.06647195667028427,
      "grad_norm": 0.1765795350074768,
      "learning_rate": 9.836128907217314e-05,
      "loss": 0.0768,
      "relation_loss": 0.010331002622842788,
      "selected_pairs": 26873.1,
      "state_query_norm": 44.81888847351074,
      "state_relation_loss": 0.0006942515959963202,
      "step": 3830,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.010331002622842788,
      "weighted_relation_loss": 0.010331002622842788
    },
    {
      "action_query_norm": 44.57098579406738,
      "action_relation_loss": 0.009965398348867893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.5921421889616463,
      "flow_matching_loss": 0.06564790047705174,
      "grad_norm": 0.16186971962451935,
      "learning_rate": 9.834726484177743e-05,
      "loss": 0.0763,
      "relation_loss": 0.010614187084138393,
      "selected_pairs": 27064.6,
      "state_query_norm": 44.81557655334473,
      "state_relation_loss": 0.0006487887818366289,
      "step": 3840,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.010614187084138393,
      "weighted_relation_loss": 0.010614187084138393
    },
    {
      "action_query_norm": 44.56662101745606,
      "action_relation_loss": 0.00978478230535984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6014967259120674,
      "flow_matching_loss": 0.0689866453409195,
      "grad_norm": 0.1949053853750229,
      "learning_rate": 9.833318186489609e-05,
      "loss": 0.0795,
      "relation_loss": 0.010468850005418062,
      "selected_pairs": 26455.8,
      "state_query_norm": 44.81415557861328,
      "state_relation_loss": 0.0006840676942374557,
      "step": 3850,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.010468850005418062,
      "weighted_relation_loss": 0.010468850005418062
    },
    {
      "action_query_norm": 44.57582588195801,
      "action_relation_loss": 0.009670669119805098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6108512628624885,
      "flow_matching_loss": 0.06356790475547314,
      "grad_norm": 0.1481783241033554,
      "learning_rate": 9.831904015864126e-05,
      "loss": 0.0739,
      "relation_loss": 0.010342993959784507,
      "selected_pairs": 27270.3,
      "state_query_norm": 44.81558418273926,
      "state_relation_loss": 0.0006723248050548136,
      "step": 3860,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.010342993959784507,
      "weighted_relation_loss": 0.010342993959784507
    },
    {
      "action_query_norm": 44.56385688781738,
      "action_relation_loss": 0.009811912290751934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.620205799812909,
      "flow_matching_loss": 0.06434832513332367,
      "grad_norm": 0.20328441262245178,
      "learning_rate": 9.830483974019645e-05,
      "loss": 0.0748,
      "relation_loss": 0.010465884301811456,
      "selected_pairs": 26848.1,
      "state_query_norm": 44.81865005493164,
      "state_relation_loss": 0.0006539719004649668,
      "step": 3870,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.010465884301811456,
      "weighted_relation_loss": 0.010465884301811456
    },
    {
      "action_query_norm": 44.56715774536133,
      "action_relation_loss": 0.009505347535014152,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6295603367633302,
      "flow_matching_loss": 0.06759745329618454,
      "grad_norm": 0.23971782624721527,
      "learning_rate": 9.82905806268165e-05,
      "loss": 0.0778,
      "relation_loss": 0.010182443261146545,
      "selected_pairs": 27089.5,
      "state_query_norm": 44.81205940246582,
      "state_relation_loss": 0.0006770958309061826,
      "step": 3880,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.010182443261146545,
      "weighted_relation_loss": 0.010182443261146545
    },
    {
      "action_query_norm": 44.56017417907715,
      "action_relation_loss": 0.010231983847916126,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.6389148737137513,
      "flow_matching_loss": 0.0667191930115223,
      "grad_norm": 0.2058243602514267,
      "learning_rate": 9.82762628358276e-05,
      "loss": 0.0776,
      "relation_loss": 0.010903441905975341,
      "selected_pairs": 27197.4,
      "state_query_norm": 44.81308784484863,
      "state_relation_loss": 0.0006714581220876425,
      "step": 3890,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.010903441905975341,
      "weighted_relation_loss": 0.010903441905975341
    },
    {
      "action_query_norm": 44.549347686767575,
      "action_relation_loss": 0.00966203911229968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.648269410664172,
      "flow_matching_loss": 0.06729542352259159,
      "grad_norm": 0.15175604820251465,
      "learning_rate": 9.826188638462718e-05,
      "loss": 0.0776,
      "relation_loss": 0.010298699513077736,
      "selected_pairs": 27010.7,
      "state_query_norm": 44.808928680419925,
      "state_relation_loss": 0.0006366603192873299,
      "step": 3900,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.010298699513077736,
      "weighted_relation_loss": 0.010298699513077736
    },
    {
      "action_query_norm": 44.548156356811525,
      "action_relation_loss": 0.009627364855259658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.657623947614593,
      "flow_matching_loss": 0.06644517034292222,
      "grad_norm": 0.1772952824831009,
      "learning_rate": 9.824745129068402e-05,
      "loss": 0.0767,
      "relation_loss": 0.010283361095935107,
      "selected_pairs": 27371.1,
      "state_query_norm": 44.80657424926758,
      "state_relation_loss": 0.0006559962290339172,
      "step": 3910,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.010283361095935107,
      "weighted_relation_loss": 0.010283361095935107
    },
    {
      "action_query_norm": 44.54182167053223,
      "action_relation_loss": 0.009685593750327825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.666978484565014,
      "flow_matching_loss": 0.06355313509702683,
      "grad_norm": 0.15309174358844757,
      "learning_rate": 9.82329575715381e-05,
      "loss": 0.0739,
      "relation_loss": 0.010347862262278795,
      "selected_pairs": 27181.1,
      "state_query_norm": 44.80865707397461,
      "state_relation_loss": 0.0006622684129979461,
      "step": 3920,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.010347862262278795,
      "weighted_relation_loss": 0.010347862262278795
    },
    {
      "action_query_norm": 44.54353256225586,
      "action_relation_loss": 0.009448414482176305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.676333021515435,
      "flow_matching_loss": 0.06631237529218197,
      "grad_norm": 0.16973040997982025,
      "learning_rate": 9.821840524480066e-05,
      "loss": 0.0764,
      "relation_loss": 0.010107984580099583,
      "selected_pairs": 26801.4,
      "state_query_norm": 44.80780944824219,
      "state_relation_loss": 0.0006595700629986823,
      "step": 3930,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010107984580099583,
      "weighted_relation_loss": 0.010107984580099583
    },
    {
      "action_query_norm": 44.53987083435059,
      "action_relation_loss": 0.009740614891052246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.685687558465856,
      "flow_matching_loss": 0.06230801306664944,
      "grad_norm": 0.13706760108470917,
      "learning_rate": 9.820379432815414e-05,
      "loss": 0.0727,
      "relation_loss": 0.010389846097677947,
      "selected_pairs": 26735.2,
      "state_query_norm": 44.80821838378906,
      "state_relation_loss": 0.0006492312182672322,
      "step": 3940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.010389846097677947,
      "weighted_relation_loss": 0.010389846097677947
    },
    {
      "action_query_norm": 44.54758911132812,
      "action_relation_loss": 0.010317089315503836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.695042095416277,
      "flow_matching_loss": 0.06677252426743507,
      "grad_norm": 0.16767409443855286,
      "learning_rate": 9.81891248393522e-05,
      "loss": 0.0778,
      "relation_loss": 0.010982921440154314,
      "selected_pairs": 26802.2,
      "state_query_norm": 44.80806465148926,
      "state_relation_loss": 0.0006658319849520922,
      "step": 3950,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010982921440154314,
      "weighted_relation_loss": 0.010982921440154314
    },
    {
      "action_query_norm": 44.533407592773436,
      "action_relation_loss": 0.010107466764748097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7043966323666977,
      "flow_matching_loss": 0.06461024358868599,
      "grad_norm": 0.15717753767967224,
      "learning_rate": 9.817439679621963e-05,
      "loss": 0.0753,
      "relation_loss": 0.010716978274285794,
      "selected_pairs": 27225.0,
      "state_query_norm": 44.80095024108887,
      "state_relation_loss": 0.0006095115852076561,
      "step": 3960,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.010716978274285794,
      "weighted_relation_loss": 0.010716978274285794
    },
    {
      "action_query_norm": 44.52951240539551,
      "action_relation_loss": 0.009795927442610264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.713751169317119,
      "flow_matching_loss": 0.06948755793273449,
      "grad_norm": 0.16907386481761932,
      "learning_rate": 9.815961021665243e-05,
      "loss": 0.0799,
      "relation_loss": 0.010408139694482088,
      "selected_pairs": 26802.7,
      "state_query_norm": 44.802076721191405,
      "state_relation_loss": 0.0006122122576925904,
      "step": 3970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.010408139694482088,
      "weighted_relation_loss": 0.010408139694482088
    },
    {
      "action_query_norm": 44.52962875366211,
      "action_relation_loss": 0.00912936432287097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.72310570626754,
      "flow_matching_loss": 0.063532555103302,
      "grad_norm": 0.19948948919773102,
      "learning_rate": 9.814476511861763e-05,
      "loss": 0.0733,
      "relation_loss": 0.009804188553243876,
      "selected_pairs": 26387.8,
      "state_query_norm": 44.79707489013672,
      "state_relation_loss": 0.0006748241896275431,
      "step": 3980,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.009804188553243876,
      "weighted_relation_loss": 0.009804188553243876
    },
    {
      "action_query_norm": 44.52130393981933,
      "action_relation_loss": 0.009792781341820956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7324602432179605,
      "flow_matching_loss": 0.06417193338274955,
      "grad_norm": 0.15868428349494934,
      "learning_rate": 9.812986152015348e-05,
      "loss": 0.0746,
      "relation_loss": 0.01042178152129054,
      "selected_pairs": 26509.3,
      "state_query_norm": 44.799790954589845,
      "state_relation_loss": 0.0006290001329034567,
      "step": 3990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.01042178152129054,
      "weighted_relation_loss": 0.01042178152129054
    },
    {
      "action_query_norm": 44.5229305267334,
      "action_relation_loss": 0.009419934265315533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7418147801683816,
      "flow_matching_loss": 0.06455154903233051,
      "grad_norm": 0.19244185090065002,
      "learning_rate": 9.811489943936922e-05,
      "loss": 0.0746,
      "relation_loss": 0.010075739491730929,
      "selected_pairs": 26943.1,
      "state_query_norm": 44.79825973510742,
      "state_relation_loss": 0.0006558052904438227,
      "step": 4000,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.010075739491730929,
      "weighted_relation_loss": 0.010075739491730929
    },
    {
      "action_query_norm": 44.519141006469724,
      "action_relation_loss": 0.009676073957234622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7511693171188027,
      "flow_matching_loss": 0.06134760119020939,
      "grad_norm": 0.17527857422828674,
      "learning_rate": 9.809987889444522e-05,
      "loss": 0.0717,
      "relation_loss": 0.010347718466073274,
      "selected_pairs": 27343.5,
      "state_query_norm": 44.79501075744629,
      "state_relation_loss": 0.000671644660178572,
      "step": 4010,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.010347718466073274,
      "weighted_relation_loss": 0.010347718466073274
    },
    {
      "action_query_norm": 44.52257652282715,
      "action_relation_loss": 0.00976315550506115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7605238540692234,
      "flow_matching_loss": 0.06348761990666389,
      "grad_norm": 0.16329637169837952,
      "learning_rate": 9.808479990363282e-05,
      "loss": 0.0739,
      "relation_loss": 0.010389250796288252,
      "selected_pairs": 26659.5,
      "state_query_norm": 44.797790145874025,
      "state_relation_loss": 0.0006260951980948448,
      "step": 4020,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.010389250796288252,
      "weighted_relation_loss": 0.010389250796288252
    },
    {
      "action_query_norm": 44.522897720336914,
      "action_relation_loss": 0.009261776227504015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7698783910196445,
      "flow_matching_loss": 0.06474573910236359,
      "grad_norm": 0.2541285753250122,
      "learning_rate": 9.806966248525445e-05,
      "loss": 0.0747,
      "relation_loss": 0.009911205898970366,
      "selected_pairs": 26875.0,
      "state_query_norm": 44.79415054321289,
      "state_relation_loss": 0.0006494296889286488,
      "step": 4030,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.009911205898970366,
      "weighted_relation_loss": 0.009911205898970366
    },
    {
      "action_query_norm": 44.516289138793944,
      "action_relation_loss": 0.009096156805753708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7792329279700656,
      "flow_matching_loss": 0.062025459855794905,
      "grad_norm": 0.22975629568099976,
      "learning_rate": 9.805446665770348e-05,
      "loss": 0.0718,
      "relation_loss": 0.009750794991850853,
      "selected_pairs": 26783.4,
      "state_query_norm": 44.78712158203125,
      "state_relation_loss": 0.0006546381395310163,
      "step": 4040,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.009750794991850853,
      "weighted_relation_loss": 0.009750794991850853
    },
    {
      "action_query_norm": 44.515032577514646,
      "action_relation_loss": 0.009170721098780633,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7885874649204867,
      "flow_matching_loss": 0.06444474644958972,
      "grad_norm": 0.2183249443769455,
      "learning_rate": 9.803921243944429e-05,
      "loss": 0.0742,
      "relation_loss": 0.009804611373692752,
      "selected_pairs": 27195.7,
      "state_query_norm": 44.790740203857425,
      "state_relation_loss": 0.000633890286553651,
      "step": 4050,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.009804611373692752,
      "weighted_relation_loss": 0.009804611373692752
    },
    {
      "action_query_norm": 44.5218921661377,
      "action_relation_loss": 0.009791109431535005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.7979420018709074,
      "flow_matching_loss": 0.0655147958546877,
      "grad_norm": 0.14232297241687775,
      "learning_rate": 9.802389984901218e-05,
      "loss": 0.0759,
      "relation_loss": 0.010426297783851624,
      "selected_pairs": 26687.4,
      "state_query_norm": 44.79026412963867,
      "state_relation_loss": 0.0006351882591843605,
      "step": 4060,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.010426297783851624,
      "weighted_relation_loss": 0.010426297783851624
    },
    {
      "action_query_norm": 44.50683517456055,
      "action_relation_loss": 0.009032589662820101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8072965388213285,
      "flow_matching_loss": 0.06406967602670192,
      "grad_norm": 0.16820605099201202,
      "learning_rate": 9.80085289050134e-05,
      "loss": 0.0737,
      "relation_loss": 0.009615070465952159,
      "selected_pairs": 26848.1,
      "state_query_norm": 44.78733367919922,
      "state_relation_loss": 0.000582480791490525,
      "step": 4070,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009615070465952159,
      "weighted_relation_loss": 0.009615070465952159
    },
    {
      "action_query_norm": 44.4965015411377,
      "action_relation_loss": 0.009652831684798002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.816651075771749,
      "flow_matching_loss": 0.06615538895130157,
      "grad_norm": 0.20343855023384094,
      "learning_rate": 9.799309962612508e-05,
      "loss": 0.0764,
      "relation_loss": 0.010260604973882436,
      "selected_pairs": 27242.0,
      "state_query_norm": 44.783185195922854,
      "state_relation_loss": 0.0006077733531128615,
      "step": 4080,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.010260604973882436,
      "weighted_relation_loss": 0.010260604973882436
    },
    {
      "action_query_norm": 44.51400604248047,
      "action_relation_loss": 0.009507251996546983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.82600561272217,
      "flow_matching_loss": 0.06384239681065082,
      "grad_norm": 0.1657855361700058,
      "learning_rate": 9.797761203109527e-05,
      "loss": 0.0739,
      "relation_loss": 0.010071383509784936,
      "selected_pairs": 26869.9,
      "state_query_norm": 44.785334777832034,
      "state_relation_loss": 0.000564131504506804,
      "step": 4090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.010071383509784936,
      "weighted_relation_loss": 0.010071383509784936
    },
    {
      "action_query_norm": 44.502970123291014,
      "action_relation_loss": 0.009158908016979694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8353601496725913,
      "flow_matching_loss": 0.06270187459886074,
      "grad_norm": 0.16098260879516602,
      "learning_rate": 9.796206613874283e-05,
      "loss": 0.0725,
      "relation_loss": 0.009762514103204011,
      "selected_pairs": 27269.9,
      "state_query_norm": 44.78124313354492,
      "state_relation_loss": 0.0006036059465259313,
      "step": 4100,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.009762514103204011,
      "weighted_relation_loss": 0.009762514103204011
    },
    {
      "action_query_norm": 44.498905181884766,
      "action_relation_loss": 0.010031059943139553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.8447146866230124,
      "flow_matching_loss": 0.06214980743825436,
      "grad_norm": 0.19425979256629944,
      "learning_rate": 9.794646196795754e-05,
      "loss": 0.0728,
      "relation_loss": 0.010625197552144527,
      "selected_pairs": 27082.6,
      "state_query_norm": 44.77988395690918,
      "state_relation_loss": 0.0005941376148257405,
      "step": 4110,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.010625197552144527,
      "weighted_relation_loss": 0.010625197552144527
    },
    {
      "action_query_norm": 44.49170608520508,
      "action_relation_loss": 0.009422142058610916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.854069223573433,
      "flow_matching_loss": 0.06391878798604012,
      "grad_norm": 0.1908654272556305,
      "learning_rate": 9.793079953769987e-05,
      "loss": 0.074,
      "relation_loss": 0.010049228835850955,
      "selected_pairs": 26350.1,
      "state_query_norm": 44.77636985778808,
      "state_relation_loss": 0.0006270867714192718,
      "step": 4120,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.010049228835850955,
      "weighted_relation_loss": 0.010049228835850955
    },
    {
      "action_query_norm": 44.49069747924805,
      "action_relation_loss": 0.009759626165032387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.863423760523854,
      "flow_matching_loss": 0.06365971118211747,
      "grad_norm": 0.1463027447462082,
      "learning_rate": 9.79150788670012e-05,
      "loss": 0.074,
      "relation_loss": 0.010369274206459523,
      "selected_pairs": 26756.2,
      "state_query_norm": 44.781114196777345,
      "state_relation_loss": 0.000609647948294878,
      "step": 4130,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.010369274206459523,
      "weighted_relation_loss": 0.010369274206459523
    },
    {
      "action_query_norm": 44.496272277832034,
      "action_relation_loss": 0.009870467148721218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.872778297474275,
      "flow_matching_loss": 0.06273616440594196,
      "grad_norm": 0.15007543563842773,
      "learning_rate": 9.78992999749636e-05,
      "loss": 0.0732,
      "relation_loss": 0.01049547316506505,
      "selected_pairs": 27103.1,
      "state_query_norm": 44.777949523925784,
      "state_relation_loss": 0.0006250060338061303,
      "step": 4140,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.01049547316506505,
      "weighted_relation_loss": 0.01049547316506505
    },
    {
      "action_query_norm": 44.48187484741211,
      "action_relation_loss": 0.00907222144305706,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.882132834424696,
      "flow_matching_loss": 0.060026868060231206,
      "grad_norm": 0.15550850331783295,
      "learning_rate": 9.788346288075994e-05,
      "loss": 0.0697,
      "relation_loss": 0.00970290470868349,
      "selected_pairs": 26664.0,
      "state_query_norm": 44.77811164855957,
      "state_relation_loss": 0.0006306832830887288,
      "step": 4150,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00970290470868349,
      "weighted_relation_loss": 0.00970290470868349
    },
    {
      "action_query_norm": 44.48792991638184,
      "action_relation_loss": 0.010006944183260202,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.891487371375117,
      "flow_matching_loss": 0.06472779139876365,
      "grad_norm": 0.15571634471416473,
      "learning_rate": 9.786756760363373e-05,
      "loss": 0.0754,
      "relation_loss": 0.010647945757955312,
      "selected_pairs": 26664.3,
      "state_query_norm": 44.77512474060059,
      "state_relation_loss": 0.0006410016445443034,
      "step": 4160,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.010647945757955312,
      "weighted_relation_loss": 0.010647945757955312
    },
    {
      "action_query_norm": 44.485909652709964,
      "action_relation_loss": 0.009128875751048326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.900841908325538,
      "flow_matching_loss": 0.06290548704564572,
      "grad_norm": 0.18107204139232635,
      "learning_rate": 9.78516141628993e-05,
      "loss": 0.0727,
      "relation_loss": 0.009778518229722977,
      "selected_pairs": 26811.7,
      "state_query_norm": 44.77510528564453,
      "state_relation_loss": 0.000649642338976264,
      "step": 4170,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.009778518229722977,
      "weighted_relation_loss": 0.009778518229722977
    },
    {
      "action_query_norm": 44.480025100708005,
      "action_relation_loss": 0.009240465937182307,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9101964452759588,
      "flow_matching_loss": 0.06367157474160194,
      "grad_norm": 0.1269778162240982,
      "learning_rate": 9.783560257794154e-05,
      "loss": 0.0736,
      "relation_loss": 0.009882145561277866,
      "selected_pairs": 26062.4,
      "state_query_norm": 44.7723445892334,
      "state_relation_loss": 0.0006416795717086643,
      "step": 4180,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.009882145561277866,
      "weighted_relation_loss": 0.009882145561277866
    },
    {
      "action_query_norm": 44.47636260986328,
      "action_relation_loss": 0.009461041027680039,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.91955098222638,
      "flow_matching_loss": 0.06407441012561321,
      "grad_norm": 0.1799089014530182,
      "learning_rate": 9.781953286821603e-05,
      "loss": 0.0742,
      "relation_loss": 0.01008949838578701,
      "selected_pairs": 26019.4,
      "state_query_norm": 44.769121170043945,
      "state_relation_loss": 0.0006284573115408421,
      "step": 4190,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.01008949838578701,
      "weighted_relation_loss": 0.01008949838578701
    },
    {
      "action_query_norm": 44.4690601348877,
      "action_relation_loss": 0.009482543263584376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9289055191768005,
      "flow_matching_loss": 0.06364769674837589,
      "grad_norm": 0.18822549283504486,
      "learning_rate": 9.780340505324901e-05,
      "loss": 0.0737,
      "relation_loss": 0.01006389344111085,
      "selected_pairs": 27131.4,
      "state_query_norm": 44.76881637573242,
      "state_relation_loss": 0.0005813503405079246,
      "step": 4200,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.01006389344111085,
      "weighted_relation_loss": 0.01006389344111085
    },
    {
      "action_query_norm": 44.46419105529785,
      "action_relation_loss": 0.00935791190713644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9382600561272216,
      "flow_matching_loss": 0.0650326233357191,
      "grad_norm": 0.11943575739860535,
      "learning_rate": 9.778721915263727e-05,
      "loss": 0.075,
      "relation_loss": 0.009958968311548234,
      "selected_pairs": 26735.3,
      "state_query_norm": 44.76153411865234,
      "state_relation_loss": 0.0006010563462041319,
      "step": 4210,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.009958968311548234,
      "weighted_relation_loss": 0.009958968311548234
    },
    {
      "action_query_norm": 44.46596908569336,
      "action_relation_loss": 0.008943013846874237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9476145930776427,
      "flow_matching_loss": 0.06127788685262203,
      "grad_norm": 0.19026072323322296,
      "learning_rate": 9.777097518604824e-05,
      "loss": 0.0709,
      "relation_loss": 0.009585418179631234,
      "selected_pairs": 26348.0,
      "state_query_norm": 44.763285064697264,
      "state_relation_loss": 0.0006424042279832065,
      "step": 4220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.009585418179631234,
      "weighted_relation_loss": 0.009585418179631234
    },
    {
      "action_query_norm": 44.46430587768555,
      "action_relation_loss": 0.008876840956509114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.956969130028064,
      "flow_matching_loss": 0.06098453365266323,
      "grad_norm": 0.21729090809822083,
      "learning_rate": 9.775467317321984e-05,
      "loss": 0.0704,
      "relation_loss": 0.009462172072380782,
      "selected_pairs": 26295.2,
      "state_query_norm": 44.766917037963864,
      "state_relation_loss": 0.0005853311915416271,
      "step": 4230,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.009462172072380782,
      "weighted_relation_loss": 0.009462172072380782
    },
    {
      "action_query_norm": 44.45735855102539,
      "action_relation_loss": 0.009296716656535864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9663236669784845,
      "flow_matching_loss": 0.06020060144364834,
      "grad_norm": 0.15933144092559814,
      "learning_rate": 9.773831313396055e-05,
      "loss": 0.0701,
      "relation_loss": 0.009940389543771744,
      "selected_pairs": 26737.3,
      "state_query_norm": 44.76322975158691,
      "state_relation_loss": 0.0006436728755943478,
      "step": 4240,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.009940389543771744,
      "weighted_relation_loss": 0.009940389543771744
    },
    {
      "action_query_norm": 44.453256607055664,
      "action_relation_loss": 0.008658955292776226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9756782039289056,
      "flow_matching_loss": 0.06242893412709236,
      "grad_norm": 0.19705018401145935,
      "learning_rate": 9.77218950881494e-05,
      "loss": 0.0717,
      "relation_loss": 0.009253784269094466,
      "selected_pairs": 26757.6,
      "state_query_norm": 44.767068481445314,
      "state_relation_loss": 0.0005948289821390062,
      "step": 4250,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.009253784269094466,
      "weighted_relation_loss": 0.009253784269094466
    },
    {
      "action_query_norm": 44.44923477172851,
      "action_relation_loss": 0.009444382740184665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9850327408793262,
      "flow_matching_loss": 0.06245002523064613,
      "grad_norm": 0.14823351800441742,
      "learning_rate": 9.770541905573583e-05,
      "loss": 0.0725,
      "relation_loss": 0.010060733929276467,
      "selected_pairs": 27221.4,
      "state_query_norm": 44.75888938903809,
      "state_relation_loss": 0.0006163511192426085,
      "step": 4260,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.010060733929276467,
      "weighted_relation_loss": 0.010060733929276467
    },
    {
      "action_query_norm": 44.4474983215332,
      "action_relation_loss": 0.008973472192883492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 3.9943872778297473,
      "flow_matching_loss": 0.06317817755043506,
      "grad_norm": 0.21489007771015167,
      "learning_rate": 9.768888505673976e-05,
      "loss": 0.0727,
      "relation_loss": 0.009538935683667659,
      "selected_pairs": 26926.1,
      "state_query_norm": 44.7634220123291,
      "state_relation_loss": 0.0005654633976519107,
      "step": 4270,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.009538935683667659,
      "weighted_relation_loss": 0.009538935683667659
    },
    {
      "action_query_norm": 44.45480461120606,
      "action_relation_loss": 0.009002350270748138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.003741814780168,
      "flow_matching_loss": 0.06187544465065002,
      "grad_norm": 0.19169530272483826,
      "learning_rate": 9.767229311125162e-05,
      "loss": 0.0715,
      "relation_loss": 0.009581542946398259,
      "selected_pairs": 24219.5,
      "state_query_norm": 44.759347915649414,
      "state_relation_loss": 0.0005791925941593945,
      "step": 4280,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.009581542946398259,
      "weighted_relation_loss": 0.009581542946398259
    },
    {
      "action_query_norm": 44.45224380493164,
      "action_relation_loss": 0.008688721433281898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.0130963517305895,
      "flow_matching_loss": 0.05916734226047993,
      "grad_norm": 0.23038411140441895,
      "learning_rate": 9.765564323943211e-05,
      "loss": 0.0685,
      "relation_loss": 0.009287677984684706,
      "selected_pairs": 27033.7,
      "state_query_norm": 44.753915786743164,
      "state_relation_loss": 0.0005989564815536141,
      "step": 4290,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.009287677984684706,
      "weighted_relation_loss": 0.009287677984684706
    },
    {
      "action_query_norm": 44.444369888305665,
      "action_relation_loss": 0.008617559168487787,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.022450888681011,
      "flow_matching_loss": 0.06097734831273556,
      "grad_norm": 0.21923932433128357,
      "learning_rate": 9.763893546151244e-05,
      "loss": 0.0702,
      "relation_loss": 0.009178998321294785,
      "selected_pairs": 27106.7,
      "state_query_norm": 44.75440216064453,
      "state_relation_loss": 0.000561439199373126,
      "step": 4300,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.009178998321294785,
      "weighted_relation_loss": 0.009178998321294785
    },
    {
      "action_query_norm": 44.44472808837891,
      "action_relation_loss": 0.008732857601717114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.031805425631431,
      "flow_matching_loss": 0.06389465443789959,
      "grad_norm": 0.20155662298202515,
      "learning_rate": 9.762216979779412e-05,
      "loss": 0.0732,
      "relation_loss": 0.009301147609949111,
      "selected_pairs": 26526.0,
      "state_query_norm": 44.7553108215332,
      "state_relation_loss": 0.0005682899965904654,
      "step": 4310,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.009301147609949111,
      "weighted_relation_loss": 0.009301147609949111
    },
    {
      "action_query_norm": 44.448992538452146,
      "action_relation_loss": 0.00929353190585971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.041159962581852,
      "flow_matching_loss": 0.06286979354918003,
      "grad_norm": 0.18210239708423615,
      "learning_rate": 9.760534626864902e-05,
      "loss": 0.0727,
      "relation_loss": 0.00987470978870988,
      "selected_pairs": 26904.8,
      "state_query_norm": 44.755977630615234,
      "state_relation_loss": 0.0005811778770294041,
      "step": 4320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00987470978870988,
      "weighted_relation_loss": 0.00987470978870988
    },
    {
      "action_query_norm": 44.43868217468262,
      "action_relation_loss": 0.008736072480678559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.050514499532273,
      "flow_matching_loss": 0.06315328292548657,
      "grad_norm": 0.15049605071544647,
      "learning_rate": 9.758846489451931e-05,
      "loss": 0.0724,
      "relation_loss": 0.009286595787853003,
      "selected_pairs": 27129.6,
      "state_query_norm": 44.748935317993165,
      "state_relation_loss": 0.0005505233362782747,
      "step": 4330,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.009286595787853003,
      "weighted_relation_loss": 0.009286595787853003
    },
    {
      "action_query_norm": 44.438916778564455,
      "action_relation_loss": 0.008793197572231293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.059869036482694,
      "flow_matching_loss": 0.05834387168288231,
      "grad_norm": 0.1527816206216812,
      "learning_rate": 9.757152569591748e-05,
      "loss": 0.0677,
      "relation_loss": 0.009367109090089799,
      "selected_pairs": 26892.2,
      "state_query_norm": 44.75033874511719,
      "state_relation_loss": 0.0005739114276366308,
      "step": 4340,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.009367109090089799,
      "weighted_relation_loss": 0.009367109090089799
    },
    {
      "action_query_norm": 44.43114013671875,
      "action_relation_loss": 0.009146018978208303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.069223573433115,
      "flow_matching_loss": 0.06161622852087021,
      "grad_norm": 0.17942702770233154,
      "learning_rate": 9.75545286934262e-05,
      "loss": 0.0713,
      "relation_loss": 0.00972856916487217,
      "selected_pairs": 26662.3,
      "state_query_norm": 44.74729919433594,
      "state_relation_loss": 0.000582550119725056,
      "step": 4350,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00972856916487217,
      "weighted_relation_loss": 0.00972856916487217
    },
    {
      "action_query_norm": 44.42569770812988,
      "action_relation_loss": 0.0087936676107347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.078578110383536,
      "flow_matching_loss": 0.061171380802989006,
      "grad_norm": 0.14303281903266907,
      "learning_rate": 9.753747390769847e-05,
      "loss": 0.0705,
      "relation_loss": 0.009367606416344643,
      "selected_pairs": 26919.9,
      "state_query_norm": 44.74500427246094,
      "state_relation_loss": 0.0005739387241192162,
      "step": 4360,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.009367606416344643,
      "weighted_relation_loss": 0.009367606416344643
    },
    {
      "action_query_norm": 44.42940559387207,
      "action_relation_loss": 0.009028889331966638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.087932647333957,
      "flow_matching_loss": 0.060187267139554024,
      "grad_norm": 0.1814042627811432,
      "learning_rate": 9.752036135945744e-05,
      "loss": 0.0698,
      "relation_loss": 0.009599499870091677,
      "selected_pairs": 26532.4,
      "state_query_norm": 44.743010711669925,
      "state_relation_loss": 0.0005706105584977195,
      "step": 4370,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.009599499870091677,
      "weighted_relation_loss": 0.009599499870091677
    },
    {
      "action_query_norm": 44.43339653015137,
      "action_relation_loss": 0.008794629899784922,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.097287184284378,
      "flow_matching_loss": 0.060745018720626834,
      "grad_norm": 0.19815506041049957,
      "learning_rate": 9.750319106949649e-05,
      "loss": 0.0702,
      "relation_loss": 0.009433153830468655,
      "selected_pairs": 26993.3,
      "state_query_norm": 44.744055557250974,
      "state_relation_loss": 0.0006385240063536912,
      "step": 4380,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.009433153830468655,
      "weighted_relation_loss": 0.009433153830468655
    },
    {
      "action_query_norm": 44.426709747314455,
      "action_relation_loss": 0.009333132999017835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.106641721234799,
      "flow_matching_loss": 0.06430534608662128,
      "grad_norm": 0.22154495120048523,
      "learning_rate": 9.748596305867913e-05,
      "loss": 0.0743,
      "relation_loss": 0.009958624187856913,
      "selected_pairs": 27151.0,
      "state_query_norm": 44.74069671630859,
      "state_relation_loss": 0.0006254912877921015,
      "step": 4390,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.009958624187856913,
      "weighted_relation_loss": 0.009958624187856913
    },
    {
      "action_query_norm": 44.41915054321289,
      "action_relation_loss": 0.008722455240786076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.11599625818522,
      "flow_matching_loss": 0.06296028755605221,
      "grad_norm": 0.16608090698719025,
      "learning_rate": 9.746867734793903e-05,
      "loss": 0.0723,
      "relation_loss": 0.009328746981918812,
      "selected_pairs": 26302.5,
      "state_query_norm": 44.738029861450194,
      "state_relation_loss": 0.0006062918051611632,
      "step": 4400,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.009328746981918812,
      "weighted_relation_loss": 0.009328746981918812
    },
    {
      "action_query_norm": 44.4186954498291,
      "action_relation_loss": 0.009086135681718588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.125350795135641,
      "flow_matching_loss": 0.060460688173770906,
      "grad_norm": 0.1689404547214508,
      "learning_rate": 9.745133395827993e-05,
      "loss": 0.0701,
      "relation_loss": 0.009683210868388414,
      "selected_pairs": 27102.3,
      "state_query_norm": 44.73851432800293,
      "state_relation_loss": 0.0005970751692075283,
      "step": 4410,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.009683210868388414,
      "weighted_relation_loss": 0.009683210868388414
    },
    {
      "action_query_norm": 44.415646362304685,
      "action_relation_loss": 0.009516546037048101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.134705332086062,
      "flow_matching_loss": 0.06055111549794674,
      "grad_norm": 0.1665840595960617,
      "learning_rate": 9.743393291077572e-05,
      "loss": 0.0707,
      "relation_loss": 0.010118338465690612,
      "selected_pairs": 26930.6,
      "state_query_norm": 44.7378231048584,
      "state_relation_loss": 0.00060179227439221,
      "step": 4420,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.010118338465690612,
      "weighted_relation_loss": 0.010118338465690612
    },
    {
      "action_query_norm": 44.4200382232666,
      "action_relation_loss": 0.008645568462088703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.144059869036482,
      "flow_matching_loss": 0.060171138122677804,
      "grad_norm": 0.14675264060497284,
      "learning_rate": 9.741647422657028e-05,
      "loss": 0.0694,
      "relation_loss": 0.009223799919709563,
      "selected_pairs": 26530.8,
      "state_query_norm": 44.73476600646973,
      "state_relation_loss": 0.0005782312306109816,
      "step": 4430,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.009223799919709563,
      "weighted_relation_loss": 0.009223799919709563
    },
    {
      "action_query_norm": 44.413483428955075,
      "action_relation_loss": 0.009065434150397778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.153414405986903,
      "flow_matching_loss": 0.059083978459239,
      "grad_norm": 0.17436249554157257,
      "learning_rate": 9.739895792687758e-05,
      "loss": 0.0687,
      "relation_loss": 0.009650598373264074,
      "selected_pairs": 26760.8,
      "state_query_norm": 44.73490753173828,
      "state_relation_loss": 0.0005851642286870629,
      "step": 4440,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.009650598373264074,
      "weighted_relation_loss": 0.009650598373264074
    },
    {
      "action_query_norm": 44.40385093688965,
      "action_relation_loss": 0.008723872527480126,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.1627689429373245,
      "flow_matching_loss": 0.05881618149578571,
      "grad_norm": 0.16392751038074493,
      "learning_rate": 9.738138403298157e-05,
      "loss": 0.0681,
      "relation_loss": 0.009286181721836328,
      "selected_pairs": 26949.0,
      "state_query_norm": 44.73393249511719,
      "state_relation_loss": 0.0005623093631584198,
      "step": 4450,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.009286181721836328,
      "weighted_relation_loss": 0.009286181721836328
    },
    {
      "action_query_norm": 44.3948787689209,
      "action_relation_loss": 0.008865694142878055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.172123479887746,
      "flow_matching_loss": 0.06315230391919613,
      "grad_norm": 0.14488714933395386,
      "learning_rate": 9.736375256623619e-05,
      "loss": 0.0726,
      "relation_loss": 0.009424334112554788,
      "selected_pairs": 27174.7,
      "state_query_norm": 44.73528137207031,
      "state_relation_loss": 0.0005586399725871161,
      "step": 4460,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.009424334112554788,
      "weighted_relation_loss": 0.009424334112554788
    },
    {
      "action_query_norm": 44.39111862182617,
      "action_relation_loss": 0.008434605924412608,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.181478016838167,
      "flow_matching_loss": 0.06259017251431942,
      "grad_norm": 0.16995033621788025,
      "learning_rate": 9.734606354806533e-05,
      "loss": 0.0716,
      "relation_loss": 0.008978092204779386,
      "selected_pairs": 26616.6,
      "state_query_norm": 44.733781814575195,
      "state_relation_loss": 0.0005434862308902666,
      "step": 4470,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.008978092204779386,
      "weighted_relation_loss": 0.008978092204779386
    },
    {
      "action_query_norm": 44.38889045715332,
      "action_relation_loss": 0.008629933325573802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.190832553788588,
      "flow_matching_loss": 0.05927357226610184,
      "grad_norm": 0.16929909586906433,
      "learning_rate": 9.73283169999628e-05,
      "loss": 0.0685,
      "relation_loss": 0.009184124041348695,
      "selected_pairs": 26854.5,
      "state_query_norm": 44.72886505126953,
      "state_relation_loss": 0.0005541906459257007,
      "step": 4480,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009184124041348695,
      "weighted_relation_loss": 0.009184124041348695
    },
    {
      "action_query_norm": 44.38515968322754,
      "action_relation_loss": 0.008391052531078459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.200187090739009,
      "flow_matching_loss": 0.05972005315124988,
      "grad_norm": 0.1671810746192932,
      "learning_rate": 9.731051294349238e-05,
      "loss": 0.0687,
      "relation_loss": 0.008958457037806511,
      "selected_pairs": 26718.2,
      "state_query_norm": 44.729618835449216,
      "state_relation_loss": 0.0005674044194165617,
      "step": 4490,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.008958457037806511,
      "weighted_relation_loss": 0.008958457037806511
    },
    {
      "action_query_norm": 44.394608688354495,
      "action_relation_loss": 0.008427533973008394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.209541627689429,
      "flow_matching_loss": 0.06084248386323452,
      "grad_norm": 0.16775131225585938,
      "learning_rate": 9.729265140028762e-05,
      "loss": 0.0698,
      "relation_loss": 0.008967610634863376,
      "selected_pairs": 26734.2,
      "state_query_norm": 44.72483329772949,
      "state_relation_loss": 0.0005400767433457077,
      "step": 4500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.008967610634863376,
      "weighted_relation_loss": 0.008967610634863376
    },
    {
      "action_query_norm": 44.39693641662598,
      "action_relation_loss": 0.0084206601139158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.21889616463985,
      "flow_matching_loss": 0.058411842212080956,
      "grad_norm": 0.15912963449954987,
      "learning_rate": 9.727473239205201e-05,
      "loss": 0.0674,
      "relation_loss": 0.00896381177008152,
      "selected_pairs": 27220.2,
      "state_query_norm": 44.72708702087402,
      "state_relation_loss": 0.0005431517958641053,
      "step": 4510,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00896381177008152,
      "weighted_relation_loss": 0.00896381177008152
    },
    {
      "action_query_norm": 44.37272338867187,
      "action_relation_loss": 0.008532963320612907,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.228250701590271,
      "flow_matching_loss": 0.06295197531580925,
      "grad_norm": 0.133110910654068,
      "learning_rate": 9.725675594055883e-05,
      "loss": 0.072,
      "relation_loss": 0.009086351469159126,
      "selected_pairs": 26849.1,
      "state_query_norm": 44.72260093688965,
      "state_relation_loss": 0.0005533881485462189,
      "step": 4520,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009086351469159126,
      "weighted_relation_loss": 0.009086351469159126
    },
    {
      "action_query_norm": 44.37254409790039,
      "action_relation_loss": 0.008932901453226805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.237605238540692,
      "flow_matching_loss": 0.061883341521024704,
      "grad_norm": 0.18515372276306152,
      "learning_rate": 9.723872206765116e-05,
      "loss": 0.0714,
      "relation_loss": 0.009505066741257905,
      "selected_pairs": 26843.8,
      "state_query_norm": 44.723287963867186,
      "state_relation_loss": 0.0005721651308704167,
      "step": 4530,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.009505066741257905,
      "weighted_relation_loss": 0.009505066741257905
    },
    {
      "action_query_norm": 44.36784439086914,
      "action_relation_loss": 0.00857691653072834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.2469597754911135,
      "flow_matching_loss": 0.06087811626493931,
      "grad_norm": 0.1878240704536438,
      "learning_rate": 9.722063079524185e-05,
      "loss": 0.07,
      "relation_loss": 0.009127980004996062,
      "selected_pairs": 26571.7,
      "state_query_norm": 44.72436447143555,
      "state_relation_loss": 0.0005510633869562298,
      "step": 4540,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.009127980004996062,
      "weighted_relation_loss": 0.009127980004996062
    },
    {
      "action_query_norm": 44.37292594909668,
      "action_relation_loss": 0.009203832689672708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.256314312441534,
      "flow_matching_loss": 0.06285121403634548,
      "grad_norm": 0.1618604063987732,
      "learning_rate": 9.720248214531351e-05,
      "loss": 0.0726,
      "relation_loss": 0.009747262857854367,
      "selected_pairs": 27244.9,
      "state_query_norm": 44.72349815368652,
      "state_relation_loss": 0.0005434301478089764,
      "step": 4550,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.009747262857854367,
      "weighted_relation_loss": 0.009747262857854367
    },
    {
      "action_query_norm": 44.37977333068848,
      "action_relation_loss": 0.00917077399790287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.265668849391955,
      "flow_matching_loss": 0.0600782074034214,
      "grad_norm": 0.19802629947662354,
      "learning_rate": 9.718427613991848e-05,
      "loss": 0.0698,
      "relation_loss": 0.009748119115829467,
      "selected_pairs": 26466.5,
      "state_query_norm": 44.72013549804687,
      "state_relation_loss": 0.0005773452401626855,
      "step": 4560,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.009748119115829467,
      "weighted_relation_loss": 0.009748119115829467
    },
    {
      "action_query_norm": 44.362457275390625,
      "action_relation_loss": 0.008240646682679653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.275023386342376,
      "flow_matching_loss": 0.061615382134914395,
      "grad_norm": 0.10688085854053497,
      "learning_rate": 9.716601280117873e-05,
      "loss": 0.0704,
      "relation_loss": 0.008824155293405056,
      "selected_pairs": 26393.0,
      "state_query_norm": 44.7194107055664,
      "state_relation_loss": 0.0005835085088619962,
      "step": 4570,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.008824155293405056,
      "weighted_relation_loss": 0.008824155293405056
    },
    {
      "action_query_norm": 44.35006942749023,
      "action_relation_loss": 0.008627076214179397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.284377923292797,
      "flow_matching_loss": 0.06093802377581596,
      "grad_norm": 0.13698184490203857,
      "learning_rate": 9.714769215128596e-05,
      "loss": 0.0702,
      "relation_loss": 0.009218303486704826,
      "selected_pairs": 27292.4,
      "state_query_norm": 44.715393447875975,
      "state_relation_loss": 0.0005912273365538568,
      "step": 4580,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.009218303486704826,
      "weighted_relation_loss": 0.009218303486704826
    },
    {
      "action_query_norm": 44.35655212402344,
      "action_relation_loss": 0.008573471009731293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.293732460243218,
      "flow_matching_loss": 0.06131633743643761,
      "grad_norm": 0.14999309182167053,
      "learning_rate": 9.712931421250152e-05,
      "loss": 0.0704,
      "relation_loss": 0.00913215707987547,
      "selected_pairs": 26826.6,
      "state_query_norm": 44.714857482910155,
      "state_relation_loss": 0.0005586861516349018,
      "step": 4590,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.00913215707987547,
      "weighted_relation_loss": 0.00913215707987547
    },
    {
      "action_query_norm": 44.35347557067871,
      "action_relation_loss": 0.008568319957703352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.303086997193639,
      "flow_matching_loss": 0.0613935437053442,
      "grad_norm": 0.14916788041591644,
      "learning_rate": 9.711087900715627e-05,
      "loss": 0.0705,
      "relation_loss": 0.009118055365979672,
      "selected_pairs": 27035.3,
      "state_query_norm": 44.71209030151367,
      "state_relation_loss": 0.0005497355043189601,
      "step": 4600,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.009118055365979672,
      "weighted_relation_loss": 0.009118055365979672
    },
    {
      "action_query_norm": 44.362090301513675,
      "action_relation_loss": 0.008772032894194126,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.31244153414406,
      "flow_matching_loss": 0.06205534674227238,
      "grad_norm": 0.13119535148143768,
      "learning_rate": 9.709238655765078e-05,
      "loss": 0.0714,
      "relation_loss": 0.009347364120185376,
      "selected_pairs": 27133.0,
      "state_query_norm": 44.712776565551756,
      "state_relation_loss": 0.0005753311328589916,
      "step": 4610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.009347364120185376,
      "weighted_relation_loss": 0.009347364120185376
    },
    {
      "action_query_norm": 44.34971466064453,
      "action_relation_loss": 0.008934357948601247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.3217960710944805,
      "flow_matching_loss": 0.05921533331274986,
      "grad_norm": 0.15949754416942596,
      "learning_rate": 9.707383688645511e-05,
      "loss": 0.0687,
      "relation_loss": 0.00952645568177104,
      "selected_pairs": 27479.4,
      "state_query_norm": 44.71306266784668,
      "state_relation_loss": 0.0005920979252550751,
      "step": 4620,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.00952645568177104,
      "weighted_relation_loss": 0.00952645568177104
    },
    {
      "action_query_norm": 44.35886192321777,
      "action_relation_loss": 0.008760489476844668,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.331150608044902,
      "flow_matching_loss": 0.061204594373703,
      "grad_norm": 0.16053755581378937,
      "learning_rate": 9.705523001610883e-05,
      "loss": 0.0706,
      "relation_loss": 0.009351794701069593,
      "selected_pairs": 27172.1,
      "state_query_norm": 44.712406158447266,
      "state_relation_loss": 0.0005913051136303693,
      "step": 4630,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.009351794701069593,
      "weighted_relation_loss": 0.009351794701069593
    },
    {
      "action_query_norm": 44.32689666748047,
      "action_relation_loss": 0.008893339056521654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.340505144995323,
      "flow_matching_loss": 0.06191463991999626,
      "grad_norm": 0.1566147655248642,
      "learning_rate": 9.703656596922107e-05,
      "loss": 0.0714,
      "relation_loss": 0.009473330853506923,
      "selected_pairs": 26385.8,
      "state_query_norm": 44.70801467895508,
      "state_relation_loss": 0.0005799918959382922,
      "step": 4640,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.009473330853506923,
      "weighted_relation_loss": 0.009473330853506923
    },
    {
      "action_query_norm": 44.33306732177734,
      "action_relation_loss": 0.008681518817320466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.349859681945744,
      "flow_matching_loss": 0.06248040869832039,
      "grad_norm": 0.16506615281105042,
      "learning_rate": 9.70178447684704e-05,
      "loss": 0.0717,
      "relation_loss": 0.009234762890264392,
      "selected_pairs": 26946.1,
      "state_query_norm": 44.70752792358398,
      "state_relation_loss": 0.0005532440030947327,
      "step": 4650,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.009234762890264392,
      "weighted_relation_loss": 0.009234762890264392
    },
    {
      "action_query_norm": 44.31895484924316,
      "action_relation_loss": 0.008966242615133523,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.359214218896165,
      "flow_matching_loss": 0.06138893105089664,
      "grad_norm": 0.13721436262130737,
      "learning_rate": 9.699906643660483e-05,
      "loss": 0.0709,
      "relation_loss": 0.009511075634509324,
      "selected_pairs": 26876.2,
      "state_query_norm": 44.70336761474609,
      "state_relation_loss": 0.0005448330863146112,
      "step": 4660,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.009511075634509324,
      "weighted_relation_loss": 0.009511075634509324
    },
    {
      "action_query_norm": 44.30943603515625,
      "action_relation_loss": 0.008411341952160001,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.368568755846586,
      "flow_matching_loss": 0.058835950866341594,
      "grad_norm": 0.18627150356769562,
      "learning_rate": 9.698023099644185e-05,
      "loss": 0.0678,
      "relation_loss": 0.008970703650265931,
      "selected_pairs": 27174.4,
      "state_query_norm": 44.70021553039551,
      "state_relation_loss": 0.0005593616690021009,
      "step": 4670,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.008970703650265931,
      "weighted_relation_loss": 0.008970703650265931
    },
    {
      "action_query_norm": 44.32429084777832,
      "action_relation_loss": 0.008404815290123224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.377923292797006,
      "flow_matching_loss": 0.062025174498558044,
      "grad_norm": 0.26194220781326294,
      "learning_rate": 9.696133847086823e-05,
      "loss": 0.071,
      "relation_loss": 0.008949615526944399,
      "selected_pairs": 26758.8,
      "state_query_norm": 44.70010185241699,
      "state_relation_loss": 0.0005448002106277272,
      "step": 4680,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.008949615526944399,
      "weighted_relation_loss": 0.008949615526944399
    },
    {
      "action_query_norm": 44.31668701171875,
      "action_relation_loss": 0.008427141420543193,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.387277829747427,
      "flow_matching_loss": 0.06255133152008056,
      "grad_norm": 0.13130992650985718,
      "learning_rate": 9.694238888284022e-05,
      "loss": 0.0715,
      "relation_loss": 0.008951335633173584,
      "selected_pairs": 27203.0,
      "state_query_norm": 44.70212593078613,
      "state_relation_loss": 0.0005241940321866423,
      "step": 4690,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.008951335633173584,
      "weighted_relation_loss": 0.008951335633173584
    },
    {
      "action_query_norm": 44.318328857421875,
      "action_relation_loss": 0.008149849250912666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.396632366697848,
      "flow_matching_loss": 0.059189239889383315,
      "grad_norm": 0.10968919098377228,
      "learning_rate": 9.692338225538333e-05,
      "loss": 0.0679,
      "relation_loss": 0.008714748825877906,
      "selected_pairs": 26989.5,
      "state_query_norm": 44.69593963623047,
      "state_relation_loss": 0.0005648995662340895,
      "step": 4700,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.008714748825877906,
      "weighted_relation_loss": 0.008714748825877906
    },
    {
      "action_query_norm": 44.31866683959961,
      "action_relation_loss": 0.00861642612144351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.4059869036482695,
      "flow_matching_loss": 0.05915864892303944,
      "grad_norm": 0.1940939724445343,
      "learning_rate": 9.690431861159241e-05,
      "loss": 0.0684,
      "relation_loss": 0.009235999453812838,
      "selected_pairs": 26391.4,
      "state_query_norm": 44.69468955993652,
      "state_relation_loss": 0.0006195733149070293,
      "step": 4710,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.009235999453812838,
      "weighted_relation_loss": 0.009235999453812838
    },
    {
      "action_query_norm": 44.329594421386716,
      "action_relation_loss": 0.008468946162611245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.415341440598691,
      "flow_matching_loss": 0.06148838996887207,
      "grad_norm": 0.1932421177625656,
      "learning_rate": 9.688519797463161e-05,
      "loss": 0.0705,
      "relation_loss": 0.00903649153187871,
      "selected_pairs": 26615.1,
      "state_query_norm": 44.69537544250488,
      "state_relation_loss": 0.0005675453459843994,
      "step": 4720,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00903649153187871,
      "weighted_relation_loss": 0.00903649153187871
    },
    {
      "action_query_norm": 44.32630233764648,
      "action_relation_loss": 0.0085566570982337,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.424695977549112,
      "flow_matching_loss": 0.056718742847442626,
      "grad_norm": 0.15519751608371735,
      "learning_rate": 9.686602036773426e-05,
      "loss": 0.0659,
      "relation_loss": 0.009152167197316884,
      "selected_pairs": 26551.4,
      "state_query_norm": 44.69639091491699,
      "state_relation_loss": 0.0005955099186394364,
      "step": 4730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.009152167197316884,
      "weighted_relation_loss": 0.009152167197316884
    },
    {
      "action_query_norm": 44.320621490478516,
      "action_relation_loss": 0.008553898520767688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.434050514499532,
      "flow_matching_loss": 0.05858287550508976,
      "grad_norm": 0.15720894932746887,
      "learning_rate": 9.684678581420302e-05,
      "loss": 0.0677,
      "relation_loss": 0.009138875640928745,
      "selected_pairs": 26685.8,
      "state_query_norm": 44.692568969726565,
      "state_relation_loss": 0.000584977213293314,
      "step": 4740,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.009138875640928745,
      "weighted_relation_loss": 0.009138875640928745
    },
    {
      "action_query_norm": 44.31777114868164,
      "action_relation_loss": 0.008624211512506008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.443405051449953,
      "flow_matching_loss": 0.05942809320986271,
      "grad_norm": 0.20909881591796875,
      "learning_rate": 9.682749433740962e-05,
      "loss": 0.0686,
      "relation_loss": 0.009196405485272407,
      "selected_pairs": 26479.2,
      "state_query_norm": 44.68649559020996,
      "state_relation_loss": 0.0005721941124647856,
      "step": 4750,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.009196405485272407,
      "weighted_relation_loss": 0.009196405485272407
    },
    {
      "action_query_norm": 44.29810638427735,
      "action_relation_loss": 0.00900779296644032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.452759588400374,
      "flow_matching_loss": 0.06169772855937481,
      "grad_norm": 0.1346811056137085,
      "learning_rate": 9.680814596079507e-05,
      "loss": 0.0713,
      "relation_loss": 0.009560128580778838,
      "selected_pairs": 26325.6,
      "state_query_norm": 44.69164276123047,
      "state_relation_loss": 0.000552335538668558,
      "step": 4760,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.009560128580778838,
      "weighted_relation_loss": 0.009560128580778838
    },
    {
      "action_query_norm": 44.30811538696289,
      "action_relation_loss": 0.008537124143913388,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.462114125350795,
      "flow_matching_loss": 0.06351399980485439,
      "grad_norm": 0.14709343016147614,
      "learning_rate": 9.678874070786945e-05,
      "loss": 0.0726,
      "relation_loss": 0.00911530777812004,
      "selected_pairs": 26525.3,
      "state_query_norm": 44.689083862304685,
      "state_relation_loss": 0.0005781836691312492,
      "step": 4770,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.00911530777812004,
      "weighted_relation_loss": 0.00911530777812004
    },
    {
      "action_query_norm": 44.314091491699216,
      "action_relation_loss": 0.0088096771389246,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.471468662301216,
      "flow_matching_loss": 0.0639607485383749,
      "grad_norm": 0.231271430850029,
      "learning_rate": 9.676927860221199e-05,
      "loss": 0.0733,
      "relation_loss": 0.009377082157880069,
      "selected_pairs": 26920.2,
      "state_query_norm": 44.68746833801269,
      "state_relation_loss": 0.0005674051586538553,
      "step": 4780,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.009377082157880069,
      "weighted_relation_loss": 0.009377082157880069
    },
    {
      "action_query_norm": 44.31471061706543,
      "action_relation_loss": 0.00878977021202445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.480823199251637,
      "flow_matching_loss": 0.06334905922412873,
      "grad_norm": 0.2204435020685196,
      "learning_rate": 9.674975966747097e-05,
      "loss": 0.0727,
      "relation_loss": 0.009361436404287816,
      "selected_pairs": 26915.8,
      "state_query_norm": 44.68417129516602,
      "state_relation_loss": 0.0005716660933103412,
      "step": 4790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.009361436404287816,
      "weighted_relation_loss": 0.009361436404287816
    },
    {
      "action_query_norm": 44.307783889770505,
      "action_relation_loss": 0.008747514896094799,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.490177736202058,
      "flow_matching_loss": 0.06268753223121167,
      "grad_norm": 0.17119720578193665,
      "learning_rate": 9.673018392736374e-05,
      "loss": 0.072,
      "relation_loss": 0.009345378447324037,
      "selected_pairs": 26436.9,
      "state_query_norm": 44.682267379760745,
      "state_relation_loss": 0.0005978633940685541,
      "step": 4800,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.009345378447324037,
      "weighted_relation_loss": 0.009345378447324037
    },
    {
      "action_query_norm": 44.30068740844727,
      "action_relation_loss": 0.008403418958187104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.499532273152479,
      "flow_matching_loss": 0.06296242512762547,
      "grad_norm": 0.2095009982585907,
      "learning_rate": 9.671055140567667e-05,
      "loss": 0.0719,
      "relation_loss": 0.008981217071413994,
      "selected_pairs": 27087.2,
      "state_query_norm": 44.68057441711426,
      "state_relation_loss": 0.0005777980666607618,
      "step": 4810,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.008981217071413994,
      "weighted_relation_loss": 0.008981217071413994
    },
    {
      "action_query_norm": 44.297922134399414,
      "action_relation_loss": 0.008794998889788986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.5088868101029,
      "flow_matching_loss": 0.059210055321455,
      "grad_norm": 0.16820713877677917,
      "learning_rate": 9.669086212626511e-05,
      "loss": 0.0686,
      "relation_loss": 0.009382979292422534,
      "selected_pairs": 27337.9,
      "state_query_norm": 44.677758407592776,
      "state_relation_loss": 0.0005879805015865713,
      "step": 4820,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.009382979292422534,
      "weighted_relation_loss": 0.009382979292422534
    },
    {
      "action_query_norm": 44.2809455871582,
      "action_relation_loss": 0.008678204799070954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.518241347053321,
      "flow_matching_loss": 0.06286666132509708,
      "grad_norm": 0.19555416703224182,
      "learning_rate": 9.667111611305341e-05,
      "loss": 0.0721,
      "relation_loss": 0.009223779663443565,
      "selected_pairs": 27014.6,
      "state_query_norm": 44.68160171508789,
      "state_relation_loss": 0.0005455749342218041,
      "step": 4830,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.009223779663443565,
      "weighted_relation_loss": 0.009223779663443565
    },
    {
      "action_query_norm": 44.290924072265625,
      "action_relation_loss": 0.008397876750677824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.527595884003742,
      "flow_matching_loss": 0.06137552559375763,
      "grad_norm": 0.16164065897464752,
      "learning_rate": 9.665131339003486e-05,
      "loss": 0.0703,
      "relation_loss": 0.008917841129004955,
      "selected_pairs": 27315.2,
      "state_query_norm": 44.67947654724121,
      "state_relation_loss": 0.0005199642677325755,
      "step": 4840,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.008917841129004955,
      "weighted_relation_loss": 0.008917841129004955
    },
    {
      "action_query_norm": 44.28368453979492,
      "action_relation_loss": 0.008898221421986818,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.536950420954163,
      "flow_matching_loss": 0.0585191298276186,
      "grad_norm": 0.17756907641887665,
      "learning_rate": 9.663145398127158e-05,
      "loss": 0.068,
      "relation_loss": 0.00943235931918025,
      "selected_pairs": 26943.8,
      "state_query_norm": 44.671793365478514,
      "state_relation_loss": 0.0005341379495803267,
      "step": 4850,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00943235931918025,
      "weighted_relation_loss": 0.00943235931918025
    },
    {
      "action_query_norm": 44.285301208496094,
      "action_relation_loss": 0.008323989901691675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.546304957904583,
      "flow_matching_loss": 0.05838267467916012,
      "grad_norm": 0.14083066582679749,
      "learning_rate": 9.661153791089467e-05,
      "loss": 0.0672,
      "relation_loss": 0.008853992167860269,
      "selected_pairs": 26363.4,
      "state_query_norm": 44.67490692138672,
      "state_relation_loss": 0.0005300025077303872,
      "step": 4860,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.008853992167860269,
      "weighted_relation_loss": 0.008853992167860269
    },
    {
      "action_query_norm": 44.27777290344238,
      "action_relation_loss": 0.008627162780612707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.555659494855004,
      "flow_matching_loss": 0.058880601078271866,
      "grad_norm": 0.13212543725967407,
      "learning_rate": 9.659156520310402e-05,
      "loss": 0.0681,
      "relation_loss": 0.009191455133259297,
      "selected_pairs": 26178.7,
      "state_query_norm": 44.67421417236328,
      "state_relation_loss": 0.0005642924283165484,
      "step": 4870,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.009191455133259297,
      "weighted_relation_loss": 0.009191455133259297
    },
    {
      "action_query_norm": 44.2738655090332,
      "action_relation_loss": 0.008329467382282019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.5650140318054255,
      "flow_matching_loss": 0.06001903004944324,
      "grad_norm": 0.16247311234474182,
      "learning_rate": 9.657153588216835e-05,
      "loss": 0.0689,
      "relation_loss": 0.008903122134506702,
      "selected_pairs": 26435.9,
      "state_query_norm": 44.66978569030762,
      "state_relation_loss": 0.000573654827894643,
      "step": 4880,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.008903122134506702,
      "weighted_relation_loss": 0.008903122134506702
    },
    {
      "action_query_norm": 44.263911437988284,
      "action_relation_loss": 0.008068238943815231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.574368568755847,
      "flow_matching_loss": 0.06133421771228313,
      "grad_norm": 0.16077832877635956,
      "learning_rate": 9.655144997242516e-05,
      "loss": 0.0699,
      "relation_loss": 0.008583759749308228,
      "selected_pairs": 27193.1,
      "state_query_norm": 44.66879501342773,
      "state_relation_loss": 0.0005155208113137632,
      "step": 4890,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.008583759749308228,
      "weighted_relation_loss": 0.008583759749308228
    },
    {
      "action_query_norm": 44.26537666320801,
      "action_relation_loss": 0.008038183627650141,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.583723105706268,
      "flow_matching_loss": 0.05929929502308369,
      "grad_norm": 0.14951717853546143,
      "learning_rate": 9.653130749828075e-05,
      "loss": 0.0679,
      "relation_loss": 0.008552493155002594,
      "selected_pairs": 27151.5,
      "state_query_norm": 44.667087936401366,
      "state_relation_loss": 0.000514309640857391,
      "step": 4900,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.008552493155002594,
      "weighted_relation_loss": 0.008552493155002594
    },
    {
      "action_query_norm": 44.26695594787598,
      "action_relation_loss": 0.00859876787289977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.593077642656689,
      "flow_matching_loss": 0.061058010533452035,
      "grad_norm": 0.1544032245874405,
      "learning_rate": 9.65111084842101e-05,
      "loss": 0.0702,
      "relation_loss": 0.009120951034128666,
      "selected_pairs": 26505.1,
      "state_query_norm": 44.66453742980957,
      "state_relation_loss": 0.000522183152497746,
      "step": 4910,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.009120951034128666,
      "weighted_relation_loss": 0.009120951034128666
    },
    {
      "action_query_norm": 44.25953979492188,
      "action_relation_loss": 0.008407329255715013,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.60243217960711,
      "flow_matching_loss": 0.060758165642619134,
      "grad_norm": 0.21154463291168213,
      "learning_rate": 9.649085295475695e-05,
      "loss": 0.0697,
      "relation_loss": 0.008910358883440495,
      "selected_pairs": 26832.7,
      "state_query_norm": 44.66914710998535,
      "state_relation_loss": 0.0005030296015320346,
      "step": 4920,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.008910358883440495,
      "weighted_relation_loss": 0.008910358883440495
    },
    {
      "action_query_norm": 44.27070693969726,
      "action_relation_loss": 0.008419192396104336,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.61178671655753,
      "flow_matching_loss": 0.05695059597492218,
      "grad_norm": 0.16888296604156494,
      "learning_rate": 9.647054093453365e-05,
      "loss": 0.0659,
      "relation_loss": 0.00892602875828743,
      "selected_pairs": 27038.6,
      "state_query_norm": 44.664939880371094,
      "state_relation_loss": 0.000506836420390755,
      "step": 4930,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00892602875828743,
      "weighted_relation_loss": 0.00892602875828743
    },
    {
      "action_query_norm": 44.256465148925784,
      "action_relation_loss": 0.008715464128181338,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.621141253507951,
      "flow_matching_loss": 0.06045155487954616,
      "grad_norm": 0.1625012904405594,
      "learning_rate": 9.645017244822123e-05,
      "loss": 0.0697,
      "relation_loss": 0.00922961886972189,
      "selected_pairs": 27036.7,
      "state_query_norm": 44.66552734375,
      "state_relation_loss": 0.0005141547764651478,
      "step": 4940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00922961886972189,
      "weighted_relation_loss": 0.00922961886972189
    },
    {
      "action_query_norm": 44.26904029846192,
      "action_relation_loss": 0.007997404597699643,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.630495790458372,
      "flow_matching_loss": 0.05950682312250137,
      "grad_norm": 0.15438248217105865,
      "learning_rate": 9.642974752056931e-05,
      "loss": 0.068,
      "relation_loss": 0.00850871168076992,
      "selected_pairs": 26715.9,
      "state_query_norm": 44.66516952514648,
      "state_relation_loss": 0.0005113070248626173,
      "step": 4950,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00850871168076992,
      "weighted_relation_loss": 0.00850871168076992
    },
    {
      "action_query_norm": 44.25282440185547,
      "action_relation_loss": 0.008463680185377598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.639850327408793,
      "flow_matching_loss": 0.06060786582529545,
      "grad_norm": 0.1562604010105133,
      "learning_rate": 9.640926617639613e-05,
      "loss": 0.0696,
      "relation_loss": 0.009002161398530006,
      "selected_pairs": 26277.7,
      "state_query_norm": 44.65870552062988,
      "state_relation_loss": 0.0005384811607655138,
      "step": 4960,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.009002161398530006,
      "weighted_relation_loss": 0.009002161398530006
    },
    {
      "action_query_norm": 44.24924163818359,
      "action_relation_loss": 0.00844856994226575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.6492048643592145,
      "flow_matching_loss": 0.05942037291824818,
      "grad_norm": 0.18847869336605072,
      "learning_rate": 9.638872844058843e-05,
      "loss": 0.0684,
      "relation_loss": 0.009007178340107202,
      "selected_pairs": 26803.0,
      "state_query_norm": 44.65931510925293,
      "state_relation_loss": 0.0005586082785157487,
      "step": 4970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.009007178340107202,
      "weighted_relation_loss": 0.009007178340107202
    },
    {
      "action_query_norm": 44.244602584838866,
      "action_relation_loss": 0.00803482080809772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.658559401309635,
      "flow_matching_loss": 0.059646034985780715,
      "grad_norm": 0.17874105274677277,
      "learning_rate": 9.63681343381015e-05,
      "loss": 0.0682,
      "relation_loss": 0.008581976313143969,
      "selected_pairs": 26901.1,
      "state_query_norm": 44.653202056884766,
      "state_relation_loss": 0.0005471553682582453,
      "step": 4980,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.008581976313143969,
      "weighted_relation_loss": 0.008581976313143969
    },
    {
      "action_query_norm": 44.251593017578124,
      "action_relation_loss": 0.008269874239340424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.667913938260056,
      "flow_matching_loss": 0.06268882155418395,
      "grad_norm": 0.21444964408874512,
      "learning_rate": 9.634748389395914e-05,
      "loss": 0.0715,
      "relation_loss": 0.00878057824447751,
      "selected_pairs": 26502.4,
      "state_query_norm": 44.650642776489256,
      "state_relation_loss": 0.0005107039323775098,
      "step": 4990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.00878057824447751,
      "weighted_relation_loss": 0.00878057824447751
    },
    {
      "action_query_norm": 44.247544860839845,
      "action_relation_loss": 0.00834330772049725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.677268475210477,
      "flow_matching_loss": 0.06033744998276234,
      "grad_norm": 0.20565660297870636,
      "learning_rate": 9.632677713325353e-05,
      "loss": 0.0692,
      "relation_loss": 0.008868380077183246,
      "selected_pairs": 27039.4,
      "state_query_norm": 44.65572280883789,
      "state_relation_loss": 0.0005250722635537386,
      "step": 5000,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.008868380077183246,
      "weighted_relation_loss": 0.008868380077183246
    },
    {
      "action_query_norm": 44.258178329467775,
      "action_relation_loss": 0.008272396121174096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.686623012160898,
      "flow_matching_loss": 0.058898765593767166,
      "grad_norm": 0.23331452906131744,
      "learning_rate": 9.63060140811454e-05,
      "loss": 0.0677,
      "relation_loss": 0.008840123098343611,
      "selected_pairs": 27536.3,
      "state_query_norm": 44.65076942443848,
      "state_relation_loss": 0.0005677269480656832,
      "step": 5010,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.008840123098343611,
      "weighted_relation_loss": 0.008840123098343611
    },
    {
      "action_query_norm": 44.242626953125,
      "action_relation_loss": 0.008780359616503119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.695977549111319,
      "flow_matching_loss": 0.05912766419351101,
      "grad_norm": 0.22962935268878937,
      "learning_rate": 9.628519476286379e-05,
      "loss": 0.0685,
      "relation_loss": 0.0093628847040236,
      "selected_pairs": 26875.0,
      "state_query_norm": 44.64540901184082,
      "state_relation_loss": 0.0005825250700581819,
      "step": 5020,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0093628847040236,
      "weighted_relation_loss": 0.0093628847040236
    },
    {
      "action_query_norm": 44.23991355895996,
      "action_relation_loss": 0.008376909373328089,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.70533208606174,
      "flow_matching_loss": 0.05909283570945263,
      "grad_norm": 0.1728779673576355,
      "learning_rate": 9.626431920370612e-05,
      "loss": 0.068,
      "relation_loss": 0.008913584426045418,
      "selected_pairs": 27154.3,
      "state_query_norm": 44.644697952270505,
      "state_relation_loss": 0.0005366750963730737,
      "step": 5030,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.008913584426045418,
      "weighted_relation_loss": 0.008913584426045418
    },
    {
      "action_query_norm": 44.239641952514646,
      "action_relation_loss": 0.008401621645316482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.714686623012161,
      "flow_matching_loss": 0.06017123199999332,
      "grad_norm": 0.21888072788715363,
      "learning_rate": 9.624338742903819e-05,
      "loss": 0.0692,
      "relation_loss": 0.0089899814222008,
      "selected_pairs": 27257.5,
      "state_query_norm": 44.64499053955078,
      "state_relation_loss": 0.0005883597070351243,
      "step": 5040,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0089899814222008,
      "weighted_relation_loss": 0.0089899814222008
    },
    {
      "action_query_norm": 44.23378791809082,
      "action_relation_loss": 0.008528594067320228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.724041159962582,
      "flow_matching_loss": 0.06065202914178371,
      "grad_norm": 0.24198387563228607,
      "learning_rate": 9.622239946429406e-05,
      "loss": 0.0697,
      "relation_loss": 0.009050773177295923,
      "selected_pairs": 26686.5,
      "state_query_norm": 44.64497299194336,
      "state_relation_loss": 0.0005221791507210582,
      "step": 5050,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.009050773177295923,
      "weighted_relation_loss": 0.009050773177295923
    },
    {
      "action_query_norm": 44.2344352722168,
      "action_relation_loss": 0.008686380041763187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.733395696913003,
      "flow_matching_loss": 0.061948714032769206,
      "grad_norm": 0.2562806308269501,
      "learning_rate": 9.620135533497609e-05,
      "loss": 0.0712,
      "relation_loss": 0.00924414126202464,
      "selected_pairs": 26639.2,
      "state_query_norm": 44.642791748046875,
      "state_relation_loss": 0.0005577612144406884,
      "step": 5060,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.00924414126202464,
      "weighted_relation_loss": 0.00924414126202464
    },
    {
      "action_query_norm": 44.22403526306152,
      "action_relation_loss": 0.008781787287443877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.742750233863424,
      "flow_matching_loss": 0.05845621526241303,
      "grad_norm": 0.15849167108535767,
      "learning_rate": 9.61802550666549e-05,
      "loss": 0.0678,
      "relation_loss": 0.009324803948402405,
      "selected_pairs": 26802.0,
      "state_query_norm": 44.6428279876709,
      "state_relation_loss": 0.0005430166842415929,
      "step": 5070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.009324803948402405,
      "weighted_relation_loss": 0.009324803948402405
    },
    {
      "action_query_norm": 44.23908233642578,
      "action_relation_loss": 0.007810530764982104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.752104770813845,
      "flow_matching_loss": 0.059096064418554306,
      "grad_norm": 0.18096210062503815,
      "learning_rate": 9.615909868496928e-05,
      "loss": 0.0675,
      "relation_loss": 0.008370291301980614,
      "selected_pairs": 27221.0,
      "state_query_norm": 44.645358657836915,
      "state_relation_loss": 0.0005597605486400425,
      "step": 5080,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008370291301980614,
      "weighted_relation_loss": 0.008370291301980614
    },
    {
      "action_query_norm": 44.21623268127441,
      "action_relation_loss": 0.00894068032503128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.761459307764266,
      "flow_matching_loss": 0.06036473847925663,
      "grad_norm": 0.1537119299173355,
      "learning_rate": 9.613788621562622e-05,
      "loss": 0.0699,
      "relation_loss": 0.009510292205959558,
      "selected_pairs": 26737.6,
      "state_query_norm": 44.63978462219238,
      "state_relation_loss": 0.0005696119129424915,
      "step": 5090,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.009510292205959558,
      "weighted_relation_loss": 0.009510292205959558
    },
    {
      "action_query_norm": 44.22448997497558,
      "action_relation_loss": 0.008422032929956913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.770813844714686,
      "flow_matching_loss": 0.06158286407589912,
      "grad_norm": 0.14892925322055817,
      "learning_rate": 9.611661768440093e-05,
      "loss": 0.0705,
      "relation_loss": 0.008939600735902786,
      "selected_pairs": 26997.2,
      "state_query_norm": 44.64187316894531,
      "state_relation_loss": 0.0005175678787054494,
      "step": 5100,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.008939600735902786,
      "weighted_relation_loss": 0.008939600735902786
    },
    {
      "action_query_norm": 44.23306083679199,
      "action_relation_loss": 0.00875004157423973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.780168381665107,
      "flow_matching_loss": 0.05731450095772743,
      "grad_norm": 0.17767523229122162,
      "learning_rate": 9.609529311713661e-05,
      "loss": 0.0666,
      "relation_loss": 0.009272908605635167,
      "selected_pairs": 27201.1,
      "state_query_norm": 44.63952522277832,
      "state_relation_loss": 0.0005228670779615641,
      "step": 5110,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.009272908605635167,
      "weighted_relation_loss": 0.009272908605635167
    },
    {
      "action_query_norm": 44.223806381225586,
      "action_relation_loss": 0.008430773904547096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.789522918615528,
      "flow_matching_loss": 0.05904968678951263,
      "grad_norm": 0.13664495944976807,
      "learning_rate": 9.607391253974466e-05,
      "loss": 0.068,
      "relation_loss": 0.00895764958113432,
      "selected_pairs": 27084.9,
      "state_query_norm": 44.63171463012695,
      "state_relation_loss": 0.0005268758512102067,
      "step": 5120,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00895764958113432,
      "weighted_relation_loss": 0.00895764958113432
    },
    {
      "action_query_norm": 44.228569412231444,
      "action_relation_loss": 0.00812761322595179,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.7988774555659495,
      "flow_matching_loss": 0.05781068466603756,
      "grad_norm": 0.135654479265213,
      "learning_rate": 9.605247597820448e-05,
      "loss": 0.0665,
      "relation_loss": 0.008645206550136208,
      "selected_pairs": 26688.3,
      "state_query_norm": 44.63320617675781,
      "state_relation_loss": 0.0005175933300051838,
      "step": 5130,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.008645206550136208,
      "weighted_relation_loss": 0.008645206550136208
    },
    {
      "action_query_norm": 44.211591720581055,
      "action_relation_loss": 0.008658537827432155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.808231992516371,
      "flow_matching_loss": 0.0590859267860651,
      "grad_norm": 0.14034371078014374,
      "learning_rate": 9.603098345856354e-05,
      "loss": 0.0683,
      "relation_loss": 0.009178223833441735,
      "selected_pairs": 26894.6,
      "state_query_norm": 44.63494453430176,
      "state_relation_loss": 0.0005196859798161313,
      "step": 5140,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.009178223833441735,
      "weighted_relation_loss": 0.009178223833441735
    },
    {
      "action_query_norm": 44.223598098754884,
      "action_relation_loss": 0.00811463724821806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.817586529466792,
      "flow_matching_loss": 0.05790490955114365,
      "grad_norm": 0.14960680902004242,
      "learning_rate": 9.600943500693725e-05,
      "loss": 0.0665,
      "relation_loss": 0.008627454377710819,
      "selected_pairs": 26767.9,
      "state_query_norm": 44.630986022949216,
      "state_relation_loss": 0.0005128172662807629,
      "step": 5150,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.008627454377710819,
      "weighted_relation_loss": 0.008627454377710819
    },
    {
      "action_query_norm": 44.2204532623291,
      "action_relation_loss": 0.0077945610973984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.826941066417213,
      "flow_matching_loss": 0.05873498134315014,
      "grad_norm": 0.15494310855865479,
      "learning_rate": 9.598783064950902e-05,
      "loss": 0.067,
      "relation_loss": 0.00830598259344697,
      "selected_pairs": 26736.7,
      "state_query_norm": 44.63420104980469,
      "state_relation_loss": 0.0005114214494824409,
      "step": 5160,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00830598259344697,
      "weighted_relation_loss": 0.00830598259344697
    },
    {
      "action_query_norm": 44.204156494140626,
      "action_relation_loss": 0.007913646707311272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.836295603367633,
      "flow_matching_loss": 0.05682870857417584,
      "grad_norm": 0.16112415492534637,
      "learning_rate": 9.596617041253018e-05,
      "loss": 0.0652,
      "relation_loss": 0.00841918382793665,
      "selected_pairs": 26747.3,
      "state_query_norm": 44.626149368286136,
      "state_relation_loss": 0.0005055371584603563,
      "step": 5170,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00841918382793665,
      "weighted_relation_loss": 0.00841918382793665
    },
    {
      "action_query_norm": 44.218848037719724,
      "action_relation_loss": 0.00844123070128262,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.845650140318054,
      "flow_matching_loss": 0.05954174250364304,
      "grad_norm": 0.12598548829555511,
      "learning_rate": 9.594445432231996e-05,
      "loss": 0.0685,
      "relation_loss": 0.00894962064921856,
      "selected_pairs": 26881.2,
      "state_query_norm": 44.622454452514646,
      "state_relation_loss": 0.000508389834431,
      "step": 5180,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00894962064921856,
      "weighted_relation_loss": 0.00894962064921856
    },
    {
      "action_query_norm": 44.20192680358887,
      "action_relation_loss": 0.00829954482614994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.855004677268475,
      "flow_matching_loss": 0.06008654125034809,
      "grad_norm": 0.1593145728111267,
      "learning_rate": 9.592268240526547e-05,
      "loss": 0.0689,
      "relation_loss": 0.00884676668792963,
      "selected_pairs": 27504.2,
      "state_query_norm": 44.62186241149902,
      "state_relation_loss": 0.0005472218472277746,
      "step": 5190,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.00884676668792963,
      "weighted_relation_loss": 0.00884676668792963
    },
    {
      "action_query_norm": 44.19908561706543,
      "action_relation_loss": 0.00860971985384822,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.864359214218896,
      "flow_matching_loss": 0.05885148085653782,
      "grad_norm": 0.13903994858264923,
      "learning_rate": 9.590085468782162e-05,
      "loss": 0.0679,
      "relation_loss": 0.00909024551510811,
      "selected_pairs": 26300.9,
      "state_query_norm": 44.62891159057617,
      "state_relation_loss": 0.00048052565543912353,
      "step": 5200,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.00909024551510811,
      "weighted_relation_loss": 0.00909024551510811
    },
    {
      "action_query_norm": 44.19598922729492,
      "action_relation_loss": 0.008017860911786556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.873713751169317,
      "flow_matching_loss": 0.05973903313279152,
      "grad_norm": 0.1287500113248825,
      "learning_rate": 9.587897119651116e-05,
      "loss": 0.0683,
      "relation_loss": 0.008530288934707642,
      "selected_pairs": 27092.1,
      "state_query_norm": 44.62654800415039,
      "state_relation_loss": 0.0005124281393364072,
      "step": 5210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.008530288934707642,
      "weighted_relation_loss": 0.008530288934707642
    },
    {
      "action_query_norm": 44.182077407836914,
      "action_relation_loss": 0.00792804709635675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.883068288119738,
      "flow_matching_loss": 0.05940866582095623,
      "grad_norm": 0.12324272841215134,
      "learning_rate": 9.585703195792459e-05,
      "loss": 0.0679,
      "relation_loss": 0.008451065234839916,
      "selected_pairs": 26572.1,
      "state_query_norm": 44.624817276000975,
      "state_relation_loss": 0.0005230181297520176,
      "step": 5220,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.008451065234839916,
      "weighted_relation_loss": 0.008451065234839916
    },
    {
      "action_query_norm": 44.18179931640625,
      "action_relation_loss": 0.008444733917713165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.892422825070159,
      "flow_matching_loss": 0.058620573952794075,
      "grad_norm": 0.1754489690065384,
      "learning_rate": 9.583503699872016e-05,
      "loss": 0.0676,
      "relation_loss": 0.008966318424791097,
      "selected_pairs": 26461.9,
      "state_query_norm": 44.616759872436525,
      "state_relation_loss": 0.0005215845361817628,
      "step": 5230,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.008966318424791097,
      "weighted_relation_loss": 0.008966318424791097
    },
    {
      "action_query_norm": 44.16719131469726,
      "action_relation_loss": 0.008454813621938229,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.90177736202058,
      "flow_matching_loss": 0.060135717689991,
      "grad_norm": 0.1795753836631775,
      "learning_rate": 9.581298634562381e-05,
      "loss": 0.0691,
      "relation_loss": 0.008990624267607928,
      "selected_pairs": 26825.0,
      "state_query_norm": 44.61704025268555,
      "state_relation_loss": 0.0005358107999200001,
      "step": 5240,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.008990624267607928,
      "weighted_relation_loss": 0.008990624267607928
    },
    {
      "action_query_norm": 44.16877822875976,
      "action_relation_loss": 0.008243430033326149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.911131898971001,
      "flow_matching_loss": 0.058799514546990395,
      "grad_norm": 0.16001452505588531,
      "learning_rate": 9.579088002542917e-05,
      "loss": 0.0676,
      "relation_loss": 0.008808355126529933,
      "selected_pairs": 27242.4,
      "state_query_norm": 44.6145320892334,
      "state_relation_loss": 0.0005649251863360405,
      "step": 5250,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.008808355126529933,
      "weighted_relation_loss": 0.008808355126529933
    },
    {
      "action_query_norm": 44.163017654418944,
      "action_relation_loss": 0.008286413131281733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.920486435921422,
      "flow_matching_loss": 0.060939985886216165,
      "grad_norm": 0.1519405096769333,
      "learning_rate": 9.57687180649975e-05,
      "loss": 0.0698,
      "relation_loss": 0.00881667472422123,
      "selected_pairs": 26881.6,
      "state_query_norm": 44.619324111938475,
      "state_relation_loss": 0.0005302616104017943,
      "step": 5260,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00881667472422123,
      "weighted_relation_loss": 0.00881667472422123
    },
    {
      "action_query_norm": 44.16362991333008,
      "action_relation_loss": 0.008026879141107201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.929840972871843,
      "flow_matching_loss": 0.05735585577785969,
      "grad_norm": 0.17241957783699036,
      "learning_rate": 9.574650049125768e-05,
      "loss": 0.0659,
      "relation_loss": 0.008535957522690296,
      "selected_pairs": 26687.0,
      "state_query_norm": 44.62096061706543,
      "state_relation_loss": 0.0005090784834465012,
      "step": 5270,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.008535957522690296,
      "weighted_relation_loss": 0.008535957522690296
    },
    {
      "action_query_norm": 44.15837020874024,
      "action_relation_loss": 0.007886184053495527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.939195509822264,
      "flow_matching_loss": 0.05899455025792122,
      "grad_norm": 0.146170511841774,
      "learning_rate": 9.572422733120614e-05,
      "loss": 0.0674,
      "relation_loss": 0.008403206057846547,
      "selected_pairs": 27336.7,
      "state_query_norm": 44.609151458740236,
      "state_relation_loss": 0.0005170222400920465,
      "step": 5280,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.008403206057846547,
      "weighted_relation_loss": 0.008403206057846547
    },
    {
      "action_query_norm": 44.165259552001956,
      "action_relation_loss": 0.008539582276716829,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.948550046772684,
      "flow_matching_loss": 0.061661620438098905,
      "grad_norm": 0.14644619822502136,
      "learning_rate": 9.570189861190689e-05,
      "loss": 0.0707,
      "relation_loss": 0.009025937132537364,
      "selected_pairs": 26801.0,
      "state_query_norm": 44.61337471008301,
      "state_relation_loss": 0.00048635479470249265,
      "step": 5290,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.009025937132537364,
      "weighted_relation_loss": 0.009025937132537364
    },
    {
      "action_query_norm": 44.17002716064453,
      "action_relation_loss": 0.008042171644046903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.9579045837231055,
      "flow_matching_loss": 0.05675834566354752,
      "grad_norm": 0.12171037495136261,
      "learning_rate": 9.56795143604914e-05,
      "loss": 0.0653,
      "relation_loss": 0.008542479854077101,
      "selected_pairs": 26387.4,
      "state_query_norm": 44.60574150085449,
      "state_relation_loss": 0.0005003082536859438,
      "step": 5300,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.008542479854077101,
      "weighted_relation_loss": 0.008542479854077101
    },
    {
      "action_query_norm": 44.153815841674806,
      "action_relation_loss": 0.008641144540160895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.967259120673527,
      "flow_matching_loss": 0.057149647176265715,
      "grad_norm": 0.14259618520736694,
      "learning_rate": 9.565707460415869e-05,
      "loss": 0.0663,
      "relation_loss": 0.009146909508854151,
      "selected_pairs": 26897.7,
      "state_query_norm": 44.60961074829102,
      "state_relation_loss": 0.0005057649308582768,
      "step": 5310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.009146909508854151,
      "weighted_relation_loss": 0.009146909508854151
    },
    {
      "action_query_norm": 44.153038787841794,
      "action_relation_loss": 0.008009305549785494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.976613657623948,
      "flow_matching_loss": 0.05584975890815258,
      "grad_norm": 0.15327191352844238,
      "learning_rate": 9.563457937017515e-05,
      "loss": 0.0643,
      "relation_loss": 0.008483346365392209,
      "selected_pairs": 26850.6,
      "state_query_norm": 44.602751922607425,
      "state_relation_loss": 0.0004740407690405846,
      "step": 5320,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008483346365392209,
      "weighted_relation_loss": 0.008483346365392209
    },
    {
      "action_query_norm": 44.16091346740723,
      "action_relation_loss": 0.008129856549203396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.985968194574369,
      "flow_matching_loss": 0.05769303143024444,
      "grad_norm": 0.12813130021095276,
      "learning_rate": 9.56120286858746e-05,
      "loss": 0.0663,
      "relation_loss": 0.00861544143408537,
      "selected_pairs": 27596.7,
      "state_query_norm": 44.606454849243164,
      "state_relation_loss": 0.00048558482667431234,
      "step": 5330,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.00861544143408537,
      "weighted_relation_loss": 0.00861544143408537
    },
    {
      "action_query_norm": 44.14870986938477,
      "action_relation_loss": 0.008718237560242415,
      "auxiliary_loss_weight": 1.0,
      "epoch": 4.99532273152479,
      "flow_matching_loss": 0.06103776581585407,
      "grad_norm": 0.13161197304725647,
      "learning_rate": 9.558942257865829e-05,
      "loss": 0.0703,
      "relation_loss": 0.009214961994439363,
      "selected_pairs": 27059.2,
      "state_query_norm": 44.602570343017575,
      "state_relation_loss": 0.0004967245098669082,
      "step": 5340,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.009214961994439363,
      "weighted_relation_loss": 0.009214961994439363
    },
    {
      "action_query_norm": 44.1424976348877,
      "action_relation_loss": 0.008021055348217487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.00467726847521,
      "flow_matching_loss": 0.05649905651807785,
      "grad_norm": 0.15919972956180573,
      "learning_rate": 9.556676107599472e-05,
      "loss": 0.0651,
      "relation_loss": 0.00855663362890482,
      "selected_pairs": 23362.0,
      "state_query_norm": 44.60202217102051,
      "state_relation_loss": 0.0005355783679988236,
      "step": 5350,
      "valid_action_samples": 210.1,
      "weighted_auxiliary_loss": 0.00855663362890482,
      "weighted_relation_loss": 0.00855663362890482
    },
    {
      "action_query_norm": 44.138419723510744,
      "action_relation_loss": 0.00793171999976039,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.014031805425631,
      "flow_matching_loss": 0.05857743993401528,
      "grad_norm": 0.16463111340999603,
      "learning_rate": 9.554404420541978e-05,
      "loss": 0.067,
      "relation_loss": 0.008472551591694355,
      "selected_pairs": 26740.2,
      "state_query_norm": 44.60019645690918,
      "state_relation_loss": 0.0005408315104432404,
      "step": 5360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.008472551591694355,
      "weighted_relation_loss": 0.008472551591694355
    },
    {
      "action_query_norm": 44.146232986450194,
      "action_relation_loss": 0.008086209651082754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.023386342376052,
      "flow_matching_loss": 0.05662926994264126,
      "grad_norm": 0.1734691560268402,
      "learning_rate": 9.55212719945366e-05,
      "loss": 0.0652,
      "relation_loss": 0.008613296272233128,
      "selected_pairs": 26661.4,
      "state_query_norm": 44.60258102416992,
      "state_relation_loss": 0.0005270866182399914,
      "step": 5370,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.008613296272233128,
      "weighted_relation_loss": 0.008613296272233128
    },
    {
      "action_query_norm": 44.14622077941895,
      "action_relation_loss": 0.007553238654509187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.032740879326473,
      "flow_matching_loss": 0.0564220804721117,
      "grad_norm": 0.19548241794109344,
      "learning_rate": 9.549844447101559e-05,
      "loss": 0.0645,
      "relation_loss": 0.008040569070726634,
      "selected_pairs": 27112.7,
      "state_query_norm": 44.59948310852051,
      "state_relation_loss": 0.00048733053845353427,
      "step": 5380,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.008040569070726634,
      "weighted_relation_loss": 0.008040569070726634
    },
    {
      "action_query_norm": 44.130619049072266,
      "action_relation_loss": 0.007940290309488774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.0420954162768945,
      "flow_matching_loss": 0.057271790131926535,
      "grad_norm": 0.18901291489601135,
      "learning_rate": 9.547556166259433e-05,
      "loss": 0.0657,
      "relation_loss": 0.008441394055262208,
      "selected_pairs": 27224.7,
      "state_query_norm": 44.594125747680664,
      "state_relation_loss": 0.0005011036788346246,
      "step": 5390,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.008441394055262208,
      "weighted_relation_loss": 0.008441394055262208
    },
    {
      "action_query_norm": 44.12945365905762,
      "action_relation_loss": 0.008436828572303056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.051449953227316,
      "flow_matching_loss": 0.05611973479390144,
      "grad_norm": 0.17584171891212463,
      "learning_rate": 9.545262359707756e-05,
      "loss": 0.0651,
      "relation_loss": 0.00894801439717412,
      "selected_pairs": 26507.0,
      "state_query_norm": 44.59790534973145,
      "state_relation_loss": 0.0005111858336022124,
      "step": 5400,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.00894801439717412,
      "weighted_relation_loss": 0.00894801439717412
    },
    {
      "action_query_norm": 44.129761123657225,
      "action_relation_loss": 0.007680081762373447,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.060804490177736,
      "flow_matching_loss": 0.05656847730278969,
      "grad_norm": 0.16217346489429474,
      "learning_rate": 9.542963030233724e-05,
      "loss": 0.0647,
      "relation_loss": 0.008166013099253178,
      "selected_pairs": 26779.8,
      "state_query_norm": 44.59459228515625,
      "state_relation_loss": 0.00048593153478577734,
      "step": 5410,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008166013099253178,
      "weighted_relation_loss": 0.008166013099253178
    },
    {
      "action_query_norm": 44.11307334899902,
      "action_relation_loss": 0.008009661408141255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.070159027128157,
      "flow_matching_loss": 0.0566330511122942,
      "grad_norm": 0.1783868670463562,
      "learning_rate": 9.540658180631237e-05,
      "loss": 0.0651,
      "relation_loss": 0.008490344975143671,
      "selected_pairs": 26569.3,
      "state_query_norm": 44.58492584228516,
      "state_relation_loss": 0.0004806835961062461,
      "step": 5420,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.008490344975143671,
      "weighted_relation_loss": 0.008490344975143671
    },
    {
      "action_query_norm": 44.13049812316895,
      "action_relation_loss": 0.007219790108501911,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.079513564078578,
      "flow_matching_loss": 0.05516791269183159,
      "grad_norm": 0.22720594704151154,
      "learning_rate": 9.538347813700904e-05,
      "loss": 0.0629,
      "relation_loss": 0.007706249784678221,
      "selected_pairs": 26971.9,
      "state_query_norm": 44.58823432922363,
      "state_relation_loss": 0.0004864597402047366,
      "step": 5430,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.007706249784678221,
      "weighted_relation_loss": 0.007706249784678221
    },
    {
      "action_query_norm": 44.10647315979004,
      "action_relation_loss": 0.008037675451487303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.088868101028999,
      "flow_matching_loss": 0.05776209197938442,
      "grad_norm": 0.19449244439601898,
      "learning_rate": 9.536031932250036e-05,
      "loss": 0.0663,
      "relation_loss": 0.008527293941006065,
      "selected_pairs": 26619.8,
      "state_query_norm": 44.589765548706055,
      "state_relation_loss": 0.0004896184575045482,
      "step": 5440,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.008527293941006065,
      "weighted_relation_loss": 0.008527293941006065
    },
    {
      "action_query_norm": 44.11114730834961,
      "action_relation_loss": 0.008055081916972994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.09822263797942,
      "flow_matching_loss": 0.05617612302303314,
      "grad_norm": 0.20847947895526886,
      "learning_rate": 9.533710539092653e-05,
      "loss": 0.0647,
      "relation_loss": 0.00854096501134336,
      "selected_pairs": 26205.6,
      "state_query_norm": 44.59085388183594,
      "state_relation_loss": 0.0004858830536250025,
      "step": 5450,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.00854096501134336,
      "weighted_relation_loss": 0.00854096501134336
    },
    {
      "action_query_norm": 44.10549049377441,
      "action_relation_loss": 0.008219106309115886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.107577174929841,
      "flow_matching_loss": 0.056797740235924724,
      "grad_norm": 0.13843649625778198,
      "learning_rate": 9.531383637049464e-05,
      "loss": 0.0655,
      "relation_loss": 0.008720156736671925,
      "selected_pairs": 26509.4,
      "state_query_norm": 44.586706924438474,
      "state_relation_loss": 0.0005010503431549296,
      "step": 5460,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.008720156736671925,
      "weighted_relation_loss": 0.008720156736671925
    },
    {
      "action_query_norm": 44.12098503112793,
      "action_relation_loss": 0.007150439824908972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.1169317118802615,
      "flow_matching_loss": 0.05904398858547211,
      "grad_norm": 0.15765373408794403,
      "learning_rate": 9.529051228947875e-05,
      "loss": 0.0667,
      "relation_loss": 0.0076436765491962435,
      "selected_pairs": 26646.4,
      "state_query_norm": 44.582188034057616,
      "state_relation_loss": 0.0004932367126457393,
      "step": 5470,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0076436765491962435,
      "weighted_relation_loss": 0.0076436765491962435
    },
    {
      "action_query_norm": 44.12167205810547,
      "action_relation_loss": 0.007403360866010189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.126286248830683,
      "flow_matching_loss": 0.05456362403929234,
      "grad_norm": 0.11901048570871353,
      "learning_rate": 9.52671331762198e-05,
      "loss": 0.0624,
      "relation_loss": 0.007852514367550611,
      "selected_pairs": 27361.6,
      "state_query_norm": 44.58390083312988,
      "state_relation_loss": 0.0004491535568377003,
      "step": 5480,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.007852514367550611,
      "weighted_relation_loss": 0.007852514367550611
    },
    {
      "action_query_norm": 44.10610580444336,
      "action_relation_loss": 0.007523299427703023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.135640785781104,
      "flow_matching_loss": 0.05697340704500675,
      "grad_norm": 0.16121447086334229,
      "learning_rate": 9.524369905912565e-05,
      "loss": 0.0649,
      "relation_loss": 0.007962117856368422,
      "selected_pairs": 26319.3,
      "state_query_norm": 44.58492660522461,
      "state_relation_loss": 0.000438818454858847,
      "step": 5490,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.007962117856368422,
      "weighted_relation_loss": 0.007962117856368422
    },
    {
      "action_query_norm": 44.112111282348636,
      "action_relation_loss": 0.007566268648952245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.144995322731525,
      "flow_matching_loss": 0.05454388670623302,
      "grad_norm": 0.15101215243339539,
      "learning_rate": 9.522020996667092e-05,
      "loss": 0.0626,
      "relation_loss": 0.008030086243525147,
      "selected_pairs": 26922.7,
      "state_query_norm": 44.58093910217285,
      "state_relation_loss": 0.00046381760912481693,
      "step": 5500,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.008030086243525147,
      "weighted_relation_loss": 0.008030086243525147
    },
    {
      "action_query_norm": 44.097638320922854,
      "action_relation_loss": 0.007778254710137844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.154349859681946,
      "flow_matching_loss": 0.05647996291518211,
      "grad_norm": 0.10527464747428894,
      "learning_rate": 9.519666592739709e-05,
      "loss": 0.0647,
      "relation_loss": 0.008242928888648748,
      "selected_pairs": 27312.2,
      "state_query_norm": 44.57475242614746,
      "state_relation_loss": 0.0004646742483600974,
      "step": 5510,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.008242928888648748,
      "weighted_relation_loss": 0.008242928888648748
    },
    {
      "action_query_norm": 44.11342658996582,
      "action_relation_loss": 0.007758952397853136,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.163704396632367,
      "flow_matching_loss": 0.05557374879717827,
      "grad_norm": 0.1535092294216156,
      "learning_rate": 9.517306696991241e-05,
      "loss": 0.0638,
      "relation_loss": 0.008233327604830264,
      "selected_pairs": 26408.9,
      "state_query_norm": 44.574900436401364,
      "state_relation_loss": 0.00047437516623176633,
      "step": 5520,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.008233327604830264,
      "weighted_relation_loss": 0.008233327604830264
    },
    {
      "action_query_norm": 44.09915885925293,
      "action_relation_loss": 0.0076206158846616745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.173058933582787,
      "flow_matching_loss": 0.057425544038414954,
      "grad_norm": 0.15126332640647888,
      "learning_rate": 9.51494131228918e-05,
      "loss": 0.0655,
      "relation_loss": 0.008116625994443894,
      "selected_pairs": 26663.8,
      "state_query_norm": 44.57641754150391,
      "state_relation_loss": 0.000496010118513368,
      "step": 5530,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.008116625994443894,
      "weighted_relation_loss": 0.008116625994443894
    },
    {
      "action_query_norm": 44.10092277526856,
      "action_relation_loss": 0.007484415313228965,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.182413470533208,
      "flow_matching_loss": 0.056301238015294075,
      "grad_norm": 0.1368400603532791,
      "learning_rate": 9.512570441507695e-05,
      "loss": 0.0643,
      "relation_loss": 0.007972338143736124,
      "selected_pairs": 27129.1,
      "state_query_norm": 44.57389373779297,
      "state_relation_loss": 0.00048792283923830836,
      "step": 5540,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.007972338143736124,
      "weighted_relation_loss": 0.007972338143736124
    },
    {
      "action_query_norm": 44.10098991394043,
      "action_relation_loss": 0.007679033325985074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.191768007483629,
      "flow_matching_loss": 0.05682532675564289,
      "grad_norm": 0.14962440729141235,
      "learning_rate": 9.510194087527615e-05,
      "loss": 0.065,
      "relation_loss": 0.008192004542797804,
      "selected_pairs": 26548.2,
      "state_query_norm": 44.56659774780273,
      "state_relation_loss": 0.0005129711906192824,
      "step": 5550,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.008192004542797804,
      "weighted_relation_loss": 0.008192004542797804
    },
    {
      "action_query_norm": 44.09565696716309,
      "action_relation_loss": 0.007725473353639245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.2011225444340505,
      "flow_matching_loss": 0.057423874735832214,
      "grad_norm": 0.16167542338371277,
      "learning_rate": 9.507812253236435e-05,
      "loss": 0.0656,
      "relation_loss": 0.008219642378389836,
      "selected_pairs": 27571.6,
      "state_query_norm": 44.570261764526364,
      "state_relation_loss": 0.0004941688908729702,
      "step": 5560,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.008219642378389836,
      "weighted_relation_loss": 0.008219642378389836
    },
    {
      "action_query_norm": 44.09221649169922,
      "action_relation_loss": 0.007645605923607945,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.210477081384472,
      "flow_matching_loss": 0.057450699433684346,
      "grad_norm": 0.18149617314338684,
      "learning_rate": 9.505424941528309e-05,
      "loss": 0.0656,
      "relation_loss": 0.00816718591377139,
      "selected_pairs": 26526.3,
      "state_query_norm": 44.567544555664064,
      "state_relation_loss": 0.0005215799959842115,
      "step": 5570,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.00816718591377139,
      "weighted_relation_loss": 0.00816718591377139
    },
    {
      "action_query_norm": 44.078913879394534,
      "action_relation_loss": 0.008200795156881213,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.219831618334893,
      "flow_matching_loss": 0.05700492113828659,
      "grad_norm": 0.157243013381958,
      "learning_rate": 9.503032155304046e-05,
      "loss": 0.0657,
      "relation_loss": 0.008668701443821192,
      "selected_pairs": 26946.0,
      "state_query_norm": 44.566341400146484,
      "state_relation_loss": 0.0004679063829826191,
      "step": 5580,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.008668701443821192,
      "weighted_relation_loss": 0.008668701443821192
    },
    {
      "action_query_norm": 44.0822151184082,
      "action_relation_loss": 0.007720569986850024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.229186155285314,
      "flow_matching_loss": 0.05857894420623779,
      "grad_norm": 0.18868926167488098,
      "learning_rate": 9.500633897471106e-05,
      "loss": 0.0668,
      "relation_loss": 0.008195242704823613,
      "selected_pairs": 27174.9,
      "state_query_norm": 44.56954078674316,
      "state_relation_loss": 0.00047467274125665424,
      "step": 5590,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.008195242704823613,
      "weighted_relation_loss": 0.008195242704823613
    },
    {
      "action_query_norm": 44.08045120239258,
      "action_relation_loss": 0.00781329027377069,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.238540692235734,
      "flow_matching_loss": 0.05758853033185005,
      "grad_norm": 0.19565355777740479,
      "learning_rate": 9.498230170943596e-05,
      "loss": 0.0659,
      "relation_loss": 0.008304927917197346,
      "selected_pairs": 26855.0,
      "state_query_norm": 44.56369590759277,
      "state_relation_loss": 0.0004916378471534699,
      "step": 5600,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008304927917197346,
      "weighted_relation_loss": 0.008304927917197346
    },
    {
      "action_query_norm": 44.079494857788085,
      "action_relation_loss": 0.007567581534385681,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.247895229186155,
      "flow_matching_loss": 0.05959913805127144,
      "grad_norm": 0.17938481271266937,
      "learning_rate": 9.495820978642275e-05,
      "loss": 0.0676,
      "relation_loss": 0.00804594405926764,
      "selected_pairs": 27106.6,
      "state_query_norm": 44.56070556640625,
      "state_relation_loss": 0.00047836248122621325,
      "step": 5610,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.00804594405926764,
      "weighted_relation_loss": 0.00804594405926764
    },
    {
      "action_query_norm": 44.0717227935791,
      "action_relation_loss": 0.007327180122956634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.257249766136576,
      "flow_matching_loss": 0.05492510125041008,
      "grad_norm": 0.165793314576149,
      "learning_rate": 9.493406323494535e-05,
      "loss": 0.0627,
      "relation_loss": 0.007809367962181568,
      "selected_pairs": 26876.7,
      "state_query_norm": 44.558001327514646,
      "state_relation_loss": 0.0004821878304937854,
      "step": 5620,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.007809367962181568,
      "weighted_relation_loss": 0.007809367962181568
    },
    {
      "action_query_norm": 44.079275894165036,
      "action_relation_loss": 0.008059040876105428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.266604303086997,
      "flow_matching_loss": 0.05567143969237805,
      "grad_norm": 0.20138201117515564,
      "learning_rate": 9.490986208434413e-05,
      "loss": 0.0642,
      "relation_loss": 0.008560617780312895,
      "selected_pairs": 26261.9,
      "state_query_norm": 44.55888633728027,
      "state_relation_loss": 0.0005015768867451698,
      "step": 5630,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.008560617780312895,
      "weighted_relation_loss": 0.008560617780312895
    },
    {
      "action_query_norm": 44.07138252258301,
      "action_relation_loss": 0.007494144747033716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.275958840037418,
      "flow_matching_loss": 0.057512130588293076,
      "grad_norm": 0.19656416773796082,
      "learning_rate": 9.488560636402577e-05,
      "loss": 0.0655,
      "relation_loss": 0.007991806091740727,
      "selected_pairs": 27202.2,
      "state_query_norm": 44.560428619384766,
      "state_relation_loss": 0.0004976613126927987,
      "step": 5640,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.007991806091740727,
      "weighted_relation_loss": 0.007991806091740727
    },
    {
      "action_query_norm": 44.063805389404294,
      "action_relation_loss": 0.007637982303276658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.285313376987839,
      "flow_matching_loss": 0.055091065168380735,
      "grad_norm": 0.16978925466537476,
      "learning_rate": 9.486129610346321e-05,
      "loss": 0.0632,
      "relation_loss": 0.008133253315463662,
      "selected_pairs": 27172.4,
      "state_query_norm": 44.55592765808105,
      "state_relation_loss": 0.0004952710616635159,
      "step": 5650,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.008133253315463662,
      "weighted_relation_loss": 0.008133253315463662
    },
    {
      "action_query_norm": 44.05714225769043,
      "action_relation_loss": 0.007836976880207657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.29466791393826,
      "flow_matching_loss": 0.058156703040003775,
      "grad_norm": 0.15500716865062714,
      "learning_rate": 9.483693133219576e-05,
      "loss": 0.0665,
      "relation_loss": 0.008325663115829229,
      "selected_pairs": 26824.9,
      "state_query_norm": 44.56164703369141,
      "state_relation_loss": 0.0004886862123385071,
      "step": 5660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.008325663115829229,
      "weighted_relation_loss": 0.008325663115829229
    },
    {
      "action_query_norm": 44.057442092895506,
      "action_relation_loss": 0.0076255563646554945,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.304022450888681,
      "flow_matching_loss": 0.059097761660814284,
      "grad_norm": 0.13702504336833954,
      "learning_rate": 9.481251207982888e-05,
      "loss": 0.0672,
      "relation_loss": 0.008093130588531495,
      "selected_pairs": 26458.3,
      "state_query_norm": 44.55420265197754,
      "state_relation_loss": 0.0004675741714891046,
      "step": 5670,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.008093130588531495,
      "weighted_relation_loss": 0.008093130588531495
    },
    {
      "action_query_norm": 44.04863662719727,
      "action_relation_loss": 0.007534305471926928,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.313376987839102,
      "flow_matching_loss": 0.05616589039564133,
      "grad_norm": 0.12529431283473969,
      "learning_rate": 9.47880383760343e-05,
      "loss": 0.0642,
      "relation_loss": 0.007991658337414265,
      "selected_pairs": 26971.7,
      "state_query_norm": 44.55230560302734,
      "state_relation_loss": 0.0004573528509354219,
      "step": 5680,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.007991658337414265,
      "weighted_relation_loss": 0.007991658337414265
    },
    {
      "action_query_norm": 44.047373580932614,
      "action_relation_loss": 0.007747919717803598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.322731524789523,
      "flow_matching_loss": 0.05518338419497013,
      "grad_norm": 0.1492680460214615,
      "learning_rate": 9.476351025054983e-05,
      "loss": 0.0634,
      "relation_loss": 0.008213713997974992,
      "selected_pairs": 27018.2,
      "state_query_norm": 44.54790687561035,
      "state_relation_loss": 0.00046579408226534726,
      "step": 5690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008213713997974992,
      "weighted_relation_loss": 0.008213713997974992
    },
    {
      "action_query_norm": 44.06197166442871,
      "action_relation_loss": 0.007690827501937747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.332086061739944,
      "flow_matching_loss": 0.05664523765444755,
      "grad_norm": 0.16806533932685852,
      "learning_rate": 9.473892773317952e-05,
      "loss": 0.0648,
      "relation_loss": 0.008179982472211123,
      "selected_pairs": 26873.0,
      "state_query_norm": 44.55258560180664,
      "state_relation_loss": 0.0004891548975137994,
      "step": 5700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.008179982472211123,
      "weighted_relation_loss": 0.008179982472211123
    },
    {
      "action_query_norm": 44.029842376708984,
      "action_relation_loss": 0.0077315786387771365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.341440598690365,
      "flow_matching_loss": 0.05403960384428501,
      "grad_norm": 0.16747891902923584,
      "learning_rate": 9.471429085379338e-05,
      "loss": 0.0623,
      "relation_loss": 0.008210837421938778,
      "selected_pairs": 26276.1,
      "state_query_norm": 44.55326156616211,
      "state_relation_loss": 0.00047925889666657897,
      "step": 5710,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.008210837421938778,
      "weighted_relation_loss": 0.008210837421938778
    },
    {
      "action_query_norm": 44.048265838623045,
      "action_relation_loss": 0.007789265038445592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.3507951356407855,
      "flow_matching_loss": 0.05411301963031292,
      "grad_norm": 0.13629819452762604,
      "learning_rate": 9.468959964232757e-05,
      "loss": 0.0624,
      "relation_loss": 0.008242757059633732,
      "selected_pairs": 27154.9,
      "state_query_norm": 44.547302627563475,
      "state_relation_loss": 0.0004534920473815873,
      "step": 5720,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.008242757059633732,
      "weighted_relation_loss": 0.008242757059633732
    },
    {
      "action_query_norm": 44.042367935180664,
      "action_relation_loss": 0.008078606892377138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.360149672591207,
      "flow_matching_loss": 0.05683009885251522,
      "grad_norm": 0.13331134617328644,
      "learning_rate": 9.466485412878425e-05,
      "loss": 0.0653,
      "relation_loss": 0.008515324210748076,
      "selected_pairs": 26553.9,
      "state_query_norm": 44.54334983825684,
      "state_relation_loss": 0.00043671735329553485,
      "step": 5730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.008515324210748076,
      "weighted_relation_loss": 0.008515324210748076
    },
    {
      "action_query_norm": 44.041735076904295,
      "action_relation_loss": 0.007573517411947251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.369504209541628,
      "flow_matching_loss": 0.05468808300793171,
      "grad_norm": 0.20532183349132538,
      "learning_rate": 9.464005434323154e-05,
      "loss": 0.0627,
      "relation_loss": 0.008031101431697607,
      "selected_pairs": 26803.3,
      "state_query_norm": 44.54201049804688,
      "state_relation_loss": 0.00045758404303342104,
      "step": 5740,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.008031101431697607,
      "weighted_relation_loss": 0.008031101431697607
    },
    {
      "action_query_norm": 44.03671455383301,
      "action_relation_loss": 0.008244621008634568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.378858746492049,
      "flow_matching_loss": 0.05677931308746338,
      "grad_norm": 0.1581134796142578,
      "learning_rate": 9.461520031580352e-05,
      "loss": 0.0655,
      "relation_loss": 0.00868772305548191,
      "selected_pairs": 27059.6,
      "state_query_norm": 44.546416473388675,
      "state_relation_loss": 0.00044310224475339053,
      "step": 5750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.00868772305548191,
      "weighted_relation_loss": 0.00868772305548191
    },
    {
      "action_query_norm": 44.03744964599609,
      "action_relation_loss": 0.007500736182555557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.38821328344247,
      "flow_matching_loss": 0.057062048092484476,
      "grad_norm": 0.15705104172229767,
      "learning_rate": 9.459029207670019e-05,
      "loss": 0.065,
      "relation_loss": 0.007980676274746657,
      "selected_pairs": 27181.1,
      "state_query_norm": 44.541939926147464,
      "state_relation_loss": 0.0004799400980118662,
      "step": 5760,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.007980676274746657,
      "weighted_relation_loss": 0.007980676274746657
    },
    {
      "action_query_norm": 44.03019561767578,
      "action_relation_loss": 0.007509702490642667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.397567820392891,
      "flow_matching_loss": 0.05723213627934456,
      "grad_norm": 0.19822436571121216,
      "learning_rate": 9.456532965618737e-05,
      "loss": 0.0652,
      "relation_loss": 0.00797896976582706,
      "selected_pairs": 26613.9,
      "state_query_norm": 44.54127655029297,
      "state_relation_loss": 0.0004692673624958843,
      "step": 5770,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00797896976582706,
      "weighted_relation_loss": 0.00797896976582706
    },
    {
      "action_query_norm": 44.02854919433594,
      "action_relation_loss": 0.007410115981474519,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.406922357343311,
      "flow_matching_loss": 0.05839674845337868,
      "grad_norm": 0.1690918207168579,
      "learning_rate": 9.454031308459681e-05,
      "loss": 0.0663,
      "relation_loss": 0.00790806943550706,
      "selected_pairs": 27269.6,
      "state_query_norm": 44.53530311584473,
      "state_relation_loss": 0.0004979535442544147,
      "step": 5780,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.00790806943550706,
      "weighted_relation_loss": 0.00790806943550706
    },
    {
      "action_query_norm": 44.034074020385745,
      "action_relation_loss": 0.007740228530019521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.416276894293732,
      "flow_matching_loss": 0.055598773434758185,
      "grad_norm": 0.15919867157936096,
      "learning_rate": 9.451524239232595e-05,
      "loss": 0.0638,
      "relation_loss": 0.008190450025722384,
      "selected_pairs": 26415.6,
      "state_query_norm": 44.53376655578613,
      "state_relation_loss": 0.00045022158592473717,
      "step": 5790,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.008190450025722384,
      "weighted_relation_loss": 0.008190450025722384
    },
    {
      "action_query_norm": 44.0165599822998,
      "action_relation_loss": 0.007035291474312544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.425631431244153,
      "flow_matching_loss": 0.056357357650995255,
      "grad_norm": 0.16336552798748016,
      "learning_rate": 9.449011760983809e-05,
      "loss": 0.0639,
      "relation_loss": 0.00751482667401433,
      "selected_pairs": 26490.2,
      "state_query_norm": 44.52896156311035,
      "state_relation_loss": 0.00047953514440450816,
      "step": 5800,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00751482667401433,
      "weighted_relation_loss": 0.00751482667401433
    },
    {
      "action_query_norm": 44.01584815979004,
      "action_relation_loss": 0.00723703415133059,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.4349859681945745,
      "flow_matching_loss": 0.05524884723126888,
      "grad_norm": 0.16435334086418152,
      "learning_rate": 9.446493876766218e-05,
      "loss": 0.063,
      "relation_loss": 0.007704639807343483,
      "selected_pairs": 27287.8,
      "state_query_norm": 44.53032760620117,
      "state_relation_loss": 0.0004676056618336588,
      "step": 5810,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.007704639807343483,
      "weighted_relation_loss": 0.007704639807343483
    },
    {
      "action_query_norm": 44.02870254516601,
      "action_relation_loss": 0.007823175424709916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.444340505144996,
      "flow_matching_loss": 0.05968627743422985,
      "grad_norm": 0.1565684974193573,
      "learning_rate": 9.44397058963929e-05,
      "loss": 0.068,
      "relation_loss": 0.008308393321931362,
      "selected_pairs": 27059.5,
      "state_query_norm": 44.534772872924805,
      "state_relation_loss": 0.00048521795542910695,
      "step": 5820,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.008308393321931362,
      "weighted_relation_loss": 0.008308393321931362
    },
    {
      "action_query_norm": 44.03500900268555,
      "action_relation_loss": 0.007616164116188884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.453695042095417,
      "flow_matching_loss": 0.05500067323446274,
      "grad_norm": 0.19516687095165253,
      "learning_rate": 9.441441902669056e-05,
      "loss": 0.0631,
      "relation_loss": 0.008094074064865708,
      "selected_pairs": 26595.7,
      "state_query_norm": 44.52711143493652,
      "state_relation_loss": 0.0004779099574079737,
      "step": 5830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.008094074064865708,
      "weighted_relation_loss": 0.008094074064865708
    },
    {
      "action_query_norm": 44.03040657043457,
      "action_relation_loss": 0.00755297471769154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.463049579045837,
      "flow_matching_loss": 0.05677470825612545,
      "grad_norm": 0.11652971804141998,
      "learning_rate": 9.43890781892811e-05,
      "loss": 0.0648,
      "relation_loss": 0.008015250694006682,
      "selected_pairs": 27014.4,
      "state_query_norm": 44.532174301147464,
      "state_relation_loss": 0.0004622759181074798,
      "step": 5840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008015250694006682,
      "weighted_relation_loss": 0.008015250694006682
    },
    {
      "action_query_norm": 44.02744483947754,
      "action_relation_loss": 0.007740182476118207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.472404115996258,
      "flow_matching_loss": 0.05502802431583405,
      "grad_norm": 0.18949733674526215,
      "learning_rate": 9.436368341495603e-05,
      "loss": 0.0632,
      "relation_loss": 0.008207641029730438,
      "selected_pairs": 26949.8,
      "state_query_norm": 44.52639465332031,
      "state_relation_loss": 0.0004674584954045713,
      "step": 5850,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.008207641029730438,
      "weighted_relation_loss": 0.008207641029730438
    },
    {
      "action_query_norm": 44.01800727844238,
      "action_relation_loss": 0.007596644852310419,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.481758652946679,
      "flow_matching_loss": 0.056259486451745035,
      "grad_norm": 0.14801648259162903,
      "learning_rate": 9.43382347345724e-05,
      "loss": 0.0643,
      "relation_loss": 0.008064033091068267,
      "selected_pairs": 27335.9,
      "state_query_norm": 44.52258796691895,
      "state_relation_loss": 0.0004673882242059335,
      "step": 5860,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.008064033091068267,
      "weighted_relation_loss": 0.008064033091068267
    },
    {
      "action_query_norm": 44.005673217773435,
      "action_relation_loss": 0.007657626643776894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.4911131898971,
      "flow_matching_loss": 0.055511243641376495,
      "grad_norm": 0.11780445277690887,
      "learning_rate": 9.431273217905273e-05,
      "loss": 0.0636,
      "relation_loss": 0.00812872899696231,
      "selected_pairs": 26948.9,
      "state_query_norm": 44.525858306884764,
      "state_relation_loss": 0.00047110234445426615,
      "step": 5870,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00812872899696231,
      "weighted_relation_loss": 0.00812872899696231
    },
    {
      "action_query_norm": 44.01648216247558,
      "action_relation_loss": 0.0072931254282593725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.500467726847521,
      "flow_matching_loss": 0.05495710223913193,
      "grad_norm": 0.171376034617424,
      "learning_rate": 9.428717577938504e-05,
      "loss": 0.0627,
      "relation_loss": 0.007746751513332128,
      "selected_pairs": 26688.4,
      "state_query_norm": 44.52122459411621,
      "state_relation_loss": 0.0004536261723842472,
      "step": 5880,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007746751513332128,
      "weighted_relation_loss": 0.007746751513332128
    },
    {
      "action_query_norm": 44.02815895080566,
      "action_relation_loss": 0.007365151913836599,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.509822263797942,
      "flow_matching_loss": 0.05703759044408798,
      "grad_norm": 0.14283797144889832,
      "learning_rate": 9.426156556662276e-05,
      "loss": 0.0648,
      "relation_loss": 0.007780860969796776,
      "selected_pairs": 26641.6,
      "state_query_norm": 44.5269157409668,
      "state_relation_loss": 0.0004157090821536258,
      "step": 5890,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.007780860969796776,
      "weighted_relation_loss": 0.007780860969796776
    },
    {
      "action_query_norm": 43.99893646240234,
      "action_relation_loss": 0.007933428138494491,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.519176800748363,
      "flow_matching_loss": 0.05642079412937164,
      "grad_norm": 0.15973617136478424,
      "learning_rate": 9.423590157188474e-05,
      "loss": 0.0648,
      "relation_loss": 0.008362584235146641,
      "selected_pairs": 26851.1,
      "state_query_norm": 44.51486320495606,
      "state_relation_loss": 0.00042915600934065876,
      "step": 5900,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.008362584235146641,
      "weighted_relation_loss": 0.008362584235146641
    },
    {
      "action_query_norm": 43.99973983764649,
      "action_relation_loss": 0.007707819528877735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.528531337698784,
      "flow_matching_loss": 0.05736280791461468,
      "grad_norm": 0.19663302600383759,
      "learning_rate": 9.421018382635513e-05,
      "loss": 0.0655,
      "relation_loss": 0.00814845347777009,
      "selected_pairs": 26630.4,
      "state_query_norm": 44.51654319763183,
      "state_relation_loss": 0.00044063398672733456,
      "step": 5910,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00814845347777009,
      "weighted_relation_loss": 0.00814845347777009
    },
    {
      "action_query_norm": 44.006268692016604,
      "action_relation_loss": 0.007553870417177677,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.537885874649205,
      "flow_matching_loss": 0.056634757295250894,
      "grad_norm": 0.15181925892829895,
      "learning_rate": 9.418441236128343e-05,
      "loss": 0.0646,
      "relation_loss": 0.007995799835771322,
      "selected_pairs": 26711.6,
      "state_query_norm": 44.52076454162598,
      "state_relation_loss": 0.00044192934001330286,
      "step": 5920,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.007995799835771322,
      "weighted_relation_loss": 0.007995799835771322
    },
    {
      "action_query_norm": 43.990858459472655,
      "action_relation_loss": 0.007500700745731592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.547240411599626,
      "flow_matching_loss": 0.05398586727678776,
      "grad_norm": 0.17182622849941254,
      "learning_rate": 9.41585872079844e-05,
      "loss": 0.0619,
      "relation_loss": 0.00792611795477569,
      "selected_pairs": 27744.1,
      "state_query_norm": 44.51334800720215,
      "state_relation_loss": 0.0004254171915818006,
      "step": 5930,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.00792611795477569,
      "weighted_relation_loss": 0.00792611795477569
    },
    {
      "action_query_norm": 43.99544105529785,
      "action_relation_loss": 0.007499447278678417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.556594948550047,
      "flow_matching_loss": 0.05505184680223465,
      "grad_norm": 0.14628566801548004,
      "learning_rate": 9.413270839783802e-05,
      "loss": 0.063,
      "relation_loss": 0.007968045491725206,
      "selected_pairs": 26741.4,
      "state_query_norm": 44.50862197875976,
      "state_relation_loss": 0.0004685982596129179,
      "step": 5940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.007968045491725206,
      "weighted_relation_loss": 0.007968045491725206
    },
    {
      "action_query_norm": 44.00515213012695,
      "action_relation_loss": 0.007688766624778509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.565949485500468,
      "flow_matching_loss": 0.057074879109859464,
      "grad_norm": 0.1391693651676178,
      "learning_rate": 9.41067759622895e-05,
      "loss": 0.0652,
      "relation_loss": 0.008126734150573611,
      "selected_pairs": 26776.6,
      "state_query_norm": 44.510520935058594,
      "state_relation_loss": 0.00043796755198854953,
      "step": 5950,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.008126734150573611,
      "weighted_relation_loss": 0.008126734150573611
    },
    {
      "action_query_norm": 44.00617218017578,
      "action_relation_loss": 0.007792674051597714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.575304022450888,
      "flow_matching_loss": 0.055615561082959175,
      "grad_norm": 0.2103864699602127,
      "learning_rate": 9.408078993284917e-05,
      "loss": 0.0639,
      "relation_loss": 0.008253970881924033,
      "selected_pairs": 26529.1,
      "state_query_norm": 44.50659866333008,
      "state_relation_loss": 0.0004612967750290409,
      "step": 5960,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.008253970881924033,
      "weighted_relation_loss": 0.008253970881924033
    },
    {
      "action_query_norm": 43.97261428833008,
      "action_relation_loss": 0.00746722356416285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.584658559401309,
      "flow_matching_loss": 0.057829960435628894,
      "grad_norm": 0.145847886800766,
      "learning_rate": 9.405475034109254e-05,
      "loss": 0.0657,
      "relation_loss": 0.00791703206487,
      "selected_pairs": 26644.7,
      "state_query_norm": 44.507655715942384,
      "state_relation_loss": 0.00044980851525906474,
      "step": 5970,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.00791703206487,
      "weighted_relation_loss": 0.00791703206487
    },
    {
      "action_query_norm": 43.9802001953125,
      "action_relation_loss": 0.007338824961334467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.5940130963517305,
      "flow_matching_loss": 0.05474920310080052,
      "grad_norm": 0.1987188458442688,
      "learning_rate": 9.402865721866015e-05,
      "loss": 0.0625,
      "relation_loss": 0.0077873004600405695,
      "selected_pairs": 26301.0,
      "state_query_norm": 44.50005187988281,
      "state_relation_loss": 0.0004484755772864446,
      "step": 5980,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0077873004600405695,
      "weighted_relation_loss": 0.0077873004600405695
    },
    {
      "action_query_norm": 44.00476760864258,
      "action_relation_loss": 0.007130998186767101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.603367633302152,
      "flow_matching_loss": 0.05436956100165844,
      "grad_norm": 0.13226206600666046,
      "learning_rate": 9.400251059725762e-05,
      "loss": 0.062,
      "relation_loss": 0.007591269351541996,
      "selected_pairs": 26803.1,
      "state_query_norm": 44.50044593811035,
      "state_relation_loss": 0.00046027111820876596,
      "step": 5990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007591269351541996,
      "weighted_relation_loss": 0.007591269351541996
    },
    {
      "action_query_norm": 43.978346252441405,
      "action_relation_loss": 0.007644092524424195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.612722170252573,
      "flow_matching_loss": 0.05598712749779224,
      "grad_norm": 0.1511753499507904,
      "learning_rate": 9.397631050865554e-05,
      "loss": 0.0641,
      "relation_loss": 0.00807322459295392,
      "selected_pairs": 26915.3,
      "state_query_norm": 44.50488471984863,
      "state_relation_loss": 0.00042913204233627764,
      "step": 6000,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00807322459295392,
      "weighted_relation_loss": 0.00807322459295392
    },
    {
      "action_query_norm": 43.98394584655762,
      "action_relation_loss": 0.007371460786089301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.622076707202994,
      "flow_matching_loss": 0.05501035638153553,
      "grad_norm": 0.16690902411937714,
      "learning_rate": 9.395005698468949e-05,
      "loss": 0.0628,
      "relation_loss": 0.007835277682170272,
      "selected_pairs": 27026.6,
      "state_query_norm": 44.494242095947264,
      "state_relation_loss": 0.0004638169426470995,
      "step": 6010,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.007835277682170272,
      "weighted_relation_loss": 0.007835277682170272
    },
    {
      "action_query_norm": 43.975325393676755,
      "action_relation_loss": 0.0074520066380500795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.631431244153415,
      "flow_matching_loss": 0.05540179572999478,
      "grad_norm": 0.2226867824792862,
      "learning_rate": 9.392375005726e-05,
      "loss": 0.0633,
      "relation_loss": 0.007880934653803707,
      "selected_pairs": 27200.2,
      "state_query_norm": 44.505043029785156,
      "state_relation_loss": 0.000428927963366732,
      "step": 6020,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.007880934653803707,
      "weighted_relation_loss": 0.007880934653803707
    },
    {
      "action_query_norm": 43.96304473876953,
      "action_relation_loss": 0.007236908655613661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.640785781103835,
      "flow_matching_loss": 0.05332280546426773,
      "grad_norm": 0.15694434940814972,
      "learning_rate": 9.389738975833242e-05,
      "loss": 0.061,
      "relation_loss": 0.007680440600961447,
      "selected_pairs": 26696.7,
      "state_query_norm": 44.4984691619873,
      "state_relation_loss": 0.0004435318725882098,
      "step": 6030,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007680440600961447,
      "weighted_relation_loss": 0.007680440600961447
    },
    {
      "action_query_norm": 43.97102661132813,
      "action_relation_loss": 0.007533442741259932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.650140318054256,
      "flow_matching_loss": 0.056073525547981264,
      "grad_norm": 0.13255195319652557,
      "learning_rate": 9.387097611993707e-05,
      "loss": 0.0641,
      "relation_loss": 0.00799925341270864,
      "selected_pairs": 26919.0,
      "state_query_norm": 44.494773864746094,
      "state_relation_loss": 0.00046581074129790066,
      "step": 6040,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00799925341270864,
      "weighted_relation_loss": 0.00799925341270864
    },
    {
      "action_query_norm": 43.96830177307129,
      "action_relation_loss": 0.007721144473180175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.659494855004677,
      "flow_matching_loss": 0.05682574920356274,
      "grad_norm": 0.11026636511087418,
      "learning_rate": 9.384450917416894e-05,
      "loss": 0.065,
      "relation_loss": 0.008170045865699648,
      "selected_pairs": 27009.5,
      "state_query_norm": 44.49930038452148,
      "state_relation_loss": 0.0004489012761041522,
      "step": 6050,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008170045865699648,
      "weighted_relation_loss": 0.008170045865699648
    },
    {
      "action_query_norm": 43.95960464477539,
      "action_relation_loss": 0.007700008386746049,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.668849391955098,
      "flow_matching_loss": 0.05706590451300144,
      "grad_norm": 0.15046018362045288,
      "learning_rate": 9.381798895318792e-05,
      "loss": 0.0652,
      "relation_loss": 0.008142413711175323,
      "selected_pairs": 26828.1,
      "state_query_norm": 44.49277877807617,
      "state_relation_loss": 0.00044240535353310404,
      "step": 6060,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.008142413711175323,
      "weighted_relation_loss": 0.008142413711175323
    },
    {
      "action_query_norm": 43.969905090332034,
      "action_relation_loss": 0.007561094872653484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.6782039289055195,
      "flow_matching_loss": 0.05424487479031086,
      "grad_norm": 0.14483270049095154,
      "learning_rate": 9.379141548921853e-05,
      "loss": 0.0623,
      "relation_loss": 0.00801080367527902,
      "selected_pairs": 27220.6,
      "state_query_norm": 44.48702163696289,
      "state_relation_loss": 0.0004497088317293674,
      "step": 6070,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00801080367527902,
      "weighted_relation_loss": 0.00801080367527902
    },
    {
      "action_query_norm": 43.94863090515137,
      "action_relation_loss": 0.007439127378165722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.68755846585594,
      "flow_matching_loss": 0.05715712159872055,
      "grad_norm": 0.1351793110370636,
      "learning_rate": 9.376478881455009e-05,
      "loss": 0.065,
      "relation_loss": 0.007885695109143854,
      "selected_pairs": 27269.6,
      "state_query_norm": 44.48666648864746,
      "state_relation_loss": 0.0004465676960535347,
      "step": 6080,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.007885695109143854,
      "weighted_relation_loss": 0.007885695109143854
    },
    {
      "action_query_norm": 43.940412521362305,
      "action_relation_loss": 0.007121992995962501,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.696913002806361,
      "flow_matching_loss": 0.05461169183254242,
      "grad_norm": 0.1672961562871933,
      "learning_rate": 9.373810896153647e-05,
      "loss": 0.0622,
      "relation_loss": 0.007564601814374327,
      "selected_pairs": 27342.2,
      "state_query_norm": 44.491024017333984,
      "state_relation_loss": 0.00044260881841182707,
      "step": 6090,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.007564601814374327,
      "weighted_relation_loss": 0.007564601814374327
    },
    {
      "action_query_norm": 43.94464340209961,
      "action_relation_loss": 0.007080580992624164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.706267539756782,
      "flow_matching_loss": 0.05655688010156155,
      "grad_norm": 0.12533599138259888,
      "learning_rate": 9.371137596259623e-05,
      "loss": 0.0641,
      "relation_loss": 0.007519886735826731,
      "selected_pairs": 27346.2,
      "state_query_norm": 44.48757781982422,
      "state_relation_loss": 0.0004393057402921841,
      "step": 6100,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.007519886735826731,
      "weighted_relation_loss": 0.007519886735826731
    },
    {
      "action_query_norm": 43.941668319702146,
      "action_relation_loss": 0.007653534132987261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.715622076707203,
      "flow_matching_loss": 0.056744784489274024,
      "grad_norm": 0.1524939239025116,
      "learning_rate": 9.368458985021248e-05,
      "loss": 0.0648,
      "relation_loss": 0.00808642990887165,
      "selected_pairs": 26277.3,
      "state_query_norm": 44.487335968017575,
      "state_relation_loss": 0.00043289576715324076,
      "step": 6110,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.00808642990887165,
      "weighted_relation_loss": 0.00808642990887165
    },
    {
      "action_query_norm": 43.9445140838623,
      "action_relation_loss": 0.007826848281547427,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.724976613657624,
      "flow_matching_loss": 0.05834831371903419,
      "grad_norm": 0.16212044656276703,
      "learning_rate": 9.365775065693287e-05,
      "loss": 0.0666,
      "relation_loss": 0.008269130252301693,
      "selected_pairs": 26673.4,
      "state_query_norm": 44.48095932006836,
      "state_relation_loss": 0.00044228195038158445,
      "step": 6120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.008269130252301693,
      "weighted_relation_loss": 0.008269130252301693
    },
    {
      "action_query_norm": 43.94439964294433,
      "action_relation_loss": 0.007443440286442638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.734331150608045,
      "flow_matching_loss": 0.05422705747187138,
      "grad_norm": 0.14697574079036713,
      "learning_rate": 9.363085841536957e-05,
      "loss": 0.0621,
      "relation_loss": 0.007876748824492097,
      "selected_pairs": 27148.0,
      "state_query_norm": 44.48177299499512,
      "state_relation_loss": 0.0004333084070822224,
      "step": 6130,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.007876748824492097,
      "weighted_relation_loss": 0.007876748824492097
    },
    {
      "action_query_norm": 43.93273582458496,
      "action_relation_loss": 0.007568924082443118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.743685687558466,
      "flow_matching_loss": 0.05472712889313698,
      "grad_norm": 0.16216464340686798,
      "learning_rate": 9.360391315819917e-05,
      "loss": 0.0628,
      "relation_loss": 0.008023258857429027,
      "selected_pairs": 27009.9,
      "state_query_norm": 44.47840194702148,
      "state_relation_loss": 0.00045433466148097067,
      "step": 6140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.008023258857429027,
      "weighted_relation_loss": 0.008023258857429027
    },
    {
      "action_query_norm": 43.92794532775879,
      "action_relation_loss": 0.007586087146773934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.7530402245088865,
      "flow_matching_loss": 0.054661748185753825,
      "grad_norm": 0.1482306569814682,
      "learning_rate": 9.357691491816269e-05,
      "loss": 0.0627,
      "relation_loss": 0.008022965071722865,
      "selected_pairs": 27107.1,
      "state_query_norm": 44.482037734985354,
      "state_relation_loss": 0.00043687793076969684,
      "step": 6150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.008022965071722865,
      "weighted_relation_loss": 0.008022965071722865
    },
    {
      "action_query_norm": 43.92081985473633,
      "action_relation_loss": 0.007292323932051659,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.762394761459308,
      "flow_matching_loss": 0.05516010597348213,
      "grad_norm": 0.14981408417224884,
      "learning_rate": 9.354986372806557e-05,
      "loss": 0.0629,
      "relation_loss": 0.007727259211242199,
      "selected_pairs": 26159.3,
      "state_query_norm": 44.47948303222656,
      "state_relation_loss": 0.0004349353926954791,
      "step": 6160,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.007727259211242199,
      "weighted_relation_loss": 0.007727259211242199
    },
    {
      "action_query_norm": 43.90093574523926,
      "action_relation_loss": 0.007551247114315629,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.771749298409729,
      "flow_matching_loss": 0.05476984865963459,
      "grad_norm": 0.14673902094364166,
      "learning_rate": 9.352275962077752e-05,
      "loss": 0.0628,
      "relation_loss": 0.007990573113784195,
      "selected_pairs": 26877.3,
      "state_query_norm": 44.47135925292969,
      "state_relation_loss": 0.0004393259616335854,
      "step": 6170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.007990573113784195,
      "weighted_relation_loss": 0.007990573113784195
    },
    {
      "action_query_norm": 43.91944046020508,
      "action_relation_loss": 0.007480135373771191,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.78110383536015,
      "flow_matching_loss": 0.05580783747136593,
      "grad_norm": 0.169485405087471,
      "learning_rate": 9.349560262923262e-05,
      "loss": 0.0637,
      "relation_loss": 0.007900180807337166,
      "selected_pairs": 26895.0,
      "state_query_norm": 44.473680114746095,
      "state_relation_loss": 0.0004200453869998455,
      "step": 6180,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007900180807337166,
      "weighted_relation_loss": 0.007900180807337166
    },
    {
      "action_query_norm": 43.92122955322266,
      "action_relation_loss": 0.0072084758430719376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.790458372310571,
      "flow_matching_loss": 0.05315229296684265,
      "grad_norm": 0.1564217209815979,
      "learning_rate": 9.346839278642913e-05,
      "loss": 0.0608,
      "relation_loss": 0.007635163469240069,
      "selected_pairs": 26651.4,
      "state_query_norm": 44.470909118652344,
      "state_relation_loss": 0.00042668755631893875,
      "step": 6190,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.007635163469240069,
      "weighted_relation_loss": 0.007635163469240069
    },
    {
      "action_query_norm": 43.93005256652832,
      "action_relation_loss": 0.007382578123360873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.799812909260991,
      "flow_matching_loss": 0.054762277752161026,
      "grad_norm": 0.13730792701244354,
      "learning_rate": 9.344113012542963e-05,
      "loss": 0.0626,
      "relation_loss": 0.007817445648834109,
      "selected_pairs": 26765.2,
      "state_query_norm": 44.47252159118652,
      "state_relation_loss": 0.00043486755748745054,
      "step": 6200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.007817445648834109,
      "weighted_relation_loss": 0.007817445648834109
    },
    {
      "action_query_norm": 43.92130012512207,
      "action_relation_loss": 0.007366056879982352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.809167446211412,
      "flow_matching_loss": 0.05621778629720211,
      "grad_norm": 0.16025537252426147,
      "learning_rate": 9.341381467936079e-05,
      "loss": 0.064,
      "relation_loss": 0.007820208836346865,
      "selected_pairs": 27359.0,
      "state_query_norm": 44.4657341003418,
      "state_relation_loss": 0.00045415191270876677,
      "step": 6210,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.007820208836346865,
      "weighted_relation_loss": 0.007820208836346865
    },
    {
      "action_query_norm": 43.925378036499026,
      "action_relation_loss": 0.00731885964050889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.818521983161833,
      "flow_matching_loss": 0.054070330783724786,
      "grad_norm": 0.16988830268383026,
      "learning_rate": 9.338644648141346e-05,
      "loss": 0.0618,
      "relation_loss": 0.007753421925008297,
      "selected_pairs": 26343.6,
      "state_query_norm": 44.472136688232425,
      "state_relation_loss": 0.0004345622641267255,
      "step": 6220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.007753421925008297,
      "weighted_relation_loss": 0.007753421925008297
    },
    {
      "action_query_norm": 43.91552543640137,
      "action_relation_loss": 0.007333415327593684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.827876520112254,
      "flow_matching_loss": 0.0551891565322876,
      "grad_norm": 0.16348019242286682,
      "learning_rate": 9.335902556484257e-05,
      "loss": 0.063,
      "relation_loss": 0.0077699556481093165,
      "selected_pairs": 27335.0,
      "state_query_norm": 44.46388893127441,
      "state_relation_loss": 0.0004365403874544427,
      "step": 6230,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0077699556481093165,
      "weighted_relation_loss": 0.0077699556481093165
    },
    {
      "action_query_norm": 43.90432968139648,
      "action_relation_loss": 0.007378217903897167,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.8372310570626755,
      "flow_matching_loss": 0.05573452934622765,
      "grad_norm": 0.14078225195407867,
      "learning_rate": 9.333155196296712e-05,
      "loss": 0.0635,
      "relation_loss": 0.007805885374546051,
      "selected_pairs": 26590.5,
      "state_query_norm": 44.46761131286621,
      "state_relation_loss": 0.0004276674007996917,
      "step": 6240,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.007805885374546051,
      "weighted_relation_loss": 0.007805885374546051
    },
    {
      "action_query_norm": 43.91507110595703,
      "action_relation_loss": 0.0072377577889710665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.846585594013097,
      "flow_matching_loss": 0.056384599581360814,
      "grad_norm": 0.18960203230381012,
      "learning_rate": 9.330402570917016e-05,
      "loss": 0.0641,
      "relation_loss": 0.007695099525153637,
      "selected_pairs": 27206.4,
      "state_query_norm": 44.46240653991699,
      "state_relation_loss": 0.0004573416488710791,
      "step": 6250,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.007695099525153637,
      "weighted_relation_loss": 0.007695099525153637
    },
    {
      "action_query_norm": 43.90101509094238,
      "action_relation_loss": 0.007360346568748355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.855940130963518,
      "flow_matching_loss": 0.05633379258215428,
      "grad_norm": 0.18315966427326202,
      "learning_rate": 9.327644683689865e-05,
      "loss": 0.0641,
      "relation_loss": 0.007786556612700224,
      "selected_pairs": 26826.8,
      "state_query_norm": 44.46368179321289,
      "state_relation_loss": 0.00042621005268301816,
      "step": 6260,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.007786556612700224,
      "weighted_relation_loss": 0.007786556612700224
    },
    {
      "action_query_norm": 43.90380363464355,
      "action_relation_loss": 0.007316926773637533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.865294667913938,
      "flow_matching_loss": 0.05410900674760342,
      "grad_norm": 0.16775421798229218,
      "learning_rate": 9.324881537966354e-05,
      "loss": 0.0619,
      "relation_loss": 0.007754534715786576,
      "selected_pairs": 26804.9,
      "state_query_norm": 44.46166763305664,
      "state_relation_loss": 0.00043760791304521264,
      "step": 6270,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007754534715786576,
      "weighted_relation_loss": 0.007754534715786576
    },
    {
      "action_query_norm": 43.902290725708006,
      "action_relation_loss": 0.007557915756478906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.874649204864359,
      "flow_matching_loss": 0.055378411710262296,
      "grad_norm": 0.16097897291183472,
      "learning_rate": 9.322113137103964e-05,
      "loss": 0.0634,
      "relation_loss": 0.008004827052354812,
      "selected_pairs": 26894.4,
      "state_query_norm": 44.46535949707031,
      "state_relation_loss": 0.0004469113191589713,
      "step": 6280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.008004827052354812,
      "weighted_relation_loss": 0.008004827052354812
    },
    {
      "action_query_norm": 43.88373222351074,
      "action_relation_loss": 0.007515379134565592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.88400374181478,
      "flow_matching_loss": 0.05653202570974827,
      "grad_norm": 0.12967325747013092,
      "learning_rate": 9.319339484466564e-05,
      "loss": 0.0645,
      "relation_loss": 0.0079400684684515,
      "selected_pairs": 26984.7,
      "state_query_norm": 44.45534248352051,
      "state_relation_loss": 0.00042468945903237907,
      "step": 6290,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0079400684684515,
      "weighted_relation_loss": 0.0079400684684515
    },
    {
      "action_query_norm": 43.91059837341309,
      "action_relation_loss": 0.0073853507172316315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.893358278765201,
      "flow_matching_loss": 0.05415202230215073,
      "grad_norm": 0.1326739490032196,
      "learning_rate": 9.316560583424403e-05,
      "loss": 0.062,
      "relation_loss": 0.007801257539540529,
      "selected_pairs": 26828.0,
      "state_query_norm": 44.45268592834473,
      "state_relation_loss": 0.0004159069008892402,
      "step": 6300,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.007801257539540529,
      "weighted_relation_loss": 0.007801257539540529
    },
    {
      "action_query_norm": 43.90029716491699,
      "action_relation_loss": 0.007345726899802685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.902712815715622,
      "flow_matching_loss": 0.05436924546957016,
      "grad_norm": 0.17393502593040466,
      "learning_rate": 9.313776437354108e-05,
      "loss": 0.0621,
      "relation_loss": 0.007760784216225147,
      "selected_pairs": 27181.5,
      "state_query_norm": 44.45618591308594,
      "state_relation_loss": 0.0004150573280639946,
      "step": 6310,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.007760784216225147,
      "weighted_relation_loss": 0.007760784216225147
    },
    {
      "action_query_norm": 43.893819427490236,
      "action_relation_loss": 0.006991861434653401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.9120673526660426,
      "flow_matching_loss": 0.053121186420321466,
      "grad_norm": 0.14313940703868866,
      "learning_rate": 9.31098704963868e-05,
      "loss": 0.0606,
      "relation_loss": 0.007434856565669179,
      "selected_pairs": 26965.1,
      "state_query_norm": 44.45252799987793,
      "state_relation_loss": 0.0004429951310157776,
      "step": 6320,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.007434856565669179,
      "weighted_relation_loss": 0.007434856565669179
    },
    {
      "action_query_norm": 43.8891300201416,
      "action_relation_loss": 0.007201653020456433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.921421889616464,
      "flow_matching_loss": 0.05299583934247494,
      "grad_norm": 0.15657047927379608,
      "learning_rate": 9.308192423667485e-05,
      "loss": 0.0606,
      "relation_loss": 0.00763589758425951,
      "selected_pairs": 26893.4,
      "state_query_norm": 44.4513126373291,
      "state_relation_loss": 0.0004342444823123515,
      "step": 6330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00763589758425951,
      "weighted_relation_loss": 0.00763589758425951
    },
    {
      "action_query_norm": 43.87865295410156,
      "action_relation_loss": 0.007099356455728412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.930776426566885,
      "flow_matching_loss": 0.055472719669342044,
      "grad_norm": 0.1261722445487976,
      "learning_rate": 9.305392562836262e-05,
      "loss": 0.063,
      "relation_loss": 0.007529757218435407,
      "selected_pairs": 27224.4,
      "state_query_norm": 44.45069007873535,
      "state_relation_loss": 0.0004304006870370358,
      "step": 6340,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.007529757218435407,
      "weighted_relation_loss": 0.007529757218435407
    },
    {
      "action_query_norm": 43.88233757019043,
      "action_relation_loss": 0.0072092027869075535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.940130963517306,
      "flow_matching_loss": 0.05375952646136284,
      "grad_norm": 0.2015656679868698,
      "learning_rate": 9.3025874705471e-05,
      "loss": 0.0614,
      "relation_loss": 0.007661747792735696,
      "selected_pairs": 26664.6,
      "state_query_norm": 44.44876899719238,
      "state_relation_loss": 0.0004525451571680605,
      "step": 6350,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.007661747792735696,
      "weighted_relation_loss": 0.007661747792735696
    },
    {
      "action_query_norm": 43.85608558654785,
      "action_relation_loss": 0.00678458851762116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.949485500467727,
      "flow_matching_loss": 0.053072234988212584,
      "grad_norm": 0.16810235381126404,
      "learning_rate": 9.299777150208457e-05,
      "loss": 0.0603,
      "relation_loss": 0.007220765203237533,
      "selected_pairs": 26341.9,
      "state_query_norm": 44.4464729309082,
      "state_relation_loss": 0.000436176749644801,
      "step": 6360,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.007220765203237533,
      "weighted_relation_loss": 0.007220765203237533
    },
    {
      "action_query_norm": 43.85762748718262,
      "action_relation_loss": 0.007458673836663365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.958840037418148,
      "flow_matching_loss": 0.05536248423159122,
      "grad_norm": 0.15828993916511536,
      "learning_rate": 9.296961605235133e-05,
      "loss": 0.0633,
      "relation_loss": 0.007894319761544467,
      "selected_pairs": 27452.1,
      "state_query_norm": 44.44828834533691,
      "state_relation_loss": 0.0004356457735411823,
      "step": 6370,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.007894319761544467,
      "weighted_relation_loss": 0.007894319761544467
    },
    {
      "action_query_norm": 43.8559009552002,
      "action_relation_loss": 0.007411936996504664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.968194574368569,
      "flow_matching_loss": 0.05484120883047581,
      "grad_norm": 0.1795727014541626,
      "learning_rate": 9.29414083904828e-05,
      "loss": 0.0627,
      "relation_loss": 0.007856181403622032,
      "selected_pairs": 27312.7,
      "state_query_norm": 44.44565353393555,
      "state_relation_loss": 0.0004442444798769429,
      "step": 6380,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.007856181403622032,
      "weighted_relation_loss": 0.007856181403622032
    },
    {
      "action_query_norm": 43.86047172546387,
      "action_relation_loss": 0.007123528374359011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.977549111318989,
      "flow_matching_loss": 0.05746663510799408,
      "grad_norm": 0.1650300770998001,
      "learning_rate": 9.291314855075397e-05,
      "loss": 0.065,
      "relation_loss": 0.007573900418356061,
      "selected_pairs": 26806.3,
      "state_query_norm": 44.43970565795898,
      "state_relation_loss": 0.00045037200034130364,
      "step": 6390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007573900418356061,
      "weighted_relation_loss": 0.007573900418356061
    },
    {
      "action_query_norm": 43.877149200439455,
      "action_relation_loss": 0.0070257417391985655,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.9869036482694105,
      "flow_matching_loss": 0.056511102989315984,
      "grad_norm": 0.1299808919429779,
      "learning_rate": 9.288483656750322e-05,
      "loss": 0.064,
      "relation_loss": 0.007466092752292752,
      "selected_pairs": 26798.4,
      "state_query_norm": 44.44053688049316,
      "state_relation_loss": 0.0004403510625706986,
      "step": 6400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.007466092752292752,
      "weighted_relation_loss": 0.007466092752292752
    },
    {
      "action_query_norm": 43.85508651733399,
      "action_relation_loss": 0.007237413804978133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 5.996258185219832,
      "flow_matching_loss": 0.05457322858273983,
      "grad_norm": 0.1374255120754242,
      "learning_rate": 9.285647247513225e-05,
      "loss": 0.0623,
      "relation_loss": 0.007698640553280711,
      "selected_pairs": 26619.8,
      "state_query_norm": 44.44068717956543,
      "state_relation_loss": 0.0004612268006894737,
      "step": 6410,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.007698640553280711,
      "weighted_relation_loss": 0.007698640553280711
    },
    {
      "action_query_norm": 43.84962539672851,
      "action_relation_loss": 0.0068875238299369815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.005612722170253,
      "flow_matching_loss": 0.052909047529101375,
      "grad_norm": 0.15499746799468994,
      "learning_rate": 9.282805630810614e-05,
      "loss": 0.0602,
      "relation_loss": 0.007332990178838372,
      "selected_pairs": 23926.7,
      "state_query_norm": 44.44168472290039,
      "state_relation_loss": 0.00044546643330249934,
      "step": 6420,
      "valid_action_samples": 212.7,
      "weighted_auxiliary_loss": 0.007332990178838372,
      "weighted_relation_loss": 0.007332990178838372
    },
    {
      "action_query_norm": 43.84440116882324,
      "action_relation_loss": 0.007339551858603954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.014967259120674,
      "flow_matching_loss": 0.05432727113366127,
      "grad_norm": 0.14830809831619263,
      "learning_rate": 9.279958810095317e-05,
      "loss": 0.0621,
      "relation_loss": 0.007798638474196195,
      "selected_pairs": 26690.3,
      "state_query_norm": 44.44087295532226,
      "state_relation_loss": 0.0004590866941725835,
      "step": 6430,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007798638474196195,
      "weighted_relation_loss": 0.007798638474196195
    },
    {
      "action_query_norm": 43.84785346984863,
      "action_relation_loss": 0.006819902127608657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.024321796071095,
      "flow_matching_loss": 0.05294747725129127,
      "grad_norm": 0.1569676548242569,
      "learning_rate": 9.277106788826494e-05,
      "loss": 0.0602,
      "relation_loss": 0.007260861434042453,
      "selected_pairs": 26732.6,
      "state_query_norm": 44.43118896484375,
      "state_relation_loss": 0.00044095939956605433,
      "step": 6440,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.007260861434042453,
      "weighted_relation_loss": 0.007260861434042453
    },
    {
      "action_query_norm": 43.83556137084961,
      "action_relation_loss": 0.006784666376188397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.033676333021515,
      "flow_matching_loss": 0.05316493473947048,
      "grad_norm": 0.10503397136926651,
      "learning_rate": 9.274249570469618e-05,
      "loss": 0.0604,
      "relation_loss": 0.007243361743167043,
      "selected_pairs": 27291.4,
      "state_query_norm": 44.43891716003418,
      "state_relation_loss": 0.0004586953698890284,
      "step": 6450,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.007243361743167043,
      "weighted_relation_loss": 0.007243361743167043
    },
    {
      "action_query_norm": 43.840915298461915,
      "action_relation_loss": 0.006798975542187691,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.043030869971936,
      "flow_matching_loss": 0.055414341017603876,
      "grad_norm": 0.15676908195018768,
      "learning_rate": 9.271387158496476e-05,
      "loss": 0.0627,
      "relation_loss": 0.007245629094541073,
      "selected_pairs": 27152.3,
      "state_query_norm": 44.43293495178223,
      "state_relation_loss": 0.0004466535552637652,
      "step": 6460,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.007245629094541073,
      "weighted_relation_loss": 0.007245629094541073
    },
    {
      "action_query_norm": 43.83823928833008,
      "action_relation_loss": 0.0068169163540005686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.052385406922357,
      "flow_matching_loss": 0.05302614979445934,
      "grad_norm": 0.1635921746492386,
      "learning_rate": 9.268519556385173e-05,
      "loss": 0.0603,
      "relation_loss": 0.007225350057706237,
      "selected_pairs": 27086.6,
      "state_query_norm": 44.43741302490234,
      "state_relation_loss": 0.00040843366587068887,
      "step": 6470,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.007225350057706237,
      "weighted_relation_loss": 0.007225350057706237
    },
    {
      "action_query_norm": 43.84207725524902,
      "action_relation_loss": 0.006837739795446396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.061739943872778,
      "flow_matching_loss": 0.0535694133490324,
      "grad_norm": 0.1380642056465149,
      "learning_rate": 9.265646767620113e-05,
      "loss": 0.0608,
      "relation_loss": 0.0072524779941886665,
      "selected_pairs": 27080.1,
      "state_query_norm": 44.43259735107422,
      "state_relation_loss": 0.00041473821038380266,
      "step": 6480,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0072524779941886665,
      "weighted_relation_loss": 0.0072524779941886665
    },
    {
      "action_query_norm": 43.84740791320801,
      "action_relation_loss": 0.006717267818748951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.0710944808231995,
      "flow_matching_loss": 0.05310473032295704,
      "grad_norm": 0.1564960926771164,
      "learning_rate": 9.262768795692006e-05,
      "loss": 0.0602,
      "relation_loss": 0.007128100888803601,
      "selected_pairs": 26996.7,
      "state_query_norm": 44.43745880126953,
      "state_relation_loss": 0.0004108330642338842,
      "step": 6490,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.007128100888803601,
      "weighted_relation_loss": 0.007128100888803601
    },
    {
      "action_query_norm": 43.851495361328126,
      "action_relation_loss": 0.006630136957392096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.080449017773621,
      "flow_matching_loss": 0.05347622223198414,
      "grad_norm": 0.1770225614309311,
      "learning_rate": 9.259885644097861e-05,
      "loss": 0.0605,
      "relation_loss": 0.007062780950218439,
      "selected_pairs": 27008.8,
      "state_query_norm": 44.4283390045166,
      "state_relation_loss": 0.0004326439375290647,
      "step": 6500,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.007062780950218439,
      "weighted_relation_loss": 0.007062780950218439
    },
    {
      "action_query_norm": 43.82093200683594,
      "action_relation_loss": 0.006688916496932507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.089803554724041,
      "flow_matching_loss": 0.05343102850019932,
      "grad_norm": 0.20026524364948273,
      "learning_rate": 9.256997316340976e-05,
      "loss": 0.0606,
      "relation_loss": 0.007123203109949827,
      "selected_pairs": 26528.5,
      "state_query_norm": 44.42891731262207,
      "state_relation_loss": 0.0004342865984654054,
      "step": 6510,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.007123203109949827,
      "weighted_relation_loss": 0.007123203109949827
    },
    {
      "action_query_norm": 43.80894012451172,
      "action_relation_loss": 0.007007652055472135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.099158091674462,
      "flow_matching_loss": 0.05225688517093659,
      "grad_norm": 0.13942275941371918,
      "learning_rate": 9.254103815930943e-05,
      "loss": 0.0597,
      "relation_loss": 0.0074260163120925425,
      "selected_pairs": 26506.6,
      "state_query_norm": 44.423432159423825,
      "state_relation_loss": 0.00041836429154500364,
      "step": 6520,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0074260163120925425,
      "weighted_relation_loss": 0.0074260163120925425
    },
    {
      "action_query_norm": 43.832081985473636,
      "action_relation_loss": 0.006878224341198802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.108512628624883,
      "flow_matching_loss": 0.05310412459075451,
      "grad_norm": 0.15298019349575043,
      "learning_rate": 9.251205146383636e-05,
      "loss": 0.0604,
      "relation_loss": 0.00732632870785892,
      "selected_pairs": 26625.1,
      "state_query_norm": 44.42648582458496,
      "state_relation_loss": 0.00044810447725467383,
      "step": 6530,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00732632870785892,
      "weighted_relation_loss": 0.00732632870785892
    },
    {
      "action_query_norm": 43.821642303466795,
      "action_relation_loss": 0.006720741745084524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.117867165575304,
      "flow_matching_loss": 0.053599404916167256,
      "grad_norm": 0.12948906421661377,
      "learning_rate": 9.248301311221216e-05,
      "loss": 0.0607,
      "relation_loss": 0.007143306033685804,
      "selected_pairs": 27504.8,
      "state_query_norm": 44.41946029663086,
      "state_relation_loss": 0.00042256425076629965,
      "step": 6540,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.007143306033685804,
      "weighted_relation_loss": 0.007143306033685804
    },
    {
      "action_query_norm": 43.82602500915527,
      "action_relation_loss": 0.00654020793735981,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.127221702525725,
      "flow_matching_loss": 0.05301373265683651,
      "grad_norm": 0.1717059463262558,
      "learning_rate": 9.245392313972115e-05,
      "loss": 0.06,
      "relation_loss": 0.006954145105555654,
      "selected_pairs": 26917.4,
      "state_query_norm": 44.41781349182129,
      "state_relation_loss": 0.0004139372205827385,
      "step": 6550,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006954145105555654,
      "weighted_relation_loss": 0.006954145105555654
    },
    {
      "action_query_norm": 43.80005645751953,
      "action_relation_loss": 0.007017776556313038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.136576239476146,
      "flow_matching_loss": 0.05342274829745293,
      "grad_norm": 0.14569537341594696,
      "learning_rate": 9.242478158171037e-05,
      "loss": 0.0609,
      "relation_loss": 0.007447565626353026,
      "selected_pairs": 26735.5,
      "state_query_norm": 44.41963272094726,
      "state_relation_loss": 0.00042978907586075363,
      "step": 6560,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.007447565626353026,
      "weighted_relation_loss": 0.007447565626353026
    },
    {
      "action_query_norm": 43.80358390808105,
      "action_relation_loss": 0.0067331480327993635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.1459307764265665,
      "flow_matching_loss": 0.055929363891482356,
      "grad_norm": 0.12372367084026337,
      "learning_rate": 9.239558847358957e-05,
      "loss": 0.0631,
      "relation_loss": 0.007147326460108161,
      "selected_pairs": 26903.7,
      "state_query_norm": 44.416419982910156,
      "state_relation_loss": 0.0004141784069361165,
      "step": 6570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007147326460108161,
      "weighted_relation_loss": 0.007147326460108161
    },
    {
      "action_query_norm": 43.78459167480469,
      "action_relation_loss": 0.006948915589600801,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.155285313376988,
      "flow_matching_loss": 0.052983229979872704,
      "grad_norm": 0.16369500756263733,
      "learning_rate": 9.236634385083114e-05,
      "loss": 0.0603,
      "relation_loss": 0.007360166078433395,
      "selected_pairs": 26816.6,
      "state_query_norm": 44.417562103271486,
      "state_relation_loss": 0.00041125050338450817,
      "step": 6580,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007360166078433395,
      "weighted_relation_loss": 0.007360166078433395
    },
    {
      "action_query_norm": 43.81523742675781,
      "action_relation_loss": 0.006766744237393141,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.164639850327409,
      "flow_matching_loss": 0.054185619205236436,
      "grad_norm": 0.1706569492816925,
      "learning_rate": 9.233704774897005e-05,
      "loss": 0.0613,
      "relation_loss": 0.007157804863527417,
      "selected_pairs": 27292.0,
      "state_query_norm": 44.41500968933106,
      "state_relation_loss": 0.0003910606203135103,
      "step": 6590,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.007157804863527417,
      "weighted_relation_loss": 0.007157804863527417
    },
    {
      "action_query_norm": 43.79345436096192,
      "action_relation_loss": 0.006666591111570597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.17399438727783,
      "flow_matching_loss": 0.05219502635300159,
      "grad_norm": 0.14766748249530792,
      "learning_rate": 9.230770020360382e-05,
      "loss": 0.0592,
      "relation_loss": 0.007049249298870563,
      "selected_pairs": 26785.9,
      "state_query_norm": 44.41730880737305,
      "state_relation_loss": 0.0003826581785688177,
      "step": 6600,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.007049249298870563,
      "weighted_relation_loss": 0.007049249298870563
    },
    {
      "action_query_norm": 43.774939346313474,
      "action_relation_loss": 0.007091930182650685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.183348924228251,
      "flow_matching_loss": 0.053130428120493886,
      "grad_norm": 0.11323659121990204,
      "learning_rate": 9.227830125039247e-05,
      "loss": 0.0606,
      "relation_loss": 0.007488295342773199,
      "selected_pairs": 27116.3,
      "state_query_norm": 44.4095157623291,
      "state_relation_loss": 0.0003963651048252359,
      "step": 6610,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.007488295342773199,
      "weighted_relation_loss": 0.007488295342773199
    },
    {
      "action_query_norm": 43.78628311157227,
      "action_relation_loss": 0.00670305984094739,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.192703461178672,
      "flow_matching_loss": 0.050668439641594884,
      "grad_norm": 0.102040596306324,
      "learning_rate": 9.224885092505853e-05,
      "loss": 0.0577,
      "relation_loss": 0.007074191560968756,
      "selected_pairs": 26898.4,
      "state_query_norm": 44.41327056884766,
      "state_relation_loss": 0.00037113165599294006,
      "step": 6620,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007074191560968756,
      "weighted_relation_loss": 0.007074191560968756
    },
    {
      "action_query_norm": 43.807012939453124,
      "action_relation_loss": 0.006685292115435004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.202057998129092,
      "flow_matching_loss": 0.05333344824612141,
      "grad_norm": 0.1224445179104805,
      "learning_rate": 9.221934926338691e-05,
      "loss": 0.0604,
      "relation_loss": 0.007074395148083568,
      "selected_pairs": 26701.0,
      "state_query_norm": 44.4072811126709,
      "state_relation_loss": 0.0003891030763043091,
      "step": 6630,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.007074395148083568,
      "weighted_relation_loss": 0.007074395148083568
    },
    {
      "action_query_norm": 43.7998779296875,
      "action_relation_loss": 0.006738976296037436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.211412535079513,
      "flow_matching_loss": 0.05216541588306427,
      "grad_norm": 0.15083497762680054,
      "learning_rate": 9.21897963012249e-05,
      "loss": 0.0593,
      "relation_loss": 0.00713791698217392,
      "selected_pairs": 26737.8,
      "state_query_norm": 44.40672225952149,
      "state_relation_loss": 0.0003989407094195485,
      "step": 6640,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00713791698217392,
      "weighted_relation_loss": 0.00713791698217392
    },
    {
      "action_query_norm": 43.787593841552734,
      "action_relation_loss": 0.007383124995976686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.220767072029934,
      "flow_matching_loss": 0.055247443914413455,
      "grad_norm": 0.11266011744737625,
      "learning_rate": 9.216019207448217e-05,
      "loss": 0.063,
      "relation_loss": 0.007801737869158387,
      "selected_pairs": 26277.3,
      "state_query_norm": 44.401454544067384,
      "state_relation_loss": 0.0004186129168374464,
      "step": 6650,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.007801737869158387,
      "weighted_relation_loss": 0.007801737869158387
    },
    {
      "action_query_norm": 43.78249626159668,
      "action_relation_loss": 0.0069774489384144545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.2301216089803555,
      "flow_matching_loss": 0.052466153725981714,
      "grad_norm": 0.12894316017627716,
      "learning_rate": 9.21305366191306e-05,
      "loss": 0.0598,
      "relation_loss": 0.007379986299201846,
      "selected_pairs": 27033.8,
      "state_query_norm": 44.4035343170166,
      "state_relation_loss": 0.0004025374219054356,
      "step": 6660,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.007379986299201846,
      "weighted_relation_loss": 0.007379986299201846
    },
    {
      "action_query_norm": 43.77484397888183,
      "action_relation_loss": 0.006722633168101311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.239476145930777,
      "flow_matching_loss": 0.055808992311358455,
      "grad_norm": 0.17000341415405273,
      "learning_rate": 9.21008299712044e-05,
      "loss": 0.0629,
      "relation_loss": 0.007119466550648212,
      "selected_pairs": 26571.6,
      "state_query_norm": 44.401818084716794,
      "state_relation_loss": 0.0003968333650846034,
      "step": 6670,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.007119466550648212,
      "weighted_relation_loss": 0.007119466550648212
    },
    {
      "action_query_norm": 43.796243286132814,
      "action_relation_loss": 0.007075605867430568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.248830682881198,
      "flow_matching_loss": 0.054834648966789246,
      "grad_norm": 0.12092134356498718,
      "learning_rate": 9.207107216679993e-05,
      "loss": 0.0623,
      "relation_loss": 0.007463844772428274,
      "selected_pairs": 26779.7,
      "state_query_norm": 44.39764823913574,
      "state_relation_loss": 0.0003882389661157504,
      "step": 6680,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.007463844772428274,
      "weighted_relation_loss": 0.007463844772428274
    },
    {
      "action_query_norm": 43.78963279724121,
      "action_relation_loss": 0.006928916275501251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.258185219831619,
      "flow_matching_loss": 0.05467057600617409,
      "grad_norm": 0.1458144634962082,
      "learning_rate": 9.204126324207574e-05,
      "loss": 0.062,
      "relation_loss": 0.007319437758997083,
      "selected_pairs": 26410.5,
      "state_query_norm": 44.394281005859376,
      "state_relation_loss": 0.0003905214718542993,
      "step": 6690,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.007319437758997083,
      "weighted_relation_loss": 0.007319437758997083
    },
    {
      "action_query_norm": 43.80736656188965,
      "action_relation_loss": 0.007053389586508274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.267539756782039,
      "flow_matching_loss": 0.05473278500139713,
      "grad_norm": 0.15125633776187897,
      "learning_rate": 9.201140323325248e-05,
      "loss": 0.0622,
      "relation_loss": 0.0074614436365664,
      "selected_pairs": 26851.0,
      "state_query_norm": 44.39058837890625,
      "state_relation_loss": 0.0004080539743881673,
      "step": 6700,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0074614436365664,
      "weighted_relation_loss": 0.0074614436365664
    },
    {
      "action_query_norm": 43.7839916229248,
      "action_relation_loss": 0.006944321468472481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.27689429373246,
      "flow_matching_loss": 0.05446554496884346,
      "grad_norm": 0.20804297924041748,
      "learning_rate": 9.198149217661287e-05,
      "loss": 0.0618,
      "relation_loss": 0.00736591899767518,
      "selected_pairs": 26691.3,
      "state_query_norm": 44.392560195922854,
      "state_relation_loss": 0.0004215975000988692,
      "step": 6710,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00736591899767518,
      "weighted_relation_loss": 0.00736591899767518
    },
    {
      "action_query_norm": 43.77158050537109,
      "action_relation_loss": 0.006924902135506272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.286248830682881,
      "flow_matching_loss": 0.053154243528842925,
      "grad_norm": 0.15510833263397217,
      "learning_rate": 9.195153010850167e-05,
      "loss": 0.0605,
      "relation_loss": 0.0073394675739109514,
      "selected_pairs": 27102.4,
      "state_query_norm": 44.39749412536621,
      "state_relation_loss": 0.00041456553735770286,
      "step": 6720,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0073394675739109514,
      "weighted_relation_loss": 0.0073394675739109514
    },
    {
      "action_query_norm": 43.78349227905274,
      "action_relation_loss": 0.006658327160403132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.295603367633302,
      "flow_matching_loss": 0.052704333513975146,
      "grad_norm": 0.1701362580060959,
      "learning_rate": 9.192151706532562e-05,
      "loss": 0.0598,
      "relation_loss": 0.007075943192467094,
      "selected_pairs": 27313.7,
      "state_query_norm": 44.3917121887207,
      "state_relation_loss": 0.00041761595057323577,
      "step": 6730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.007075943192467094,
      "weighted_relation_loss": 0.007075943192467094
    },
    {
      "action_query_norm": 43.76839599609375,
      "action_relation_loss": 0.006716764625161886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.304957904583723,
      "flow_matching_loss": 0.05071825683116913,
      "grad_norm": 0.19155500829219818,
      "learning_rate": 9.189145308355337e-05,
      "loss": 0.0578,
      "relation_loss": 0.007107934076339006,
      "selected_pairs": 27173.3,
      "state_query_norm": 44.39212837219238,
      "state_relation_loss": 0.0003911694715498015,
      "step": 6740,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.007107934076339006,
      "weighted_relation_loss": 0.007107934076339006
    },
    {
      "action_query_norm": 43.756409072875975,
      "action_relation_loss": 0.006666597072035074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.314312441534144,
      "flow_matching_loss": 0.051644813269376755,
      "grad_norm": 0.15433238446712494,
      "learning_rate": 9.186133819971555e-05,
      "loss": 0.0587,
      "relation_loss": 0.007054030755534768,
      "selected_pairs": 26939.5,
      "state_query_norm": 44.38532409667969,
      "state_relation_loss": 0.00038743368932045994,
      "step": 6750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.007054030755534768,
      "weighted_relation_loss": 0.007054030755534768
    },
    {
      "action_query_norm": 43.74357528686524,
      "action_relation_loss": 0.006489688344299793,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.323666978484565,
      "flow_matching_loss": 0.053207167983055116,
      "grad_norm": 0.09876024723052979,
      "learning_rate": 9.183117245040455e-05,
      "loss": 0.0601,
      "relation_loss": 0.006878706160932779,
      "selected_pairs": 26715.1,
      "state_query_norm": 44.389642333984376,
      "state_relation_loss": 0.0003890178370056674,
      "step": 6760,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006878706160932779,
      "weighted_relation_loss": 0.006878706160932779
    },
    {
      "action_query_norm": 43.75431594848633,
      "action_relation_loss": 0.006900997925549745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.333021515434986,
      "flow_matching_loss": 0.05402936898171902,
      "grad_norm": 0.1182493343949318,
      "learning_rate": 9.180095587227458e-05,
      "loss": 0.0613,
      "relation_loss": 0.007314682053402066,
      "selected_pairs": 27057.8,
      "state_query_norm": 44.3797966003418,
      "state_relation_loss": 0.0004136840609135106,
      "step": 6770,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.007314682053402066,
      "weighted_relation_loss": 0.007314682053402066
    },
    {
      "action_query_norm": 43.74781761169434,
      "action_relation_loss": 0.007042235229164362,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.342376052385407,
      "flow_matching_loss": 0.054570694267749784,
      "grad_norm": 0.13436101377010345,
      "learning_rate": 9.177068850204167e-05,
      "loss": 0.062,
      "relation_loss": 0.0074641487095505,
      "selected_pairs": 26408.9,
      "state_query_norm": 44.38068046569824,
      "state_relation_loss": 0.0004219134687446058,
      "step": 6780,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0074641487095505,
      "weighted_relation_loss": 0.0074641487095505
    },
    {
      "action_query_norm": 43.77018814086914,
      "action_relation_loss": 0.006448138225823641,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.351730589335828,
      "flow_matching_loss": 0.05172903910279274,
      "grad_norm": 0.12418577820062637,
      "learning_rate": 9.174037037648351e-05,
      "loss": 0.0586,
      "relation_loss": 0.006883240072056651,
      "selected_pairs": 26945.2,
      "state_query_norm": 44.38049507141113,
      "state_relation_loss": 0.0004351019160822034,
      "step": 6790,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006883240072056651,
      "weighted_relation_loss": 0.006883240072056651
    },
    {
      "action_query_norm": 43.75904808044434,
      "action_relation_loss": 0.006639908254146576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.361085126286249,
      "flow_matching_loss": 0.051051155850291255,
      "grad_norm": 0.13405078649520874,
      "learning_rate": 9.171000153243947e-05,
      "loss": 0.0581,
      "relation_loss": 0.007077562343329191,
      "selected_pairs": 26706.7,
      "state_query_norm": 44.37896461486817,
      "state_relation_loss": 0.00043765411246567967,
      "step": 6800,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.007077562343329191,
      "weighted_relation_loss": 0.007077562343329191
    },
    {
      "action_query_norm": 43.758414840698244,
      "action_relation_loss": 0.006846842681989073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.37043966323667,
      "flow_matching_loss": 0.054413894191384315,
      "grad_norm": 0.15364670753479004,
      "learning_rate": 9.167958200681057e-05,
      "loss": 0.0617,
      "relation_loss": 0.007275105686858296,
      "selected_pairs": 26779.7,
      "state_query_norm": 44.37834243774414,
      "state_relation_loss": 0.0004282630165107548,
      "step": 6810,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.007275105686858296,
      "weighted_relation_loss": 0.007275105686858296
    },
    {
      "action_query_norm": 43.756465911865234,
      "action_relation_loss": 0.0065449916291981936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.37979420018709,
      "flow_matching_loss": 0.053637844324111936,
      "grad_norm": 0.18818363547325134,
      "learning_rate": 9.164911183655943e-05,
      "loss": 0.0606,
      "relation_loss": 0.00695250891149044,
      "selected_pairs": 26875.8,
      "state_query_norm": 44.38003921508789,
      "state_relation_loss": 0.0004075173055753112,
      "step": 6820,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00695250891149044,
      "weighted_relation_loss": 0.00695250891149044
    },
    {
      "action_query_norm": 43.76157913208008,
      "action_relation_loss": 0.006537463515996933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.3891487371375115,
      "flow_matching_loss": 0.05095654278993607,
      "grad_norm": 0.15369227528572083,
      "learning_rate": 9.161859105871013e-05,
      "loss": 0.0579,
      "relation_loss": 0.006970538385212421,
      "selected_pairs": 27458.2,
      "state_query_norm": 44.379323196411136,
      "state_relation_loss": 0.0004330748139182106,
      "step": 6830,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.006970538385212421,
      "weighted_relation_loss": 0.006970538385212421
    },
    {
      "action_query_norm": 43.75357513427734,
      "action_relation_loss": 0.006531739048659802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.398503274087933,
      "flow_matching_loss": 0.05139389336109161,
      "grad_norm": 0.13809335231781006,
      "learning_rate": 9.158801971034832e-05,
      "loss": 0.0583,
      "relation_loss": 0.006953336670994759,
      "selected_pairs": 27055.6,
      "state_query_norm": 44.37565116882324,
      "state_relation_loss": 0.0004215976718114689,
      "step": 6840,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.006953336670994759,
      "weighted_relation_loss": 0.006953336670994759
    },
    {
      "action_query_norm": 43.74530944824219,
      "action_relation_loss": 0.006440305197611451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.407857811038354,
      "flow_matching_loss": 0.050352757796645164,
      "grad_norm": 0.10591834783554077,
      "learning_rate": 9.155739782862107e-05,
      "loss": 0.0572,
      "relation_loss": 0.00683058057911694,
      "selected_pairs": 26551.6,
      "state_query_norm": 44.37723655700684,
      "state_relation_loss": 0.000390275297104381,
      "step": 6850,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.00683058057911694,
      "weighted_relation_loss": 0.00683058057911694
    },
    {
      "action_query_norm": 43.7309009552002,
      "action_relation_loss": 0.006568486522883177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.417212347988775,
      "flow_matching_loss": 0.051409750059247014,
      "grad_norm": 0.11192331463098526,
      "learning_rate": 9.152672545073687e-05,
      "loss": 0.0584,
      "relation_loss": 0.006968137389048934,
      "selected_pairs": 26784.5,
      "state_query_norm": 44.37295761108398,
      "state_relation_loss": 0.000399650790495798,
      "step": 6860,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.006968137389048934,
      "weighted_relation_loss": 0.006968137389048934
    },
    {
      "action_query_norm": 43.732495880126955,
      "action_relation_loss": 0.006819267105311155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.426566884939196,
      "flow_matching_loss": 0.053970422223210336,
      "grad_norm": 0.12213995307683945,
      "learning_rate": 9.149600261396551e-05,
      "loss": 0.0612,
      "relation_loss": 0.0072010734118521215,
      "selected_pairs": 26546.5,
      "state_query_norm": 44.37446899414063,
      "state_relation_loss": 0.0003818062163190916,
      "step": 6870,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0072010734118521215,
      "weighted_relation_loss": 0.0072010734118521215
    },
    {
      "action_query_norm": 43.72285118103027,
      "action_relation_loss": 0.006636080471798778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.435921421889616,
      "flow_matching_loss": 0.04994846433401108,
      "grad_norm": 0.13380779325962067,
      "learning_rate": 9.146522935563817e-05,
      "loss": 0.057,
      "relation_loss": 0.007050401391461491,
      "selected_pairs": 27434.6,
      "state_query_norm": 44.37203254699707,
      "state_relation_loss": 0.000414320916752331,
      "step": 6880,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.007050401391461491,
      "weighted_relation_loss": 0.007050401391461491
    },
    {
      "action_query_norm": 43.7162425994873,
      "action_relation_loss": 0.006572076445445418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.445275958840037,
      "flow_matching_loss": 0.05064353533089161,
      "grad_norm": 0.1044616848230362,
      "learning_rate": 9.143440571314723e-05,
      "loss": 0.0576,
      "relation_loss": 0.006965007167309523,
      "selected_pairs": 26661.8,
      "state_query_norm": 44.372083282470705,
      "state_relation_loss": 0.00039293069276027384,
      "step": 6890,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.006965007167309523,
      "weighted_relation_loss": 0.006965007167309523
    },
    {
      "action_query_norm": 43.72359848022461,
      "action_relation_loss": 0.0065607157535851005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.454630495790458,
      "flow_matching_loss": 0.04976914674043655,
      "grad_norm": 0.12650108337402344,
      "learning_rate": 9.140353172394636e-05,
      "loss": 0.0567,
      "relation_loss": 0.006962293153628707,
      "selected_pairs": 26711.1,
      "state_query_norm": 44.36912803649902,
      "state_relation_loss": 0.0004015774291474372,
      "step": 6900,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006962293153628707,
      "weighted_relation_loss": 0.006962293153628707
    },
    {
      "action_query_norm": 43.71282196044922,
      "action_relation_loss": 0.006439852248877287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.463985032740879,
      "flow_matching_loss": 0.052982740104198456,
      "grad_norm": 0.16676552593708038,
      "learning_rate": 9.137260742555032e-05,
      "loss": 0.0598,
      "relation_loss": 0.00682919411920011,
      "selected_pairs": 27269.6,
      "state_query_norm": 44.368543243408205,
      "state_relation_loss": 0.00038934184703975916,
      "step": 6910,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.00682919411920011,
      "weighted_relation_loss": 0.00682919411920011
    },
    {
      "action_query_norm": 43.7261661529541,
      "action_relation_loss": 0.007018094509840011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.4733395696913005,
      "flow_matching_loss": 0.054323868826031685,
      "grad_norm": 0.1400856226682663,
      "learning_rate": 9.13416328555351e-05,
      "loss": 0.0618,
      "relation_loss": 0.007433746615424753,
      "selected_pairs": 26964.6,
      "state_query_norm": 44.37123107910156,
      "state_relation_loss": 0.0004156520706601441,
      "step": 6920,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.007433746615424753,
      "weighted_relation_loss": 0.007433746615424753
    },
    {
      "action_query_norm": 43.71648979187012,
      "action_relation_loss": 0.006435311818495393,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.482694106641722,
      "flow_matching_loss": 0.05150684677064419,
      "grad_norm": 0.15526732802391052,
      "learning_rate": 9.131060805153769e-05,
      "loss": 0.0583,
      "relation_loss": 0.006841960875317454,
      "selected_pairs": 26418.6,
      "state_query_norm": 44.36124153137207,
      "state_relation_loss": 0.0004066490655532107,
      "step": 6930,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.006841960875317454,
      "weighted_relation_loss": 0.006841960875317454
    },
    {
      "action_query_norm": 43.71054458618164,
      "action_relation_loss": 0.006632676487788558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.492048643592142,
      "flow_matching_loss": 0.05307513512670994,
      "grad_norm": 0.12811622023582458,
      "learning_rate": 9.127953305125618e-05,
      "loss": 0.0601,
      "relation_loss": 0.007018877705559134,
      "selected_pairs": 26249.2,
      "state_query_norm": 44.360969924926756,
      "state_relation_loss": 0.00038620118575636295,
      "step": 6940,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.007018877705559134,
      "weighted_relation_loss": 0.007018877705559134
    },
    {
      "action_query_norm": 43.701084518432616,
      "action_relation_loss": 0.006122930999845266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.501403180542563,
      "flow_matching_loss": 0.053330506384372714,
      "grad_norm": 0.14946000277996063,
      "learning_rate": 9.124840789244958e-05,
      "loss": 0.0599,
      "relation_loss": 0.006538109015673399,
      "selected_pairs": 26640.2,
      "state_query_norm": 44.36044921875,
      "state_relation_loss": 0.0004151780332904309,
      "step": 6950,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.006538109015673399,
      "weighted_relation_loss": 0.006538109015673399
    },
    {
      "action_query_norm": 43.70185813903809,
      "action_relation_loss": 0.0068824514746665955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.510757717492984,
      "flow_matching_loss": 0.05263781882822514,
      "grad_norm": 0.14473900198936462,
      "learning_rate": 9.121723261293794e-05,
      "loss": 0.0599,
      "relation_loss": 0.007269434118643403,
      "selected_pairs": 26802.6,
      "state_query_norm": 44.358486557006835,
      "state_relation_loss": 0.0003869826527079567,
      "step": 6960,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.007269434118643403,
      "weighted_relation_loss": 0.007269434118643403
    },
    {
      "action_query_norm": 43.72706146240235,
      "action_relation_loss": 0.006368119176477194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.520112254443405,
      "flow_matching_loss": 0.05042549073696136,
      "grad_norm": 0.12551471590995789,
      "learning_rate": 9.118600725060214e-05,
      "loss": 0.0572,
      "relation_loss": 0.006760485703125596,
      "selected_pairs": 27102.4,
      "state_query_norm": 44.359622955322266,
      "state_relation_loss": 0.0003923665615729988,
      "step": 6970,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.006760485703125596,
      "weighted_relation_loss": 0.006760485703125596
    },
    {
      "action_query_norm": 43.704829406738284,
      "action_relation_loss": 0.006463345838710666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.529466791393826,
      "flow_matching_loss": 0.05245669186115265,
      "grad_norm": 0.14297184348106384,
      "learning_rate": 9.115473184338392e-05,
      "loss": 0.0593,
      "relation_loss": 0.006835080916061997,
      "selected_pairs": 27131.1,
      "state_query_norm": 44.358414840698245,
      "state_relation_loss": 0.0003717351355589926,
      "step": 6980,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.006835080916061997,
      "weighted_relation_loss": 0.006835080916061997
    },
    {
      "action_query_norm": 43.70576629638672,
      "action_relation_loss": 0.006495900778099894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.538821328344247,
      "flow_matching_loss": 0.05338845364749432,
      "grad_norm": 0.2040201723575592,
      "learning_rate": 9.11234064292859e-05,
      "loss": 0.0603,
      "relation_loss": 0.006867722049355507,
      "selected_pairs": 26278.2,
      "state_query_norm": 44.35702323913574,
      "state_relation_loss": 0.0003718212770763785,
      "step": 6990,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.006867722049355507,
      "weighted_relation_loss": 0.006867722049355507
    },
    {
      "action_query_norm": 43.70815086364746,
      "action_relation_loss": 0.006513972766697407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.5481758652946676,
      "flow_matching_loss": 0.05322741307318211,
      "grad_norm": 0.17812882363796234,
      "learning_rate": 9.109203104637137e-05,
      "loss": 0.0601,
      "relation_loss": 0.006891952781006694,
      "selected_pairs": 26967.1,
      "state_query_norm": 44.35193023681641,
      "state_relation_loss": 0.00037798005796503277,
      "step": 7000,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.006891952781006694,
      "weighted_relation_loss": 0.006891952781006694
    },
    {
      "action_query_norm": 43.68952980041504,
      "action_relation_loss": 0.007114670239388943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.557530402245089,
      "flow_matching_loss": 0.05452901087701321,
      "grad_norm": 0.1730414181947708,
      "learning_rate": 9.10606057327644e-05,
      "loss": 0.062,
      "relation_loss": 0.007510599261149764,
      "selected_pairs": 27085.2,
      "state_query_norm": 44.350609588623044,
      "state_relation_loss": 0.00039592895773239435,
      "step": 7010,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.007510599261149764,
      "weighted_relation_loss": 0.007510599261149764
    },
    {
      "action_query_norm": 43.67240333557129,
      "action_relation_loss": 0.0064134927000850436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.56688493919551,
      "flow_matching_loss": 0.052308255806565285,
      "grad_norm": 0.15238092839717865,
      "learning_rate": 9.10291305266497e-05,
      "loss": 0.0591,
      "relation_loss": 0.006792191229760647,
      "selected_pairs": 26570.0,
      "state_query_norm": 44.354444885253905,
      "state_relation_loss": 0.0003786985675105825,
      "step": 7020,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.006792191229760647,
      "weighted_relation_loss": 0.006792191229760647
    },
    {
      "action_query_norm": 43.68705825805664,
      "action_relation_loss": 0.006581665063276887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.576239476145931,
      "flow_matching_loss": 0.05394665077328682,
      "grad_norm": 0.2354922741651535,
      "learning_rate": 9.099760546627261e-05,
      "loss": 0.0609,
      "relation_loss": 0.0069951106328517195,
      "selected_pairs": 26781.5,
      "state_query_norm": 44.34524345397949,
      "state_relation_loss": 0.0004134456132305786,
      "step": 7030,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0069951106328517195,
      "weighted_relation_loss": 0.0069951106328517195
    },
    {
      "action_query_norm": 43.68865547180176,
      "action_relation_loss": 0.006064770463854075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.585594013096352,
      "flow_matching_loss": 0.053738201037049294,
      "grad_norm": 0.19122366607189178,
      "learning_rate": 9.096603058993907e-05,
      "loss": 0.0602,
      "relation_loss": 0.006470524612814188,
      "selected_pairs": 26627.0,
      "state_query_norm": 44.34684104919434,
      "state_relation_loss": 0.00040575416933279483,
      "step": 7040,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.006470524612814188,
      "weighted_relation_loss": 0.006470524612814188
    },
    {
      "action_query_norm": 43.67397842407227,
      "action_relation_loss": 0.006833198107779026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.594948550046773,
      "flow_matching_loss": 0.053746432065963745,
      "grad_norm": 0.20359869301319122,
      "learning_rate": 9.093440593601552e-05,
      "loss": 0.061,
      "relation_loss": 0.007253956701606512,
      "selected_pairs": 26485.3,
      "state_query_norm": 44.34759292602539,
      "state_relation_loss": 0.00042075865203514693,
      "step": 7050,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.007253956701606512,
      "weighted_relation_loss": 0.007253956701606512
    },
    {
      "action_query_norm": 43.68113899230957,
      "action_relation_loss": 0.006457907008007168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.604303086997193,
      "flow_matching_loss": 0.049683669582009315,
      "grad_norm": 0.13451963663101196,
      "learning_rate": 9.090273154292889e-05,
      "loss": 0.0566,
      "relation_loss": 0.006914901593700051,
      "selected_pairs": 26897.0,
      "state_query_norm": 44.339643478393555,
      "state_relation_loss": 0.0004569945333059877,
      "step": 7060,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006914901593700051,
      "weighted_relation_loss": 0.006914901593700051
    },
    {
      "action_query_norm": 43.69424209594727,
      "action_relation_loss": 0.006537413178011775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.613657623947614,
      "flow_matching_loss": 0.05352145321667194,
      "grad_norm": 0.17427393794059753,
      "learning_rate": 9.087100744916656e-05,
      "loss": 0.0605,
      "relation_loss": 0.006941023794934153,
      "selected_pairs": 26925.3,
      "state_query_norm": 44.343666076660156,
      "state_relation_loss": 0.00040361059946008026,
      "step": 7070,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006941023794934153,
      "weighted_relation_loss": 0.006941023794934153
    },
    {
      "action_query_norm": 43.692893981933594,
      "action_relation_loss": 0.0066293831914663315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.6230121608980355,
      "flow_matching_loss": 0.053369054943323134,
      "grad_norm": 0.12834906578063965,
      "learning_rate": 9.08392336932763e-05,
      "loss": 0.0604,
      "relation_loss": 0.007040307112038136,
      "selected_pairs": 27554.9,
      "state_query_norm": 44.34640350341797,
      "state_relation_loss": 0.00041092392639257013,
      "step": 7080,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.007040307112038136,
      "weighted_relation_loss": 0.007040307112038136
    },
    {
      "action_query_norm": 43.68196487426758,
      "action_relation_loss": 0.006599127035588026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.632366697848457,
      "flow_matching_loss": 0.05243967175483703,
      "grad_norm": 0.13752715289592743,
      "learning_rate": 9.080741031386619e-05,
      "loss": 0.0594,
      "relation_loss": 0.007010248489677906,
      "selected_pairs": 26898.2,
      "state_query_norm": 44.3479866027832,
      "state_relation_loss": 0.00041112142789643256,
      "step": 7090,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007010248489677906,
      "weighted_relation_loss": 0.007010248489677906
    },
    {
      "action_query_norm": 43.69433441162109,
      "action_relation_loss": 0.006167812319472432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.641721234798878,
      "flow_matching_loss": 0.0530134167522192,
      "grad_norm": 0.11702179908752441,
      "learning_rate": 9.077553734960468e-05,
      "loss": 0.0596,
      "relation_loss": 0.006579452101141215,
      "selected_pairs": 26696.2,
      "state_query_norm": 44.33974342346191,
      "state_relation_loss": 0.00041163982241414486,
      "step": 7100,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006579452101141215,
      "weighted_relation_loss": 0.006579452101141215
    },
    {
      "action_query_norm": 43.65559387207031,
      "action_relation_loss": 0.006674376409500837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.651075771749299,
      "flow_matching_loss": 0.04847705252468586,
      "grad_norm": 0.13616275787353516,
      "learning_rate": 9.074361483922041e-05,
      "loss": 0.0556,
      "relation_loss": 0.007074948633089662,
      "selected_pairs": 27149.6,
      "state_query_norm": 44.3374740600586,
      "state_relation_loss": 0.0004005722439615056,
      "step": 7110,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.007074948633089662,
      "weighted_relation_loss": 0.007074948633089662
    },
    {
      "action_query_norm": 43.6639461517334,
      "action_relation_loss": 0.006626031827181578,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.66043030869972,
      "flow_matching_loss": 0.05103137493133545,
      "grad_norm": 0.1305621713399887,
      "learning_rate": 9.071164282150224e-05,
      "loss": 0.058,
      "relation_loss": 0.007015626085922122,
      "selected_pairs": 26895.7,
      "state_query_norm": 44.33593978881836,
      "state_relation_loss": 0.00038959430239629,
      "step": 7120,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.007015626085922122,
      "weighted_relation_loss": 0.007015626085922122
    },
    {
      "action_query_norm": 43.6537971496582,
      "action_relation_loss": 0.006175319477915764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.66978484565014,
      "flow_matching_loss": 0.05109506994485855,
      "grad_norm": 0.13246014714241028,
      "learning_rate": 9.067962133529919e-05,
      "loss": 0.0576,
      "relation_loss": 0.006547423172742128,
      "selected_pairs": 26642.1,
      "state_query_norm": 44.337181854248044,
      "state_relation_loss": 0.000372103726840578,
      "step": 7130,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.006547423172742128,
      "weighted_relation_loss": 0.006547423172742128
    },
    {
      "action_query_norm": 43.641253280639646,
      "action_relation_loss": 0.006466809241101146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.679139382600561,
      "flow_matching_loss": 0.05063442550599575,
      "grad_norm": 0.14145781099796295,
      "learning_rate": 9.064755041952036e-05,
      "loss": 0.0575,
      "relation_loss": 0.006866326136514544,
      "selected_pairs": 26922.8,
      "state_query_norm": 44.33143539428711,
      "state_relation_loss": 0.00039951685757841913,
      "step": 7140,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006866326136514544,
      "weighted_relation_loss": 0.006866326136514544
    },
    {
      "action_query_norm": 43.64274978637695,
      "action_relation_loss": 0.00667327013798058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.688493919550982,
      "flow_matching_loss": 0.05246102921664715,
      "grad_norm": 0.1704777181148529,
      "learning_rate": 9.061543011313497e-05,
      "loss": 0.0596,
      "relation_loss": 0.007097053667530417,
      "selected_pairs": 26389.7,
      "state_query_norm": 44.32696418762207,
      "state_relation_loss": 0.00042378358193673193,
      "step": 7150,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.007097053667530417,
      "weighted_relation_loss": 0.007097053667530417
    },
    {
      "action_query_norm": 43.65634002685547,
      "action_relation_loss": 0.006673441361635923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.697848456501403,
      "flow_matching_loss": 0.052810323983430864,
      "grad_norm": 0.1081700474023819,
      "learning_rate": 9.05832604551722e-05,
      "loss": 0.0599,
      "relation_loss": 0.007113587995991111,
      "selected_pairs": 27289.1,
      "state_query_norm": 44.329203033447264,
      "state_relation_loss": 0.0004401465936098248,
      "step": 7160,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.007113587995991111,
      "weighted_relation_loss": 0.007113587995991111
    },
    {
      "action_query_norm": 43.655819702148435,
      "action_relation_loss": 0.0063011626247316595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.7072029934518245,
      "flow_matching_loss": 0.05008523017168045,
      "grad_norm": 0.09569405019283295,
      "learning_rate": 9.055104148472124e-05,
      "loss": 0.0568,
      "relation_loss": 0.006705809058621526,
      "selected_pairs": 26709.2,
      "state_query_norm": 44.32837982177735,
      "state_relation_loss": 0.0004046464542625472,
      "step": 7170,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006705809058621526,
      "weighted_relation_loss": 0.006705809058621526
    },
    {
      "action_query_norm": 43.63466567993164,
      "action_relation_loss": 0.006351092457771301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.716557530402245,
      "flow_matching_loss": 0.05217468999326229,
      "grad_norm": 0.15823668241500854,
      "learning_rate": 9.051877324093114e-05,
      "loss": 0.0589,
      "relation_loss": 0.006771228555589914,
      "selected_pairs": 26963.6,
      "state_query_norm": 44.32973785400391,
      "state_relation_loss": 0.0004201360890874639,
      "step": 7180,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.006771228555589914,
      "weighted_relation_loss": 0.006771228555589914
    },
    {
      "action_query_norm": 43.63949851989746,
      "action_relation_loss": 0.006201302260160446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.725912067352666,
      "flow_matching_loss": 0.049984697997570035,
      "grad_norm": 0.20654639601707458,
      "learning_rate": 9.04864557630109e-05,
      "loss": 0.0566,
      "relation_loss": 0.006619548471644521,
      "selected_pairs": 26618.7,
      "state_query_norm": 44.32631950378418,
      "state_relation_loss": 0.00041824622603598983,
      "step": 7190,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.006619548471644521,
      "weighted_relation_loss": 0.006619548471644521
    },
    {
      "action_query_norm": 43.63776054382324,
      "action_relation_loss": 0.006308030337095261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.735266604303087,
      "flow_matching_loss": 0.05519898384809494,
      "grad_norm": 0.14731310307979584,
      "learning_rate": 9.045408909022928e-05,
      "loss": 0.0619,
      "relation_loss": 0.006716013373807073,
      "selected_pairs": 27004.5,
      "state_query_norm": 44.322716522216794,
      "state_relation_loss": 0.0004079830425325781,
      "step": 7200,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.006716013373807073,
      "weighted_relation_loss": 0.006716013373807073
    },
    {
      "action_query_norm": 43.65450210571289,
      "action_relation_loss": 0.0064496822655200955,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.744621141253508,
      "flow_matching_loss": 0.052130066230893134,
      "grad_norm": 0.2659134268760681,
      "learning_rate": 9.042167326191484e-05,
      "loss": 0.059,
      "relation_loss": 0.00686056767590344,
      "selected_pairs": 26989.9,
      "state_query_norm": 44.33084297180176,
      "state_relation_loss": 0.0004108854598598555,
      "step": 7210,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00686056767590344,
      "weighted_relation_loss": 0.00686056767590344
    },
    {
      "action_query_norm": 43.63700141906738,
      "action_relation_loss": 0.006352364644408226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.753975678203929,
      "flow_matching_loss": 0.053149442002177236,
      "grad_norm": 0.19089831411838531,
      "learning_rate": 9.038920831745587e-05,
      "loss": 0.0599,
      "relation_loss": 0.006762922322377562,
      "selected_pairs": 27083.9,
      "state_query_norm": 44.31668663024902,
      "state_relation_loss": 0.00041055766050703826,
      "step": 7220,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.006762922322377562,
      "weighted_relation_loss": 0.006762922322377562
    },
    {
      "action_query_norm": 43.64890022277832,
      "action_relation_loss": 0.006114320456981659,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.76333021515435,
      "flow_matching_loss": 0.05261532925069332,
      "grad_norm": 0.13430167734622955,
      "learning_rate": 9.035669429630036e-05,
      "loss": 0.0591,
      "relation_loss": 0.006491498230025173,
      "selected_pairs": 26840.5,
      "state_query_norm": 44.315002059936525,
      "state_relation_loss": 0.000377177819609642,
      "step": 7230,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006491498230025173,
      "weighted_relation_loss": 0.006491498230025173
    },
    {
      "action_query_norm": 43.603768539428714,
      "action_relation_loss": 0.0062743295449763535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.772684752104771,
      "flow_matching_loss": 0.05360403768718243,
      "grad_norm": 0.13787592947483063,
      "learning_rate": 9.032413123795588e-05,
      "loss": 0.0602,
      "relation_loss": 0.006626548524945974,
      "selected_pairs": 26490.1,
      "state_query_norm": 44.31579933166504,
      "state_relation_loss": 0.0003522190439980477,
      "step": 7240,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.006626548524945974,
      "weighted_relation_loss": 0.006626548524945974
    },
    {
      "action_query_norm": 43.62190589904785,
      "action_relation_loss": 0.006032337155193091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.7820392890551915,
      "flow_matching_loss": 0.051414640992879866,
      "grad_norm": 0.1528172791004181,
      "learning_rate": 9.029151918198962e-05,
      "loss": 0.0578,
      "relation_loss": 0.006426968006417155,
      "selected_pairs": 26466.0,
      "state_query_norm": 44.31224250793457,
      "state_relation_loss": 0.0003946308541344479,
      "step": 7250,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.006426968006417155,
      "weighted_relation_loss": 0.006426968006417155
    },
    {
      "action_query_norm": 43.6368221282959,
      "action_relation_loss": 0.006257679592818022,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.791393826005613,
      "flow_matching_loss": 0.0513981357216835,
      "grad_norm": 0.14482460916042328,
      "learning_rate": 9.025885816802833e-05,
      "loss": 0.0581,
      "relation_loss": 0.0066529251635074615,
      "selected_pairs": 26639.9,
      "state_query_norm": 44.31619377136231,
      "state_relation_loss": 0.00039524553285446017,
      "step": 7260,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0066529251635074615,
      "weighted_relation_loss": 0.0066529251635074615
    },
    {
      "action_query_norm": 43.62471389770508,
      "action_relation_loss": 0.006443737028166652,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.800748362956034,
      "flow_matching_loss": 0.052535079047083856,
      "grad_norm": 0.16399717330932617,
      "learning_rate": 9.02261482357582e-05,
      "loss": 0.0594,
      "relation_loss": 0.0068476570304483175,
      "selected_pairs": 26826.3,
      "state_query_norm": 44.316368865966794,
      "state_relation_loss": 0.00040391999937128276,
      "step": 7270,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0068476570304483175,
      "weighted_relation_loss": 0.0068476570304483175
    },
    {
      "action_query_norm": 43.624200439453126,
      "action_relation_loss": 0.006059239571914077,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.810102899906455,
      "flow_matching_loss": 0.05101130791008472,
      "grad_norm": 0.17438916862010956,
      "learning_rate": 9.019338942492485e-05,
      "loss": 0.0575,
      "relation_loss": 0.0064460319932550195,
      "selected_pairs": 26438.1,
      "state_query_norm": 44.304948043823245,
      "state_relation_loss": 0.00038679239514749497,
      "step": 7280,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0064460319932550195,
      "weighted_relation_loss": 0.0064460319932550195
    },
    {
      "action_query_norm": 43.62727317810059,
      "action_relation_loss": 0.006402757950127125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.819457436856876,
      "flow_matching_loss": 0.053443627431988716,
      "grad_norm": 0.17170938849449158,
      "learning_rate": 9.01605817753334e-05,
      "loss": 0.0602,
      "relation_loss": 0.0067882628180086614,
      "selected_pairs": 27134.6,
      "state_query_norm": 44.31164283752442,
      "state_relation_loss": 0.0003855048242257908,
      "step": 7290,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0067882628180086614,
      "weighted_relation_loss": 0.0067882628180086614
    },
    {
      "action_query_norm": 43.63328056335449,
      "action_relation_loss": 0.006268796836957335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.828811973807296,
      "flow_matching_loss": 0.05377762317657471,
      "grad_norm": 0.161981999874115,
      "learning_rate": 9.012772532684818e-05,
      "loss": 0.0604,
      "relation_loss": 0.006627411767840386,
      "selected_pairs": 26850.7,
      "state_query_norm": 44.30688629150391,
      "state_relation_loss": 0.0003586148755857721,
      "step": 7300,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006627411767840386,
      "weighted_relation_loss": 0.006627411767840386
    },
    {
      "action_query_norm": 43.60507316589356,
      "action_relation_loss": 0.006367624644190073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.838166510757717,
      "flow_matching_loss": 0.053394101187586786,
      "grad_norm": 0.22129853069782257,
      "learning_rate": 9.00948201193929e-05,
      "loss": 0.0602,
      "relation_loss": 0.006759726256132126,
      "selected_pairs": 26733.4,
      "state_query_norm": 44.309099960327146,
      "state_relation_loss": 0.00039210161776281893,
      "step": 7310,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006759726256132126,
      "weighted_relation_loss": 0.006759726256132126
    },
    {
      "action_query_norm": 43.605111312866214,
      "action_relation_loss": 0.006387083744630218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.847521047708138,
      "flow_matching_loss": 0.052593175694346427,
      "grad_norm": 0.13999220728874207,
      "learning_rate": 9.006186619295047e-05,
      "loss": 0.0594,
      "relation_loss": 0.006766820419579744,
      "selected_pairs": 26828.8,
      "state_query_norm": 44.304757308959964,
      "state_relation_loss": 0.00037973660801071674,
      "step": 7320,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006766820419579744,
      "weighted_relation_loss": 0.006766820419579744
    },
    {
      "action_query_norm": 43.58758010864258,
      "action_relation_loss": 0.006660006521269679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.856875584658559,
      "flow_matching_loss": 0.053844019025564197,
      "grad_norm": 0.15886016190052032,
      "learning_rate": 9.002886358756305e-05,
      "loss": 0.0609,
      "relation_loss": 0.007036220189183951,
      "selected_pairs": 27571.3,
      "state_query_norm": 44.30408821105957,
      "state_relation_loss": 0.00037621366791427133,
      "step": 7330,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.007036220189183951,
      "weighted_relation_loss": 0.007036220189183951
    },
    {
      "action_query_norm": 43.59145965576172,
      "action_relation_loss": 0.006123990099877119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.8662301216089805,
      "flow_matching_loss": 0.051048650220036504,
      "grad_norm": 0.16263249516487122,
      "learning_rate": 8.999581234333188e-05,
      "loss": 0.0576,
      "relation_loss": 0.0065146115142852064,
      "selected_pairs": 27056.8,
      "state_query_norm": 44.30025863647461,
      "state_relation_loss": 0.0003906214056769386,
      "step": 7340,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0065146115142852064,
      "weighted_relation_loss": 0.0065146115142852064
    },
    {
      "action_query_norm": 43.60072135925293,
      "action_relation_loss": 0.006203605607151985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.875584658559402,
      "flow_matching_loss": 0.04985675737261772,
      "grad_norm": 0.20019419491291046,
      "learning_rate": 8.996271250041734e-05,
      "loss": 0.0565,
      "relation_loss": 0.006600017054006457,
      "selected_pairs": 26802.9,
      "state_query_norm": 44.29436988830567,
      "state_relation_loss": 0.0003964114875998348,
      "step": 7350,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.006600017054006457,
      "weighted_relation_loss": 0.006600017054006457
    },
    {
      "action_query_norm": 43.591261291503905,
      "action_relation_loss": 0.00631485884077847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.884939195509823,
      "flow_matching_loss": 0.05251405350863934,
      "grad_norm": 0.15108731389045715,
      "learning_rate": 8.99295640990389e-05,
      "loss": 0.0592,
      "relation_loss": 0.006676772562786937,
      "selected_pairs": 27599.3,
      "state_query_norm": 44.30175094604492,
      "state_relation_loss": 0.0003619136434281245,
      "step": 7360,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.006676772562786937,
      "weighted_relation_loss": 0.006676772562786937
    },
    {
      "action_query_norm": 43.611694717407225,
      "action_relation_loss": 0.006295843422412873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.894293732460243,
      "flow_matching_loss": 0.05210021995007992,
      "grad_norm": 0.14164195954799652,
      "learning_rate": 8.989636717947494e-05,
      "loss": 0.0588,
      "relation_loss": 0.006655699713155628,
      "selected_pairs": 27337.1,
      "state_query_norm": 44.29655151367187,
      "state_relation_loss": 0.00035985640424769373,
      "step": 7370,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.006655699713155628,
      "weighted_relation_loss": 0.006655699713155628
    },
    {
      "action_query_norm": 43.59158210754394,
      "action_relation_loss": 0.0060938969720155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.903648269410664,
      "flow_matching_loss": 0.050044070556759836,
      "grad_norm": 0.13534332811832428,
      "learning_rate": 8.986312178206291e-05,
      "loss": 0.0565,
      "relation_loss": 0.006484280666336417,
      "selected_pairs": 26689.2,
      "state_query_norm": 44.29750785827637,
      "state_relation_loss": 0.000390383685589768,
      "step": 7380,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006484280666336417,
      "weighted_relation_loss": 0.006484280666336417
    },
    {
      "action_query_norm": 43.586494064331056,
      "action_relation_loss": 0.006148045836016536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.913002806361085,
      "flow_matching_loss": 0.05057233721017838,
      "grad_norm": 0.1820174753665924,
      "learning_rate": 8.982982794719904e-05,
      "loss": 0.0571,
      "relation_loss": 0.006520496914163232,
      "selected_pairs": 26234.2,
      "state_query_norm": 44.29037780761719,
      "state_relation_loss": 0.000372451136354357,
      "step": 7390,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.006520496914163232,
      "weighted_relation_loss": 0.006520496914163232
    },
    {
      "action_query_norm": 43.59878082275391,
      "action_relation_loss": 0.005773114925250411,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.922357343311506,
      "flow_matching_loss": 0.05259327590465546,
      "grad_norm": 0.15335537493228912,
      "learning_rate": 8.979648571533852e-05,
      "loss": 0.0587,
      "relation_loss": 0.006125276768580079,
      "selected_pairs": 27414.5,
      "state_query_norm": 44.29578857421875,
      "state_relation_loss": 0.00035216186370234936,
      "step": 7400,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.006125276768580079,
      "weighted_relation_loss": 0.006125276768580079
    },
    {
      "action_query_norm": 43.56861534118652,
      "action_relation_loss": 0.006284711882472038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.931711880261927,
      "flow_matching_loss": 0.049066074192523956,
      "grad_norm": 0.11916718631982803,
      "learning_rate": 8.976309512699528e-05,
      "loss": 0.0557,
      "relation_loss": 0.00661532636731863,
      "selected_pairs": 26938.9,
      "state_query_norm": 44.291995239257815,
      "state_relation_loss": 0.0003306145197711885,
      "step": 7410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.00661532636731863,
      "weighted_relation_loss": 0.00661532636731863
    },
    {
      "action_query_norm": 43.57400398254394,
      "action_relation_loss": 0.0064982619602233175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.9410664172123475,
      "flow_matching_loss": 0.05030223578214645,
      "grad_norm": 0.14640460908412933,
      "learning_rate": 8.972965622274205e-05,
      "loss": 0.0572,
      "relation_loss": 0.006863438664004207,
      "selected_pairs": 26712.5,
      "state_query_norm": 44.277852630615236,
      "state_relation_loss": 0.00036517675325740127,
      "step": 7420,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006863438664004207,
      "weighted_relation_loss": 0.006863438664004207
    },
    {
      "action_query_norm": 43.573107147216795,
      "action_relation_loss": 0.0059314107988029715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.950420954162769,
      "flow_matching_loss": 0.05061840005218983,
      "grad_norm": 0.139078289270401,
      "learning_rate": 8.969616904321026e-05,
      "loss": 0.0569,
      "relation_loss": 0.006271378370001912,
      "selected_pairs": 26823.7,
      "state_query_norm": 44.289798736572266,
      "state_relation_loss": 0.0003399675653781742,
      "step": 7430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.006271378370001912,
      "weighted_relation_loss": 0.006271378370001912
    },
    {
      "action_query_norm": 43.56813430786133,
      "action_relation_loss": 0.0066055514384061095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.95977549111319,
      "flow_matching_loss": 0.04981100633740425,
      "grad_norm": 0.1608988344669342,
      "learning_rate": 8.966263362908998e-05,
      "loss": 0.0568,
      "relation_loss": 0.006948178354650736,
      "selected_pairs": 26712.4,
      "state_query_norm": 44.285464477539065,
      "state_relation_loss": 0.000342626916244626,
      "step": 7440,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006948178354650736,
      "weighted_relation_loss": 0.006948178354650736
    },
    {
      "action_query_norm": 43.556173706054686,
      "action_relation_loss": 0.006428881222382188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.969130028063611,
      "flow_matching_loss": 0.053565800189971924,
      "grad_norm": 0.13106699287891388,
      "learning_rate": 8.962905002112989e-05,
      "loss": 0.0604,
      "relation_loss": 0.006798069970682263,
      "selected_pairs": 26691.8,
      "state_query_norm": 44.28373374938965,
      "state_relation_loss": 0.0003691886755404994,
      "step": 7450,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006798069970682263,
      "weighted_relation_loss": 0.006798069970682263
    },
    {
      "action_query_norm": 43.55745620727539,
      "action_relation_loss": 0.006601135293021798,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.978484565014032,
      "flow_matching_loss": 0.050289322063326834,
      "grad_norm": 0.11905795335769653,
      "learning_rate": 8.959541826013724e-05,
      "loss": 0.0573,
      "relation_loss": 0.006968747638165951,
      "selected_pairs": 26746.0,
      "state_query_norm": 44.27297019958496,
      "state_relation_loss": 0.00036761232477147134,
      "step": 7460,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006968747638165951,
      "weighted_relation_loss": 0.006968747638165951
    },
    {
      "action_query_norm": 43.56153106689453,
      "action_relation_loss": 0.006307610450312496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.987839101964453,
      "flow_matching_loss": 0.04872550517320633,
      "grad_norm": 0.14985044300556183,
      "learning_rate": 8.956173838697781e-05,
      "loss": 0.0554,
      "relation_loss": 0.006676748022437095,
      "selected_pairs": 26899.5,
      "state_query_norm": 44.27656936645508,
      "state_relation_loss": 0.0003691375808557495,
      "step": 7470,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006676748022437095,
      "weighted_relation_loss": 0.006676748022437095
    },
    {
      "action_query_norm": 43.55925674438477,
      "action_relation_loss": 0.006026756344363093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 6.997193638914874,
      "flow_matching_loss": 0.05182146802544594,
      "grad_norm": 0.12235304713249207,
      "learning_rate": 8.952801044257581e-05,
      "loss": 0.0582,
      "relation_loss": 0.006376396864652634,
      "selected_pairs": 26529.9,
      "state_query_norm": 44.278476333618165,
      "state_relation_loss": 0.0003496405406622216,
      "step": 7480,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.006376396864652634,
      "weighted_relation_loss": 0.006376396864652634
    },
    {
      "action_query_norm": 43.54938735961914,
      "action_relation_loss": 0.006111092353239656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.006548175865294,
      "flow_matching_loss": 0.05262735895812511,
      "grad_norm": 0.14119422435760498,
      "learning_rate": 8.949423446791387e-05,
      "loss": 0.0591,
      "relation_loss": 0.006503481790423394,
      "selected_pairs": 24532.8,
      "state_query_norm": 44.27404670715332,
      "state_relation_loss": 0.00039238953613676133,
      "step": 7490,
      "valid_action_samples": 215.2,
      "weighted_auxiliary_loss": 0.006503481790423394,
      "weighted_relation_loss": 0.006503481790423394
    },
    {
      "action_query_norm": 43.55806312561035,
      "action_relation_loss": 0.0059967918787151575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.015902712815715,
      "flow_matching_loss": 0.048252522200345996,
      "grad_norm": 0.14016462862491608,
      "learning_rate": 8.9460410504033e-05,
      "loss": 0.0546,
      "relation_loss": 0.006363831553608179,
      "selected_pairs": 26719.0,
      "state_query_norm": 44.27694511413574,
      "state_relation_loss": 0.00036703962832689284,
      "step": 7500,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.006363831553608179,
      "weighted_relation_loss": 0.006363831553608179
    },
    {
      "action_query_norm": 43.57368927001953,
      "action_relation_loss": 0.006076685432344675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.0252572497661365,
      "flow_matching_loss": 0.04872053787112236,
      "grad_norm": 0.1265110969543457,
      "learning_rate": 8.942653859203248e-05,
      "loss": 0.0552,
      "relation_loss": 0.006496957689523697,
      "selected_pairs": 27063.8,
      "state_query_norm": 44.26921157836914,
      "state_relation_loss": 0.0004202722368063405,
      "step": 7510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.006496957689523697,
      "weighted_relation_loss": 0.006496957689523697
    },
    {
      "action_query_norm": 43.55252265930176,
      "action_relation_loss": 0.005764368921518326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.034611786716558,
      "flow_matching_loss": 0.049311622977256775,
      "grad_norm": 0.1630803346633911,
      "learning_rate": 8.93926187730699e-05,
      "loss": 0.0555,
      "relation_loss": 0.006149688828736543,
      "selected_pairs": 27530.2,
      "state_query_norm": 44.264081954956055,
      "state_relation_loss": 0.000385319915949367,
      "step": 7520,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.006149688828736543,
      "weighted_relation_loss": 0.006149688828736543
    },
    {
      "action_query_norm": 43.53257942199707,
      "action_relation_loss": 0.005990970600396395,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.043966323666979,
      "flow_matching_loss": 0.05129353851079941,
      "grad_norm": 0.14351193606853485,
      "learning_rate": 8.935865108836102e-05,
      "loss": 0.0577,
      "relation_loss": 0.006368854502215982,
      "selected_pairs": 26868.9,
      "state_query_norm": 44.259802627563474,
      "state_relation_loss": 0.00037788396002724767,
      "step": 7530,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.006368854502215982,
      "weighted_relation_loss": 0.006368854502215982
    },
    {
      "action_query_norm": 43.52682952880859,
      "action_relation_loss": 0.006129305716603995,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.0533208606174,
      "flow_matching_loss": 0.05165530256927013,
      "grad_norm": 0.19962522387504578,
      "learning_rate": 8.932463557917981e-05,
      "loss": 0.0582,
      "relation_loss": 0.006504360353574157,
      "selected_pairs": 27058.6,
      "state_query_norm": 44.26646957397461,
      "state_relation_loss": 0.0003750546544324607,
      "step": 7540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.006504360353574157,
      "weighted_relation_loss": 0.006504360353574157
    },
    {
      "action_query_norm": 43.538062286376956,
      "action_relation_loss": 0.00587067254818976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.06267539756782,
      "flow_matching_loss": 0.04919847175478935,
      "grad_norm": 0.1637241542339325,
      "learning_rate": 8.929057228685829e-05,
      "loss": 0.0554,
      "relation_loss": 0.006245912844315171,
      "selected_pairs": 26162.2,
      "state_query_norm": 44.263811111450195,
      "state_relation_loss": 0.0003752402524696663,
      "step": 7550,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.006245912844315171,
      "weighted_relation_loss": 0.006245912844315171
    },
    {
      "action_query_norm": 43.540885543823244,
      "action_relation_loss": 0.0058673887979239225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.072029934518241,
      "flow_matching_loss": 0.04661035835742951,
      "grad_norm": 0.18103082478046417,
      "learning_rate": 8.925646125278658e-05,
      "loss": 0.0528,
      "relation_loss": 0.006229475373402238,
      "selected_pairs": 26683.5,
      "state_query_norm": 44.26259231567383,
      "state_relation_loss": 0.0003620865725679323,
      "step": 7560,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006229475373402238,
      "weighted_relation_loss": 0.006229475373402238
    },
    {
      "action_query_norm": 43.53704986572266,
      "action_relation_loss": 0.0058266724925488235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.081384471468662,
      "flow_matching_loss": 0.05108252502977848,
      "grad_norm": 0.12495385110378265,
      "learning_rate": 8.92223025184128e-05,
      "loss": 0.0573,
      "relation_loss": 0.006173052918165922,
      "selected_pairs": 26848.8,
      "state_query_norm": 44.261177444458006,
      "state_relation_loss": 0.00034638033539522437,
      "step": 7570,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006173052918165922,
      "weighted_relation_loss": 0.006173052918165922
    },
    {
      "action_query_norm": 43.53496551513672,
      "action_relation_loss": 0.005663750600069762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.090739008419083,
      "flow_matching_loss": 0.04807907454669476,
      "grad_norm": 0.10054165869951248,
      "learning_rate": 8.918809612524303e-05,
      "loss": 0.0541,
      "relation_loss": 0.006031820178031921,
      "selected_pairs": 26575.9,
      "state_query_norm": 44.26022453308106,
      "state_relation_loss": 0.0003680696536321193,
      "step": 7580,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.006031820178031921,
      "weighted_relation_loss": 0.006031820178031921
    },
    {
      "action_query_norm": 43.544730758666994,
      "action_relation_loss": 0.0061195445712655784,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.100093545369504,
      "flow_matching_loss": 0.048573162779212,
      "grad_norm": 0.16347071528434753,
      "learning_rate": 8.91538421148413e-05,
      "loss": 0.0551,
      "relation_loss": 0.006486164219677448,
      "selected_pairs": 26421.2,
      "state_query_norm": 44.252661895751956,
      "state_relation_loss": 0.00036661959311459216,
      "step": 7590,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.006486164219677448,
      "weighted_relation_loss": 0.006486164219677448
    },
    {
      "action_query_norm": 43.55088844299316,
      "action_relation_loss": 0.00602897247299552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1094480823199255,
      "flow_matching_loss": 0.04890548288822174,
      "grad_norm": 0.0912206843495369,
      "learning_rate": 8.91195405288294e-05,
      "loss": 0.0553,
      "relation_loss": 0.006390761304646731,
      "selected_pairs": 26797.9,
      "state_query_norm": 44.25759162902832,
      "state_relation_loss": 0.0003617887705331668,
      "step": 7600,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.006390761304646731,
      "weighted_relation_loss": 0.006390761304646731
    },
    {
      "action_query_norm": 43.52466316223145,
      "action_relation_loss": 0.005975332763046026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.118802619270346,
      "flow_matching_loss": 0.050148694217205046,
      "grad_norm": 0.1406974494457245,
      "learning_rate": 8.908519140888705e-05,
      "loss": 0.0565,
      "relation_loss": 0.006337573518976569,
      "selected_pairs": 26929.9,
      "state_query_norm": 44.25783157348633,
      "state_relation_loss": 0.00036224075593054294,
      "step": 7610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.006337573518976569,
      "weighted_relation_loss": 0.006337573518976569
    },
    {
      "action_query_norm": 43.521596908569336,
      "action_relation_loss": 0.0060242309235036375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.128157156220767,
      "flow_matching_loss": 0.049383503198623654,
      "grad_norm": 0.17729641497135162,
      "learning_rate": 8.905079479675164e-05,
      "loss": 0.0558,
      "relation_loss": 0.006377565115690232,
      "selected_pairs": 26339.0,
      "state_query_norm": 44.256282043457034,
      "state_relation_loss": 0.00035333425039425493,
      "step": 7620,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.006377565115690232,
      "weighted_relation_loss": 0.006377565115690232
    },
    {
      "action_query_norm": 43.543789291381835,
      "action_relation_loss": 0.005958688911050558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.137511693171188,
      "flow_matching_loss": 0.0479870043694973,
      "grad_norm": 0.1776260882616043,
      "learning_rate": 8.901635073421831e-05,
      "loss": 0.0543,
      "relation_loss": 0.0063352904748171564,
      "selected_pairs": 27088.2,
      "state_query_norm": 44.253784561157225,
      "state_relation_loss": 0.0003766015695873648,
      "step": 7630,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0063352904748171564,
      "weighted_relation_loss": 0.0063352904748171564
    },
    {
      "action_query_norm": 43.527339935302734,
      "action_relation_loss": 0.0059939509257674215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.146866230121609,
      "flow_matching_loss": 0.049339502677321435,
      "grad_norm": 0.11443939805030823,
      "learning_rate": 8.898185926313983e-05,
      "loss": 0.0557,
      "relation_loss": 0.0063676610589027405,
      "selected_pairs": 26827.2,
      "state_query_norm": 44.246952819824216,
      "state_relation_loss": 0.00037371014477685096,
      "step": 7640,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0063676610589027405,
      "weighted_relation_loss": 0.0063676610589027405
    },
    {
      "action_query_norm": 43.52555541992187,
      "action_relation_loss": 0.0056128958240151405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.15622076707203,
      "flow_matching_loss": 0.048354306071996686,
      "grad_norm": 0.13916480541229248,
      "learning_rate": 8.894732042542659e-05,
      "loss": 0.0543,
      "relation_loss": 0.005994708742946386,
      "selected_pairs": 27271.9,
      "state_query_norm": 44.24513549804688,
      "state_relation_loss": 0.0003818129742285237,
      "step": 7650,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005994708742946386,
      "weighted_relation_loss": 0.005994708742946386
    },
    {
      "action_query_norm": 43.526593399047854,
      "action_relation_loss": 0.006037887558341027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.165575304022451,
      "flow_matching_loss": 0.04999626912176609,
      "grad_norm": 0.12251188606023788,
      "learning_rate": 8.891273426304654e-05,
      "loss": 0.0564,
      "relation_loss": 0.006423807051032781,
      "selected_pairs": 26946.4,
      "state_query_norm": 44.24552268981934,
      "state_relation_loss": 0.0003859195334371179,
      "step": 7660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006423807051032781,
      "weighted_relation_loss": 0.006423807051032781
    },
    {
      "action_query_norm": 43.51315422058106,
      "action_relation_loss": 0.005715219536796212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1749298409728715,
      "flow_matching_loss": 0.04782939851284027,
      "grad_norm": 0.1603185385465622,
      "learning_rate": 8.887810081802515e-05,
      "loss": 0.0539,
      "relation_loss": 0.006093339063227177,
      "selected_pairs": 26694.4,
      "state_query_norm": 44.24037399291992,
      "state_relation_loss": 0.00037811954971402886,
      "step": 7670,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006093339063227177,
      "weighted_relation_loss": 0.006093339063227177
    },
    {
      "action_query_norm": 43.50674743652344,
      "action_relation_loss": 0.006257286062464118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.1842843779232926,
      "flow_matching_loss": 0.0488457378000021,
      "grad_norm": 0.1707419902086258,
      "learning_rate": 8.884342013244529e-05,
      "loss": 0.0555,
      "relation_loss": 0.006627061916515231,
      "selected_pairs": 27314.0,
      "state_query_norm": 44.242313385009766,
      "state_relation_loss": 0.0003697758715134114,
      "step": 7680,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.006627061916515231,
      "weighted_relation_loss": 0.006627061916515231
    },
    {
      "action_query_norm": 43.5134895324707,
      "action_relation_loss": 0.00602090461179614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.193638914873714,
      "flow_matching_loss": 0.051371942088007924,
      "grad_norm": 0.1627969741821289,
      "learning_rate": 8.880869224844725e-05,
      "loss": 0.0578,
      "relation_loss": 0.006379178259521723,
      "selected_pairs": 26902.5,
      "state_query_norm": 44.24578628540039,
      "state_relation_loss": 0.00035827368265017867,
      "step": 7690,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.006379178259521723,
      "weighted_relation_loss": 0.006379178259521723
    },
    {
      "action_query_norm": 43.50856399536133,
      "action_relation_loss": 0.005970504134893417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.202993451824135,
      "flow_matching_loss": 0.050541899725794794,
      "grad_norm": 0.17353017628192902,
      "learning_rate": 8.877391720822875e-05,
      "loss": 0.0569,
      "relation_loss": 0.00631998092867434,
      "selected_pairs": 26876.9,
      "state_query_norm": 44.24085311889648,
      "state_relation_loss": 0.00034947679087053983,
      "step": 7700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00631998092867434,
      "weighted_relation_loss": 0.00631998092867434
    },
    {
      "action_query_norm": 43.50540618896484,
      "action_relation_loss": 0.006010890891775489,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.212347988774556,
      "flow_matching_loss": 0.0494708314538002,
      "grad_norm": 0.14818966388702393,
      "learning_rate": 8.873909505404468e-05,
      "loss": 0.0558,
      "relation_loss": 0.006355317542329431,
      "selected_pairs": 26846.2,
      "state_query_norm": 44.23800621032715,
      "state_relation_loss": 0.00034442661271896215,
      "step": 7710,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006355317542329431,
      "weighted_relation_loss": 0.006355317542329431
    },
    {
      "action_query_norm": 43.50413284301758,
      "action_relation_loss": 0.005531033873558045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.221702525724977,
      "flow_matching_loss": 0.047774603590369225,
      "grad_norm": 0.13166667520999908,
      "learning_rate": 8.870422582820726e-05,
      "loss": 0.0537,
      "relation_loss": 0.005917236814275384,
      "selected_pairs": 26801.9,
      "state_query_norm": 44.23707313537598,
      "state_relation_loss": 0.0003862029290758073,
      "step": 7720,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005917236814275384,
      "weighted_relation_loss": 0.005917236814275384
    },
    {
      "action_query_norm": 43.516635513305665,
      "action_relation_loss": 0.006001623906195163,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.231057062675397,
      "flow_matching_loss": 0.050580498203635214,
      "grad_norm": 0.19468897581100464,
      "learning_rate": 8.866930957308588e-05,
      "loss": 0.0569,
      "relation_loss": 0.006365481298416853,
      "selected_pairs": 26943.4,
      "state_query_norm": 44.238145446777345,
      "state_relation_loss": 0.0003638574649812654,
      "step": 7730,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.006365481298416853,
      "weighted_relation_loss": 0.006365481298416853
    },
    {
      "action_query_norm": 43.51035041809082,
      "action_relation_loss": 0.005761670926585793,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.240411599625818,
      "flow_matching_loss": 0.048285913094878195,
      "grad_norm": 0.19483573734760284,
      "learning_rate": 8.86343463311071e-05,
      "loss": 0.0544,
      "relation_loss": 0.006104824226349592,
      "selected_pairs": 26987.8,
      "state_query_norm": 44.22932510375976,
      "state_relation_loss": 0.0003431532881222665,
      "step": 7740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.006104824226349592,
      "weighted_relation_loss": 0.006104824226349592
    },
    {
      "action_query_norm": 43.50794982910156,
      "action_relation_loss": 0.006196057563647628,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.249766136576239,
      "flow_matching_loss": 0.05096806734800339,
      "grad_norm": 0.19733865559101105,
      "learning_rate": 8.859933614475452e-05,
      "loss": 0.0575,
      "relation_loss": 0.006572640221565962,
      "selected_pairs": 27200.3,
      "state_query_norm": 44.23340759277344,
      "state_relation_loss": 0.0003765825240407139,
      "step": 7750,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.006572640221565962,
      "weighted_relation_loss": 0.006572640221565962
    },
    {
      "action_query_norm": 43.48115501403809,
      "action_relation_loss": 0.005853065010160208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.2591206735266605,
      "flow_matching_loss": 0.051800908893346785,
      "grad_norm": 0.17009155452251434,
      "learning_rate": 8.85642790565689e-05,
      "loss": 0.058,
      "relation_loss": 0.006206188537180424,
      "selected_pairs": 27312.9,
      "state_query_norm": 44.22821273803711,
      "state_relation_loss": 0.0003531235066475347,
      "step": 7760,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.006206188537180424,
      "weighted_relation_loss": 0.006206188537180424
    },
    {
      "action_query_norm": 43.454682159423825,
      "action_relation_loss": 0.005790727213025093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.268475210477082,
      "flow_matching_loss": 0.051579761505126956,
      "grad_norm": 0.13748620450496674,
      "learning_rate": 8.852917510914782e-05,
      "loss": 0.0577,
      "relation_loss": 0.006143357558175922,
      "selected_pairs": 27549.8,
      "state_query_norm": 44.22510757446289,
      "state_relation_loss": 0.0003526303276885301,
      "step": 7770,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.006143357558175922,
      "weighted_relation_loss": 0.006143357558175922
    },
    {
      "action_query_norm": 43.46779174804688,
      "action_relation_loss": 0.006054859701544046,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.277829747427503,
      "flow_matching_loss": 0.04976009726524353,
      "grad_norm": 0.10645550489425659,
      "learning_rate": 8.84940243451459e-05,
      "loss": 0.0562,
      "relation_loss": 0.006413512863218784,
      "selected_pairs": 27103.3,
      "state_query_norm": 44.22856979370117,
      "state_relation_loss": 0.00035865322570316494,
      "step": 7780,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.006413512863218784,
      "weighted_relation_loss": 0.006413512863218784
    },
    {
      "action_query_norm": 43.492257308959964,
      "action_relation_loss": 0.005839149001985788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.287184284377924,
      "flow_matching_loss": 0.05208116173744202,
      "grad_norm": 0.12510448694229126,
      "learning_rate": 8.845882680727469e-05,
      "loss": 0.0583,
      "relation_loss": 0.006202531140297651,
      "selected_pairs": 27064.9,
      "state_query_norm": 44.22684860229492,
      "state_relation_loss": 0.00036338215577416124,
      "step": 7790,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.006202531140297651,
      "weighted_relation_loss": 0.006202531140297651
    },
    {
      "action_query_norm": 43.47873420715332,
      "action_relation_loss": 0.006067987717688084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.296538821328344,
      "flow_matching_loss": 0.04986803866922855,
      "grad_norm": 0.17451153695583344,
      "learning_rate": 8.842358253830246e-05,
      "loss": 0.0563,
      "relation_loss": 0.00643425895832479,
      "selected_pairs": 26664.8,
      "state_query_norm": 44.226480102539064,
      "state_relation_loss": 0.0003662713250378147,
      "step": 7800,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00643425895832479,
      "weighted_relation_loss": 0.00643425895832479
    },
    {
      "action_query_norm": 43.47661094665527,
      "action_relation_loss": 0.006020184047520161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.305893358278765,
      "flow_matching_loss": 0.05161513760685921,
      "grad_norm": 0.17877434194087982,
      "learning_rate": 8.838829158105434e-05,
      "loss": 0.058,
      "relation_loss": 0.006384608009830118,
      "selected_pairs": 26546.3,
      "state_query_norm": 44.21848831176758,
      "state_relation_loss": 0.00036442397395148874,
      "step": 7810,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.006384608009830118,
      "weighted_relation_loss": 0.006384608009830118
    },
    {
      "action_query_norm": 43.48195953369141,
      "action_relation_loss": 0.006008056318387389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.315247895229186,
      "flow_matching_loss": 0.047982240468263625,
      "grad_norm": 0.14720413088798523,
      "learning_rate": 8.835295397841217e-05,
      "loss": 0.0544,
      "relation_loss": 0.006371916411444545,
      "selected_pairs": 26577.2,
      "state_query_norm": 44.22291145324707,
      "state_relation_loss": 0.0003638601745478809,
      "step": 7820,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.006371916411444545,
      "weighted_relation_loss": 0.006371916411444545
    },
    {
      "action_query_norm": 43.471374130249025,
      "action_relation_loss": 0.006057794205844402,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.324602432179607,
      "flow_matching_loss": 0.047831473499536516,
      "grad_norm": 0.149617001414299,
      "learning_rate": 8.831756977331447e-05,
      "loss": 0.0543,
      "relation_loss": 0.006431535910815,
      "selected_pairs": 26386.2,
      "state_query_norm": 44.222298812866214,
      "state_relation_loss": 0.0003737417107913643,
      "step": 7830,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.006431535910815,
      "weighted_relation_loss": 0.006431535910815
    },
    {
      "action_query_norm": 43.468461227416995,
      "action_relation_loss": 0.005835955403745175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.333956969130028,
      "flow_matching_loss": 0.05323540903627873,
      "grad_norm": 0.15312862396240234,
      "learning_rate": 8.828213900875638e-05,
      "loss": 0.0594,
      "relation_loss": 0.006203641416504979,
      "selected_pairs": 26689.0,
      "state_query_norm": 44.21633033752441,
      "state_relation_loss": 0.00036768594000022856,
      "step": 7840,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006203641416504979,
      "weighted_relation_loss": 0.006203641416504979
    },
    {
      "action_query_norm": 43.459790802001955,
      "action_relation_loss": 0.005649760412052274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.343311506080449,
      "flow_matching_loss": 0.0497539471834898,
      "grad_norm": 0.12501387298107147,
      "learning_rate": 8.824666172778964e-05,
      "loss": 0.0558,
      "relation_loss": 0.006010461086407304,
      "selected_pairs": 26765.9,
      "state_query_norm": 44.21207160949707,
      "state_relation_loss": 0.0003607006830861792,
      "step": 7850,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.006010461086407304,
      "weighted_relation_loss": 0.006010461086407304
    },
    {
      "action_query_norm": 43.46673698425293,
      "action_relation_loss": 0.006158857326954603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.35266604303087,
      "flow_matching_loss": 0.05063841007649898,
      "grad_norm": 0.15008506178855896,
      "learning_rate": 8.821113797352245e-05,
      "loss": 0.0572,
      "relation_loss": 0.006534096598625183,
      "selected_pairs": 26852.2,
      "state_query_norm": 44.214768981933595,
      "state_relation_loss": 0.0003752393851755187,
      "step": 7860,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006534096598625183,
      "weighted_relation_loss": 0.006534096598625183
    },
    {
      "action_query_norm": 43.46484375,
      "action_relation_loss": 0.0060466394759714605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.362020579981291,
      "flow_matching_loss": 0.048755685612559316,
      "grad_norm": 0.136449933052063,
      "learning_rate": 8.817556778911956e-05,
      "loss": 0.0552,
      "relation_loss": 0.006409980123862624,
      "selected_pairs": 26671.0,
      "state_query_norm": 44.219702911376956,
      "state_relation_loss": 0.0003633406711742282,
      "step": 7870,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.006409980123862624,
      "weighted_relation_loss": 0.006409980123862624
    },
    {
      "action_query_norm": 43.46007041931152,
      "action_relation_loss": 0.005687102722004056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.371375116931712,
      "flow_matching_loss": 0.0469893641769886,
      "grad_norm": 0.145870178937912,
      "learning_rate": 8.81399512178021e-05,
      "loss": 0.0531,
      "relation_loss": 0.0060685592703521255,
      "selected_pairs": 26755.8,
      "state_query_norm": 44.20895576477051,
      "state_relation_loss": 0.00038145653961692006,
      "step": 7880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0060685592703521255,
      "weighted_relation_loss": 0.0060685592703521255
    },
    {
      "action_query_norm": 43.45289344787598,
      "action_relation_loss": 0.00581656564027071,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.380729653882133,
      "flow_matching_loss": 0.047639507427811624,
      "grad_norm": 0.13382253050804138,
      "learning_rate": 8.810428830284752e-05,
      "loss": 0.0538,
      "relation_loss": 0.006156529067084193,
      "selected_pairs": 26500.4,
      "state_query_norm": 44.20793685913086,
      "state_relation_loss": 0.0003399634297238663,
      "step": 7890,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.006156529067084193,
      "weighted_relation_loss": 0.006156529067084193
    },
    {
      "action_query_norm": 43.45623207092285,
      "action_relation_loss": 0.005724583007395268,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.390084190832554,
      "flow_matching_loss": 0.04861746653914452,
      "grad_norm": 0.1349562704563141,
      "learning_rate": 8.806857908758967e-05,
      "loss": 0.0547,
      "relation_loss": 0.0060762483160942795,
      "selected_pairs": 26536.6,
      "state_query_norm": 44.20352325439453,
      "state_relation_loss": 0.00035166525631211696,
      "step": 7900,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0060762483160942795,
      "weighted_relation_loss": 0.0060762483160942795
    },
    {
      "action_query_norm": 43.43456192016602,
      "action_relation_loss": 0.005639436235651374,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.399438727782975,
      "flow_matching_loss": 0.048199590668082236,
      "grad_norm": 0.12542402744293213,
      "learning_rate": 8.80328236154186e-05,
      "loss": 0.0542,
      "relation_loss": 0.005971774971112609,
      "selected_pairs": 26556.1,
      "state_query_norm": 44.2058609008789,
      "state_relation_loss": 0.0003323387179989368,
      "step": 7910,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005971774971112609,
      "weighted_relation_loss": 0.005971774971112609
    },
    {
      "action_query_norm": 43.43721008300781,
      "action_relation_loss": 0.005576744768768549,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.408793264733395,
      "flow_matching_loss": 0.04926833175122738,
      "grad_norm": 0.1365237683057785,
      "learning_rate": 8.799702192978056e-05,
      "loss": 0.0552,
      "relation_loss": 0.005910018645226956,
      "selected_pairs": 25996.9,
      "state_query_norm": 44.2066650390625,
      "state_relation_loss": 0.000333273911383003,
      "step": 7920,
      "valid_action_samples": 228.4,
      "weighted_auxiliary_loss": 0.005910018645226956,
      "weighted_relation_loss": 0.005910018645226956
    },
    {
      "action_query_norm": 43.434150314331056,
      "action_relation_loss": 0.005854792799800635,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.4181478016838165,
      "flow_matching_loss": 0.04789751544594765,
      "grad_norm": 0.1297348290681839,
      "learning_rate": 8.7961174074178e-05,
      "loss": 0.0541,
      "relation_loss": 0.0061805086676031355,
      "selected_pairs": 26708.8,
      "state_query_norm": 44.2059684753418,
      "state_relation_loss": 0.0003257158532505855,
      "step": 7930,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0061805086676031355,
      "weighted_relation_loss": 0.0061805086676031355
    },
    {
      "action_query_norm": 43.43715705871582,
      "action_relation_loss": 0.005759506532922387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.427502338634238,
      "flow_matching_loss": 0.04646976739168167,
      "grad_norm": 0.11065036803483963,
      "learning_rate": 8.792528009216942e-05,
      "loss": 0.0525,
      "relation_loss": 0.006079491041600704,
      "selected_pairs": 26739.5,
      "state_query_norm": 44.206816864013675,
      "state_relation_loss": 0.00031998447084333745,
      "step": 7940,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006079491041600704,
      "weighted_relation_loss": 0.006079491041600704
    },
    {
      "action_query_norm": 43.43337135314941,
      "action_relation_loss": 0.0059046127833426,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.436856875584659,
      "flow_matching_loss": 0.04973817765712738,
      "grad_norm": 0.13777494430541992,
      "learning_rate": 8.788934002736943e-05,
      "loss": 0.056,
      "relation_loss": 0.006263691326603293,
      "selected_pairs": 26670.3,
      "state_query_norm": 44.189432525634764,
      "state_relation_loss": 0.00035907852870877834,
      "step": 7950,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.006263691326603293,
      "weighted_relation_loss": 0.006263691326603293
    },
    {
      "action_query_norm": 43.44487075805664,
      "action_relation_loss": 0.005789294466376304,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.44621141253508,
      "flow_matching_loss": 0.05019766576588154,
      "grad_norm": 0.10322604328393936,
      "learning_rate": 8.785335392344857e-05,
      "loss": 0.0563,
      "relation_loss": 0.006127710780128836,
      "selected_pairs": 26644.4,
      "state_query_norm": 44.20276794433594,
      "state_relation_loss": 0.00033841639815364034,
      "step": 7960,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.006127710780128836,
      "weighted_relation_loss": 0.006127710780128836
    },
    {
      "action_query_norm": 43.45174865722656,
      "action_relation_loss": 0.005747606977820396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.4555659494855,
      "flow_matching_loss": 0.04942052476108074,
      "grad_norm": 0.14478513598442078,
      "learning_rate": 8.781732182413336e-05,
      "loss": 0.0555,
      "relation_loss": 0.00609310888685286,
      "selected_pairs": 26787.1,
      "state_query_norm": 44.19245223999023,
      "state_relation_loss": 0.0003455019585089758,
      "step": 7970,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.00609310888685286,
      "weighted_relation_loss": 0.00609310888685286
    },
    {
      "action_query_norm": 43.4489242553711,
      "action_relation_loss": 0.0056494992226362225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.464920486435921,
      "flow_matching_loss": 0.04982063136994839,
      "grad_norm": 0.11276188492774963,
      "learning_rate": 8.778124377320618e-05,
      "loss": 0.0558,
      "relation_loss": 0.006002359976992011,
      "selected_pairs": 26275.4,
      "state_query_norm": 44.19621696472168,
      "state_relation_loss": 0.0003528607776388526,
      "step": 7980,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.006002359976992011,
      "weighted_relation_loss": 0.006002359976992011
    },
    {
      "action_query_norm": 43.445777130126956,
      "action_relation_loss": 0.005839629471302033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.474275023386342,
      "flow_matching_loss": 0.04806035459041595,
      "grad_norm": 0.127560093998909,
      "learning_rate": 8.774511981450529e-05,
      "loss": 0.0543,
      "relation_loss": 0.006190181383863092,
      "selected_pairs": 26989.2,
      "state_query_norm": 44.20588493347168,
      "state_relation_loss": 0.0003505519562168047,
      "step": 7990,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.006190181383863092,
      "weighted_relation_loss": 0.006190181383863092
    },
    {
      "action_query_norm": 43.44046821594238,
      "action_relation_loss": 0.0055622969754040245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.483629560336763,
      "flow_matching_loss": 0.051319151371717456,
      "grad_norm": 0.10864349454641342,
      "learning_rate": 8.770894999192467e-05,
      "loss": 0.0572,
      "relation_loss": 0.005911046825349331,
      "selected_pairs": 26742.7,
      "state_query_norm": 44.19379920959473,
      "state_relation_loss": 0.00034874990524258465,
      "step": 8000,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005911046825349331,
      "weighted_relation_loss": 0.005911046825349331
    },
    {
      "action_query_norm": 43.44916687011719,
      "action_relation_loss": 0.00569970109499991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.492984097287184,
      "flow_matching_loss": 0.04871971867978573,
      "grad_norm": 0.1463811844587326,
      "learning_rate": 8.767273434941412e-05,
      "loss": 0.0548,
      "relation_loss": 0.006053097173571586,
      "selected_pairs": 26966.4,
      "state_query_norm": 44.189362335205075,
      "state_relation_loss": 0.0003533960960339755,
      "step": 8010,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.006053097173571586,
      "weighted_relation_loss": 0.006053097173571586
    },
    {
      "action_query_norm": 43.437312316894534,
      "action_relation_loss": 0.00608223294839263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.5023386342376055,
      "flow_matching_loss": 0.051475830003619195,
      "grad_norm": 0.14335897564888,
      "learning_rate": 8.7636472930979e-05,
      "loss": 0.0579,
      "relation_loss": 0.006446232786402106,
      "selected_pairs": 26736.2,
      "state_query_norm": 44.19231910705567,
      "state_relation_loss": 0.0003639998933067545,
      "step": 8020,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.006446232786402106,
      "weighted_relation_loss": 0.006446232786402106
    },
    {
      "action_query_norm": 43.42927551269531,
      "action_relation_loss": 0.0056179664097726345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.511693171188027,
      "flow_matching_loss": 0.047309853881597516,
      "grad_norm": 0.1451500803232193,
      "learning_rate": 8.76001657806804e-05,
      "loss": 0.0533,
      "relation_loss": 0.005980340810492635,
      "selected_pairs": 27082.3,
      "state_query_norm": 44.18317832946777,
      "state_relation_loss": 0.00036237452586647124,
      "step": 8030,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005980340810492635,
      "weighted_relation_loss": 0.005980340810492635
    },
    {
      "action_query_norm": 43.41172370910645,
      "action_relation_loss": 0.005732928961515426,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.521047708138447,
      "flow_matching_loss": 0.04914323575794697,
      "grad_norm": 0.13115575909614563,
      "learning_rate": 8.756381294263491e-05,
      "loss": 0.0552,
      "relation_loss": 0.006056242017075419,
      "selected_pairs": 26593.9,
      "state_query_norm": 44.18758316040039,
      "state_relation_loss": 0.0003233130235457793,
      "step": 8040,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.006056242017075419,
      "weighted_relation_loss": 0.006056242017075419
    },
    {
      "action_query_norm": 43.40832405090332,
      "action_relation_loss": 0.006021651998162269,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.530402245088868,
      "flow_matching_loss": 0.04820681773126125,
      "grad_norm": 0.12961575388908386,
      "learning_rate": 8.752741446101464e-05,
      "loss": 0.0546,
      "relation_loss": 0.006362974923104048,
      "selected_pairs": 26991.3,
      "state_query_norm": 44.18642349243164,
      "state_relation_loss": 0.00034132291038986297,
      "step": 8050,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.006362974923104048,
      "weighted_relation_loss": 0.006362974923104048
    },
    {
      "action_query_norm": 43.41553802490235,
      "action_relation_loss": 0.005694916471838951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.539756782039289,
      "flow_matching_loss": 0.04774831458926201,
      "grad_norm": 0.14316326379776,
      "learning_rate": 8.749097038004722e-05,
      "loss": 0.0538,
      "relation_loss": 0.006053094705566764,
      "selected_pairs": 27456.2,
      "state_query_norm": 44.186039733886716,
      "state_relation_loss": 0.0003581782220862806,
      "step": 8060,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.006053094705566764,
      "weighted_relation_loss": 0.006053094705566764
    },
    {
      "action_query_norm": 43.42279472351074,
      "action_relation_loss": 0.005576863046735525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.54911131898971,
      "flow_matching_loss": 0.04815117716789245,
      "grad_norm": 0.14321987330913544,
      "learning_rate": 8.745448074401561e-05,
      "loss": 0.0541,
      "relation_loss": 0.005909721786156297,
      "selected_pairs": 26803.1,
      "state_query_norm": 44.183477783203124,
      "state_relation_loss": 0.00033285870449617505,
      "step": 8070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005909721786156297,
      "weighted_relation_loss": 0.005909721786156297
    },
    {
      "action_query_norm": 43.420305252075195,
      "action_relation_loss": 0.005815601814538241,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.558465855940131,
      "flow_matching_loss": 0.0493345845490694,
      "grad_norm": 0.17183557152748108,
      "learning_rate": 8.741794559725818e-05,
      "loss": 0.0555,
      "relation_loss": 0.0061460454016923904,
      "selected_pairs": 27639.3,
      "state_query_norm": 44.17973098754883,
      "state_relation_loss": 0.00033044365991372613,
      "step": 8080,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0061460454016923904,
      "weighted_relation_loss": 0.0061460454016923904
    },
    {
      "action_query_norm": 43.4058536529541,
      "action_relation_loss": 0.005914829205721617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.567820392890552,
      "flow_matching_loss": 0.05056015960872173,
      "grad_norm": 0.13811837136745453,
      "learning_rate": 8.738136498416856e-05,
      "loss": 0.0568,
      "relation_loss": 0.006247008638456464,
      "selected_pairs": 26410.0,
      "state_query_norm": 44.16874656677246,
      "state_relation_loss": 0.00033217941236216573,
      "step": 8090,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.006247008638456464,
      "weighted_relation_loss": 0.006247008638456464
    },
    {
      "action_query_norm": 43.401427841186525,
      "action_relation_loss": 0.005560670606791973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.5771749298409725,
      "flow_matching_loss": 0.04719761051237583,
      "grad_norm": 0.12229588627815247,
      "learning_rate": 8.734473894919564e-05,
      "loss": 0.0531,
      "relation_loss": 0.005900799063965678,
      "selected_pairs": 27361.5,
      "state_query_norm": 44.1740966796875,
      "state_relation_loss": 0.0003401284251594916,
      "step": 8100,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.005900799063965678,
      "weighted_relation_loss": 0.005900799063965678
    },
    {
      "action_query_norm": 43.3788444519043,
      "action_relation_loss": 0.006119791837409139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.586529466791394,
      "flow_matching_loss": 0.04819631353020668,
      "grad_norm": 0.15366649627685547,
      "learning_rate": 8.730806753684353e-05,
      "loss": 0.0547,
      "relation_loss": 0.006465612957254052,
      "selected_pairs": 26233.7,
      "state_query_norm": 44.17488098144531,
      "state_relation_loss": 0.0003458211576798931,
      "step": 8110,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.006465612957254052,
      "weighted_relation_loss": 0.006465612957254052
    },
    {
      "action_query_norm": 43.40079040527344,
      "action_relation_loss": 0.005979448324069381,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.595884003741815,
      "flow_matching_loss": 0.05028989873826504,
      "grad_norm": 0.1573653370141983,
      "learning_rate": 8.727135079167144e-05,
      "loss": 0.0566,
      "relation_loss": 0.0063179376069456335,
      "selected_pairs": 26593.6,
      "state_query_norm": 44.1758228302002,
      "state_relation_loss": 0.000338489274145104,
      "step": 8120,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0063179376069456335,
      "weighted_relation_loss": 0.0063179376069456335
    },
    {
      "action_query_norm": 43.38109016418457,
      "action_relation_loss": 0.005751015804708004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.605238540692236,
      "flow_matching_loss": 0.04833018556237221,
      "grad_norm": 0.12054584175348282,
      "learning_rate": 8.723458875829368e-05,
      "loss": 0.0544,
      "relation_loss": 0.006088179023936391,
      "selected_pairs": 27317.2,
      "state_query_norm": 44.16715431213379,
      "state_relation_loss": 0.00033716316684149204,
      "step": 8130,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.006088179023936391,
      "weighted_relation_loss": 0.006088179023936391
    },
    {
      "action_query_norm": 43.3971607208252,
      "action_relation_loss": 0.005911216605454683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.614593077642657,
      "flow_matching_loss": 0.04930594153702259,
      "grad_norm": 0.12709547579288483,
      "learning_rate": 8.719778148137958e-05,
      "loss": 0.0556,
      "relation_loss": 0.006249718787148595,
      "selected_pairs": 27218.6,
      "state_query_norm": 44.17415199279785,
      "state_relation_loss": 0.00033850221370812505,
      "step": 8140,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.006249718787148595,
      "weighted_relation_loss": 0.006249718787148595
    },
    {
      "action_query_norm": 43.39917068481445,
      "action_relation_loss": 0.005567954201251268,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.623947614593078,
      "flow_matching_loss": 0.05128689259290695,
      "grad_norm": 0.14865291118621826,
      "learning_rate": 8.716092900565346e-05,
      "loss": 0.0572,
      "relation_loss": 0.005894651729613542,
      "selected_pairs": 27339.6,
      "state_query_norm": 44.17418212890625,
      "state_relation_loss": 0.0003266975312726572,
      "step": 8150,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.005894651729613542,
      "weighted_relation_loss": 0.005894651729613542
    },
    {
      "action_query_norm": 43.37378044128418,
      "action_relation_loss": 0.005791318882256746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.633302151543498,
      "flow_matching_loss": 0.04833954609930515,
      "grad_norm": 0.17818289995193481,
      "learning_rate": 8.712403137589456e-05,
      "loss": 0.0545,
      "relation_loss": 0.0061305140145123,
      "selected_pairs": 26443.3,
      "state_query_norm": 44.15860099792481,
      "state_relation_loss": 0.00033919503330253064,
      "step": 8160,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0061305140145123,
      "weighted_relation_loss": 0.0061305140145123
    },
    {
      "action_query_norm": 43.38366851806641,
      "action_relation_loss": 0.005578132532536984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.642656688493919,
      "flow_matching_loss": 0.0454193789511919,
      "grad_norm": 0.11627157032489777,
      "learning_rate": 8.708708863693697e-05,
      "loss": 0.0513,
      "relation_loss": 0.005920421238988638,
      "selected_pairs": 26592.6,
      "state_query_norm": 44.153174591064456,
      "state_relation_loss": 0.00034228867443744095,
      "step": 8170,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.005920421238988638,
      "weighted_relation_loss": 0.005920421238988638
    },
    {
      "action_query_norm": 43.37848739624023,
      "action_relation_loss": 0.005457812175154686,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.65201122544434,
      "flow_matching_loss": 0.046919526532292366,
      "grad_norm": 0.154556006193161,
      "learning_rate": 8.70501008336696e-05,
      "loss": 0.0527,
      "relation_loss": 0.005785214528441429,
      "selected_pairs": 26625.7,
      "state_query_norm": 44.15934104919434,
      "state_relation_loss": 0.00032740232127252965,
      "step": 8180,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005785214528441429,
      "weighted_relation_loss": 0.005785214528441429
    },
    {
      "action_query_norm": 43.36198463439941,
      "action_relation_loss": 0.005979861039668322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.6613657623947615,
      "flow_matching_loss": 0.04883704334497452,
      "grad_norm": 0.10916616022586823,
      "learning_rate": 8.701306801103611e-05,
      "loss": 0.0552,
      "relation_loss": 0.006316413544118405,
      "selected_pairs": 26687.9,
      "state_query_norm": 44.15739631652832,
      "state_relation_loss": 0.00033655244624242187,
      "step": 8190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.006316413544118405,
      "weighted_relation_loss": 0.006316413544118405
    },
    {
      "action_query_norm": 43.37220916748047,
      "action_relation_loss": 0.0058976980857551094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.670720299345183,
      "flow_matching_loss": 0.05072715207934379,
      "grad_norm": 0.17696405947208405,
      "learning_rate": 8.697599021403489e-05,
      "loss": 0.057,
      "relation_loss": 0.006237948033958674,
      "selected_pairs": 26757.4,
      "state_query_norm": 44.15495262145996,
      "state_relation_loss": 0.00034024992492049934,
      "step": 8200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.006237948033958674,
      "weighted_relation_loss": 0.006237948033958674
    },
    {
      "action_query_norm": 43.37020721435547,
      "action_relation_loss": 0.0057137522380799055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.680074836295604,
      "flow_matching_loss": 0.04855022542178631,
      "grad_norm": 0.1744249165058136,
      "learning_rate": 8.693886748771896e-05,
      "loss": 0.0546,
      "relation_loss": 0.006078543514013291,
      "selected_pairs": 26844.6,
      "state_query_norm": 44.15210494995117,
      "state_relation_loss": 0.00036479124682955446,
      "step": 8210,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.006078543514013291,
      "weighted_relation_loss": 0.006078543514013291
    },
    {
      "action_query_norm": 43.37037200927735,
      "action_relation_loss": 0.0056789205875247715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.689429373246024,
      "flow_matching_loss": 0.04532674625515938,
      "grad_norm": 0.0952664390206337,
      "learning_rate": 8.690169987719593e-05,
      "loss": 0.0513,
      "relation_loss": 0.00601778132840991,
      "selected_pairs": 26694.2,
      "state_query_norm": 44.15265998840332,
      "state_relation_loss": 0.00033886072051245717,
      "step": 8220,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00601778132840991,
      "weighted_relation_loss": 0.00601778132840991
    },
    {
      "action_query_norm": 43.36976928710938,
      "action_relation_loss": 0.006033993232995272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.698783910196445,
      "flow_matching_loss": 0.049701114743947984,
      "grad_norm": 0.1361154019832611,
      "learning_rate": 8.686448742762791e-05,
      "loss": 0.0561,
      "relation_loss": 0.006384985521435738,
      "selected_pairs": 27404.6,
      "state_query_norm": 44.155325698852536,
      "state_relation_loss": 0.00035099225933663545,
      "step": 8230,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.006384985521435738,
      "weighted_relation_loss": 0.006384985521435738
    },
    {
      "action_query_norm": 43.364118576049805,
      "action_relation_loss": 0.005606643483042717,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.708138447146866,
      "flow_matching_loss": 0.04563350193202496,
      "grad_norm": 0.09529143571853638,
      "learning_rate": 8.682723018423161e-05,
      "loss": 0.0516,
      "relation_loss": 0.005948556773364544,
      "selected_pairs": 26593.1,
      "state_query_norm": 44.146893692016604,
      "state_relation_loss": 0.0003419133514398709,
      "step": 8240,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.005948556773364544,
      "weighted_relation_loss": 0.005948556773364544
    },
    {
      "action_query_norm": 43.361310195922854,
      "action_relation_loss": 0.005525015527382493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.717492984097287,
      "flow_matching_loss": 0.04622501730918884,
      "grad_norm": 0.15881748497486115,
      "learning_rate": 8.678992819227804e-05,
      "loss": 0.0521,
      "relation_loss": 0.00586432246491313,
      "selected_pairs": 26852.4,
      "state_query_norm": 44.151258850097655,
      "state_relation_loss": 0.00033930692879948767,
      "step": 8250,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.00586432246491313,
      "weighted_relation_loss": 0.00586432246491313
    },
    {
      "action_query_norm": 43.35429420471191,
      "action_relation_loss": 0.00548310331068933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.726847521047708,
      "flow_matching_loss": 0.04764912165701389,
      "grad_norm": 0.15188752114772797,
      "learning_rate": 8.675258149709265e-05,
      "loss": 0.0535,
      "relation_loss": 0.005829891795292497,
      "selected_pairs": 26673.4,
      "state_query_norm": 44.13902893066406,
      "state_relation_loss": 0.00034678851952776315,
      "step": 8260,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.005829891795292497,
      "weighted_relation_loss": 0.005829891795292497
    },
    {
      "action_query_norm": 43.350775146484374,
      "action_relation_loss": 0.005560228461399674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.736202057998129,
      "flow_matching_loss": 0.045935691148042676,
      "grad_norm": 0.11567998677492142,
      "learning_rate": 8.67151901440552e-05,
      "loss": 0.0518,
      "relation_loss": 0.00589366196654737,
      "selected_pairs": 26626.7,
      "state_query_norm": 44.141313552856445,
      "state_relation_loss": 0.00033343348186463115,
      "step": 8270,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00589366196654737,
      "weighted_relation_loss": 0.00589366196654737
    },
    {
      "action_query_norm": 43.34155464172363,
      "action_relation_loss": 0.005731506645679474,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.74555659494855,
      "flow_matching_loss": 0.04810691811144352,
      "grad_norm": 0.1440669447183609,
      "learning_rate": 8.667775417859971e-05,
      "loss": 0.0542,
      "relation_loss": 0.006065547466278076,
      "selected_pairs": 27274.3,
      "state_query_norm": 44.14329071044922,
      "state_relation_loss": 0.00033404076530132445,
      "step": 8280,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.006065547466278076,
      "weighted_relation_loss": 0.006065547466278076
    },
    {
      "action_query_norm": 43.34480857849121,
      "action_relation_loss": 0.005685587134212256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.754911131898971,
      "flow_matching_loss": 0.046323123201727864,
      "grad_norm": 0.14092214405536652,
      "learning_rate": 8.664027364621441e-05,
      "loss": 0.0524,
      "relation_loss": 0.006031195959076285,
      "selected_pairs": 27181.9,
      "state_query_norm": 44.13786735534668,
      "state_relation_loss": 0.00034560874628368764,
      "step": 8290,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.006031195959076285,
      "weighted_relation_loss": 0.006031195959076285
    },
    {
      "action_query_norm": 43.36500205993652,
      "action_relation_loss": 0.005649807630106807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.764265668849392,
      "flow_matching_loss": 0.04622671753168106,
      "grad_norm": 0.14978404343128204,
      "learning_rate": 8.660274859244168e-05,
      "loss": 0.0522,
      "relation_loss": 0.0059962878469377754,
      "selected_pairs": 27406.4,
      "state_query_norm": 44.13841285705566,
      "state_relation_loss": 0.0003464802197413519,
      "step": 8300,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0059962878469377754,
      "weighted_relation_loss": 0.0059962878469377754
    },
    {
      "action_query_norm": 43.358903121948245,
      "action_relation_loss": 0.005557329533621669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.773620205799813,
      "flow_matching_loss": 0.04781529940664768,
      "grad_norm": 0.1475232094526291,
      "learning_rate": 8.656517906287799e-05,
      "loss": 0.0537,
      "relation_loss": 0.005905397562310099,
      "selected_pairs": 27082.1,
      "state_query_norm": 44.13456268310547,
      "state_relation_loss": 0.00034806803159881383,
      "step": 8310,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005905397562310099,
      "weighted_relation_loss": 0.005905397562310099
    },
    {
      "action_query_norm": 43.3277027130127,
      "action_relation_loss": 0.005616369005292654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.782974742750234,
      "flow_matching_loss": 0.04910811968147755,
      "grad_norm": 0.14765411615371704,
      "learning_rate": 8.652756510317387e-05,
      "loss": 0.055,
      "relation_loss": 0.005930189369246363,
      "selected_pairs": 27179.3,
      "state_query_norm": 44.14152565002441,
      "state_relation_loss": 0.00031382028246298435,
      "step": 8320,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.005930189369246363,
      "weighted_relation_loss": 0.005930189369246363
    },
    {
      "action_query_norm": 43.34097595214844,
      "action_relation_loss": 0.005590955074876547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.792329279700655,
      "flow_matching_loss": 0.04689528495073318,
      "grad_norm": 0.13899008929729462,
      "learning_rate": 8.648990675903382e-05,
      "loss": 0.0528,
      "relation_loss": 0.005918074212968349,
      "selected_pairs": 27013.8,
      "state_query_norm": 44.138546752929685,
      "state_relation_loss": 0.0003271191759267822,
      "step": 8330,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.005918074212968349,
      "weighted_relation_loss": 0.005918074212968349
    },
    {
      "action_query_norm": 43.33363761901855,
      "action_relation_loss": 0.00575435752980411,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.801683816651076,
      "flow_matching_loss": 0.04826635383069515,
      "grad_norm": 0.14553126692771912,
      "learning_rate": 8.64522040762163e-05,
      "loss": 0.0544,
      "relation_loss": 0.00609274972230196,
      "selected_pairs": 27175.6,
      "state_query_norm": 44.1339729309082,
      "state_relation_loss": 0.0003383922768989578,
      "step": 8340,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.00609274972230196,
      "weighted_relation_loss": 0.00609274972230196
    },
    {
      "action_query_norm": 43.34918441772461,
      "action_relation_loss": 0.005783038027584553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8110383536014965,
      "flow_matching_loss": 0.04657166227698326,
      "grad_norm": 0.12525223195552826,
      "learning_rate": 8.64144571005336e-05,
      "loss": 0.0527,
      "relation_loss": 0.006125221680849791,
      "selected_pairs": 26490.7,
      "state_query_norm": 44.13686637878418,
      "state_relation_loss": 0.00034218372020404785,
      "step": 8350,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.006125221680849791,
      "weighted_relation_loss": 0.006125221680849791
    },
    {
      "action_query_norm": 43.33890838623047,
      "action_relation_loss": 0.0054940832778811455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8203928905519176,
      "flow_matching_loss": 0.04698893092572689,
      "grad_norm": 0.1681271344423294,
      "learning_rate": 8.637666587785184e-05,
      "loss": 0.0528,
      "relation_loss": 0.005840314645320177,
      "selected_pairs": 26507.8,
      "state_query_norm": 44.12945137023926,
      "state_relation_loss": 0.0003462313732597977,
      "step": 8360,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.005840314645320177,
      "weighted_relation_loss": 0.005840314645320177
    },
    {
      "action_query_norm": 43.32399139404297,
      "action_relation_loss": 0.005233243061229587,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.829747427502339,
      "flow_matching_loss": 0.04886344410479069,
      "grad_norm": 0.155411496758461,
      "learning_rate": 8.633883045409096e-05,
      "loss": 0.0544,
      "relation_loss": 0.005577915674075484,
      "selected_pairs": 26873.5,
      "state_query_norm": 44.12381362915039,
      "state_relation_loss": 0.0003446726448601112,
      "step": 8370,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005577915674075484,
      "weighted_relation_loss": 0.005577915674075484
    },
    {
      "action_query_norm": 43.32217559814453,
      "action_relation_loss": 0.005658365320414305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.83910196445276,
      "flow_matching_loss": 0.04782414622604847,
      "grad_norm": 0.15121373534202576,
      "learning_rate": 8.630095087522457e-05,
      "loss": 0.0538,
      "relation_loss": 0.005991380475461483,
      "selected_pairs": 26778.5,
      "state_query_norm": 44.12671127319336,
      "state_relation_loss": 0.00033301514631602915,
      "step": 8380,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005991380475461483,
      "weighted_relation_loss": 0.005991380475461483
    },
    {
      "action_query_norm": 43.31742973327637,
      "action_relation_loss": 0.00567414602264762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.848456501403181,
      "flow_matching_loss": 0.050271684676408766,
      "grad_norm": 0.1600276678800583,
      "learning_rate": 8.62630271872799e-05,
      "loss": 0.0563,
      "relation_loss": 0.005988317215815187,
      "selected_pairs": 26247.1,
      "state_query_norm": 44.125614166259766,
      "state_relation_loss": 0.00031417118152603506,
      "step": 8390,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.005988317215815187,
      "weighted_relation_loss": 0.005988317215815187
    },
    {
      "action_query_norm": 43.3225040435791,
      "action_relation_loss": 0.005356709938496351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.857811038353601,
      "flow_matching_loss": 0.04613002277910709,
      "grad_norm": 0.15910162031650543,
      "learning_rate": 8.62250594363378e-05,
      "loss": 0.0518,
      "relation_loss": 0.005693492060527205,
      "selected_pairs": 26954.1,
      "state_query_norm": 44.123421096801756,
      "state_relation_loss": 0.0003367821016581729,
      "step": 8400,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.005693492060527205,
      "weighted_relation_loss": 0.005693492060527205
    },
    {
      "action_query_norm": 43.31966857910156,
      "action_relation_loss": 0.005444920249283313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.867165575304022,
      "flow_matching_loss": 0.04661489948630333,
      "grad_norm": 0.09779112040996552,
      "learning_rate": 8.61870476685327e-05,
      "loss": 0.0524,
      "relation_loss": 0.005751287704333663,
      "selected_pairs": 26431.7,
      "state_query_norm": 44.13056945800781,
      "state_relation_loss": 0.0003063674841541797,
      "step": 8410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.005751287704333663,
      "weighted_relation_loss": 0.005751287704333663
    },
    {
      "action_query_norm": 43.31761245727539,
      "action_relation_loss": 0.005149397067725658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.876520112254443,
      "flow_matching_loss": 0.04589583314955235,
      "grad_norm": 0.11577881872653961,
      "learning_rate": 8.614899193005248e-05,
      "loss": 0.0514,
      "relation_loss": 0.005467539187520742,
      "selected_pairs": 27131.2,
      "state_query_norm": 44.123466873168944,
      "state_relation_loss": 0.00031814219255466014,
      "step": 8420,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.005467539187520742,
      "weighted_relation_loss": 0.005467539187520742
    },
    {
      "action_query_norm": 43.30231246948242,
      "action_relation_loss": 0.005385503545403481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.885874649204864,
      "flow_matching_loss": 0.04688060022890568,
      "grad_norm": 0.11586689203977585,
      "learning_rate": 8.611089226713842e-05,
      "loss": 0.0526,
      "relation_loss": 0.005704958317801356,
      "selected_pairs": 27084.8,
      "state_query_norm": 44.126329803466795,
      "state_relation_loss": 0.0003194547811290249,
      "step": 8430,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005704958317801356,
      "weighted_relation_loss": 0.005704958317801356
    },
    {
      "action_query_norm": 43.31388854980469,
      "action_relation_loss": 0.005695479596033693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.8952291861552855,
      "flow_matching_loss": 0.04567302167415619,
      "grad_norm": 0.13364215195178986,
      "learning_rate": 8.607274872608521e-05,
      "loss": 0.0517,
      "relation_loss": 0.006020595412701368,
      "selected_pairs": 27010.7,
      "state_query_norm": 44.105675506591794,
      "state_relation_loss": 0.0003251158748753369,
      "step": 8440,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.006020595412701368,
      "weighted_relation_loss": 0.006020595412701368
    },
    {
      "action_query_norm": 43.30330963134766,
      "action_relation_loss": 0.005450105527415872,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.904583723105707,
      "flow_matching_loss": 0.045834996923804286,
      "grad_norm": 0.13525041937828064,
      "learning_rate": 8.603456135324088e-05,
      "loss": 0.0516,
      "relation_loss": 0.005780978547409177,
      "selected_pairs": 26807.3,
      "state_query_norm": 44.1138240814209,
      "state_relation_loss": 0.0003308730287244543,
      "step": 8450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.005780978547409177,
      "weighted_relation_loss": 0.005780978547409177
    },
    {
      "action_query_norm": 43.29062690734863,
      "action_relation_loss": 0.005543269403278828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.913938260056128,
      "flow_matching_loss": 0.046328244358301164,
      "grad_norm": 0.15758824348449707,
      "learning_rate": 8.599633019500665e-05,
      "loss": 0.0522,
      "relation_loss": 0.005863533820956945,
      "selected_pairs": 26547.9,
      "state_query_norm": 44.11823959350586,
      "state_relation_loss": 0.0003202643507393077,
      "step": 8460,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005863533820956945,
      "weighted_relation_loss": 0.005863533820956945
    },
    {
      "action_query_norm": 43.28967781066895,
      "action_relation_loss": 0.00527576832100749,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.923292797006548,
      "flow_matching_loss": 0.04777024872601032,
      "grad_norm": 0.14008109271526337,
      "learning_rate": 8.595805529783702e-05,
      "loss": 0.0534,
      "relation_loss": 0.00560315428301692,
      "selected_pairs": 27523.1,
      "state_query_norm": 44.12001190185547,
      "state_relation_loss": 0.00032738602021709085,
      "step": 8470,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.00560315428301692,
      "weighted_relation_loss": 0.00560315428301692
    },
    {
      "action_query_norm": 43.31319618225098,
      "action_relation_loss": 0.00547883128747344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.932647333956969,
      "flow_matching_loss": 0.049157334864139555,
      "grad_norm": 0.09474657475948334,
      "learning_rate": 8.591973670823958e-05,
      "loss": 0.055,
      "relation_loss": 0.005810821475461125,
      "selected_pairs": 26875.1,
      "state_query_norm": 44.105909729003905,
      "state_relation_loss": 0.0003319901443319395,
      "step": 8480,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005810821475461125,
      "weighted_relation_loss": 0.005810821475461125
    },
    {
      "action_query_norm": 43.29974555969238,
      "action_relation_loss": 0.005436201766133309,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.94200187090739,
      "flow_matching_loss": 0.04657632447779179,
      "grad_norm": 0.11250871419906616,
      "learning_rate": 8.588137447277503e-05,
      "loss": 0.0524,
      "relation_loss": 0.005775084299966693,
      "selected_pairs": 27265.4,
      "state_query_norm": 44.109106826782224,
      "state_relation_loss": 0.0003388825280126184,
      "step": 8490,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005775084299966693,
      "weighted_relation_loss": 0.005775084299966693
    },
    {
      "action_query_norm": 43.30502853393555,
      "action_relation_loss": 0.00558337545953691,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.951356407857811,
      "flow_matching_loss": 0.04720558077096939,
      "grad_norm": 0.12459256500005722,
      "learning_rate": 8.584296863805709e-05,
      "loss": 0.0531,
      "relation_loss": 0.005914074927568435,
      "selected_pairs": 26437.1,
      "state_query_norm": 44.110464859008786,
      "state_relation_loss": 0.00033069939818233254,
      "step": 8500,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.005914074927568435,
      "weighted_relation_loss": 0.005914074927568435
    },
    {
      "action_query_norm": 43.31530876159668,
      "action_relation_loss": 0.005507964873686433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.960710944808232,
      "flow_matching_loss": 0.04604918584227562,
      "grad_norm": 0.15955986082553864,
      "learning_rate": 8.580451925075248e-05,
      "loss": 0.0519,
      "relation_loss": 0.0058425032999366525,
      "selected_pairs": 27197.1,
      "state_query_norm": 44.10766487121582,
      "state_relation_loss": 0.00033453852811362597,
      "step": 8510,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0058425032999366525,
      "weighted_relation_loss": 0.0058425032999366525
    },
    {
      "action_query_norm": 43.28312301635742,
      "action_relation_loss": 0.005678969575092197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.9700654817586525,
      "flow_matching_loss": 0.04855979941785336,
      "grad_norm": 0.13802367448806763,
      "learning_rate": 8.576602635758085e-05,
      "loss": 0.0546,
      "relation_loss": 0.006003642966970801,
      "selected_pairs": 26778.9,
      "state_query_norm": 44.11481819152832,
      "state_relation_loss": 0.0003246733540436253,
      "step": 8520,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.006003642966970801,
      "weighted_relation_loss": 0.006003642966970801
    },
    {
      "action_query_norm": 43.29033203125,
      "action_relation_loss": 0.005458386475220322,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.979420018709074,
      "flow_matching_loss": 0.04633636474609375,
      "grad_norm": 0.13042820990085602,
      "learning_rate": 8.572749000531469e-05,
      "loss": 0.0521,
      "relation_loss": 0.005770194809883833,
      "selected_pairs": 26967.7,
      "state_query_norm": 44.10846786499023,
      "state_relation_loss": 0.00031180833466351034,
      "step": 8530,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.005770194809883833,
      "weighted_relation_loss": 0.005770194809883833
    },
    {
      "action_query_norm": 43.27749252319336,
      "action_relation_loss": 0.005613719904795289,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.988774555659495,
      "flow_matching_loss": 0.046996939927339554,
      "grad_norm": 0.13255396485328674,
      "learning_rate": 8.568891024077926e-05,
      "loss": 0.0529,
      "relation_loss": 0.005934846960008145,
      "selected_pairs": 26921.0,
      "state_query_norm": 44.10331802368164,
      "state_relation_loss": 0.000321127095958218,
      "step": 8540,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005934846960008145,
      "weighted_relation_loss": 0.005934846960008145
    },
    {
      "action_query_norm": 43.27320899963379,
      "action_relation_loss": 0.0051665240433067085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 7.998129092609916,
      "flow_matching_loss": 0.04787727333605289,
      "grad_norm": 0.1653246432542801,
      "learning_rate": 8.565028711085265e-05,
      "loss": 0.0534,
      "relation_loss": 0.005484803533181548,
      "selected_pairs": 27362.5,
      "state_query_norm": 44.09983520507812,
      "state_relation_loss": 0.00031827950733713805,
      "step": 8550,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.005484803533181548,
      "weighted_relation_loss": 0.005484803533181548
    },
    {
      "action_query_norm": 43.27480583190918,
      "action_relation_loss": 0.0052103182766586546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.007483629560337,
      "flow_matching_loss": 0.04633924812078476,
      "grad_norm": 0.15672898292541504,
      "learning_rate": 8.561162066246561e-05,
      "loss": 0.0519,
      "relation_loss": 0.005520252883434296,
      "selected_pairs": 24460.0,
      "state_query_norm": 44.0998031616211,
      "state_relation_loss": 0.00030993454856798053,
      "step": 8560,
      "valid_action_samples": 214.9,
      "weighted_auxiliary_loss": 0.005520252883434296,
      "weighted_relation_loss": 0.005520252883434296
    },
    {
      "action_query_norm": 43.262942504882815,
      "action_relation_loss": 0.0053221032489091154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.016838166510757,
      "flow_matching_loss": 0.04670960903167724,
      "grad_norm": 0.1848088949918747,
      "learning_rate": 8.557291094260152e-05,
      "loss": 0.0523,
      "relation_loss": 0.005634018778800964,
      "selected_pairs": 27131.2,
      "state_query_norm": 44.110273361206055,
      "state_relation_loss": 0.0003119154542218894,
      "step": 8570,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.005634018778800964,
      "weighted_relation_loss": 0.005634018778800964
    },
    {
      "action_query_norm": 43.25875968933106,
      "action_relation_loss": 0.005251833191141486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.026192703461179,
      "flow_matching_loss": 0.04585273526608944,
      "grad_norm": 0.20195704698562622,
      "learning_rate": 8.553415799829629e-05,
      "loss": 0.0514,
      "relation_loss": 0.005564551567658782,
      "selected_pairs": 26905.3,
      "state_query_norm": 44.09705352783203,
      "state_relation_loss": 0.000312718388158828,
      "step": 8580,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005564551567658782,
      "weighted_relation_loss": 0.005564551567658782
    },
    {
      "action_query_norm": 43.240243148803714,
      "action_relation_loss": 0.005107136722654104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.0355472404116,
      "flow_matching_loss": 0.04888825342059135,
      "grad_norm": 0.1980530172586441,
      "learning_rate": 8.549536187663847e-05,
      "loss": 0.0543,
      "relation_loss": 0.005407796381041408,
      "selected_pairs": 26179.7,
      "state_query_norm": 44.09567451477051,
      "state_relation_loss": 0.00030065953906159847,
      "step": 8590,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.005407796381041408,
      "weighted_relation_loss": 0.005407796381041408
    },
    {
      "action_query_norm": 43.23395347595215,
      "action_relation_loss": 0.0052199896425008776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.044901777362021,
      "flow_matching_loss": 0.047707848623394965,
      "grad_norm": 0.15464769303798676,
      "learning_rate": 8.545652262476897e-05,
      "loss": 0.0532,
      "relation_loss": 0.005539508210495115,
      "selected_pairs": 27154.8,
      "state_query_norm": 44.09718284606934,
      "state_relation_loss": 0.00031951862038113175,
      "step": 8600,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.005539508210495115,
      "weighted_relation_loss": 0.005539508210495115
    },
    {
      "action_query_norm": 43.25869255065918,
      "action_relation_loss": 0.005170166306197643,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.054256314312441,
      "flow_matching_loss": 0.045898812264204024,
      "grad_norm": 0.12044212967157364,
      "learning_rate": 8.541764028988115e-05,
      "loss": 0.0514,
      "relation_loss": 0.005479849968105554,
      "selected_pairs": 27336.8,
      "state_query_norm": 44.091240310668944,
      "state_relation_loss": 0.00030968356877565385,
      "step": 8610,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.005479849968105554,
      "weighted_relation_loss": 0.005479849968105554
    },
    {
      "action_query_norm": 43.23971252441406,
      "action_relation_loss": 0.00553157408721745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.063610851262862,
      "flow_matching_loss": 0.04534098282456398,
      "grad_norm": 0.12542405724525452,
      "learning_rate": 8.537871491922071e-05,
      "loss": 0.0512,
      "relation_loss": 0.0058306203689426185,
      "selected_pairs": 27225.2,
      "state_query_norm": 44.08582649230957,
      "state_relation_loss": 0.00029904627299401907,
      "step": 8620,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0058306203689426185,
      "weighted_relation_loss": 0.0058306203689426185
    },
    {
      "action_query_norm": 43.272351837158205,
      "action_relation_loss": 0.005064129736274481,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.072965388213284,
      "flow_matching_loss": 0.04646853022277355,
      "grad_norm": 0.12573674321174622,
      "learning_rate": 8.533974656008566e-05,
      "loss": 0.0518,
      "relation_loss": 0.00537352841347456,
      "selected_pairs": 27011.4,
      "state_query_norm": 44.08203964233398,
      "state_relation_loss": 0.0003093986742896959,
      "step": 8630,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.00537352841347456,
      "weighted_relation_loss": 0.00537352841347456
    },
    {
      "action_query_norm": 43.27324905395508,
      "action_relation_loss": 0.005237653804942966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.082319925163704,
      "flow_matching_loss": 0.048231182247400285,
      "grad_norm": 0.11367553472518921,
      "learning_rate": 8.530073525982621e-05,
      "loss": 0.0538,
      "relation_loss": 0.005540161579847336,
      "selected_pairs": 26873.3,
      "state_query_norm": 44.09097557067871,
      "state_relation_loss": 0.00030250770796556024,
      "step": 8640,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005540161579847336,
      "weighted_relation_loss": 0.005540161579847336
    },
    {
      "action_query_norm": 43.243841552734374,
      "action_relation_loss": 0.004991016956046224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.091674462114126,
      "flow_matching_loss": 0.04634991735219955,
      "grad_norm": 0.1282375156879425,
      "learning_rate": 8.526168106584476e-05,
      "loss": 0.0517,
      "relation_loss": 0.005302185332402587,
      "selected_pairs": 26410.9,
      "state_query_norm": 44.08402442932129,
      "state_relation_loss": 0.00031116837926674633,
      "step": 8650,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.005302185332402587,
      "weighted_relation_loss": 0.005302185332402587
    },
    {
      "action_query_norm": 43.24400787353515,
      "action_relation_loss": 0.005078992061316967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.101028999064546,
      "flow_matching_loss": 0.04615008980035782,
      "grad_norm": 0.15907208621501923,
      "learning_rate": 8.522258402559586e-05,
      "loss": 0.0515,
      "relation_loss": 0.005376626318320632,
      "selected_pairs": 26229.1,
      "state_query_norm": 44.08747100830078,
      "state_relation_loss": 0.0002976342977490276,
      "step": 8660,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.005376626318320632,
      "weighted_relation_loss": 0.005376626318320632
    },
    {
      "action_query_norm": 43.24919319152832,
      "action_relation_loss": 0.005332494759932161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.110383536014968,
      "flow_matching_loss": 0.04566706456243992,
      "grad_norm": 0.10301949083805084,
      "learning_rate": 8.518344418658613e-05,
      "loss": 0.0513,
      "relation_loss": 0.005640464555472136,
      "selected_pairs": 26738.3,
      "state_query_norm": 44.085292434692384,
      "state_relation_loss": 0.00030796980718150735,
      "step": 8670,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005640464555472136,
      "weighted_relation_loss": 0.005640464555472136
    },
    {
      "action_query_norm": 43.22433700561523,
      "action_relation_loss": 0.005162450857460499,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.119738072965388,
      "flow_matching_loss": 0.043083003535866735,
      "grad_norm": 0.14987055957317352,
      "learning_rate": 8.51442615963741e-05,
      "loss": 0.0485,
      "relation_loss": 0.005457599693909287,
      "selected_pairs": 27067.9,
      "state_query_norm": 44.07732925415039,
      "state_relation_loss": 0.0002951488073449582,
      "step": 8680,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.005457599693909287,
      "weighted_relation_loss": 0.005457599693909287
    },
    {
      "action_query_norm": 43.244034957885745,
      "action_relation_loss": 0.004991720197722316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.129092609915809,
      "flow_matching_loss": 0.04485543444752693,
      "grad_norm": 0.11821799725294113,
      "learning_rate": 8.510503630257034e-05,
      "loss": 0.0502,
      "relation_loss": 0.005296548688784241,
      "selected_pairs": 27263.8,
      "state_query_norm": 44.075110626220706,
      "state_relation_loss": 0.0003048285230761394,
      "step": 8690,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005296548688784241,
      "weighted_relation_loss": 0.005296548688784241
    },
    {
      "action_query_norm": 43.26199645996094,
      "action_relation_loss": 0.005048930831253529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.13844714686623,
      "flow_matching_loss": 0.04700503386557102,
      "grad_norm": 0.13858941197395325,
      "learning_rate": 8.50657683528373e-05,
      "loss": 0.0524,
      "relation_loss": 0.005354734882712364,
      "selected_pairs": 27174.1,
      "state_query_norm": 44.07235527038574,
      "state_relation_loss": 0.00030580399034079163,
      "step": 8700,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.005354734882712364,
      "weighted_relation_loss": 0.005354734882712364
    },
    {
      "action_query_norm": 43.27197761535645,
      "action_relation_loss": 0.005168966110795736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.14780168381665,
      "flow_matching_loss": 0.047639339417219165,
      "grad_norm": 0.13245870172977448,
      "learning_rate": 8.502645779488922e-05,
      "loss": 0.0531,
      "relation_loss": 0.00547198774293065,
      "selected_pairs": 27053.9,
      "state_query_norm": 44.072505187988284,
      "state_relation_loss": 0.00030302162631414833,
      "step": 8710,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.00547198774293065,
      "weighted_relation_loss": 0.00547198774293065
    },
    {
      "action_query_norm": 43.23750114440918,
      "action_relation_loss": 0.0050106881186366085,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.157156220767073,
      "flow_matching_loss": 0.04669430740177631,
      "grad_norm": 0.11997365951538086,
      "learning_rate": 8.498710467649213e-05,
      "loss": 0.052,
      "relation_loss": 0.00533675430342555,
      "selected_pairs": 26684.5,
      "state_query_norm": 44.068138122558594,
      "state_relation_loss": 0.0003260662517277524,
      "step": 8720,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00533675430342555,
      "weighted_relation_loss": 0.00533675430342555
    },
    {
      "action_query_norm": 43.21562538146973,
      "action_relation_loss": 0.005278390552848578,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.166510757717493,
      "flow_matching_loss": 0.04682014286518097,
      "grad_norm": 0.11167114228010178,
      "learning_rate": 8.494770904546381e-05,
      "loss": 0.0524,
      "relation_loss": 0.0055931746028363705,
      "selected_pairs": 27179.5,
      "state_query_norm": 44.064187622070314,
      "state_relation_loss": 0.0003147840587189421,
      "step": 8730,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0055931746028363705,
      "weighted_relation_loss": 0.0055931746028363705
    },
    {
      "action_query_norm": 43.21525344848633,
      "action_relation_loss": 0.005099353985860944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.175865294667913,
      "flow_matching_loss": 0.04633586257696152,
      "grad_norm": 0.11405301094055176,
      "learning_rate": 8.490827094967363e-05,
      "loss": 0.0518,
      "relation_loss": 0.005418563587591052,
      "selected_pairs": 26640.7,
      "state_query_norm": 44.06014137268066,
      "state_relation_loss": 0.0003192096104612574,
      "step": 8740,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.005418563587591052,
      "weighted_relation_loss": 0.005418563587591052
    },
    {
      "action_query_norm": 43.25154800415039,
      "action_relation_loss": 0.005154468142427504,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.185219831618335,
      "flow_matching_loss": 0.04597040191292763,
      "grad_norm": 0.11830084770917892,
      "learning_rate": 8.486879043704263e-05,
      "loss": 0.0514,
      "relation_loss": 0.005472940066829324,
      "selected_pairs": 26914.1,
      "state_query_norm": 44.067970275878906,
      "state_relation_loss": 0.0003184719302225858,
      "step": 8750,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005472940066829324,
      "weighted_relation_loss": 0.005472940066829324
    },
    {
      "action_query_norm": 43.21731834411621,
      "action_relation_loss": 0.0050872181542217735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.194574368568755,
      "flow_matching_loss": 0.04626720137894154,
      "grad_norm": 0.13105729222297668,
      "learning_rate": 8.482926755554332e-05,
      "loss": 0.0517,
      "relation_loss": 0.0053901846520602705,
      "selected_pairs": 26615.2,
      "state_query_norm": 44.06802978515625,
      "state_relation_loss": 0.0003029664367204532,
      "step": 8760,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0053901846520602705,
      "weighted_relation_loss": 0.0053901846520602705
    },
    {
      "action_query_norm": 43.19396133422852,
      "action_relation_loss": 0.005294197145849467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.203928905519177,
      "flow_matching_loss": 0.04708000756800175,
      "grad_norm": 0.1356760859489441,
      "learning_rate": 8.478970235319975e-05,
      "loss": 0.0527,
      "relation_loss": 0.005594029603525996,
      "selected_pairs": 26764.3,
      "state_query_norm": 44.05846633911133,
      "state_relation_loss": 0.0002998324576765299,
      "step": 8770,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.005594029603525996,
      "weighted_relation_loss": 0.005594029603525996
    },
    {
      "action_query_norm": 43.239379501342775,
      "action_relation_loss": 0.005136481951922178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.213283442469598,
      "flow_matching_loss": 0.04381522238254547,
      "grad_norm": 0.11310999095439911,
      "learning_rate": 8.475009487808737e-05,
      "loss": 0.0493,
      "relation_loss": 0.005450222501531243,
      "selected_pairs": 26619.6,
      "state_query_norm": 44.06527595520019,
      "state_relation_loss": 0.0003137405816232786,
      "step": 8780,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005450222501531243,
      "weighted_relation_loss": 0.005450222501531243
    },
    {
      "action_query_norm": 43.22764663696289,
      "action_relation_loss": 0.005017907451838255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.22263797942002,
      "flow_matching_loss": 0.04436940215528011,
      "grad_norm": 0.11981162428855896,
      "learning_rate": 8.471044517833298e-05,
      "loss": 0.0497,
      "relation_loss": 0.0053451125044375655,
      "selected_pairs": 27043.0,
      "state_query_norm": 44.06510238647461,
      "state_relation_loss": 0.0003272050991654396,
      "step": 8790,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0053451125044375655,
      "weighted_relation_loss": 0.0053451125044375655
    },
    {
      "action_query_norm": 43.21464462280274,
      "action_relation_loss": 0.005183472717180848,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.23199251637044,
      "flow_matching_loss": 0.04364344403147698,
      "grad_norm": 0.09926335513591766,
      "learning_rate": 8.467075330211474e-05,
      "loss": 0.0491,
      "relation_loss": 0.005504464684054256,
      "selected_pairs": 26878.7,
      "state_query_norm": 44.05738754272461,
      "state_relation_loss": 0.0003209919494111091,
      "step": 8800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005504464684054256,
      "weighted_relation_loss": 0.005504464684054256
    },
    {
      "action_query_norm": 43.2083984375,
      "action_relation_loss": 0.005086236400529742,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.24134705332086,
      "flow_matching_loss": 0.04540758654475212,
      "grad_norm": 0.1331862509250641,
      "learning_rate": 8.463101929766198e-05,
      "loss": 0.0508,
      "relation_loss": 0.005393364932388067,
      "selected_pairs": 27102.2,
      "state_query_norm": 44.055796432495114,
      "state_relation_loss": 0.0003071284969337285,
      "step": 8810,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005393364932388067,
      "weighted_relation_loss": 0.005393364932388067
    },
    {
      "action_query_norm": 43.21920661926269,
      "action_relation_loss": 0.005120036331936717,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.250701590271282,
      "flow_matching_loss": 0.0464565210044384,
      "grad_norm": 0.11762115359306335,
      "learning_rate": 8.459124321325528e-05,
      "loss": 0.0519,
      "relation_loss": 0.0054593908600509165,
      "selected_pairs": 26433.8,
      "state_query_norm": 44.053897094726565,
      "state_relation_loss": 0.00033935451065190134,
      "step": 8820,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0054593908600509165,
      "weighted_relation_loss": 0.0054593908600509165
    },
    {
      "action_query_norm": 43.20773544311523,
      "action_relation_loss": 0.005227813590317964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.260056127221702,
      "flow_matching_loss": 0.04687047190964222,
      "grad_norm": 0.10760888457298279,
      "learning_rate": 8.455142509722634e-05,
      "loss": 0.0524,
      "relation_loss": 0.005534298205748201,
      "selected_pairs": 26970.5,
      "state_query_norm": 44.055796432495114,
      "state_relation_loss": 0.0003064845892367885,
      "step": 8830,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.005534298205748201,
      "weighted_relation_loss": 0.005534298205748201
    },
    {
      "action_query_norm": 43.217960357666016,
      "action_relation_loss": 0.004984092200174928,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.269410664172124,
      "flow_matching_loss": 0.045401034876704216,
      "grad_norm": 0.17854483425617218,
      "learning_rate": 8.451156499795792e-05,
      "loss": 0.0507,
      "relation_loss": 0.005303400848060846,
      "selected_pairs": 26829.3,
      "state_query_norm": 44.05112838745117,
      "state_relation_loss": 0.00031930865370668473,
      "step": 8840,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005303400848060846,
      "weighted_relation_loss": 0.005303400848060846
    },
    {
      "action_query_norm": 43.19009170532227,
      "action_relation_loss": 0.005301251541823149,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.278765201122544,
      "flow_matching_loss": 0.0460806917399168,
      "grad_norm": 0.12293041497468948,
      "learning_rate": 8.447166296388381e-05,
      "loss": 0.0517,
      "relation_loss": 0.005595350544899702,
      "selected_pairs": 26342.9,
      "state_query_norm": 44.05284042358399,
      "state_relation_loss": 0.0002940990147180855,
      "step": 8850,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.005595350544899702,
      "weighted_relation_loss": 0.005595350544899702
    },
    {
      "action_query_norm": 43.164886474609375,
      "action_relation_loss": 0.0054650697857141495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.288119738072965,
      "flow_matching_loss": 0.047912179678678515,
      "grad_norm": 0.14825645089149475,
      "learning_rate": 8.443171904348872e-05,
      "loss": 0.0537,
      "relation_loss": 0.005775812128558755,
      "selected_pairs": 27356.5,
      "state_query_norm": 44.0450984954834,
      "state_relation_loss": 0.0003107423865003511,
      "step": 8860,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.005775812128558755,
      "weighted_relation_loss": 0.005775812128558755
    },
    {
      "action_query_norm": 43.20614128112793,
      "action_relation_loss": 0.005099262436851859,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.297474275023387,
      "flow_matching_loss": 0.04633716344833374,
      "grad_norm": 0.1339755803346634,
      "learning_rate": 8.43917332853083e-05,
      "loss": 0.0517,
      "relation_loss": 0.005402587680146098,
      "selected_pairs": 27200.2,
      "state_query_norm": 44.04477119445801,
      "state_relation_loss": 0.00030332516180351374,
      "step": 8870,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.005402587680146098,
      "weighted_relation_loss": 0.005402587680146098
    },
    {
      "action_query_norm": 43.20040321350098,
      "action_relation_loss": 0.005063460441306233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.306828811973807,
      "flow_matching_loss": 0.043965579569339754,
      "grad_norm": 0.13809092342853546,
      "learning_rate": 8.435170573792901e-05,
      "loss": 0.0493,
      "relation_loss": 0.00538313090801239,
      "selected_pairs": 26805.0,
      "state_query_norm": 44.04411697387695,
      "state_relation_loss": 0.00031967049289960414,
      "step": 8880,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.00538313090801239,
      "weighted_relation_loss": 0.00538313090801239
    },
    {
      "action_query_norm": 43.17486991882324,
      "action_relation_loss": 0.005115572130307555,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.316183348924229,
      "flow_matching_loss": 0.04603819325566292,
      "grad_norm": 0.14114069938659668,
      "learning_rate": 8.431163644998808e-05,
      "loss": 0.0515,
      "relation_loss": 0.005447502341121435,
      "selected_pairs": 27124.4,
      "state_query_norm": 44.047926330566405,
      "state_relation_loss": 0.00033193031267728654,
      "step": 8890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.005447502341121435,
      "weighted_relation_loss": 0.005447502341121435
    },
    {
      "action_query_norm": 43.19563026428223,
      "action_relation_loss": 0.005171372881159186,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.325537885874649,
      "flow_matching_loss": 0.04544043131172657,
      "grad_norm": 0.16032887995243073,
      "learning_rate": 8.42715254701735e-05,
      "loss": 0.0509,
      "relation_loss": 0.005488748196512461,
      "selected_pairs": 26869.0,
      "state_query_norm": 44.03320159912109,
      "state_relation_loss": 0.00031737537938170135,
      "step": 8900,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005488748196512461,
      "weighted_relation_loss": 0.005488748196512461
    },
    {
      "action_query_norm": 43.1842643737793,
      "action_relation_loss": 0.005005862005054951,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.334892422825071,
      "flow_matching_loss": 0.04865771122276783,
      "grad_norm": 0.16380047798156738,
      "learning_rate": 8.42313728472239e-05,
      "loss": 0.054,
      "relation_loss": 0.005331090232357383,
      "selected_pairs": 27316.6,
      "state_query_norm": 44.03963203430176,
      "state_relation_loss": 0.00032522832334507256,
      "step": 8910,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.005331090232357383,
      "weighted_relation_loss": 0.005331090232357383
    },
    {
      "action_query_norm": 43.15929832458496,
      "action_relation_loss": 0.0051907694432884455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.344246959775491,
      "flow_matching_loss": 0.04597634002566338,
      "grad_norm": 0.1730412244796753,
      "learning_rate": 8.419117862992844e-05,
      "loss": 0.0515,
      "relation_loss": 0.005509176384657621,
      "selected_pairs": 26898.3,
      "state_query_norm": 44.042798614501955,
      "state_relation_loss": 0.0003184069588314742,
      "step": 8920,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005509176384657621,
      "weighted_relation_loss": 0.005509176384657621
    },
    {
      "action_query_norm": 43.18649559020996,
      "action_relation_loss": 0.005368362972512841,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.353601496725911,
      "flow_matching_loss": 0.045639445260167125,
      "grad_norm": 0.17897962033748627,
      "learning_rate": 8.415094286712694e-05,
      "loss": 0.0513,
      "relation_loss": 0.005698756314814091,
      "selected_pairs": 26485.4,
      "state_query_norm": 44.046573638916016,
      "state_relation_loss": 0.0003303933248389512,
      "step": 8930,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005698756314814091,
      "weighted_relation_loss": 0.005698756314814091
    },
    {
      "action_query_norm": 43.196193313598634,
      "action_relation_loss": 0.005044457456097007,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.362956033676333,
      "flow_matching_loss": 0.0440243698656559,
      "grad_norm": 0.14193156361579895,
      "learning_rate": 8.411066560770965e-05,
      "loss": 0.0494,
      "relation_loss": 0.005395944882184267,
      "selected_pairs": 26767.9,
      "state_query_norm": 44.0377700805664,
      "state_relation_loss": 0.000351487408624962,
      "step": 8940,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.005395944882184267,
      "weighted_relation_loss": 0.005395944882184267
    },
    {
      "action_query_norm": 43.16396789550781,
      "action_relation_loss": 0.004841528041288257,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.372310570626754,
      "flow_matching_loss": 0.04449300281703472,
      "grad_norm": 0.1288224309682846,
      "learning_rate": 8.407034690061721e-05,
      "loss": 0.0497,
      "relation_loss": 0.005167336296290159,
      "selected_pairs": 26492.7,
      "state_query_norm": 44.03855400085449,
      "state_relation_loss": 0.00032580824627075343,
      "step": 8950,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005167336296290159,
      "weighted_relation_loss": 0.005167336296290159
    },
    {
      "action_query_norm": 43.151252365112306,
      "action_relation_loss": 0.005344419414177537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.381665107577176,
      "flow_matching_loss": 0.04454375244677067,
      "grad_norm": 0.091457799077034,
      "learning_rate": 8.402998679484067e-05,
      "loss": 0.0502,
      "relation_loss": 0.005663305893540383,
      "selected_pairs": 27612.4,
      "state_query_norm": 44.036072158813475,
      "state_relation_loss": 0.0003188865113770589,
      "step": 8960,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.005663305893540383,
      "weighted_relation_loss": 0.005663305893540383
    },
    {
      "action_query_norm": 43.143974685668944,
      "action_relation_loss": 0.00496071120724082,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.391019644527596,
      "flow_matching_loss": 0.04348017498850822,
      "grad_norm": 0.12917186319828033,
      "learning_rate": 8.398958533942135e-05,
      "loss": 0.0488,
      "relation_loss": 0.0052736957091838125,
      "selected_pairs": 26857.1,
      "state_query_norm": 44.041838836669925,
      "state_relation_loss": 0.0003129845135845244,
      "step": 8970,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0052736957091838125,
      "weighted_relation_loss": 0.0052736957091838125
    },
    {
      "action_query_norm": 43.13588752746582,
      "action_relation_loss": 0.004980239272117615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.400374181478018,
      "flow_matching_loss": 0.04348979741334915,
      "grad_norm": 0.12793681025505066,
      "learning_rate": 8.394914258345083e-05,
      "loss": 0.0488,
      "relation_loss": 0.005291888583451509,
      "selected_pairs": 26212.4,
      "state_query_norm": 44.04083404541016,
      "state_relation_loss": 0.0003116492822300643,
      "step": 8980,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.005291888583451509,
      "weighted_relation_loss": 0.005291888583451509
    },
    {
      "action_query_norm": 43.16169052124023,
      "action_relation_loss": 0.005193193117156625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.409728718428438,
      "flow_matching_loss": 0.04584285281598568,
      "grad_norm": 0.10956867039203644,
      "learning_rate": 8.390865857607089e-05,
      "loss": 0.0513,
      "relation_loss": 0.005496887909248471,
      "selected_pairs": 26691.9,
      "state_query_norm": 44.03979377746582,
      "state_relation_loss": 0.00030369477171916516,
      "step": 8990,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005496887909248471,
      "weighted_relation_loss": 0.005496887909248471
    },
    {
      "action_query_norm": 43.14594421386719,
      "action_relation_loss": 0.00523554072715342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.419083255378858,
      "flow_matching_loss": 0.04805075414478779,
      "grad_norm": 0.1186748668551445,
      "learning_rate": 8.386813336647339e-05,
      "loss": 0.0536,
      "relation_loss": 0.0055484742857515815,
      "selected_pairs": 26603.5,
      "state_query_norm": 44.031084060668945,
      "state_relation_loss": 0.0003129335585981607,
      "step": 9000,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0055484742857515815,
      "weighted_relation_loss": 0.0055484742857515815
    },
    {
      "action_query_norm": 43.14426383972168,
      "action_relation_loss": 0.005233167763799429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.42843779232928,
      "flow_matching_loss": 0.04556596539914608,
      "grad_norm": 0.12791487574577332,
      "learning_rate": 8.38275670039003e-05,
      "loss": 0.0511,
      "relation_loss": 0.0055317634250968695,
      "selected_pairs": 26645.4,
      "state_query_norm": 44.03639602661133,
      "state_relation_loss": 0.00029859565838705746,
      "step": 9010,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0055317634250968695,
      "weighted_relation_loss": 0.0055317634250968695
    },
    {
      "action_query_norm": 43.15347175598144,
      "action_relation_loss": 0.0050461394246667625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.4377923292797,
      "flow_matching_loss": 0.044663918763399126,
      "grad_norm": 0.13180790841579437,
      "learning_rate": 8.378695953764356e-05,
      "loss": 0.05,
      "relation_loss": 0.005344007862731814,
      "selected_pairs": 26778.5,
      "state_query_norm": 44.02792892456055,
      "state_relation_loss": 0.0002978685399284586,
      "step": 9020,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005344007862731814,
      "weighted_relation_loss": 0.005344007862731814
    },
    {
      "action_query_norm": 43.15244331359863,
      "action_relation_loss": 0.005501069966703654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.447146866230122,
      "flow_matching_loss": 0.04449840635061264,
      "grad_norm": 0.14828446507453918,
      "learning_rate": 8.374631101704509e-05,
      "loss": 0.0503,
      "relation_loss": 0.005830807005986571,
      "selected_pairs": 26071.9,
      "state_query_norm": 44.02259559631348,
      "state_relation_loss": 0.00032973699271678927,
      "step": 9030,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.005830807005986571,
      "weighted_relation_loss": 0.005830807005986571
    },
    {
      "action_query_norm": 43.128532028198244,
      "action_relation_loss": 0.005351653415709734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.456501403180543,
      "flow_matching_loss": 0.04447641745209694,
      "grad_norm": 0.1741601675748825,
      "learning_rate": 8.370562149149665e-05,
      "loss": 0.0502,
      "relation_loss": 0.00567577457986772,
      "selected_pairs": 26664.1,
      "state_query_norm": 44.02373428344727,
      "state_relation_loss": 0.0003241211932618171,
      "step": 9040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00567577457986772,
      "weighted_relation_loss": 0.00567577457986772
    },
    {
      "action_query_norm": 43.11932907104492,
      "action_relation_loss": 0.005135882971808314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.465855940130963,
      "flow_matching_loss": 0.0441493634134531,
      "grad_norm": 0.1816817820072174,
      "learning_rate": 8.366489101043989e-05,
      "loss": 0.0496,
      "relation_loss": 0.005437664408236742,
      "selected_pairs": 26919.4,
      "state_query_norm": 44.02636871337891,
      "state_relation_loss": 0.00030178141314536334,
      "step": 9050,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.005437664408236742,
      "weighted_relation_loss": 0.005437664408236742
    },
    {
      "action_query_norm": 43.147253799438474,
      "action_relation_loss": 0.004990569408982992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.475210477081385,
      "flow_matching_loss": 0.043797390162944795,
      "grad_norm": 0.13303478062152863,
      "learning_rate": 8.362411962336614e-05,
      "loss": 0.0491,
      "relation_loss": 0.005291010485962033,
      "selected_pairs": 26577.9,
      "state_query_norm": 44.02032661437988,
      "state_relation_loss": 0.0003004411468282342,
      "step": 9060,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.005291010485962033,
      "weighted_relation_loss": 0.005291010485962033
    },
    {
      "action_query_norm": 43.09594383239746,
      "action_relation_loss": 0.00532454471103847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.484565014031805,
      "flow_matching_loss": 0.048794709518551825,
      "grad_norm": 0.15970204770565033,
      "learning_rate": 8.358330737981651e-05,
      "loss": 0.0544,
      "relation_loss": 0.005611108988523484,
      "selected_pairs": 26577.8,
      "state_query_norm": 44.016563415527344,
      "state_relation_loss": 0.0002865643531549722,
      "step": 9070,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.005611108988523484,
      "weighted_relation_loss": 0.005611108988523484
    },
    {
      "action_query_norm": 43.10431594848633,
      "action_relation_loss": 0.0049984019715338945,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.493919550982227,
      "flow_matching_loss": 0.048008454591035844,
      "grad_norm": 0.11834815889596939,
      "learning_rate": 8.35424543293817e-05,
      "loss": 0.0533,
      "relation_loss": 0.0052909843623638155,
      "selected_pairs": 27126.9,
      "state_query_norm": 44.014582824707034,
      "state_relation_loss": 0.0002925823122495785,
      "step": 9080,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0052909843623638155,
      "weighted_relation_loss": 0.0052909843623638155
    },
    {
      "action_query_norm": 43.12400512695312,
      "action_relation_loss": 0.004963069781661033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.503274087932647,
      "flow_matching_loss": 0.04549392014741897,
      "grad_norm": 0.1793319582939148,
      "learning_rate": 8.350156052170206e-05,
      "loss": 0.0508,
      "relation_loss": 0.005279894080013036,
      "selected_pairs": 26829.3,
      "state_query_norm": 44.013369750976565,
      "state_relation_loss": 0.0003168243332765996,
      "step": 9090,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005279894080013036,
      "weighted_relation_loss": 0.005279894080013036
    },
    {
      "action_query_norm": 43.12558822631836,
      "action_relation_loss": 0.005371610354632139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.512628624883067,
      "flow_matching_loss": 0.04649207517504692,
      "grad_norm": 0.1412477195262909,
      "learning_rate": 8.346062600646738e-05,
      "loss": 0.0522,
      "relation_loss": 0.005692934151738882,
      "selected_pairs": 27296.8,
      "state_query_norm": 44.010184860229494,
      "state_relation_loss": 0.0003213238378521055,
      "step": 9100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005692934151738882,
      "weighted_relation_loss": 0.005692934151738882
    },
    {
      "action_query_norm": 43.12104034423828,
      "action_relation_loss": 0.005064993910491466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.52198316183349,
      "flow_matching_loss": 0.049724673852324486,
      "grad_norm": 0.15579761564731598,
      "learning_rate": 8.341965083341697e-05,
      "loss": 0.0551,
      "relation_loss": 0.005389686953276396,
      "selected_pairs": 26893.0,
      "state_query_norm": 44.01201133728027,
      "state_relation_loss": 0.0003246930806199089,
      "step": 9110,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.005389686953276396,
      "weighted_relation_loss": 0.005389686953276396
    },
    {
      "action_query_norm": 43.11345176696777,
      "action_relation_loss": 0.004970647906884551,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.53133769878391,
      "flow_matching_loss": 0.04383740723133087,
      "grad_norm": 0.11463232338428497,
      "learning_rate": 8.337863505233953e-05,
      "loss": 0.0491,
      "relation_loss": 0.005286638671532273,
      "selected_pairs": 26481.3,
      "state_query_norm": 44.00727310180664,
      "state_relation_loss": 0.000315990750095807,
      "step": 9120,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005286638671532273,
      "weighted_relation_loss": 0.005286638671532273
    },
    {
      "action_query_norm": 43.135429763793944,
      "action_relation_loss": 0.005022984556853771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.540692235734332,
      "flow_matching_loss": 0.04378755949437618,
      "grad_norm": 0.11602769792079926,
      "learning_rate": 8.33375787130731e-05,
      "loss": 0.0491,
      "relation_loss": 0.005349212372675538,
      "selected_pairs": 26686.3,
      "state_query_norm": 44.01239280700683,
      "state_relation_loss": 0.0003262277517933398,
      "step": 9130,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.005349212372675538,
      "weighted_relation_loss": 0.005349212372675538
    },
    {
      "action_query_norm": 43.15616569519043,
      "action_relation_loss": 0.004987041186541319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.550046772684752,
      "flow_matching_loss": 0.046131661534309386,
      "grad_norm": 0.13194240629673004,
      "learning_rate": 8.3296481865505e-05,
      "loss": 0.0514,
      "relation_loss": 0.005296760471537709,
      "selected_pairs": 26624.9,
      "state_query_norm": 44.00533180236816,
      "state_relation_loss": 0.00030971923260949554,
      "step": 9140,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.005296760471537709,
      "weighted_relation_loss": 0.005296760471537709
    },
    {
      "action_query_norm": 43.14071044921875,
      "action_relation_loss": 0.004761851392686367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.559401309635174,
      "flow_matching_loss": 0.04342733323574066,
      "grad_norm": 0.12803708016872406,
      "learning_rate": 8.325534455957181e-05,
      "loss": 0.0485,
      "relation_loss": 0.005073163704946637,
      "selected_pairs": 27106.1,
      "state_query_norm": 44.00399475097656,
      "state_relation_loss": 0.00031131234136410056,
      "step": 9150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.005073163704946637,
      "weighted_relation_loss": 0.005073163704946637
    },
    {
      "action_query_norm": 43.13405838012695,
      "action_relation_loss": 0.004856486106291413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.568755846585594,
      "flow_matching_loss": 0.04344975985586643,
      "grad_norm": 0.17473767697811127,
      "learning_rate": 8.321416684525917e-05,
      "loss": 0.0486,
      "relation_loss": 0.005157025530934334,
      "selected_pairs": 27087.5,
      "state_query_norm": 43.9972526550293,
      "state_relation_loss": 0.0003005393809871748,
      "step": 9160,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005157025530934334,
      "weighted_relation_loss": 0.005157025530934334
    },
    {
      "action_query_norm": 43.13232421875,
      "action_relation_loss": 0.0048268782440572975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.578110383536014,
      "flow_matching_loss": 0.04706941582262516,
      "grad_norm": 0.20019349455833435,
      "learning_rate": 8.317294877260193e-05,
      "loss": 0.0522,
      "relation_loss": 0.005141147226095199,
      "selected_pairs": 27199.0,
      "state_query_norm": 44.00111465454101,
      "state_relation_loss": 0.0003142689703963697,
      "step": 9170,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.005141147226095199,
      "weighted_relation_loss": 0.005141147226095199
    },
    {
      "action_query_norm": 43.142219924926756,
      "action_relation_loss": 0.004921241756528616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.587464920486436,
      "flow_matching_loss": 0.042644485086202624,
      "grad_norm": 0.1379099041223526,
      "learning_rate": 8.313169039168394e-05,
      "loss": 0.0479,
      "relation_loss": 0.005229341145604849,
      "selected_pairs": 27482.4,
      "state_query_norm": 44.00117073059082,
      "state_relation_loss": 0.0003080994065385312,
      "step": 9180,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.005229341145604849,
      "weighted_relation_loss": 0.005229341145604849
    },
    {
      "action_query_norm": 43.126190185546875,
      "action_relation_loss": 0.0048228813335299495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.596819457436856,
      "flow_matching_loss": 0.04645279720425606,
      "grad_norm": 0.15111514925956726,
      "learning_rate": 8.3090391752638e-05,
      "loss": 0.0516,
      "relation_loss": 0.005127568496391177,
      "selected_pairs": 26741.5,
      "state_query_norm": 43.992984008789065,
      "state_relation_loss": 0.00030468710465356705,
      "step": 9190,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.005127568496391177,
      "weighted_relation_loss": 0.005127568496391177
    },
    {
      "action_query_norm": 43.11558837890625,
      "action_relation_loss": 0.004875652678310871,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.606173994387278,
      "flow_matching_loss": 0.043102262541651726,
      "grad_norm": 0.16812902688980103,
      "learning_rate": 8.304905290564586e-05,
      "loss": 0.0483,
      "relation_loss": 0.005197589471936226,
      "selected_pairs": 27266.2,
      "state_query_norm": 43.990953063964845,
      "state_relation_loss": 0.0003219368198188022,
      "step": 9200,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.005197589471936226,
      "weighted_relation_loss": 0.005197589471936226
    },
    {
      "action_query_norm": 43.113085174560545,
      "action_relation_loss": 0.004926435835659504,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.615528531337699,
      "flow_matching_loss": 0.04278726987540722,
      "grad_norm": 0.22250047326087952,
      "learning_rate": 8.300767390093813e-05,
      "loss": 0.048,
      "relation_loss": 0.005212605325505137,
      "selected_pairs": 27952.5,
      "state_query_norm": 43.99501457214355,
      "state_relation_loss": 0.00028616950585274025,
      "step": 9210,
      "valid_action_samples": 236.9,
      "weighted_auxiliary_loss": 0.005212605325505137,
      "weighted_relation_loss": 0.005212605325505137
    },
    {
      "action_query_norm": 43.10600204467774,
      "action_relation_loss": 0.004873039713129401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.62488306828812,
      "flow_matching_loss": 0.04367922581732273,
      "grad_norm": 0.17356178164482117,
      "learning_rate": 8.296625478879417e-05,
      "loss": 0.0488,
      "relation_loss": 0.00515715079382062,
      "selected_pairs": 26711.9,
      "state_query_norm": 43.9880126953125,
      "state_relation_loss": 0.00028411107487045226,
      "step": 9220,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.00515715079382062,
      "weighted_relation_loss": 0.00515715079382062
    },
    {
      "action_query_norm": 43.13269691467285,
      "action_relation_loss": 0.0051347331143915655,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.63423760523854,
      "flow_matching_loss": 0.04439784586429596,
      "grad_norm": 0.14266736805438995,
      "learning_rate": 8.292479561954214e-05,
      "loss": 0.0498,
      "relation_loss": 0.005432360712438822,
      "selected_pairs": 27202.2,
      "state_query_norm": 43.99099235534668,
      "state_relation_loss": 0.00029762757767457513,
      "step": 9230,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.005432360712438822,
      "weighted_relation_loss": 0.005432360712438822
    },
    {
      "action_query_norm": 43.11391677856445,
      "action_relation_loss": 0.00484047643840313,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.643592142188961,
      "flow_matching_loss": 0.0441760566085577,
      "grad_norm": 0.10519833117723465,
      "learning_rate": 8.288329644355883e-05,
      "loss": 0.0493,
      "relation_loss": 0.005141444178298116,
      "selected_pairs": 26848.0,
      "state_query_norm": 43.98957557678223,
      "state_relation_loss": 0.0003009676089277491,
      "step": 9240,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.005141444178298116,
      "weighted_relation_loss": 0.005141444178298116
    },
    {
      "action_query_norm": 43.1237434387207,
      "action_relation_loss": 0.0047794448677450415,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.652946679139383,
      "flow_matching_loss": 0.044907935336232184,
      "grad_norm": 0.1162748783826828,
      "learning_rate": 8.284175731126965e-05,
      "loss": 0.05,
      "relation_loss": 0.005087054194882512,
      "selected_pairs": 27155.9,
      "state_query_norm": 43.982242584228516,
      "state_relation_loss": 0.0003076093882555142,
      "step": 9250,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.005087054194882512,
      "weighted_relation_loss": 0.005087054194882512
    },
    {
      "action_query_norm": 43.10916709899902,
      "action_relation_loss": 0.00481202439405024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.662301216089803,
      "flow_matching_loss": 0.04430035874247551,
      "grad_norm": 0.14471836388111115,
      "learning_rate": 8.280017827314854e-05,
      "loss": 0.0494,
      "relation_loss": 0.0051048495806753635,
      "selected_pairs": 26880.4,
      "state_query_norm": 43.983652114868164,
      "state_relation_loss": 0.0002928252040874213,
      "step": 9260,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0051048495806753635,
      "weighted_relation_loss": 0.0051048495806753635
    },
    {
      "action_query_norm": 43.11765060424805,
      "action_relation_loss": 0.004725975962355733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.671655753040225,
      "flow_matching_loss": 0.04732888787984848,
      "grad_norm": 0.15534333884716034,
      "learning_rate": 8.2758559379718e-05,
      "loss": 0.0523,
      "relation_loss": 0.005008913204073906,
      "selected_pairs": 26986.5,
      "state_query_norm": 43.98317451477051,
      "state_relation_loss": 0.0002829372970154509,
      "step": 9270,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.005008913204073906,
      "weighted_relation_loss": 0.005008913204073906
    },
    {
      "action_query_norm": 43.099124908447266,
      "action_relation_loss": 0.005005443841218949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.681010289990645,
      "flow_matching_loss": 0.04602639079093933,
      "grad_norm": 0.12768115103244781,
      "learning_rate": 8.271690068154887e-05,
      "loss": 0.0513,
      "relation_loss": 0.005294672260060906,
      "selected_pairs": 26872.2,
      "state_query_norm": 43.984528350830075,
      "state_relation_loss": 0.00028922836063429715,
      "step": 9280,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.005294672260060906,
      "weighted_relation_loss": 0.005294672260060906
    },
    {
      "action_query_norm": 43.07581825256348,
      "action_relation_loss": 0.005018869554623962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.690364826941066,
      "flow_matching_loss": 0.04508408270776272,
      "grad_norm": 0.19075040519237518,
      "learning_rate": 8.26752022292604e-05,
      "loss": 0.0504,
      "relation_loss": 0.005315048899501562,
      "selected_pairs": 27291.2,
      "state_query_norm": 43.9798080444336,
      "state_relation_loss": 0.0002961793070426211,
      "step": 9290,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005315048899501562,
      "weighted_relation_loss": 0.005315048899501562
    },
    {
      "action_query_norm": 43.09407539367676,
      "action_relation_loss": 0.005076065380126238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.699719363891488,
      "flow_matching_loss": 0.04442860074341297,
      "grad_norm": 0.1279761642217636,
      "learning_rate": 8.263346407352018e-05,
      "loss": 0.0498,
      "relation_loss": 0.005377977713942527,
      "selected_pairs": 26551.2,
      "state_query_norm": 43.97418022155762,
      "state_relation_loss": 0.0003019124109414406,
      "step": 9300,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005377977713942527,
      "weighted_relation_loss": 0.005377977713942527
    },
    {
      "action_query_norm": 43.108877944946286,
      "action_relation_loss": 0.005228499136865139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.709073900841908,
      "flow_matching_loss": 0.0467925101518631,
      "grad_norm": 0.1338217407464981,
      "learning_rate": 8.259168626504395e-05,
      "loss": 0.0523,
      "relation_loss": 0.005517008109018206,
      "selected_pairs": 27293.3,
      "state_query_norm": 43.98537483215332,
      "state_relation_loss": 0.0002885089605115354,
      "step": 9310,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005517008109018206,
      "weighted_relation_loss": 0.005517008109018206
    },
    {
      "action_query_norm": 43.10457992553711,
      "action_relation_loss": 0.0047630007378757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.71842843779233,
      "flow_matching_loss": 0.04492508806288242,
      "grad_norm": 0.10996563732624054,
      "learning_rate": 8.254986885459572e-05,
      "loss": 0.05,
      "relation_loss": 0.005057297134771943,
      "selected_pairs": 26667.1,
      "state_query_norm": 43.98046035766602,
      "state_relation_loss": 0.0002942963968962431,
      "step": 9320,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.005057297134771943,
      "weighted_relation_loss": 0.005057297134771943
    },
    {
      "action_query_norm": 43.073639678955075,
      "action_relation_loss": 0.004763386026024818,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.72778297474275,
      "flow_matching_loss": 0.04546650946140289,
      "grad_norm": 0.11660440266132355,
      "learning_rate": 8.250801189298757e-05,
      "loss": 0.0505,
      "relation_loss": 0.005077844392508269,
      "selected_pairs": 26792.7,
      "state_query_norm": 43.968240356445314,
      "state_relation_loss": 0.00031445839558728037,
      "step": 9330,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.005077844392508269,
      "weighted_relation_loss": 0.005077844392508269
    },
    {
      "action_query_norm": 43.06318321228027,
      "action_relation_loss": 0.004912282573059201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.737137511693172,
      "flow_matching_loss": 0.04413933046162129,
      "grad_norm": 0.1509198695421219,
      "learning_rate": 8.246611543107967e-05,
      "loss": 0.0494,
      "relation_loss": 0.005240510497242212,
      "selected_pairs": 26315.6,
      "state_query_norm": 43.9737850189209,
      "state_relation_loss": 0.00032822789216879755,
      "step": 9340,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.005240510497242212,
      "weighted_relation_loss": 0.005240510497242212
    },
    {
      "action_query_norm": 43.05999488830567,
      "action_relation_loss": 0.005008561071008444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.746492048643592,
      "flow_matching_loss": 0.044563844427466394,
      "grad_norm": 0.14304645359516144,
      "learning_rate": 8.242417951978018e-05,
      "loss": 0.0499,
      "relation_loss": 0.005311698699370026,
      "selected_pairs": 26851.8,
      "state_query_norm": 43.97073936462402,
      "state_relation_loss": 0.0003031375992577523,
      "step": 9350,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.005311698699370026,
      "weighted_relation_loss": 0.005311698699370026
    },
    {
      "action_query_norm": 43.078104400634764,
      "action_relation_loss": 0.004810494882985949,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.755846585594012,
      "flow_matching_loss": 0.04506068676710129,
      "grad_norm": 0.134433314204216,
      "learning_rate": 8.238220421004517e-05,
      "loss": 0.0502,
      "relation_loss": 0.005124754691496492,
      "selected_pairs": 26414.5,
      "state_query_norm": 43.961225509643555,
      "state_relation_loss": 0.0003142598347039893,
      "step": 9360,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.005124754691496492,
      "weighted_relation_loss": 0.005124754691496492
    },
    {
      "action_query_norm": 43.077264404296876,
      "action_relation_loss": 0.004890680825337767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.765201122544434,
      "flow_matching_loss": 0.046203787997365,
      "grad_norm": 0.13749633729457855,
      "learning_rate": 8.234018955287859e-05,
      "loss": 0.0514,
      "relation_loss": 0.005212346790358424,
      "selected_pairs": 27082.6,
      "state_query_norm": 43.963748931884766,
      "state_relation_loss": 0.00032166595628950744,
      "step": 9370,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005212346790358424,
      "weighted_relation_loss": 0.005212346790358424
    },
    {
      "action_query_norm": 43.07114028930664,
      "action_relation_loss": 0.00486113247461617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.774555659494855,
      "flow_matching_loss": 0.044611381366848946,
      "grad_norm": 0.15504594147205353,
      "learning_rate": 8.229813559933224e-05,
      "loss": 0.0498,
      "relation_loss": 0.00518394117243588,
      "selected_pairs": 27225.4,
      "state_query_norm": 43.95924949645996,
      "state_relation_loss": 0.00032280868908856063,
      "step": 9380,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00518394117243588,
      "weighted_relation_loss": 0.00518394117243588
    },
    {
      "action_query_norm": 43.07227554321289,
      "action_relation_loss": 0.004716491559520364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.783910196445277,
      "flow_matching_loss": 0.04855128191411495,
      "grad_norm": 0.17607539892196655,
      "learning_rate": 8.22560424005056e-05,
      "loss": 0.0536,
      "relation_loss": 0.0050228070467710495,
      "selected_pairs": 26422.0,
      "state_query_norm": 43.96553344726563,
      "state_relation_loss": 0.00030631544359494,
      "step": 9390,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0050228070467710495,
      "weighted_relation_loss": 0.0050228070467710495
    },
    {
      "action_query_norm": 43.07264633178711,
      "action_relation_loss": 0.005253935046494007,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.793264733395697,
      "flow_matching_loss": 0.04323320649564266,
      "grad_norm": 0.1541174203157425,
      "learning_rate": 8.22139100075459e-05,
      "loss": 0.0488,
      "relation_loss": 0.005572812538594008,
      "selected_pairs": 27243.0,
      "state_query_norm": 43.9654655456543,
      "state_relation_loss": 0.00031887752702459694,
      "step": 9400,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.005572812538594008,
      "weighted_relation_loss": 0.005572812538594008
    },
    {
      "action_query_norm": 43.073215866088866,
      "action_relation_loss": 0.004759753169491887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.802619270346117,
      "flow_matching_loss": 0.04616542123258114,
      "grad_norm": 0.14836761355400085,
      "learning_rate": 8.217173847164799e-05,
      "loss": 0.0512,
      "relation_loss": 0.005061005474999547,
      "selected_pairs": 26548.0,
      "state_query_norm": 43.96200065612793,
      "state_relation_loss": 0.000301252311328426,
      "step": 9410,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005061005474999547,
      "weighted_relation_loss": 0.005061005474999547
    },
    {
      "action_query_norm": 43.062617492675784,
      "action_relation_loss": 0.00471025281585753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.811973807296539,
      "flow_matching_loss": 0.04377520903944969,
      "grad_norm": 0.11882156878709793,
      "learning_rate": 8.212952784405422e-05,
      "loss": 0.0488,
      "relation_loss": 0.005025090137496591,
      "selected_pairs": 26180.6,
      "state_query_norm": 43.96050796508789,
      "state_relation_loss": 0.00031483735365327447,
      "step": 9420,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.005025090137496591,
      "weighted_relation_loss": 0.005025090137496591
    },
    {
      "action_query_norm": 43.050424575805664,
      "action_relation_loss": 0.00501984735019505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.82132834424696,
      "flow_matching_loss": 0.04414389617741108,
      "grad_norm": 0.13373228907585144,
      "learning_rate": 8.208727817605451e-05,
      "loss": 0.0495,
      "relation_loss": 0.005330607667565346,
      "selected_pairs": 26478.8,
      "state_query_norm": 43.9655574798584,
      "state_relation_loss": 0.0003107604454271495,
      "step": 9430,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.005330607667565346,
      "weighted_relation_loss": 0.005330607667565346
    },
    {
      "action_query_norm": 43.052011489868164,
      "action_relation_loss": 0.004567633615806699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.830682881197381,
      "flow_matching_loss": 0.04490711614489555,
      "grad_norm": 0.12235713005065918,
      "learning_rate": 8.204498951898618e-05,
      "loss": 0.0498,
      "relation_loss": 0.004889118950814009,
      "selected_pairs": 26415.1,
      "state_query_norm": 43.95136070251465,
      "state_relation_loss": 0.00032148528553079816,
      "step": 9440,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004889118950814009,
      "weighted_relation_loss": 0.004889118950814009
    },
    {
      "action_query_norm": 43.06283721923828,
      "action_relation_loss": 0.0049927349667996165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.840037418147801,
      "flow_matching_loss": 0.044830378517508505,
      "grad_norm": 0.11691263318061829,
      "learning_rate": 8.200266192423395e-05,
      "loss": 0.0501,
      "relation_loss": 0.005308585148304701,
      "selected_pairs": 26411.2,
      "state_query_norm": 43.95293846130371,
      "state_relation_loss": 0.0003158502251608297,
      "step": 9450,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.005308585148304701,
      "weighted_relation_loss": 0.005308585148304701
    },
    {
      "action_query_norm": 43.06407241821289,
      "action_relation_loss": 0.004802700690925121,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.849391955098223,
      "flow_matching_loss": 0.04513336345553398,
      "grad_norm": 0.1363372802734375,
      "learning_rate": 8.196029544322981e-05,
      "loss": 0.0502,
      "relation_loss": 0.005105103505775333,
      "selected_pairs": 27088.9,
      "state_query_norm": 43.95101432800293,
      "state_relation_loss": 0.0003024028759682551,
      "step": 9460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.005105103505775333,
      "weighted_relation_loss": 0.005105103505775333
    },
    {
      "action_query_norm": 43.054264068603516,
      "action_relation_loss": 0.004975810274481773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.858746492048644,
      "flow_matching_loss": 0.04575463384389877,
      "grad_norm": 0.13657401502132416,
      "learning_rate": 8.191789012745305e-05,
      "loss": 0.051,
      "relation_loss": 0.005270827282220125,
      "selected_pairs": 27009.4,
      "state_query_norm": 43.95051651000976,
      "state_relation_loss": 0.00029501701646950096,
      "step": 9470,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.005270827282220125,
      "weighted_relation_loss": 0.005270827282220125
    },
    {
      "action_query_norm": 43.04610023498535,
      "action_relation_loss": 0.004893549485132098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.868101028999064,
      "flow_matching_loss": 0.04409634917974472,
      "grad_norm": 0.1262483149766922,
      "learning_rate": 8.187544602843014e-05,
      "loss": 0.0493,
      "relation_loss": 0.005201801611110568,
      "selected_pairs": 27410.5,
      "state_query_norm": 43.94687919616699,
      "state_relation_loss": 0.00030825213179923593,
      "step": 9480,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.005201801611110568,
      "weighted_relation_loss": 0.005201801611110568
    },
    {
      "action_query_norm": 43.039694213867186,
      "action_relation_loss": 0.004811166832223534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.877455565949486,
      "flow_matching_loss": 0.045385951176285744,
      "grad_norm": 0.15472622215747833,
      "learning_rate": 8.183296319773466e-05,
      "loss": 0.0505,
      "relation_loss": 0.005120034050196409,
      "selected_pairs": 26856.2,
      "state_query_norm": 43.93949203491211,
      "state_relation_loss": 0.0003088671859586611,
      "step": 9490,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.005120034050196409,
      "weighted_relation_loss": 0.005120034050196409
    },
    {
      "action_query_norm": 43.038639831542966,
      "action_relation_loss": 0.004638693248853088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.886810102899906,
      "flow_matching_loss": 0.043238714709877966,
      "grad_norm": 0.1517510861158371,
      "learning_rate": 8.179044168698721e-05,
      "loss": 0.0482,
      "relation_loss": 0.004933308297768235,
      "selected_pairs": 27456.7,
      "state_query_norm": 43.938298034667966,
      "state_relation_loss": 0.0002946150052594021,
      "step": 9500,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.004933308297768235,
      "weighted_relation_loss": 0.004933308297768235
    },
    {
      "action_query_norm": 43.024953079223636,
      "action_relation_loss": 0.004868335649371147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.896164639850328,
      "flow_matching_loss": 0.04084088280797005,
      "grad_norm": 0.15656669437885284,
      "learning_rate": 8.174788154785548e-05,
      "loss": 0.046,
      "relation_loss": 0.00517624020576477,
      "selected_pairs": 26688.8,
      "state_query_norm": 43.944847869873044,
      "state_relation_loss": 0.0003079045738559216,
      "step": 9510,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00517624020576477,
      "weighted_relation_loss": 0.00517624020576477
    },
    {
      "action_query_norm": 43.0299690246582,
      "action_relation_loss": 0.004711396340280771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.905519176800748,
      "flow_matching_loss": 0.04271137788891792,
      "grad_norm": 0.1219518631696701,
      "learning_rate": 8.170528283205403e-05,
      "loss": 0.0477,
      "relation_loss": 0.004994767950847745,
      "selected_pairs": 26759.9,
      "state_query_norm": 43.947933959960935,
      "state_relation_loss": 0.00028337159019429234,
      "step": 9520,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004994767950847745,
      "weighted_relation_loss": 0.004994767950847745
    },
    {
      "action_query_norm": 43.03015899658203,
      "action_relation_loss": 0.004816030710935592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.91487371375117,
      "flow_matching_loss": 0.0444343876093626,
      "grad_norm": 0.15145745873451233,
      "learning_rate": 8.166264559134435e-05,
      "loss": 0.0495,
      "relation_loss": 0.005105334892868995,
      "selected_pairs": 26708.5,
      "state_query_norm": 43.943620681762695,
      "state_relation_loss": 0.00028930417320225386,
      "step": 9530,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.005105334892868995,
      "weighted_relation_loss": 0.005105334892868995
    },
    {
      "action_query_norm": 43.022718048095705,
      "action_relation_loss": 0.00459470390342176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.92422825070159,
      "flow_matching_loss": 0.04296342208981514,
      "grad_norm": 0.13135862350463867,
      "learning_rate": 8.161996987753467e-05,
      "loss": 0.0479,
      "relation_loss": 0.004888777295127511,
      "selected_pairs": 26968.2,
      "state_query_norm": 43.94250831604004,
      "state_relation_loss": 0.0002940734411822632,
      "step": 9540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.004888777295127511,
      "weighted_relation_loss": 0.004888777295127511
    },
    {
      "action_query_norm": 43.003628158569335,
      "action_relation_loss": 0.005059671681374311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.93358278765201,
      "flow_matching_loss": 0.04475867971777916,
      "grad_norm": 0.1176440566778183,
      "learning_rate": 8.157725574248e-05,
      "loss": 0.0501,
      "relation_loss": 0.005348570458590984,
      "selected_pairs": 26371.3,
      "state_query_norm": 43.93775825500488,
      "state_relation_loss": 0.00028889884997624905,
      "step": 9550,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.005348570458590984,
      "weighted_relation_loss": 0.005348570458590984
    },
    {
      "action_query_norm": 43.005818939208986,
      "action_relation_loss": 0.004767106939107179,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.942937324602433,
      "flow_matching_loss": 0.04380059503018856,
      "grad_norm": 0.15902842581272125,
      "learning_rate": 8.153450323808209e-05,
      "loss": 0.0488,
      "relation_loss": 0.005043411953374743,
      "selected_pairs": 26551.3,
      "state_query_norm": 43.935638427734375,
      "state_relation_loss": 0.00027630494732875375,
      "step": 9560,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.005043411953374743,
      "weighted_relation_loss": 0.005043411953374743
    },
    {
      "action_query_norm": 42.99824104309082,
      "action_relation_loss": 0.004720151424407959,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.952291861552853,
      "flow_matching_loss": 0.04153163023293018,
      "grad_norm": 0.14778825640678406,
      "learning_rate": 8.149171241628923e-05,
      "loss": 0.0466,
      "relation_loss": 0.005023396899923682,
      "selected_pairs": 26832.9,
      "state_query_norm": 43.930898666381836,
      "state_relation_loss": 0.0003032455628272146,
      "step": 9570,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.005023396899923682,
      "weighted_relation_loss": 0.005023396899923682
    },
    {
      "action_query_norm": 43.0095947265625,
      "action_relation_loss": 0.004659124137833715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.961646398503275,
      "flow_matching_loss": 0.04484748691320419,
      "grad_norm": 0.15531529486179352,
      "learning_rate": 8.144888332909631e-05,
      "loss": 0.0498,
      "relation_loss": 0.0049481111578643325,
      "selected_pairs": 26706.4,
      "state_query_norm": 43.937903594970706,
      "state_relation_loss": 0.0002889870316721499,
      "step": 9580,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0049481111578643325,
      "weighted_relation_loss": 0.0049481111578643325
    },
    {
      "action_query_norm": 43.00441627502441,
      "action_relation_loss": 0.00473461146466434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.971000935453695,
      "flow_matching_loss": 0.043571889773011206,
      "grad_norm": 0.1533534824848175,
      "learning_rate": 8.14060160285447e-05,
      "loss": 0.0486,
      "relation_loss": 0.005026043206453323,
      "selected_pairs": 27293.5,
      "state_query_norm": 43.93689727783203,
      "state_relation_loss": 0.0002914317650720477,
      "step": 9590,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.005026043206453323,
      "weighted_relation_loss": 0.005026043206453323
    },
    {
      "action_query_norm": 42.98540153503418,
      "action_relation_loss": 0.004812979605048895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.980355472404115,
      "flow_matching_loss": 0.043251019343733786,
      "grad_norm": 0.12838119268417358,
      "learning_rate": 8.136311056672223e-05,
      "loss": 0.0483,
      "relation_loss": 0.005092442454770208,
      "selected_pairs": 27040.8,
      "state_query_norm": 43.931685256958005,
      "state_relation_loss": 0.00027946287009399383,
      "step": 9600,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.005092442454770208,
      "weighted_relation_loss": 0.005092442454770208
    },
    {
      "action_query_norm": 43.002770233154294,
      "action_relation_loss": 0.004809805518016219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.989710009354537,
      "flow_matching_loss": 0.04154081828892231,
      "grad_norm": 0.09908411651849747,
      "learning_rate": 8.132016699576308e-05,
      "loss": 0.0467,
      "relation_loss": 0.0051298199221491815,
      "selected_pairs": 26944.2,
      "state_query_norm": 43.93040733337402,
      "state_relation_loss": 0.00032001440995372833,
      "step": 9610,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0051298199221491815,
      "weighted_relation_loss": 0.0051298199221491815
    },
    {
      "action_query_norm": 43.00549430847168,
      "action_relation_loss": 0.00460859127342701,
      "auxiliary_loss_weight": 1.0,
      "epoch": 8.999064546304957,
      "flow_matching_loss": 0.04462701492011547,
      "grad_norm": 0.12661167979240417,
      "learning_rate": 8.127718536784771e-05,
      "loss": 0.0495,
      "relation_loss": 0.004919590195640922,
      "selected_pairs": 26407.7,
      "state_query_norm": 43.930855178833006,
      "state_relation_loss": 0.00031099891639314593,
      "step": 9620,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004919590195640922,
      "weighted_relation_loss": 0.004919590195640922
    },
    {
      "action_query_norm": 42.99584197998047,
      "action_relation_loss": 0.004428567830473185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.00841908325538,
      "flow_matching_loss": 0.04302779473364353,
      "grad_norm": 0.12503094971179962,
      "learning_rate": 8.12341657352029e-05,
      "loss": 0.0478,
      "relation_loss": 0.0047359921969473365,
      "selected_pairs": 24759.8,
      "state_query_norm": 43.91821479797363,
      "state_relation_loss": 0.0003074243577430025,
      "step": 9630,
      "valid_action_samples": 216.1,
      "weighted_auxiliary_loss": 0.0047359921969473365,
      "weighted_relation_loss": 0.0047359921969473365
    },
    {
      "action_query_norm": 42.99203605651856,
      "action_relation_loss": 0.004579762974753976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.0177736202058,
      "flow_matching_loss": 0.04225019477307797,
      "grad_norm": 0.11754634231328964,
      "learning_rate": 8.119110815010151e-05,
      "loss": 0.0471,
      "relation_loss": 0.004870037455111742,
      "selected_pairs": 26692.2,
      "state_query_norm": 43.92426300048828,
      "state_relation_loss": 0.00029027444834355264,
      "step": 9640,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004870037455111742,
      "weighted_relation_loss": 0.004870037455111742
    },
    {
      "action_query_norm": 42.99719123840332,
      "action_relation_loss": 0.0045875328127294775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.027128157156222,
      "flow_matching_loss": 0.04314125292003155,
      "grad_norm": 0.12206841260194778,
      "learning_rate": 8.114801266486261e-05,
      "loss": 0.048,
      "relation_loss": 0.004866323806345463,
      "selected_pairs": 26873.4,
      "state_query_norm": 43.91410789489746,
      "state_relation_loss": 0.00027879105473402885,
      "step": 9650,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004866323806345463,
      "weighted_relation_loss": 0.004866323806345463
    },
    {
      "action_query_norm": 42.97133026123047,
      "action_relation_loss": 0.004602113435976207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.036482694106642,
      "flow_matching_loss": 0.0442662063986063,
      "grad_norm": 0.14635974168777466,
      "learning_rate": 8.110487933185122e-05,
      "loss": 0.0492,
      "relation_loss": 0.004899736354127526,
      "selected_pairs": 26883.5,
      "state_query_norm": 43.9197265625,
      "state_relation_loss": 0.00029762284248135984,
      "step": 9660,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004899736354127526,
      "weighted_relation_loss": 0.004899736354127526
    },
    {
      "action_query_norm": 42.968010330200194,
      "action_relation_loss": 0.004636077955365181,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.045837231057062,
      "flow_matching_loss": 0.04203653931617737,
      "grad_norm": 0.15058647096157074,
      "learning_rate": 8.106170820347847e-05,
      "loss": 0.0469,
      "relation_loss": 0.004906423529610038,
      "selected_pairs": 26850.6,
      "state_query_norm": 43.91480140686035,
      "state_relation_loss": 0.00027034554514102636,
      "step": 9670,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004906423529610038,
      "weighted_relation_loss": 0.004906423529610038
    },
    {
      "action_query_norm": 42.97418975830078,
      "action_relation_loss": 0.004434381891041994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.055191768007484,
      "flow_matching_loss": 0.040643992647528646,
      "grad_norm": 0.1459079086780548,
      "learning_rate": 8.101849933220132e-05,
      "loss": 0.0454,
      "relation_loss": 0.004709259420633316,
      "selected_pairs": 27031.8,
      "state_query_norm": 43.91674995422363,
      "state_relation_loss": 0.00027487751794978975,
      "step": 9680,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004709259420633316,
      "weighted_relation_loss": 0.004709259420633316
    },
    {
      "action_query_norm": 42.993883514404295,
      "action_relation_loss": 0.004509340412914753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.064546304957904,
      "flow_matching_loss": 0.042374931275844574,
      "grad_norm": 0.15992633998394012,
      "learning_rate": 8.097525277052264e-05,
      "loss": 0.0472,
      "relation_loss": 0.00477921161800623,
      "selected_pairs": 27428.1,
      "state_query_norm": 43.92283248901367,
      "state_relation_loss": 0.00026987123710568993,
      "step": 9690,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.00477921161800623,
      "weighted_relation_loss": 0.00477921161800623
    },
    {
      "action_query_norm": 42.97308464050293,
      "action_relation_loss": 0.004556716978549957,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.073900841908326,
      "flow_matching_loss": 0.04277292899787426,
      "grad_norm": 0.12435667216777802,
      "learning_rate": 8.093196857099106e-05,
      "loss": 0.0476,
      "relation_loss": 0.004832978826016188,
      "selected_pairs": 27039.1,
      "state_query_norm": 43.91232261657715,
      "state_relation_loss": 0.00027626181836239996,
      "step": 9700,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004832978826016188,
      "weighted_relation_loss": 0.004832978826016188
    },
    {
      "action_query_norm": 42.979862213134766,
      "action_relation_loss": 0.004546291753649711,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.083255378858746,
      "flow_matching_loss": 0.041689975559711455,
      "grad_norm": 0.10829045623540878,
      "learning_rate": 8.088864678620096e-05,
      "loss": 0.0465,
      "relation_loss": 0.004829169297590852,
      "selected_pairs": 27078.4,
      "state_query_norm": 43.9027400970459,
      "state_relation_loss": 0.00028287745226407424,
      "step": 9710,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.004829169297590852,
      "weighted_relation_loss": 0.004829169297590852
    },
    {
      "action_query_norm": 42.97715301513672,
      "action_relation_loss": 0.004479422839358449,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.092609915809167,
      "flow_matching_loss": 0.043893634155392644,
      "grad_norm": 0.11097166687250137,
      "learning_rate": 8.084528746879243e-05,
      "loss": 0.0486,
      "relation_loss": 0.004754338040947914,
      "selected_pairs": 26664.5,
      "state_query_norm": 43.9124698638916,
      "state_relation_loss": 0.0002749151652096771,
      "step": 9720,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004754338040947914,
      "weighted_relation_loss": 0.004754338040947914
    },
    {
      "action_query_norm": 42.99383888244629,
      "action_relation_loss": 0.004370023496448994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.101964452759589,
      "flow_matching_loss": 0.042486618459224704,
      "grad_norm": 0.1034969612956047,
      "learning_rate": 8.080189067145108e-05,
      "loss": 0.0471,
      "relation_loss": 0.004655383760109544,
      "selected_pairs": 27248.0,
      "state_query_norm": 43.90597229003906,
      "state_relation_loss": 0.00028536025347420947,
      "step": 9730,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.004655383760109544,
      "weighted_relation_loss": 0.004655383760109544
    },
    {
      "action_query_norm": 42.96929092407227,
      "action_relation_loss": 0.004551885183900595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.111318989710009,
      "flow_matching_loss": 0.04322241023182869,
      "grad_norm": 0.161651611328125,
      "learning_rate": 8.075845644690813e-05,
      "loss": 0.0481,
      "relation_loss": 0.004828588105738163,
      "selected_pairs": 26988.2,
      "state_query_norm": 43.898867797851565,
      "state_relation_loss": 0.0002767028912785463,
      "step": 9740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004828588105738163,
      "weighted_relation_loss": 0.004828588105738163
    },
    {
      "action_query_norm": 42.940644073486325,
      "action_relation_loss": 0.004558901069685817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.12067352666043,
      "flow_matching_loss": 0.0422446858137846,
      "grad_norm": 0.13091643154621124,
      "learning_rate": 8.07149848479403e-05,
      "loss": 0.0471,
      "relation_loss": 0.0048458601348102094,
      "selected_pairs": 26454.8,
      "state_query_norm": 43.90405693054199,
      "state_relation_loss": 0.0002869590622140095,
      "step": 9750,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0048458601348102094,
      "weighted_relation_loss": 0.0048458601348102094
    },
    {
      "action_query_norm": 42.94532661437988,
      "action_relation_loss": 0.0045190256554633376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.130028063610851,
      "flow_matching_loss": 0.043508315458893776,
      "grad_norm": 0.1755535900592804,
      "learning_rate": 8.067147592736962e-05,
      "loss": 0.0483,
      "relation_loss": 0.00479498696513474,
      "selected_pairs": 27063.0,
      "state_query_norm": 43.904822158813474,
      "state_relation_loss": 0.00027596128056757153,
      "step": 9760,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.00479498696513474,
      "weighted_relation_loss": 0.00479498696513474
    },
    {
      "action_query_norm": 42.972708511352536,
      "action_relation_loss": 0.004588420130312442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.139382600561273,
      "flow_matching_loss": 0.04159580208361149,
      "grad_norm": 0.1127086877822876,
      "learning_rate": 8.062792973806357e-05,
      "loss": 0.0465,
      "relation_loss": 0.004884116817265749,
      "selected_pairs": 26458.3,
      "state_query_norm": 43.90934028625488,
      "state_relation_loss": 0.00029569666658062487,
      "step": 9770,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.004884116817265749,
      "weighted_relation_loss": 0.004884116817265749
    },
    {
      "action_query_norm": 42.9714054107666,
      "action_relation_loss": 0.004525377880781889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.148737137511693,
      "flow_matching_loss": 0.04148213304579258,
      "grad_norm": 0.14068906009197235,
      "learning_rate": 8.058434633293485e-05,
      "loss": 0.0463,
      "relation_loss": 0.004815569939091801,
      "selected_pairs": 26735.5,
      "state_query_norm": 43.89520797729492,
      "state_relation_loss": 0.0002901919855503365,
      "step": 9780,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004815569939091801,
      "weighted_relation_loss": 0.004815569939091801
    },
    {
      "action_query_norm": 42.9486888885498,
      "action_relation_loss": 0.004560828674584627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.158091674462113,
      "flow_matching_loss": 0.04095632433891296,
      "grad_norm": 0.13602006435394287,
      "learning_rate": 8.054072576494143e-05,
      "loss": 0.0458,
      "relation_loss": 0.0048482690937817095,
      "selected_pairs": 26499.6,
      "state_query_norm": 43.8963119506836,
      "state_relation_loss": 0.0002874404439353384,
      "step": 9790,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0048482690937817095,
      "weighted_relation_loss": 0.0048482690937817095
    },
    {
      "action_query_norm": 42.94476470947266,
      "action_relation_loss": 0.004665436083450913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.167446211412535,
      "flow_matching_loss": 0.043487843871116635,
      "grad_norm": 0.12494390457868576,
      "learning_rate": 8.04970680870864e-05,
      "loss": 0.0484,
      "relation_loss": 0.004949446115642786,
      "selected_pairs": 27030.0,
      "state_query_norm": 43.907853317260745,
      "state_relation_loss": 0.0002840100845787674,
      "step": 9800,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004949446115642786,
      "weighted_relation_loss": 0.004949446115642786
    },
    {
      "action_query_norm": 42.94003524780273,
      "action_relation_loss": 0.004508730582892894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.176800748362956,
      "flow_matching_loss": 0.04226373098790646,
      "grad_norm": 0.12301168590784073,
      "learning_rate": 8.045337335241793e-05,
      "loss": 0.0471,
      "relation_loss": 0.0047928932122886184,
      "selected_pairs": 26800.4,
      "state_query_norm": 43.88762016296387,
      "state_relation_loss": 0.0002841626032022759,
      "step": 9810,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0047928932122886184,
      "weighted_relation_loss": 0.0047928932122886184
    },
    {
      "action_query_norm": 42.942547988891604,
      "action_relation_loss": 0.004599036276340484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.186155285313378,
      "flow_matching_loss": 0.04183984883129597,
      "grad_norm": 0.1324826329946518,
      "learning_rate": 8.040964161402932e-05,
      "loss": 0.0467,
      "relation_loss": 0.004884950304403901,
      "selected_pairs": 26989.0,
      "state_query_norm": 43.893002319335935,
      "state_relation_loss": 0.00028591402369784193,
      "step": 9820,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004884950304403901,
      "weighted_relation_loss": 0.004884950304403901
    },
    {
      "action_query_norm": 42.934668731689456,
      "action_relation_loss": 0.004461403051391244,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.195509822263798,
      "flow_matching_loss": 0.04209759533405304,
      "grad_norm": 0.13949503004550934,
      "learning_rate": 8.036587292505869e-05,
      "loss": 0.0469,
      "relation_loss": 0.004752853047102689,
      "selected_pairs": 27173.4,
      "state_query_norm": 43.89203109741211,
      "state_relation_loss": 0.00029144997388357296,
      "step": 9830,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.004752853047102689,
      "weighted_relation_loss": 0.004752853047102689
    },
    {
      "action_query_norm": 42.96077308654785,
      "action_relation_loss": 0.004344241507351398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.204864359214218,
      "flow_matching_loss": 0.043160290271043775,
      "grad_norm": 0.13170740008354187,
      "learning_rate": 8.032206733868912e-05,
      "loss": 0.0478,
      "relation_loss": 0.004643059195950627,
      "selected_pairs": 27010.4,
      "state_query_norm": 43.89578514099121,
      "state_relation_loss": 0.0002988176987855695,
      "step": 9840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004643059195950627,
      "weighted_relation_loss": 0.004643059195950627
    },
    {
      "action_query_norm": 42.942770385742186,
      "action_relation_loss": 0.004624425759539008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.21421889616464,
      "flow_matching_loss": 0.04411791600286961,
      "grad_norm": 0.1378062665462494,
      "learning_rate": 8.027822490814858e-05,
      "loss": 0.049,
      "relation_loss": 0.004909850237891078,
      "selected_pairs": 26600.0,
      "state_query_norm": 43.89479370117188,
      "state_relation_loss": 0.00028542446671053767,
      "step": 9850,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004909850237891078,
      "weighted_relation_loss": 0.004909850237891078
    },
    {
      "action_query_norm": 42.94640846252442,
      "action_relation_loss": 0.004578444128856063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.22357343311506,
      "flow_matching_loss": 0.04486177340149879,
      "grad_norm": 0.16760790348052979,
      "learning_rate": 8.023434568670972e-05,
      "loss": 0.0497,
      "relation_loss": 0.004869875404983759,
      "selected_pairs": 27033.0,
      "state_query_norm": 43.88798446655274,
      "state_relation_loss": 0.0002914313139626756,
      "step": 9860,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004869875404983759,
      "weighted_relation_loss": 0.004869875404983759
    },
    {
      "action_query_norm": 42.9619815826416,
      "action_relation_loss": 0.004399803467094898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.232927970065482,
      "flow_matching_loss": 0.04414447695016861,
      "grad_norm": 0.13000711798667908,
      "learning_rate": 8.019042972768992e-05,
      "loss": 0.0489,
      "relation_loss": 0.004715035296976566,
      "selected_pairs": 27059.4,
      "state_query_norm": 43.88212203979492,
      "state_relation_loss": 0.00031523180659860374,
      "step": 9870,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004715035296976566,
      "weighted_relation_loss": 0.004715035296976566
    },
    {
      "action_query_norm": 42.93482933044434,
      "action_relation_loss": 0.004472045181319118,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.242282507015902,
      "flow_matching_loss": 0.04027652144432068,
      "grad_norm": 0.16552820801734924,
      "learning_rate": 8.014647708445124e-05,
      "loss": 0.045,
      "relation_loss": 0.004767857817932963,
      "selected_pairs": 27130.3,
      "state_query_norm": 43.87439956665039,
      "state_relation_loss": 0.0002958125493023545,
      "step": 9880,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004767857817932963,
      "weighted_relation_loss": 0.004767857817932963
    },
    {
      "action_query_norm": 42.95859031677246,
      "action_relation_loss": 0.004439353290945291,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.251637043966324,
      "flow_matching_loss": 0.04125928319990635,
      "grad_norm": 0.12477488815784454,
      "learning_rate": 8.010248781040027e-05,
      "loss": 0.046,
      "relation_loss": 0.004739679023623466,
      "selected_pairs": 26738.0,
      "state_query_norm": 43.88942451477051,
      "state_relation_loss": 0.00030032576760277154,
      "step": 9890,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004739679023623466,
      "weighted_relation_loss": 0.004739679023623466
    },
    {
      "action_query_norm": 42.92060546875,
      "action_relation_loss": 0.004586373781785369,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.260991580916745,
      "flow_matching_loss": 0.0434944961220026,
      "grad_norm": 0.1337260752916336,
      "learning_rate": 8.005846195898814e-05,
      "loss": 0.0484,
      "relation_loss": 0.004885459505021572,
      "selected_pairs": 26387.5,
      "state_query_norm": 43.86948471069336,
      "state_relation_loss": 0.0002990856795804575,
      "step": 9900,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.004885459505021572,
      "weighted_relation_loss": 0.004885459505021572
    },
    {
      "action_query_norm": 42.92077217102051,
      "action_relation_loss": 0.004659823235124349,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.270346117867165,
      "flow_matching_loss": 0.04027962498366833,
      "grad_norm": 0.14900453388690948,
      "learning_rate": 8.00143995837104e-05,
      "loss": 0.0452,
      "relation_loss": 0.0049340090248733755,
      "selected_pairs": 27361.5,
      "state_query_norm": 43.87471580505371,
      "state_relation_loss": 0.0002741858246736228,
      "step": 9910,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0049340090248733755,
      "weighted_relation_loss": 0.0049340090248733755
    },
    {
      "action_query_norm": 42.93149261474609,
      "action_relation_loss": 0.004692488186992705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.279700654817587,
      "flow_matching_loss": 0.04204538874328136,
      "grad_norm": 0.1536063849925995,
      "learning_rate": 7.997030073810698e-05,
      "loss": 0.047,
      "relation_loss": 0.004971117386594415,
      "selected_pairs": 27387.0,
      "state_query_norm": 43.86530723571777,
      "state_relation_loss": 0.0002786291908705607,
      "step": 9920,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.004971117386594415,
      "weighted_relation_loss": 0.004971117386594415
    },
    {
      "action_query_norm": 42.90607032775879,
      "action_relation_loss": 0.004402164788916707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.289055191768007,
      "flow_matching_loss": 0.04369921162724495,
      "grad_norm": 0.14987130463123322,
      "learning_rate": 7.992616547576217e-05,
      "loss": 0.0484,
      "relation_loss": 0.004671708773821592,
      "selected_pairs": 26413.0,
      "state_query_norm": 43.886264419555665,
      "state_relation_loss": 0.000269543967442587,
      "step": 9930,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004671708773821592,
      "weighted_relation_loss": 0.004671708773821592
    },
    {
      "action_query_norm": 42.93026885986328,
      "action_relation_loss": 0.0044357896549627185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.298409728718429,
      "flow_matching_loss": 0.041881751641631125,
      "grad_norm": 0.09554803371429443,
      "learning_rate": 7.988199385030445e-05,
      "loss": 0.0466,
      "relation_loss": 0.004706341633573174,
      "selected_pairs": 26960.3,
      "state_query_norm": 43.875919342041016,
      "state_relation_loss": 0.00027055200189352036,
      "step": 9940,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.004706341633573174,
      "weighted_relation_loss": 0.004706341633573174
    },
    {
      "action_query_norm": 42.90527191162109,
      "action_relation_loss": 0.004325008299201727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.30776426566885,
      "flow_matching_loss": 0.04117053337395191,
      "grad_norm": 0.1129649430513382,
      "learning_rate": 7.98377859154065e-05,
      "loss": 0.0458,
      "relation_loss": 0.00460748104378581,
      "selected_pairs": 26821.6,
      "state_query_norm": 43.868700408935545,
      "state_relation_loss": 0.0002824727474944666,
      "step": 9950,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.00460748104378581,
      "weighted_relation_loss": 0.00460748104378581
    },
    {
      "action_query_norm": 42.900186920166014,
      "action_relation_loss": 0.004496192838996649,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.31711880261927,
      "flow_matching_loss": 0.042604440450668336,
      "grad_norm": 0.13594524562358856,
      "learning_rate": 7.979354172478516e-05,
      "loss": 0.0474,
      "relation_loss": 0.004786836449056864,
      "selected_pairs": 26922.8,
      "state_query_norm": 43.86592178344726,
      "state_relation_loss": 0.0002906435751356184,
      "step": 9960,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004786836449056864,
      "weighted_relation_loss": 0.004786836449056864
    },
    {
      "action_query_norm": 42.898063278198244,
      "action_relation_loss": 0.004471360705792904,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.326473339569691,
      "flow_matching_loss": 0.04230794459581375,
      "grad_norm": 0.13624314963817596,
      "learning_rate": 7.974926133220127e-05,
      "loss": 0.0471,
      "relation_loss": 0.004759079031646252,
      "selected_pairs": 26801.1,
      "state_query_norm": 43.86941223144531,
      "state_relation_loss": 0.00028771842480637135,
      "step": 9970,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.004759079031646252,
      "weighted_relation_loss": 0.004759079031646252
    },
    {
      "action_query_norm": 42.90232391357422,
      "action_relation_loss": 0.004316081712022424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.335827876520112,
      "flow_matching_loss": 0.043906936421990395,
      "grad_norm": 0.11128062009811401,
      "learning_rate": 7.970494479145967e-05,
      "loss": 0.0485,
      "relation_loss": 0.004606863204389811,
      "selected_pairs": 26986.7,
      "state_query_norm": 43.860130310058594,
      "state_relation_loss": 0.000290781504008919,
      "step": 9980,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004606863204389811,
      "weighted_relation_loss": 0.004606863204389811
    },
    {
      "action_query_norm": 42.92453117370606,
      "action_relation_loss": 0.0045378259383141994,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.345182413470534,
      "flow_matching_loss": 0.04210234731435776,
      "grad_norm": 0.1451738327741623,
      "learning_rate": 7.966059215640919e-05,
      "loss": 0.0469,
      "relation_loss": 0.004824037849903107,
      "selected_pairs": 27013.6,
      "state_query_norm": 43.86476135253906,
      "state_relation_loss": 0.0002862118708435446,
      "step": 9990,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004824037849903107,
      "weighted_relation_loss": 0.004824037849903107
    },
    {
      "action_query_norm": 42.91828155517578,
      "action_relation_loss": 0.004664387693628669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.354536950420954,
      "flow_matching_loss": 0.04083852171897888,
      "grad_norm": 0.14055536687374115,
      "learning_rate": 7.961620348094241e-05,
      "loss": 0.0458,
      "relation_loss": 0.004938376229256391,
      "selected_pairs": 26443.5,
      "state_query_norm": 43.85973587036133,
      "state_relation_loss": 0.00027398857928346844,
      "step": 10000,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004938376229256391,
      "weighted_relation_loss": 0.004938376229256391
    },
    {
      "action_query_norm": 42.89957046508789,
      "action_relation_loss": 0.004592484422028064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.363891487371376,
      "flow_matching_loss": 0.04471989534795284,
      "grad_norm": 0.12717503309249878,
      "learning_rate": 7.957177881899577e-05,
      "loss": 0.0496,
      "relation_loss": 0.004858666146174073,
      "selected_pairs": 25962.7,
      "state_query_norm": 43.85670928955078,
      "state_relation_loss": 0.0002661817692569457,
      "step": 10010,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.004858666146174073,
      "weighted_relation_loss": 0.004858666146174073
    },
    {
      "action_query_norm": 42.88125152587891,
      "action_relation_loss": 0.00443004744593054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.373246024321796,
      "flow_matching_loss": 0.042516285181045534,
      "grad_norm": 0.12347987294197083,
      "learning_rate": 7.952731822454944e-05,
      "loss": 0.0472,
      "relation_loss": 0.004698798293247819,
      "selected_pairs": 27070.1,
      "state_query_norm": 43.85017967224121,
      "state_relation_loss": 0.0002687507861992344,
      "step": 10020,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004698798293247819,
      "weighted_relation_loss": 0.004698798293247819
    },
    {
      "action_query_norm": 42.87261810302734,
      "action_relation_loss": 0.004729294171556831,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.382600561272216,
      "flow_matching_loss": 0.04407857656478882,
      "grad_norm": 0.10827614367008209,
      "learning_rate": 7.948282175162722e-05,
      "loss": 0.0491,
      "relation_loss": 0.004985256725922227,
      "selected_pairs": 26922.1,
      "state_query_norm": 43.865441513061526,
      "state_relation_loss": 0.000255962491792161,
      "step": 10030,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004985256725922227,
      "weighted_relation_loss": 0.004985256725922227
    },
    {
      "action_query_norm": 42.88226547241211,
      "action_relation_loss": 0.004399646911770106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.391955098222638,
      "flow_matching_loss": 0.04129451587796211,
      "grad_norm": 0.1265108585357666,
      "learning_rate": 7.943828945429652e-05,
      "loss": 0.046,
      "relation_loss": 0.004660655651241541,
      "selected_pairs": 26778.7,
      "state_query_norm": 43.85275573730469,
      "state_relation_loss": 0.000261008671077434,
      "step": 10040,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004660655651241541,
      "weighted_relation_loss": 0.004660655651241541
    },
    {
      "action_query_norm": 42.877708435058594,
      "action_relation_loss": 0.004374241386540234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.401309635173059,
      "flow_matching_loss": 0.042824941128492354,
      "grad_norm": 0.12271279096603394,
      "learning_rate": 7.939372138666827e-05,
      "loss": 0.0474,
      "relation_loss": 0.004623217601329088,
      "selected_pairs": 26604.3,
      "state_query_norm": 43.851247787475586,
      "state_relation_loss": 0.0002489761827746406,
      "step": 10050,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004623217601329088,
      "weighted_relation_loss": 0.004623217601329088
    },
    {
      "action_query_norm": 42.84829483032227,
      "action_relation_loss": 0.004189975350163877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.41066417212348,
      "flow_matching_loss": 0.03964447230100632,
      "grad_norm": 0.15350207686424255,
      "learning_rate": 7.934911760289692e-05,
      "loss": 0.0441,
      "relation_loss": 0.004456431092694402,
      "selected_pairs": 26763.5,
      "state_query_norm": 43.84638366699219,
      "state_relation_loss": 0.00026645576435839756,
      "step": 10060,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004456431092694402,
      "weighted_relation_loss": 0.004456431092694402
    },
    {
      "action_query_norm": 42.837982177734375,
      "action_relation_loss": 0.004618962854146957,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.4200187090739,
      "flow_matching_loss": 0.0434703279286623,
      "grad_norm": 0.1174330785870552,
      "learning_rate": 7.930447815718022e-05,
      "loss": 0.0484,
      "relation_loss": 0.00488522183150053,
      "selected_pairs": 26992.9,
      "state_query_norm": 43.84844818115234,
      "state_relation_loss": 0.0002662590748514049,
      "step": 10070,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00488522183150053,
      "weighted_relation_loss": 0.00488522183150053
    },
    {
      "action_query_norm": 42.83904838562012,
      "action_relation_loss": 0.004630227666348219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.429373246024323,
      "flow_matching_loss": 0.04030882380902767,
      "grad_norm": 0.12402574717998505,
      "learning_rate": 7.925980310375934e-05,
      "loss": 0.0452,
      "relation_loss": 0.004889183724299073,
      "selected_pairs": 26988.3,
      "state_query_norm": 43.84844551086426,
      "state_relation_loss": 0.00025895611906889826,
      "step": 10080,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004889183724299073,
      "weighted_relation_loss": 0.004889183724299073
    },
    {
      "action_query_norm": 42.85981063842773,
      "action_relation_loss": 0.004256684798747301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.438727782974743,
      "flow_matching_loss": 0.04193015918135643,
      "grad_norm": 0.13044199347496033,
      "learning_rate": 7.921509249691865e-05,
      "loss": 0.0465,
      "relation_loss": 0.004524347931146621,
      "selected_pairs": 26924.7,
      "state_query_norm": 43.846867370605466,
      "state_relation_loss": 0.00026766305381897837,
      "step": 10090,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004524347931146621,
      "weighted_relation_loss": 0.004524347931146621
    },
    {
      "action_query_norm": 42.8919563293457,
      "action_relation_loss": 0.004545858222991228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.448082319925163,
      "flow_matching_loss": 0.04275795109570026,
      "grad_norm": 0.16737322509288788,
      "learning_rate": 7.917034639098579e-05,
      "loss": 0.0476,
      "relation_loss": 0.004809100879356265,
      "selected_pairs": 26757.3,
      "state_query_norm": 43.83726081848145,
      "state_relation_loss": 0.000263242753862869,
      "step": 10100,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004809100879356265,
      "weighted_relation_loss": 0.004809100879356265
    },
    {
      "action_query_norm": 42.86476631164551,
      "action_relation_loss": 0.004465323104523122,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.457436856875585,
      "flow_matching_loss": 0.04261722043156624,
      "grad_norm": 0.14321830868721008,
      "learning_rate": 7.912556484033146e-05,
      "loss": 0.0473,
      "relation_loss": 0.004721287311986089,
      "selected_pairs": 26947.2,
      "state_query_norm": 43.837927627563474,
      "state_relation_loss": 0.0002559642176493071,
      "step": 10110,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004721287311986089,
      "weighted_relation_loss": 0.004721287311986089
    },
    {
      "action_query_norm": 42.858219146728516,
      "action_relation_loss": 0.004613534105010331,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.466791393826005,
      "flow_matching_loss": 0.042460684478282926,
      "grad_norm": 0.1650298833847046,
      "learning_rate": 7.908074789936952e-05,
      "loss": 0.0473,
      "relation_loss": 0.004881726624444127,
      "selected_pairs": 26439.1,
      "state_query_norm": 43.8361873626709,
      "state_relation_loss": 0.0002681925645447336,
      "step": 10120,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004881726624444127,
      "weighted_relation_loss": 0.004881726624444127
    },
    {
      "action_query_norm": 42.86565399169922,
      "action_relation_loss": 0.004307612124830484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.476145930776427,
      "flow_matching_loss": 0.04238546676933765,
      "grad_norm": 0.14282546937465668,
      "learning_rate": 7.903589562255673e-05,
      "loss": 0.0469,
      "relation_loss": 0.004562503844499588,
      "selected_pairs": 27100.8,
      "state_query_norm": 43.841373443603516,
      "state_relation_loss": 0.00025489175168331714,
      "step": 10130,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.004562503844499588,
      "weighted_relation_loss": 0.004562503844499588
    },
    {
      "action_query_norm": 42.8339485168457,
      "action_relation_loss": 0.004430148098617792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.485500467726848,
      "flow_matching_loss": 0.04125153943896294,
      "grad_norm": 0.11926688253879547,
      "learning_rate": 7.899100806439285e-05,
      "loss": 0.0459,
      "relation_loss": 0.004686045693233609,
      "selected_pairs": 26595.2,
      "state_query_norm": 43.84441871643067,
      "state_relation_loss": 0.00025589763972675426,
      "step": 10140,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004686045693233609,
      "weighted_relation_loss": 0.004686045693233609
    },
    {
      "action_query_norm": 42.83758277893067,
      "action_relation_loss": 0.0046459479257464405,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.494855004677268,
      "flow_matching_loss": 0.04193579852581024,
      "grad_norm": 0.12374287843704224,
      "learning_rate": 7.894608527942049e-05,
      "loss": 0.0468,
      "relation_loss": 0.004909548861905933,
      "selected_pairs": 26553.8,
      "state_query_norm": 43.837839889526364,
      "state_relation_loss": 0.0002636009405250661,
      "step": 10150,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.004909548861905933,
      "weighted_relation_loss": 0.004909548861905933
    },
    {
      "action_query_norm": 42.84360885620117,
      "action_relation_loss": 0.004300077236257494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.50420954162769,
      "flow_matching_loss": 0.04049925245344639,
      "grad_norm": 0.12359768152236938,
      "learning_rate": 7.89011273222251e-05,
      "loss": 0.0451,
      "relation_loss": 0.004559343820437789,
      "selected_pairs": 26991.3,
      "state_query_norm": 43.832861709594724,
      "state_relation_loss": 0.00025926656526280566,
      "step": 10160,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004559343820437789,
      "weighted_relation_loss": 0.004559343820437789
    },
    {
      "action_query_norm": 42.83513984680176,
      "action_relation_loss": 0.004344647284597158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.51356407857811,
      "flow_matching_loss": 0.04325905069708824,
      "grad_norm": 0.14130209386348724,
      "learning_rate": 7.885613424743482e-05,
      "loss": 0.0479,
      "relation_loss": 0.004608963429927826,
      "selected_pairs": 26551.2,
      "state_query_norm": 43.82863311767578,
      "state_relation_loss": 0.00026431617297930645,
      "step": 10170,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.004608963429927826,
      "weighted_relation_loss": 0.004608963429927826
    },
    {
      "action_query_norm": 42.80644989013672,
      "action_relation_loss": 0.0044141374062746765,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.522918615528532,
      "flow_matching_loss": 0.044044113531708715,
      "grad_norm": 0.13286323845386505,
      "learning_rate": 7.881110610972044e-05,
      "loss": 0.0487,
      "relation_loss": 0.004677898203954101,
      "selected_pairs": 27037.7,
      "state_query_norm": 43.833290100097656,
      "state_relation_loss": 0.0002637607976794243,
      "step": 10180,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004677898203954101,
      "weighted_relation_loss": 0.004677898203954101
    },
    {
      "action_query_norm": 42.799403762817384,
      "action_relation_loss": 0.004492791835218668,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.532273152478952,
      "flow_matching_loss": 0.04232778735458851,
      "grad_norm": 0.12505018711090088,
      "learning_rate": 7.876604296379544e-05,
      "loss": 0.0471,
      "relation_loss": 0.004747076984494924,
      "selected_pairs": 26849.8,
      "state_query_norm": 43.82361068725586,
      "state_relation_loss": 0.0002542852016631514,
      "step": 10190,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004747076984494924,
      "weighted_relation_loss": 0.004747076984494924
    },
    {
      "action_query_norm": 42.826616287231445,
      "action_relation_loss": 0.004357847152277827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.541627689429372,
      "flow_matching_loss": 0.04289374612271786,
      "grad_norm": 0.10576240718364716,
      "learning_rate": 7.872094486441579e-05,
      "loss": 0.0475,
      "relation_loss": 0.004627980384975671,
      "selected_pairs": 26728.3,
      "state_query_norm": 43.8254581451416,
      "state_relation_loss": 0.00027013318322133273,
      "step": 10200,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004627980384975671,
      "weighted_relation_loss": 0.004627980384975671
    },
    {
      "action_query_norm": 42.847076416015625,
      "action_relation_loss": 0.004226854559965432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.550982226379794,
      "flow_matching_loss": 0.042223941534757614,
      "grad_norm": 0.130096435546875,
      "learning_rate": 7.867581186637991e-05,
      "loss": 0.0467,
      "relation_loss": 0.004480723617598414,
      "selected_pairs": 27325.5,
      "state_query_norm": 43.822421646118165,
      "state_relation_loss": 0.00025386904599145055,
      "step": 10210,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.004480723617598414,
      "weighted_relation_loss": 0.004480723617598414
    },
    {
      "action_query_norm": 42.82803153991699,
      "action_relation_loss": 0.004321979032829404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.560336763330215,
      "flow_matching_loss": 0.041551883518695834,
      "grad_norm": 0.15348851680755615,
      "learning_rate": 7.863064402452866e-05,
      "loss": 0.0461,
      "relation_loss": 0.004586932482197881,
      "selected_pairs": 26875.3,
      "state_query_norm": 43.81255111694336,
      "state_relation_loss": 0.0002649534842930734,
      "step": 10220,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004586932482197881,
      "weighted_relation_loss": 0.004586932482197881
    },
    {
      "action_query_norm": 42.79876899719238,
      "action_relation_loss": 0.004234418948180973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.569691300280637,
      "flow_matching_loss": 0.041791925206780434,
      "grad_norm": 0.11193431913852692,
      "learning_rate": 7.858544139374524e-05,
      "loss": 0.0463,
      "relation_loss": 0.004494536318816245,
      "selected_pairs": 26479.4,
      "state_query_norm": 43.81941299438476,
      "state_relation_loss": 0.00026011734589701516,
      "step": 10230,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.004494536318816245,
      "weighted_relation_loss": 0.004494536318816245
    },
    {
      "action_query_norm": 42.80351257324219,
      "action_relation_loss": 0.004482110729441047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.579045837231057,
      "flow_matching_loss": 0.04250505864620209,
      "grad_norm": 0.1567823588848114,
      "learning_rate": 7.854020402895508e-05,
      "loss": 0.0472,
      "relation_loss": 0.0047420862596482035,
      "selected_pairs": 27347.9,
      "state_query_norm": 43.81709289550781,
      "state_relation_loss": 0.0002599755738629028,
      "step": 10240,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0047420862596482035,
      "weighted_relation_loss": 0.0047420862596482035
    },
    {
      "action_query_norm": 42.80457954406738,
      "action_relation_loss": 0.004412411758676171,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.588400374181479,
      "flow_matching_loss": 0.042984592542052266,
      "grad_norm": 0.10827602446079254,
      "learning_rate": 7.849493198512587e-05,
      "loss": 0.0477,
      "relation_loss": 0.0046766570303589106,
      "selected_pairs": 26225.9,
      "state_query_norm": 43.82726287841797,
      "state_relation_loss": 0.000264245287689846,
      "step": 10250,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0046766570303589106,
      "weighted_relation_loss": 0.0046766570303589106
    },
    {
      "action_query_norm": 42.82985916137695,
      "action_relation_loss": 0.004224053747020662,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.597754911131899,
      "flow_matching_loss": 0.040792576223611834,
      "grad_norm": 0.09865312278270721,
      "learning_rate": 7.84496253172674e-05,
      "loss": 0.0453,
      "relation_loss": 0.004485962446779013,
      "selected_pairs": 26416.1,
      "state_query_norm": 43.80390930175781,
      "state_relation_loss": 0.0002619087477796711,
      "step": 10260,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.004485962446779013,
      "weighted_relation_loss": 0.004485962446779013
    },
    {
      "action_query_norm": 42.81447982788086,
      "action_relation_loss": 0.0043743288610130545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.60710944808232,
      "flow_matching_loss": 0.04152282029390335,
      "grad_norm": 0.12942078709602356,
      "learning_rate": 7.840428408043155e-05,
      "loss": 0.0462,
      "relation_loss": 0.004646791610866785,
      "selected_pairs": 26923.1,
      "state_query_norm": 43.8084098815918,
      "state_relation_loss": 0.0002724627280258574,
      "step": 10270,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004646791610866785,
      "weighted_relation_loss": 0.004646791610866785
    },
    {
      "action_query_norm": 42.77681350708008,
      "action_relation_loss": 0.004496187856420875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.616463985032741,
      "flow_matching_loss": 0.041417348012328145,
      "grad_norm": 0.1652030348777771,
      "learning_rate": 7.835890832971218e-05,
      "loss": 0.0462,
      "relation_loss": 0.0047643448691815134,
      "selected_pairs": 26710.8,
      "state_query_norm": 43.813012313842776,
      "state_relation_loss": 0.0002681570025742985,
      "step": 10280,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0047643448691815134,
      "weighted_relation_loss": 0.0047643448691815134
    },
    {
      "action_query_norm": 42.76826629638672,
      "action_relation_loss": 0.00436277121771127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.625818521983161,
      "flow_matching_loss": 0.041854899004101755,
      "grad_norm": 0.13662728667259216,
      "learning_rate": 7.831349812024512e-05,
      "loss": 0.0465,
      "relation_loss": 0.00463915872387588,
      "selected_pairs": 27224.6,
      "state_query_norm": 43.799767303466794,
      "state_relation_loss": 0.00027638752362690866,
      "step": 10290,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00463915872387588,
      "weighted_relation_loss": 0.00463915872387588
    },
    {
      "action_query_norm": 42.77440338134765,
      "action_relation_loss": 0.004354127706028521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.635173058933583,
      "flow_matching_loss": 0.040659360215067865,
      "grad_norm": 0.1346285492181778,
      "learning_rate": 7.826805350720807e-05,
      "loss": 0.0453,
      "relation_loss": 0.004619510052725673,
      "selected_pairs": 27132.2,
      "state_query_norm": 43.81759910583496,
      "state_relation_loss": 0.0002653822200954892,
      "step": 10300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004619510052725673,
      "weighted_relation_loss": 0.004619510052725673
    },
    {
      "action_query_norm": 42.780562210083005,
      "action_relation_loss": 0.004396856459788978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.644527595884004,
      "flow_matching_loss": 0.04450401589274407,
      "grad_norm": 0.16373883187770844,
      "learning_rate": 7.822257454582049e-05,
      "loss": 0.0492,
      "relation_loss": 0.004679303127340972,
      "selected_pairs": 27266.0,
      "state_query_norm": 43.80197944641113,
      "state_relation_loss": 0.0002824466559104621,
      "step": 10310,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004679303127340972,
      "weighted_relation_loss": 0.004679303127340972
    },
    {
      "action_query_norm": 42.806786346435544,
      "action_relation_loss": 0.004463608469814062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.653882132834426,
      "flow_matching_loss": 0.04156087003648281,
      "grad_norm": 0.17781062424182892,
      "learning_rate": 7.817706129134363e-05,
      "loss": 0.0463,
      "relation_loss": 0.004748366726562381,
      "selected_pairs": 26299.1,
      "state_query_norm": 43.80660896301269,
      "state_relation_loss": 0.00028475824947236104,
      "step": 10320,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.004748366726562381,
      "weighted_relation_loss": 0.004748366726562381
    },
    {
      "action_query_norm": 42.76921768188477,
      "action_relation_loss": 0.004443688318133354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.663236669784846,
      "flow_matching_loss": 0.04381221905350685,
      "grad_norm": 0.1817813515663147,
      "learning_rate": 7.813151379908036e-05,
      "loss": 0.0485,
      "relation_loss": 0.004728380544111133,
      "selected_pairs": 27251.1,
      "state_query_norm": 43.80030632019043,
      "state_relation_loss": 0.0002846922288881615,
      "step": 10330,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.004728380544111133,
      "weighted_relation_loss": 0.004728380544111133
    },
    {
      "action_query_norm": 42.78659782409668,
      "action_relation_loss": 0.00433144005946815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.672591206735266,
      "flow_matching_loss": 0.04170345515012741,
      "grad_norm": 0.1488935649394989,
      "learning_rate": 7.808593212437523e-05,
      "loss": 0.0463,
      "relation_loss": 0.004626624472439289,
      "selected_pairs": 27365.9,
      "state_query_norm": 43.81595993041992,
      "state_relation_loss": 0.0002951843984192237,
      "step": 10340,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.004626624472439289,
      "weighted_relation_loss": 0.004626624472439289
    },
    {
      "action_query_norm": 42.768608474731444,
      "action_relation_loss": 0.004352360125631094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.681945743685688,
      "flow_matching_loss": 0.04124853052198887,
      "grad_norm": 0.11880116164684296,
      "learning_rate": 7.804031632261421e-05,
      "loss": 0.0459,
      "relation_loss": 0.004650145908817649,
      "selected_pairs": 25797.4,
      "state_query_norm": 43.79492378234863,
      "state_relation_loss": 0.00029778569878544656,
      "step": 10350,
      "valid_action_samples": 227.6,
      "weighted_auxiliary_loss": 0.004650145908817649,
      "weighted_relation_loss": 0.004650145908817649
    },
    {
      "action_query_norm": 42.79274711608887,
      "action_relation_loss": 0.0041416778229177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.691300280636108,
      "flow_matching_loss": 0.04387366473674774,
      "grad_norm": 0.1321437507867813,
      "learning_rate": 7.799466644922484e-05,
      "loss": 0.0483,
      "relation_loss": 0.004409707523882389,
      "selected_pairs": 26994.8,
      "state_query_norm": 43.80315742492676,
      "state_relation_loss": 0.0002680297155166045,
      "step": 10360,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004409707523882389,
      "weighted_relation_loss": 0.004409707523882389
    },
    {
      "action_query_norm": 42.76879997253418,
      "action_relation_loss": 0.004134837444871664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.70065481758653,
      "flow_matching_loss": 0.04159581698477268,
      "grad_norm": 0.15790429711341858,
      "learning_rate": 7.794898255967602e-05,
      "loss": 0.046,
      "relation_loss": 0.0043961668852716684,
      "selected_pairs": 26622.7,
      "state_query_norm": 43.803399658203126,
      "state_relation_loss": 0.00026132943603442984,
      "step": 10370,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0043961668852716684,
      "weighted_relation_loss": 0.0043961668852716684
    },
    {
      "action_query_norm": 42.76380386352539,
      "action_relation_loss": 0.0044529679464176295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.71000935453695,
      "flow_matching_loss": 0.043109501525759694,
      "grad_norm": 0.1264084428548813,
      "learning_rate": 7.790326470947795e-05,
      "loss": 0.0478,
      "relation_loss": 0.004712895583361387,
      "selected_pairs": 27182.3,
      "state_query_norm": 43.79126052856445,
      "state_relation_loss": 0.0002599276413093321,
      "step": 10380,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.004712895583361387,
      "weighted_relation_loss": 0.004712895583361387
    },
    {
      "action_query_norm": 42.759425735473634,
      "action_relation_loss": 0.004239356378093362,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.71936389148737,
      "flow_matching_loss": 0.04218449965119362,
      "grad_norm": 0.1374433934688568,
      "learning_rate": 7.785751295418217e-05,
      "loss": 0.0467,
      "relation_loss": 0.004507356556132436,
      "selected_pairs": 27687.7,
      "state_query_norm": 43.78810157775879,
      "state_relation_loss": 0.00026800015039043503,
      "step": 10390,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.004507356556132436,
      "weighted_relation_loss": 0.004507356556132436
    },
    {
      "action_query_norm": 42.75069007873535,
      "action_relation_loss": 0.004328173818066716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.728718428437793,
      "flow_matching_loss": 0.04122941456735134,
      "grad_norm": 0.16447806358337402,
      "learning_rate": 7.781172734938136e-05,
      "loss": 0.0458,
      "relation_loss": 0.004582772636786104,
      "selected_pairs": 26716.3,
      "state_query_norm": 43.793372344970706,
      "state_relation_loss": 0.0002545988303609192,
      "step": 10400,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.004582772636786104,
      "weighted_relation_loss": 0.004582772636786104
    },
    {
      "action_query_norm": 42.75471305847168,
      "action_relation_loss": 0.004117040918208659,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.738072965388213,
      "flow_matching_loss": 0.04108301028609276,
      "grad_norm": 0.11893278360366821,
      "learning_rate": 7.776590795070933e-05,
      "loss": 0.0455,
      "relation_loss": 0.004382959986105561,
      "selected_pairs": 26946.6,
      "state_query_norm": 43.78115234375,
      "state_relation_loss": 0.0002659190431586467,
      "step": 10410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004382959986105561,
      "weighted_relation_loss": 0.004382959986105561
    },
    {
      "action_query_norm": 42.75924186706543,
      "action_relation_loss": 0.004039914370514453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.747427502338635,
      "flow_matching_loss": 0.04109371416270733,
      "grad_norm": 0.15005788207054138,
      "learning_rate": 7.772005481384099e-05,
      "loss": 0.0454,
      "relation_loss": 0.00430756420828402,
      "selected_pairs": 26877.7,
      "state_query_norm": 43.792313766479495,
      "state_relation_loss": 0.00026764985523186624,
      "step": 10420,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00430756420828402,
      "weighted_relation_loss": 0.00430756420828402
    },
    {
      "action_query_norm": 42.75942535400391,
      "action_relation_loss": 0.004191679111681878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.756782039289055,
      "flow_matching_loss": 0.040294500067830086,
      "grad_norm": 0.1391831636428833,
      "learning_rate": 7.767416799449223e-05,
      "loss": 0.0447,
      "relation_loss": 0.004445119621232152,
      "selected_pairs": 26782.8,
      "state_query_norm": 43.78228950500488,
      "state_relation_loss": 0.00025344042805954816,
      "step": 10430,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004445119621232152,
      "weighted_relation_loss": 0.004445119621232152
    },
    {
      "action_query_norm": 42.730506896972656,
      "action_relation_loss": 0.004209901811555028,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.766136576239477,
      "flow_matching_loss": 0.04058828186243772,
      "grad_norm": 0.15555165708065033,
      "learning_rate": 7.762824754841984e-05,
      "loss": 0.0451,
      "relation_loss": 0.0044773325324058534,
      "selected_pairs": 27200.9,
      "state_query_norm": 43.78750801086426,
      "state_relation_loss": 0.00026743074849946427,
      "step": 10440,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0044773325324058534,
      "weighted_relation_loss": 0.0044773325324058534
    },
    {
      "action_query_norm": 42.718492889404295,
      "action_relation_loss": 0.004047039919532836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.775491113189897,
      "flow_matching_loss": 0.04168836250901222,
      "grad_norm": 0.1303788721561432,
      "learning_rate": 7.758229353142152e-05,
      "loss": 0.046,
      "relation_loss": 0.004325089487247169,
      "selected_pairs": 27150.5,
      "state_query_norm": 43.78758850097656,
      "state_relation_loss": 0.00027804953278973696,
      "step": 10450,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.004325089487247169,
      "weighted_relation_loss": 0.004325089487247169
    },
    {
      "action_query_norm": 42.73207740783691,
      "action_relation_loss": 0.004216941655613482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.784845650140317,
      "flow_matching_loss": 0.04007614925503731,
      "grad_norm": 0.14466513693332672,
      "learning_rate": 7.753630599933572e-05,
      "loss": 0.0445,
      "relation_loss": 0.00447176790330559,
      "selected_pairs": 26671.7,
      "state_query_norm": 43.778154373168945,
      "state_relation_loss": 0.00025482621131232006,
      "step": 10460,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00447176790330559,
      "weighted_relation_loss": 0.00447176790330559
    },
    {
      "action_query_norm": 42.73256645202637,
      "action_relation_loss": 0.004107381450012326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.79420018709074,
      "flow_matching_loss": 0.04200241528451443,
      "grad_norm": 0.12192001938819885,
      "learning_rate": 7.749028500804161e-05,
      "loss": 0.0464,
      "relation_loss": 0.004367510322481394,
      "selected_pairs": 27389.4,
      "state_query_norm": 43.795684814453125,
      "state_relation_loss": 0.00026012885064119473,
      "step": 10470,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.004367510322481394,
      "weighted_relation_loss": 0.004367510322481394
    },
    {
      "action_query_norm": 42.73618087768555,
      "action_relation_loss": 0.004131031036376953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.80355472404116,
      "flow_matching_loss": 0.04178224727511406,
      "grad_norm": 0.11753581464290619,
      "learning_rate": 7.744423061345906e-05,
      "loss": 0.0462,
      "relation_loss": 0.004380707838572561,
      "selected_pairs": 27174.5,
      "state_query_norm": 43.78521575927734,
      "state_relation_loss": 0.00024967679928522557,
      "step": 10480,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.004380707838572561,
      "weighted_relation_loss": 0.004380707838572561
    },
    {
      "action_query_norm": 42.73416328430176,
      "action_relation_loss": 0.004070009500719607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.812909260991582,
      "flow_matching_loss": 0.04105192497372627,
      "grad_norm": 0.12060369551181793,
      "learning_rate": 7.73981428715485e-05,
      "loss": 0.0454,
      "relation_loss": 0.004325189301744104,
      "selected_pairs": 26873.1,
      "state_query_norm": 43.77817306518555,
      "state_relation_loss": 0.0002551798170316033,
      "step": 10490,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004325189301744104,
      "weighted_relation_loss": 0.004325189301744104
    },
    {
      "action_query_norm": 42.746969604492186,
      "action_relation_loss": 0.00427303109318018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.822263797942002,
      "flow_matching_loss": 0.04318785779178143,
      "grad_norm": 0.17213964462280273,
      "learning_rate": 7.735202183831085e-05,
      "loss": 0.0477,
      "relation_loss": 0.004532286804169417,
      "selected_pairs": 27359.2,
      "state_query_norm": 43.78772392272949,
      "state_relation_loss": 0.0002592557677417062,
      "step": 10500,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.004532286804169417,
      "weighted_relation_loss": 0.004532286804169417
    },
    {
      "action_query_norm": 42.72409248352051,
      "action_relation_loss": 0.004252314288169146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.831618334892422,
      "flow_matching_loss": 0.0409439317882061,
      "grad_norm": 0.14112365245819092,
      "learning_rate": 7.730586756978757e-05,
      "loss": 0.0454,
      "relation_loss": 0.004505132697522641,
      "selected_pairs": 26695.9,
      "state_query_norm": 43.77989654541015,
      "state_relation_loss": 0.0002528184064431116,
      "step": 10510,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004505132697522641,
      "weighted_relation_loss": 0.004505132697522641
    },
    {
      "action_query_norm": 42.697439956665036,
      "action_relation_loss": 0.004246857319958508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.840972871842844,
      "flow_matching_loss": 0.039973160997033116,
      "grad_norm": 0.1043865829706192,
      "learning_rate": 7.72596801220604e-05,
      "loss": 0.0445,
      "relation_loss": 0.0045229494106024505,
      "selected_pairs": 27227.3,
      "state_query_norm": 43.77484283447266,
      "state_relation_loss": 0.00027609207900241016,
      "step": 10520,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0045229494106024505,
      "weighted_relation_loss": 0.0045229494106024505
    },
    {
      "action_query_norm": 42.71171226501465,
      "action_relation_loss": 0.004170241486281157,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.850327408793264,
      "flow_matching_loss": 0.04092327244579792,
      "grad_norm": 0.14965088665485382,
      "learning_rate": 7.721345955125147e-05,
      "loss": 0.0454,
      "relation_loss": 0.0044422589242458345,
      "selected_pairs": 27309.7,
      "state_query_norm": 43.7708553314209,
      "state_relation_loss": 0.0002720174306887202,
      "step": 10530,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0044422589242458345,
      "weighted_relation_loss": 0.0044422589242458345
    },
    {
      "action_query_norm": 42.71886215209961,
      "action_relation_loss": 0.004237601184286177,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.859681945743686,
      "flow_matching_loss": 0.0386045578867197,
      "grad_norm": 0.09903240948915482,
      "learning_rate": 7.71672059135231e-05,
      "loss": 0.0431,
      "relation_loss": 0.004532198561355472,
      "selected_pairs": 26618.8,
      "state_query_norm": 43.763519287109375,
      "state_relation_loss": 0.000294597442552913,
      "step": 10540,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.004532198561355472,
      "weighted_relation_loss": 0.004532198561355472
    },
    {
      "action_query_norm": 42.71483459472656,
      "action_relation_loss": 0.004252003226429224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.869036482694106,
      "flow_matching_loss": 0.04303879551589489,
      "grad_norm": 0.10589133203029633,
      "learning_rate": 7.712091926507787e-05,
      "loss": 0.0476,
      "relation_loss": 0.004534424096345901,
      "selected_pairs": 27201.6,
      "state_query_norm": 43.76914978027344,
      "state_relation_loss": 0.00028242100670468063,
      "step": 10550,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.004534424096345901,
      "weighted_relation_loss": 0.004534424096345901
    },
    {
      "action_query_norm": 42.71662979125976,
      "action_relation_loss": 0.004044552682898939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.878391019644528,
      "flow_matching_loss": 0.03880609460175037,
      "grad_norm": 0.11569442600011826,
      "learning_rate": 7.70745996621584e-05,
      "loss": 0.0431,
      "relation_loss": 0.0043163958471268415,
      "selected_pairs": 26745.2,
      "state_query_norm": 43.762398910522464,
      "state_relation_loss": 0.0002718431642279029,
      "step": 10560,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0043163958471268415,
      "weighted_relation_loss": 0.0043163958471268415
    },
    {
      "action_query_norm": 42.70237503051758,
      "action_relation_loss": 0.00422780318185687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.887745556594949,
      "flow_matching_loss": 0.040765869244933126,
      "grad_norm": 0.1293005645275116,
      "learning_rate": 7.702824716104735e-05,
      "loss": 0.0452,
      "relation_loss": 0.004482174338772893,
      "selected_pairs": 26614.4,
      "state_query_norm": 43.77906036376953,
      "state_relation_loss": 0.0002543711452744901,
      "step": 10570,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.004482174338772893,
      "weighted_relation_loss": 0.004482174338772893
    },
    {
      "action_query_norm": 42.70243034362793,
      "action_relation_loss": 0.003976036491803825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.897100093545369,
      "flow_matching_loss": 0.042002249509096146,
      "grad_norm": 0.11708196252584457,
      "learning_rate": 7.698186181806743e-05,
      "loss": 0.0462,
      "relation_loss": 0.004223417001776397,
      "selected_pairs": 26298.9,
      "state_query_norm": 43.77141151428223,
      "state_relation_loss": 0.00024738051142776386,
      "step": 10580,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.004223417001776397,
      "weighted_relation_loss": 0.004223417001776397
    },
    {
      "action_query_norm": 42.69042854309082,
      "action_relation_loss": 0.004283651942387223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.90645463049579,
      "flow_matching_loss": 0.04135846346616745,
      "grad_norm": 0.12242907285690308,
      "learning_rate": 7.693544368958116e-05,
      "loss": 0.0459,
      "relation_loss": 0.004525176947936416,
      "selected_pairs": 26553.1,
      "state_query_norm": 43.76890106201172,
      "state_relation_loss": 0.00024152491823770106,
      "step": 10590,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.004525176947936416,
      "weighted_relation_loss": 0.004525176947936416
    },
    {
      "action_query_norm": 42.683020401000974,
      "action_relation_loss": 0.004192743636667729,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.915809167446211,
      "flow_matching_loss": 0.04195283465087414,
      "grad_norm": 0.0949411615729332,
      "learning_rate": 7.688899283199096e-05,
      "loss": 0.0464,
      "relation_loss": 0.004446094855666161,
      "selected_pairs": 26923.4,
      "state_query_norm": 43.76077499389648,
      "state_relation_loss": 0.00025335123063996434,
      "step": 10600,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004446094855666161,
      "weighted_relation_loss": 0.004446094855666161
    },
    {
      "action_query_norm": 42.69474220275879,
      "action_relation_loss": 0.004226749762892723,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.925163704396633,
      "flow_matching_loss": 0.0411827553063631,
      "grad_norm": 0.12707768380641937,
      "learning_rate": 7.684250930173901e-05,
      "loss": 0.0457,
      "relation_loss": 0.00449319239705801,
      "selected_pairs": 26340.3,
      "state_query_norm": 43.76073455810547,
      "state_relation_loss": 0.00026644266763469203,
      "step": 10610,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.00449319239705801,
      "weighted_relation_loss": 0.00449319239705801
    },
    {
      "action_query_norm": 42.70837326049805,
      "action_relation_loss": 0.0041118782944977285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.934518241347053,
      "flow_matching_loss": 0.04235273562371731,
      "grad_norm": 0.10269718617200851,
      "learning_rate": 7.679599315530718e-05,
      "loss": 0.0467,
      "relation_loss": 0.004391689482145011,
      "selected_pairs": 26771.3,
      "state_query_norm": 43.76316833496094,
      "state_relation_loss": 0.00027981118764728306,
      "step": 10620,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004391689482145011,
      "weighted_relation_loss": 0.004391689482145011
    },
    {
      "action_query_norm": 42.66003646850586,
      "action_relation_loss": 0.004070205101743341,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.943872778297475,
      "flow_matching_loss": 0.04218262620270252,
      "grad_norm": 0.1220046728849411,
      "learning_rate": 7.674944444921695e-05,
      "loss": 0.0465,
      "relation_loss": 0.0043249648530036214,
      "selected_pairs": 26344.7,
      "state_query_norm": 43.75275421142578,
      "state_relation_loss": 0.00025475982401985676,
      "step": 10630,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0043249648530036214,
      "weighted_relation_loss": 0.0043249648530036214
    },
    {
      "action_query_norm": 42.68260307312012,
      "action_relation_loss": 0.0041212357580661775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.953227315247895,
      "flow_matching_loss": 0.03937334381043911,
      "grad_norm": 0.09660127758979797,
      "learning_rate": 7.670286324002944e-05,
      "loss": 0.0437,
      "relation_loss": 0.004373274836689233,
      "selected_pairs": 27124.6,
      "state_query_norm": 43.75393218994141,
      "state_relation_loss": 0.0002520390655263327,
      "step": 10640,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004373274836689233,
      "weighted_relation_loss": 0.004373274836689233
    },
    {
      "action_query_norm": 42.66484184265137,
      "action_relation_loss": 0.004257298982702195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.962581852198316,
      "flow_matching_loss": 0.040782219171524046,
      "grad_norm": 0.13911551237106323,
      "learning_rate": 7.665624958434514e-05,
      "loss": 0.0453,
      "relation_loss": 0.0045071737840771675,
      "selected_pairs": 26850.8,
      "state_query_norm": 43.75753669738769,
      "state_relation_loss": 0.0002498747708159499,
      "step": 10650,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0045071737840771675,
      "weighted_relation_loss": 0.0045071737840771675
    },
    {
      "action_query_norm": 42.67475280761719,
      "action_relation_loss": 0.004266744758933782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.971936389148738,
      "flow_matching_loss": 0.04190982393920421,
      "grad_norm": 0.1309448480606079,
      "learning_rate": 7.66096035388041e-05,
      "loss": 0.0464,
      "relation_loss": 0.004515205137431621,
      "selected_pairs": 27058.9,
      "state_query_norm": 43.757833099365236,
      "state_relation_loss": 0.0002484603566699661,
      "step": 10660,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004515205137431621,
      "weighted_relation_loss": 0.004515205137431621
    },
    {
      "action_query_norm": 42.66763038635254,
      "action_relation_loss": 0.0042975931894034146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.981290926099158,
      "flow_matching_loss": 0.040353655070066455,
      "grad_norm": 0.11679382622241974,
      "learning_rate": 7.656292516008563e-05,
      "loss": 0.0449,
      "relation_loss": 0.004545724717900157,
      "selected_pairs": 25962.6,
      "state_query_norm": 43.75206451416015,
      "state_relation_loss": 0.00024813152122078465,
      "step": 10670,
      "valid_action_samples": 228.3,
      "weighted_auxiliary_loss": 0.004545724717900157,
      "weighted_relation_loss": 0.004545724717900157
    },
    {
      "action_query_norm": 42.680423736572266,
      "action_relation_loss": 0.004010506812483073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 9.99064546304958,
      "flow_matching_loss": 0.04111814796924591,
      "grad_norm": 0.10590868443250656,
      "learning_rate": 7.651621450490837e-05,
      "loss": 0.0454,
      "relation_loss": 0.004256102326326072,
      "selected_pairs": 27103.3,
      "state_query_norm": 43.74783058166504,
      "state_relation_loss": 0.0002455954396282323,
      "step": 10680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.004256102326326072,
      "weighted_relation_loss": 0.004256102326326072
    },
    {
      "action_query_norm": 42.66926918029785,
      "action_relation_loss": 0.0039207364432513716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.0,
      "flow_matching_loss": 0.040429700165987015,
      "grad_norm": 0.22839227318763733,
      "learning_rate": 7.646947163003017e-05,
      "loss": 0.0446,
      "relation_loss": 0.004161565494723618,
      "selected_pairs": 23529.8,
      "state_query_norm": 43.74670906066895,
      "state_relation_loss": 0.0002408290223684162,
      "step": 10690,
      "valid_action_samples": 210.5,
      "weighted_auxiliary_loss": 0.004161565494723618,
      "weighted_relation_loss": 0.004161565494723618
    },
    {
      "action_query_norm": 42.673394775390626,
      "action_relation_loss": 0.0037463878747075797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.00935453695042,
      "flow_matching_loss": 0.03678142558783293,
      "grad_norm": 0.11696640402078629,
      "learning_rate": 7.642269659224803e-05,
      "loss": 0.0408,
      "relation_loss": 0.003999861888587475,
      "selected_pairs": 27135.3,
      "state_query_norm": 43.752266311645506,
      "state_relation_loss": 0.0002534740357077681,
      "step": 10700,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.003999861888587475,
      "weighted_relation_loss": 0.003999861888587475
    },
    {
      "action_query_norm": 42.660757064819336,
      "action_relation_loss": 0.0038866572314873336,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.018709073900842,
      "flow_matching_loss": 0.0387756172567606,
      "grad_norm": 0.11891204118728638,
      "learning_rate": 7.637588944839803e-05,
      "loss": 0.0429,
      "relation_loss": 0.004127214872278273,
      "selected_pairs": 26921.8,
      "state_query_norm": 43.73726615905762,
      "state_relation_loss": 0.00024055765097728,
      "step": 10710,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004127214872278273,
      "weighted_relation_loss": 0.004127214872278273
    },
    {
      "action_query_norm": 42.6497314453125,
      "action_relation_loss": 0.003794079879298806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.028063610851262,
      "flow_matching_loss": 0.03777414746582508,
      "grad_norm": 0.1098790243268013,
      "learning_rate": 7.632905025535528e-05,
      "loss": 0.0418,
      "relation_loss": 0.004043947020545602,
      "selected_pairs": 26046.4,
      "state_query_norm": 43.73707313537598,
      "state_relation_loss": 0.0002498671630746685,
      "step": 10720,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.004043947020545602,
      "weighted_relation_loss": 0.004043947020545602
    },
    {
      "action_query_norm": 42.6587776184082,
      "action_relation_loss": 0.003977627912536264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.037418147801684,
      "flow_matching_loss": 0.03830584585666656,
      "grad_norm": 0.12396496534347534,
      "learning_rate": 7.628217907003378e-05,
      "loss": 0.0425,
      "relation_loss": 0.004238005820661783,
      "selected_pairs": 26847.7,
      "state_query_norm": 43.74082145690918,
      "state_relation_loss": 0.00026037790230475365,
      "step": 10730,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.004238005820661783,
      "weighted_relation_loss": 0.004238005820661783
    },
    {
      "action_query_norm": 42.6450080871582,
      "action_relation_loss": 0.0038903277600184083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.046772684752105,
      "flow_matching_loss": 0.04041842967271805,
      "grad_norm": 0.10513374209403992,
      "learning_rate": 7.623527594938649e-05,
      "loss": 0.0446,
      "relation_loss": 0.004153427411802113,
      "selected_pairs": 26786.3,
      "state_query_norm": 43.74650192260742,
      "state_relation_loss": 0.00026309972017770635,
      "step": 10740,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004153427411802113,
      "weighted_relation_loss": 0.004153427411802113
    },
    {
      "action_query_norm": 42.64039421081543,
      "action_relation_loss": 0.0038729713298380373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.056127221702527,
      "flow_matching_loss": 0.03814585693180561,
      "grad_norm": 0.10992752760648727,
      "learning_rate": 7.618834095040509e-05,
      "loss": 0.0423,
      "relation_loss": 0.00413621524348855,
      "selected_pairs": 26896.0,
      "state_query_norm": 43.74156150817871,
      "state_relation_loss": 0.0002632439369335771,
      "step": 10750,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00413621524348855,
      "weighted_relation_loss": 0.00413621524348855
    },
    {
      "action_query_norm": 42.62705268859863,
      "action_relation_loss": 0.0038443847559392454,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.065481758652947,
      "flow_matching_loss": 0.038378430902957915,
      "grad_norm": 0.1497085690498352,
      "learning_rate": 7.614137413012e-05,
      "loss": 0.0425,
      "relation_loss": 0.004103823332116008,
      "selected_pairs": 27458.3,
      "state_query_norm": 43.7304801940918,
      "state_relation_loss": 0.0002594385819975287,
      "step": 10760,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.004103823332116008,
      "weighted_relation_loss": 0.004103823332116008
    },
    {
      "action_query_norm": 42.66132431030273,
      "action_relation_loss": 0.004011736460961401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.074836295603367,
      "flow_matching_loss": 0.04266808480024338,
      "grad_norm": 0.10288290679454803,
      "learning_rate": 7.609437554560042e-05,
      "loss": 0.0469,
      "relation_loss": 0.0042700625490397215,
      "selected_pairs": 26650.3,
      "state_query_norm": 43.73513298034668,
      "state_relation_loss": 0.00025832615938270465,
      "step": 10770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0042700625490397215,
      "weighted_relation_loss": 0.0042700625490397215
    },
    {
      "action_query_norm": 42.6436580657959,
      "action_relation_loss": 0.0038731839507818223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.084190832553789,
      "flow_matching_loss": 0.04024676457047462,
      "grad_norm": 0.11455005407333374,
      "learning_rate": 7.604734525395398e-05,
      "loss": 0.0444,
      "relation_loss": 0.004109698208048939,
      "selected_pairs": 26591.3,
      "state_query_norm": 43.73085708618164,
      "state_relation_loss": 0.00023651422379771246,
      "step": 10780,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004109698208048939,
      "weighted_relation_loss": 0.004109698208048939
    },
    {
      "action_query_norm": 42.65212821960449,
      "action_relation_loss": 0.003955822624266148,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.09354536950421,
      "flow_matching_loss": 0.03932825550436973,
      "grad_norm": 0.11649639159440994,
      "learning_rate": 7.600028331232699e-05,
      "loss": 0.0435,
      "relation_loss": 0.004210053128190339,
      "selected_pairs": 26884.9,
      "state_query_norm": 43.73414344787598,
      "state_relation_loss": 0.00025423051265534015,
      "step": 10790,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004210053128190339,
      "weighted_relation_loss": 0.004210053128190339
    },
    {
      "action_query_norm": 42.63546867370606,
      "action_relation_loss": 0.004184771445579827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.102899906454631,
      "flow_matching_loss": 0.04183328971266746,
      "grad_norm": 0.11281179636716843,
      "learning_rate": 7.595318977790408e-05,
      "loss": 0.0463,
      "relation_loss": 0.004426701040938496,
      "selected_pairs": 27106.9,
      "state_query_norm": 43.73264656066895,
      "state_relation_loss": 0.00024192963901441543,
      "step": 10800,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.004426701040938496,
      "weighted_relation_loss": 0.004426701040938496
    },
    {
      "action_query_norm": 42.60693244934082,
      "action_relation_loss": 0.004093350027687848,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.112254443405051,
      "flow_matching_loss": 0.0383406724780798,
      "grad_norm": 0.1612478792667389,
      "learning_rate": 7.590606470790836e-05,
      "loss": 0.0427,
      "relation_loss": 0.004341133730486035,
      "selected_pairs": 26969.6,
      "state_query_norm": 43.728563690185545,
      "state_relation_loss": 0.0002477837100741453,
      "step": 10810,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.004341133730486035,
      "weighted_relation_loss": 0.004341133730486035
    },
    {
      "action_query_norm": 42.63794860839844,
      "action_relation_loss": 0.0038950821151956918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.121608980355472,
      "flow_matching_loss": 0.04084248393774033,
      "grad_norm": 0.19911807775497437,
      "learning_rate": 7.585890815960125e-05,
      "loss": 0.045,
      "relation_loss": 0.004145640996284783,
      "selected_pairs": 27032.6,
      "state_query_norm": 43.7252197265625,
      "state_relation_loss": 0.0002505588956410065,
      "step": 10820,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004145640996284783,
      "weighted_relation_loss": 0.004145640996284783
    },
    {
      "action_query_norm": 42.60751876831055,
      "action_relation_loss": 0.003902742499485612,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.130963517305894,
      "flow_matching_loss": 0.03854049816727638,
      "grad_norm": 0.17376025021076202,
      "learning_rate": 7.581172019028238e-05,
      "loss": 0.0427,
      "relation_loss": 0.004145219596102834,
      "selected_pairs": 27123.2,
      "state_query_norm": 43.72225685119629,
      "state_relation_loss": 0.00024247711553471163,
      "step": 10830,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004145219596102834,
      "weighted_relation_loss": 0.004145219596102834
    },
    {
      "action_query_norm": 42.57868194580078,
      "action_relation_loss": 0.00417117727920413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.140318054256314,
      "flow_matching_loss": 0.041412489116191865,
      "grad_norm": 0.13195113837718964,
      "learning_rate": 7.576450085728959e-05,
      "loss": 0.0458,
      "relation_loss": 0.004411392053589225,
      "selected_pairs": 27161.6,
      "state_query_norm": 43.727803802490236,
      "state_relation_loss": 0.00024021476274356247,
      "step": 10840,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.004411392053589225,
      "weighted_relation_loss": 0.004411392053589225
    },
    {
      "action_query_norm": 42.620396041870116,
      "action_relation_loss": 0.003825902845710516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.149672591206736,
      "flow_matching_loss": 0.03971050754189491,
      "grad_norm": 0.1317906230688095,
      "learning_rate": 7.571725021799885e-05,
      "loss": 0.0438,
      "relation_loss": 0.004074063058942557,
      "selected_pairs": 27010.4,
      "state_query_norm": 43.71568145751953,
      "state_relation_loss": 0.0002481602248735726,
      "step": 10850,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004074063058942557,
      "weighted_relation_loss": 0.004074063058942557
    },
    {
      "action_query_norm": 42.63962821960449,
      "action_relation_loss": 0.004042672575451434,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.159027128157156,
      "flow_matching_loss": 0.03955089710652828,
      "grad_norm": 0.14057163894176483,
      "learning_rate": 7.566996832982409e-05,
      "loss": 0.0438,
      "relation_loss": 0.0042820533504709605,
      "selected_pairs": 26663.3,
      "state_query_norm": 43.718742752075194,
      "state_relation_loss": 0.0002393807328189723,
      "step": 10860,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0042820533504709605,
      "weighted_relation_loss": 0.0042820533504709605
    },
    {
      "action_query_norm": 42.617667007446286,
      "action_relation_loss": 0.003943368582986296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.168381665107578,
      "flow_matching_loss": 0.03667128644883633,
      "grad_norm": 0.12333831191062927,
      "learning_rate": 7.56226552502173e-05,
      "loss": 0.0408,
      "relation_loss": 0.004177195369265974,
      "selected_pairs": 26916.8,
      "state_query_norm": 43.719977951049806,
      "state_relation_loss": 0.00023382674116874115,
      "step": 10870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004177195369265974,
      "weighted_relation_loss": 0.004177195369265974
    },
    {
      "action_query_norm": 42.61392250061035,
      "action_relation_loss": 0.0040694852359592915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.177736202057998,
      "flow_matching_loss": 0.03921726904809475,
      "grad_norm": 0.12439458072185516,
      "learning_rate": 7.557531103666833e-05,
      "loss": 0.0435,
      "relation_loss": 0.0043186625000089405,
      "selected_pairs": 26753.4,
      "state_query_norm": 43.71254692077637,
      "state_relation_loss": 0.0002491772567736916,
      "step": 10880,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0043186625000089405,
      "weighted_relation_loss": 0.0043186625000089405
    },
    {
      "action_query_norm": 42.63844528198242,
      "action_relation_loss": 0.0038025174057111145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.187090739008418,
      "flow_matching_loss": 0.03986913152039051,
      "grad_norm": 0.12693193554878235,
      "learning_rate": 7.552793574670485e-05,
      "loss": 0.0439,
      "relation_loss": 0.0040620154235512015,
      "selected_pairs": 27129.9,
      "state_query_norm": 43.72552261352539,
      "state_relation_loss": 0.00025949806004064155,
      "step": 10890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0040620154235512015,
      "weighted_relation_loss": 0.0040620154235512015
    },
    {
      "action_query_norm": 42.64218635559082,
      "action_relation_loss": 0.003871276346035302,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.19644527595884,
      "flow_matching_loss": 0.04173144213855266,
      "grad_norm": 0.132273331284523,
      "learning_rate": 7.54805294378923e-05,
      "loss": 0.0459,
      "relation_loss": 0.004122107545845211,
      "selected_pairs": 27207.3,
      "state_query_norm": 43.70915184020996,
      "state_relation_loss": 0.00025083119689952584,
      "step": 10900,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.004122107545845211,
      "weighted_relation_loss": 0.004122107545845211
    },
    {
      "action_query_norm": 42.60288963317871,
      "action_relation_loss": 0.004026259714737535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.20579981290926,
      "flow_matching_loss": 0.04143388010561466,
      "grad_norm": 0.15402653813362122,
      "learning_rate": 7.543309216783384e-05,
      "loss": 0.0457,
      "relation_loss": 0.004262605123221875,
      "selected_pairs": 27387.4,
      "state_query_norm": 43.7215934753418,
      "state_relation_loss": 0.0002363454390433617,
      "step": 10910,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.004262605123221875,
      "weighted_relation_loss": 0.004262605123221875
    },
    {
      "action_query_norm": 42.629281997680664,
      "action_relation_loss": 0.0037761081708595156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.215154349859683,
      "flow_matching_loss": 0.039451354369521144,
      "grad_norm": 0.14059874415397644,
      "learning_rate": 7.53856239941702e-05,
      "loss": 0.0435,
      "relation_loss": 0.004014265979640186,
      "selected_pairs": 26578.0,
      "state_query_norm": 43.70568008422852,
      "state_relation_loss": 0.00023815780150471255,
      "step": 10920,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004014265979640186,
      "weighted_relation_loss": 0.004014265979640186
    },
    {
      "action_query_norm": 42.62460021972656,
      "action_relation_loss": 0.003982979175634682,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.224508886810103,
      "flow_matching_loss": 0.04004985503852367,
      "grad_norm": 0.12248000502586365,
      "learning_rate": 7.533812497457972e-05,
      "loss": 0.0443,
      "relation_loss": 0.004238756443373859,
      "selected_pairs": 26759.7,
      "state_query_norm": 43.706179809570315,
      "state_relation_loss": 0.000255777218262665,
      "step": 10930,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004238756443373859,
      "weighted_relation_loss": 0.004238756443373859
    },
    {
      "action_query_norm": 42.58618812561035,
      "action_relation_loss": 0.0038787663681432603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.233863423760523,
      "flow_matching_loss": 0.03962991535663605,
      "grad_norm": 0.14201046526432037,
      "learning_rate": 7.529059516677814e-05,
      "loss": 0.0438,
      "relation_loss": 0.004120836965739727,
      "selected_pairs": 27034.5,
      "state_query_norm": 43.70640983581543,
      "state_relation_loss": 0.000242070596141275,
      "step": 10940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004120836965739727,
      "weighted_relation_loss": 0.004120836965739727
    },
    {
      "action_query_norm": 42.599521255493165,
      "action_relation_loss": 0.003983854781836271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.243217960710945,
      "flow_matching_loss": 0.04074309021234512,
      "grad_norm": 0.13809539377689362,
      "learning_rate": 7.524303462851872e-05,
      "loss": 0.045,
      "relation_loss": 0.004228973714634776,
      "selected_pairs": 27127.5,
      "state_query_norm": 43.699554443359375,
      "state_relation_loss": 0.0002451189197017811,
      "step": 10950,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.004228973714634776,
      "weighted_relation_loss": 0.004228973714634776
    },
    {
      "action_query_norm": 42.62411003112793,
      "action_relation_loss": 0.00399454515427351,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.252572497661365,
      "flow_matching_loss": 0.04050636999309063,
      "grad_norm": 0.13255397975444794,
      "learning_rate": 7.519544341759192e-05,
      "loss": 0.0447,
      "relation_loss": 0.004241223866119981,
      "selected_pairs": 27010.5,
      "state_query_norm": 43.70036659240723,
      "state_relation_loss": 0.00024667871912242845,
      "step": 10960,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004241223866119981,
      "weighted_relation_loss": 0.004241223866119981
    },
    {
      "action_query_norm": 42.58299179077149,
      "action_relation_loss": 0.0039333835244178775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.261927034611787,
      "flow_matching_loss": 0.040290015190839766,
      "grad_norm": 0.14270949363708496,
      "learning_rate": 7.514782159182562e-05,
      "loss": 0.0445,
      "relation_loss": 0.0041836870834231375,
      "selected_pairs": 26504.1,
      "state_query_norm": 43.6956901550293,
      "state_relation_loss": 0.0002503035590052605,
      "step": 10970,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0041836870834231375,
      "weighted_relation_loss": 0.0041836870834231375
    },
    {
      "action_query_norm": 42.59684791564941,
      "action_relation_loss": 0.00393531140871346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.271281571562207,
      "flow_matching_loss": 0.041044576838612554,
      "grad_norm": 0.15059764683246613,
      "learning_rate": 7.510016920908481e-05,
      "loss": 0.0452,
      "relation_loss": 0.004167464561760426,
      "selected_pairs": 27034.4,
      "state_query_norm": 43.69803237915039,
      "state_relation_loss": 0.00023215320106828585,
      "step": 10980,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004167464561760426,
      "weighted_relation_loss": 0.004167464561760426
    },
    {
      "action_query_norm": 42.5934139251709,
      "action_relation_loss": 0.003894702484831214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.28063610851263,
      "flow_matching_loss": 0.03963598236441612,
      "grad_norm": 0.13394778966903687,
      "learning_rate": 7.505248632727158e-05,
      "loss": 0.0438,
      "relation_loss": 0.004126827046275139,
      "selected_pairs": 26257.4,
      "state_query_norm": 43.69881935119629,
      "state_relation_loss": 0.000232124503236264,
      "step": 10990,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.004126827046275139,
      "weighted_relation_loss": 0.004126827046275139
    },
    {
      "action_query_norm": 42.570002746582034,
      "action_relation_loss": 0.003926285658963025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.28999064546305,
      "flow_matching_loss": 0.0420623816549778,
      "grad_norm": 0.16183732450008392,
      "learning_rate": 7.50047730043252e-05,
      "loss": 0.0462,
      "relation_loss": 0.00417479791212827,
      "selected_pairs": 26094.0,
      "state_query_norm": 43.69133682250977,
      "state_relation_loss": 0.00024851224152371286,
      "step": 11000,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.00417479791212827,
      "weighted_relation_loss": 0.00417479791212827
    },
    {
      "action_query_norm": 42.57895355224609,
      "action_relation_loss": 0.0038481988944113255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.29934518241347,
      "flow_matching_loss": 0.03883303999900818,
      "grad_norm": 0.1304682195186615,
      "learning_rate": 7.495702929822183e-05,
      "loss": 0.0429,
      "relation_loss": 0.004087693290784955,
      "selected_pairs": 26578.5,
      "state_query_norm": 43.69822731018066,
      "state_relation_loss": 0.00023949438327690585,
      "step": 11010,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004087693290784955,
      "weighted_relation_loss": 0.004087693290784955
    },
    {
      "action_query_norm": 42.58634948730469,
      "action_relation_loss": 0.004253161395899952,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.308699719363892,
      "flow_matching_loss": 0.03828379474580288,
      "grad_norm": 0.12221948802471161,
      "learning_rate": 7.490925526697455e-05,
      "loss": 0.0428,
      "relation_loss": 0.004509346280246973,
      "selected_pairs": 27011.2,
      "state_query_norm": 43.68860626220703,
      "state_relation_loss": 0.00025618485087761653,
      "step": 11020,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.004509346280246973,
      "weighted_relation_loss": 0.004509346280246973
    },
    {
      "action_query_norm": 42.549167251586915,
      "action_relation_loss": 0.0038879490457475185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.318054256314312,
      "flow_matching_loss": 0.041597338393330574,
      "grad_norm": 0.12853145599365234,
      "learning_rate": 7.486145096863334e-05,
      "loss": 0.0458,
      "relation_loss": 0.004155722470022738,
      "selected_pairs": 27032.0,
      "state_query_norm": 43.684588241577146,
      "state_relation_loss": 0.00026777345628943295,
      "step": 11030,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.004155722470022738,
      "weighted_relation_loss": 0.004155722470022738
    },
    {
      "action_query_norm": 42.532950973510744,
      "action_relation_loss": 0.004047180619090795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.327408793264734,
      "flow_matching_loss": 0.041412287205457685,
      "grad_norm": 0.1209360733628273,
      "learning_rate": 7.481361646128491e-05,
      "loss": 0.0457,
      "relation_loss": 0.004321592207998037,
      "selected_pairs": 26665.8,
      "state_query_norm": 43.676554107666014,
      "state_relation_loss": 0.0002744116573012434,
      "step": 11040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004321592207998037,
      "weighted_relation_loss": 0.004321592207998037
    },
    {
      "action_query_norm": 42.56552925109863,
      "action_relation_loss": 0.0038962622871622442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.336763330215154,
      "flow_matching_loss": 0.042125994339585304,
      "grad_norm": 0.08398967236280441,
      "learning_rate": 7.476575180305271e-05,
      "loss": 0.0463,
      "relation_loss": 0.0041467079659923915,
      "selected_pairs": 27012.6,
      "state_query_norm": 43.6801399230957,
      "state_relation_loss": 0.00025044570356840267,
      "step": 11050,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0041467079659923915,
      "weighted_relation_loss": 0.0041467079659923915
    },
    {
      "action_query_norm": 42.57226028442383,
      "action_relation_loss": 0.004024270898662507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.346117867165574,
      "flow_matching_loss": 0.03872951306402683,
      "grad_norm": 0.13818098604679108,
      "learning_rate": 7.471785705209682e-05,
      "loss": 0.043,
      "relation_loss": 0.004275072179734707,
      "selected_pairs": 26667.4,
      "state_query_norm": 43.685231399536136,
      "state_relation_loss": 0.0002508013538317755,
      "step": 11060,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004275072179734707,
      "weighted_relation_loss": 0.004275072179734707
    },
    {
      "action_query_norm": 42.54271965026855,
      "action_relation_loss": 0.003887351485900581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.355472404115996,
      "flow_matching_loss": 0.04150015339255333,
      "grad_norm": 0.15757544338703156,
      "learning_rate": 7.466993226661387e-05,
      "loss": 0.0456,
      "relation_loss": 0.0041450494667515155,
      "selected_pairs": 26899.9,
      "state_query_norm": 43.686234664916995,
      "state_relation_loss": 0.0002576979895820841,
      "step": 11070,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0041450494667515155,
      "weighted_relation_loss": 0.0041450494667515155
    },
    {
      "action_query_norm": 42.586029052734375,
      "action_relation_loss": 0.0040568435098975895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.364826941066417,
      "flow_matching_loss": 0.04167223758995533,
      "grad_norm": 0.13074401021003723,
      "learning_rate": 7.462197750483699e-05,
      "loss": 0.046,
      "relation_loss": 0.004308830806985497,
      "selected_pairs": 27222.4,
      "state_query_norm": 43.67835998535156,
      "state_relation_loss": 0.00025198737275786696,
      "step": 11080,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.004308830806985497,
      "weighted_relation_loss": 0.004308830806985497
    },
    {
      "action_query_norm": 42.577070999145505,
      "action_relation_loss": 0.0038117868592962624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.374181478016839,
      "flow_matching_loss": 0.03796698357909918,
      "grad_norm": 0.13533152639865875,
      "learning_rate": 7.457399282503574e-05,
      "loss": 0.042,
      "relation_loss": 0.004071207088418305,
      "selected_pairs": 27004.6,
      "state_query_norm": 43.67687454223633,
      "state_relation_loss": 0.00025942025386029857,
      "step": 11090,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.004071207088418305,
      "weighted_relation_loss": 0.004071207088418305
    },
    {
      "action_query_norm": 42.59277229309082,
      "action_relation_loss": 0.003800738020800054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.383536014967259,
      "flow_matching_loss": 0.040378966554999354,
      "grad_norm": 0.17553609609603882,
      "learning_rate": 7.452597828551604e-05,
      "loss": 0.0444,
      "relation_loss": 0.004056604159995913,
      "selected_pairs": 27295.1,
      "state_query_norm": 43.68165245056152,
      "state_relation_loss": 0.0002558661479270086,
      "step": 11100,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.004056604159995913,
      "weighted_relation_loss": 0.004056604159995913
    },
    {
      "action_query_norm": 42.58711318969726,
      "action_relation_loss": 0.0037359467707574366,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.39289055191768,
      "flow_matching_loss": 0.03717151544988155,
      "grad_norm": 0.14053159952163696,
      "learning_rate": 7.447793394462006e-05,
      "loss": 0.0411,
      "relation_loss": 0.003972625406458974,
      "selected_pairs": 26667.7,
      "state_query_norm": 43.68346824645996,
      "state_relation_loss": 0.000236678562941961,
      "step": 11110,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003972625406458974,
      "weighted_relation_loss": 0.003972625406458974
    },
    {
      "action_query_norm": 42.55723114013672,
      "action_relation_loss": 0.004141738568432629,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.402245088868101,
      "flow_matching_loss": 0.03992633186280727,
      "grad_norm": 0.09892483055591583,
      "learning_rate": 7.442985986072624e-05,
      "loss": 0.0443,
      "relation_loss": 0.004387310752645135,
      "selected_pairs": 27247.6,
      "state_query_norm": 43.67517776489258,
      "state_relation_loss": 0.0002455721391015686,
      "step": 11120,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.004387310752645135,
      "weighted_relation_loss": 0.004387310752645135
    },
    {
      "action_query_norm": 42.54767761230469,
      "action_relation_loss": 0.003919997485354543,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.411599625818521,
      "flow_matching_loss": 0.037986254692077635,
      "grad_norm": 0.13509836792945862,
      "learning_rate": 7.438175609224908e-05,
      "loss": 0.0421,
      "relation_loss": 0.004152872948907316,
      "selected_pairs": 26073.2,
      "state_query_norm": 43.68203773498535,
      "state_relation_loss": 0.00023287548392545432,
      "step": 11130,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.004152872948907316,
      "weighted_relation_loss": 0.004152872948907316
    },
    {
      "action_query_norm": 42.54707374572754,
      "action_relation_loss": 0.0038994948845356703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.420954162768943,
      "flow_matching_loss": 0.041212021932005884,
      "grad_norm": 0.12493203580379486,
      "learning_rate": 7.433362269763924e-05,
      "loss": 0.0454,
      "relation_loss": 0.004147809674032032,
      "selected_pairs": 26759.0,
      "state_query_norm": 43.6765495300293,
      "state_relation_loss": 0.00024831479095155373,
      "step": 11140,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.004147809674032032,
      "weighted_relation_loss": 0.004147809674032032
    },
    {
      "action_query_norm": 42.5290584564209,
      "action_relation_loss": 0.003988396469503641,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.430308699719363,
      "flow_matching_loss": 0.042917919158935544,
      "grad_norm": 0.13837026059627533,
      "learning_rate": 7.428545973538329e-05,
      "loss": 0.0471,
      "relation_loss": 0.004230892495252192,
      "selected_pairs": 26951.3,
      "state_query_norm": 43.665902328491214,
      "state_relation_loss": 0.0002424960141070187,
      "step": 11150,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004230892495252192,
      "weighted_relation_loss": 0.004230892495252192
    },
    {
      "action_query_norm": 42.537746047973634,
      "action_relation_loss": 0.003929149499163032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.439663236669785,
      "flow_matching_loss": 0.03869785889983177,
      "grad_norm": 0.08087056130170822,
      "learning_rate": 7.42372672640038e-05,
      "loss": 0.0429,
      "relation_loss": 0.004165740823373199,
      "selected_pairs": 26686.2,
      "state_query_norm": 43.67084732055664,
      "state_relation_loss": 0.0002365913533139974,
      "step": 11160,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004165740823373199,
      "weighted_relation_loss": 0.004165740823373199
    },
    {
      "action_query_norm": 42.54490013122559,
      "action_relation_loss": 0.0036608227295801044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.449017773620206,
      "flow_matching_loss": 0.037953941337764266,
      "grad_norm": 0.11889032274484634,
      "learning_rate": 7.418904534205917e-05,
      "loss": 0.0418,
      "relation_loss": 0.00389165049418807,
      "selected_pairs": 26969.3,
      "state_query_norm": 43.65970153808594,
      "state_relation_loss": 0.0002308277616975829,
      "step": 11170,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.00389165049418807,
      "weighted_relation_loss": 0.00389165049418807
    },
    {
      "action_query_norm": 42.49825553894043,
      "action_relation_loss": 0.003915617289021611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.458372310570628,
      "flow_matching_loss": 0.04098211526870728,
      "grad_norm": 0.12902480363845825,
      "learning_rate": 7.414079402814356e-05,
      "loss": 0.0451,
      "relation_loss": 0.004150992841459811,
      "selected_pairs": 26328.1,
      "state_query_norm": 43.67050704956055,
      "state_relation_loss": 0.00023537555098300799,
      "step": 11180,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.004150992841459811,
      "weighted_relation_loss": 0.004150992841459811
    },
    {
      "action_query_norm": 42.5231746673584,
      "action_relation_loss": 0.003972138604149222,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.467726847521048,
      "flow_matching_loss": 0.042163175716996196,
      "grad_norm": 0.10194742679595947,
      "learning_rate": 7.409251338088687e-05,
      "loss": 0.0464,
      "relation_loss": 0.004210592084564269,
      "selected_pairs": 26691.4,
      "state_query_norm": 43.669833755493165,
      "state_relation_loss": 0.00023845352116040884,
      "step": 11190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.004210592084564269,
      "weighted_relation_loss": 0.004210592084564269
    },
    {
      "action_query_norm": 42.501482009887695,
      "action_relation_loss": 0.0037943133385851978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.477081384471468,
      "flow_matching_loss": 0.038275615125894544,
      "grad_norm": 0.11818605661392212,
      "learning_rate": 7.404420345895467e-05,
      "loss": 0.0423,
      "relation_loss": 0.004043357283808291,
      "selected_pairs": 26949.8,
      "state_query_norm": 43.671560287475586,
      "state_relation_loss": 0.00024904398596845565,
      "step": 11200,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.004043357283808291,
      "weighted_relation_loss": 0.004043357283808291
    },
    {
      "action_query_norm": 42.51612510681152,
      "action_relation_loss": 0.0039986535673961045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.48643592142189,
      "flow_matching_loss": 0.040045468509197234,
      "grad_norm": 0.11597684770822525,
      "learning_rate": 7.399586432104804e-05,
      "loss": 0.0443,
      "relation_loss": 0.004254984715953469,
      "selected_pairs": 26594.2,
      "state_query_norm": 43.665856170654294,
      "state_relation_loss": 0.00025633122422732415,
      "step": 11210,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.004254984715953469,
      "weighted_relation_loss": 0.004254984715953469
    },
    {
      "action_query_norm": 42.544498443603516,
      "action_relation_loss": 0.003698730980977416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.49579045837231,
      "flow_matching_loss": 0.039299705252051356,
      "grad_norm": 0.12825846672058105,
      "learning_rate": 7.394749602590358e-05,
      "loss": 0.0432,
      "relation_loss": 0.003944159601815045,
      "selected_pairs": 27272.6,
      "state_query_norm": 43.66533317565918,
      "state_relation_loss": 0.0002454285815474577,
      "step": 11220,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003944159601815045,
      "weighted_relation_loss": 0.003944159601815045
    },
    {
      "action_query_norm": 42.5229907989502,
      "action_relation_loss": 0.003973377193324268,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.505144995322732,
      "flow_matching_loss": 0.038304343819618225,
      "grad_norm": 0.1238190159201622,
      "learning_rate": 7.389909863229336e-05,
      "loss": 0.0425,
      "relation_loss": 0.004226588574238121,
      "selected_pairs": 27263.8,
      "state_query_norm": 43.656486129760744,
      "state_relation_loss": 0.0002532113969209604,
      "step": 11230,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004226588574238121,
      "weighted_relation_loss": 0.004226588574238121
    },
    {
      "action_query_norm": 42.510806274414065,
      "action_relation_loss": 0.0038533891318365933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.514499532273152,
      "flow_matching_loss": 0.03970620296895504,
      "grad_norm": 0.11103831231594086,
      "learning_rate": 7.385067219902477e-05,
      "loss": 0.0438,
      "relation_loss": 0.004109020135365427,
      "selected_pairs": 27270.3,
      "state_query_norm": 43.64999656677246,
      "state_relation_loss": 0.0002556310122599825,
      "step": 11240,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004109020135365427,
      "weighted_relation_loss": 0.004109020135365427
    },
    {
      "action_query_norm": 42.53361167907715,
      "action_relation_loss": 0.003904156479984522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.523854069223573,
      "flow_matching_loss": 0.038330187648534776,
      "grad_norm": 0.1579604595899582,
      "learning_rate": 7.380221678494049e-05,
      "loss": 0.0425,
      "relation_loss": 0.004171757074072957,
      "selected_pairs": 27061.9,
      "state_query_norm": 43.655304718017575,
      "state_relation_loss": 0.0002676005387911573,
      "step": 11250,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.004171757074072957,
      "weighted_relation_loss": 0.004171757074072957
    },
    {
      "action_query_norm": 42.4927474975586,
      "action_relation_loss": 0.003801622963510454,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.533208606173995,
      "flow_matching_loss": 0.03844508975744247,
      "grad_norm": 0.10808455944061279,
      "learning_rate": 7.37537324489184e-05,
      "loss": 0.0425,
      "relation_loss": 0.004050125856883824,
      "selected_pairs": 26575.1,
      "state_query_norm": 43.65817413330078,
      "state_relation_loss": 0.0002485028773662634,
      "step": 11260,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.004050125856883824,
      "weighted_relation_loss": 0.004050125856883824
    },
    {
      "action_query_norm": 42.49844131469727,
      "action_relation_loss": 0.003884954168461263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.542563143124415,
      "flow_matching_loss": 0.039468009024858475,
      "grad_norm": 0.11291804909706116,
      "learning_rate": 7.370521924987155e-05,
      "loss": 0.0436,
      "relation_loss": 0.004142424440942705,
      "selected_pairs": 26226.8,
      "state_query_norm": 43.65622482299805,
      "state_relation_loss": 0.00025747026083990934,
      "step": 11270,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.004142424440942705,
      "weighted_relation_loss": 0.004142424440942705
    },
    {
      "action_query_norm": 42.542389678955075,
      "action_relation_loss": 0.0039891953812912105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.551917680074837,
      "flow_matching_loss": 0.04105084352195263,
      "grad_norm": 0.12619861960411072,
      "learning_rate": 7.365667724674806e-05,
      "loss": 0.0453,
      "relation_loss": 0.004248790629208088,
      "selected_pairs": 26668.4,
      "state_query_norm": 43.65003509521485,
      "state_relation_loss": 0.0002595952042611316,
      "step": 11280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004248790629208088,
      "weighted_relation_loss": 0.004248790629208088
    },
    {
      "action_query_norm": 42.51394081115723,
      "action_relation_loss": 0.003908019093796611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.561272217025257,
      "flow_matching_loss": 0.038909167796373365,
      "grad_norm": 0.14675575494766235,
      "learning_rate": 7.360810649853105e-05,
      "loss": 0.0431,
      "relation_loss": 0.004167931084521115,
      "selected_pairs": 26665.5,
      "state_query_norm": 43.64744911193848,
      "state_relation_loss": 0.00025991200818680226,
      "step": 11290,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004167931084521115,
      "weighted_relation_loss": 0.004167931084521115
    },
    {
      "action_query_norm": 42.48341102600098,
      "action_relation_loss": 0.0038052772404626014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.570626753975677,
      "flow_matching_loss": 0.03783904481679201,
      "grad_norm": 0.12933826446533203,
      "learning_rate": 7.355950706423853e-05,
      "loss": 0.0419,
      "relation_loss": 0.004087700857780874,
      "selected_pairs": 26735.5,
      "state_query_norm": 43.63308219909668,
      "state_relation_loss": 0.00028242359112482516,
      "step": 11300,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004087700857780874,
      "weighted_relation_loss": 0.004087700857780874
    },
    {
      "action_query_norm": 42.478716659545896,
      "action_relation_loss": 0.003956886846572161,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.5799812909261,
      "flow_matching_loss": 0.038245727121829984,
      "grad_norm": 0.11112893372774124,
      "learning_rate": 7.351087900292342e-05,
      "loss": 0.0424,
      "relation_loss": 0.0041994888102635745,
      "selected_pairs": 26895.8,
      "state_query_norm": 43.649071884155276,
      "state_relation_loss": 0.00024260199425043538,
      "step": 11310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0041994888102635745,
      "weighted_relation_loss": 0.0041994888102635745
    },
    {
      "action_query_norm": 42.46148338317871,
      "action_relation_loss": 0.0036986125633120537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.58933582787652,
      "flow_matching_loss": 0.037250956520438194,
      "grad_norm": 0.13643145561218262,
      "learning_rate": 7.346222237367339e-05,
      "loss": 0.0412,
      "relation_loss": 0.003955610608682036,
      "selected_pairs": 26762.2,
      "state_query_norm": 43.65101165771485,
      "state_relation_loss": 0.00025699809193611143,
      "step": 11320,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.003955610608682036,
      "weighted_relation_loss": 0.003955610608682036
    },
    {
      "action_query_norm": 42.4746753692627,
      "action_relation_loss": 0.003776721633039415,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.598690364826941,
      "flow_matching_loss": 0.03948550187051296,
      "grad_norm": 0.12979422509670258,
      "learning_rate": 7.341353723561081e-05,
      "loss": 0.0435,
      "relation_loss": 0.004015261260792613,
      "selected_pairs": 27264.6,
      "state_query_norm": 43.651815032958986,
      "state_relation_loss": 0.00023853956226957962,
      "step": 11330,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.004015261260792613,
      "weighted_relation_loss": 0.004015261260792613
    },
    {
      "action_query_norm": 42.4785717010498,
      "action_relation_loss": 0.00376555344555527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.608044901777362,
      "flow_matching_loss": 0.039218299090862274,
      "grad_norm": 0.11788928508758545,
      "learning_rate": 7.336482364789277e-05,
      "loss": 0.0432,
      "relation_loss": 0.003998031746596098,
      "selected_pairs": 26961.9,
      "state_query_norm": 43.65436935424805,
      "state_relation_loss": 0.00023247830395121128,
      "step": 11340,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003998031746596098,
      "weighted_relation_loss": 0.003998031746596098
    },
    {
      "action_query_norm": 42.48552055358887,
      "action_relation_loss": 0.003780337795615196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.617399438727784,
      "flow_matching_loss": 0.039116973988711834,
      "grad_norm": 0.1184115782380104,
      "learning_rate": 7.331608166971082e-05,
      "loss": 0.0431,
      "relation_loss": 0.004019887535832822,
      "selected_pairs": 26433.9,
      "state_query_norm": 43.64521255493164,
      "state_relation_loss": 0.0002395496834651567,
      "step": 11350,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.004019887535832822,
      "weighted_relation_loss": 0.004019887535832822
    },
    {
      "action_query_norm": 42.4797176361084,
      "action_relation_loss": 0.0037265227176249026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.626753975678204,
      "flow_matching_loss": 0.03867080956697464,
      "grad_norm": 0.12608341872692108,
      "learning_rate": 7.326731136029108e-05,
      "loss": 0.0426,
      "relation_loss": 0.003956550499424338,
      "selected_pairs": 26737.2,
      "state_query_norm": 43.645348358154294,
      "state_relation_loss": 0.0002300277876202017,
      "step": 11360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003956550499424338,
      "weighted_relation_loss": 0.003956550499424338
    },
    {
      "action_query_norm": 42.46860809326172,
      "action_relation_loss": 0.003916680975817144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.636108512628624,
      "flow_matching_loss": 0.03998575024306774,
      "grad_norm": 0.14907924830913544,
      "learning_rate": 7.321851277889408e-05,
      "loss": 0.0441,
      "relation_loss": 0.004152754019014537,
      "selected_pairs": 26780.4,
      "state_query_norm": 43.63792953491211,
      "state_relation_loss": 0.0002360730417422019,
      "step": 11370,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.004152754019014537,
      "weighted_relation_loss": 0.004152754019014537
    },
    {
      "action_query_norm": 42.4685775756836,
      "action_relation_loss": 0.003884552512317896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.645463049579046,
      "flow_matching_loss": 0.03974514715373516,
      "grad_norm": 0.13331425189971924,
      "learning_rate": 7.316968598481469e-05,
      "loss": 0.0439,
      "relation_loss": 0.004134142235852778,
      "selected_pairs": 26923.5,
      "state_query_norm": 43.628187561035155,
      "state_relation_loss": 0.0002495897701010108,
      "step": 11380,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.004134142235852778,
      "weighted_relation_loss": 0.004134142235852778
    },
    {
      "action_query_norm": 42.4740406036377,
      "action_relation_loss": 0.0037504769396036865,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.654817586529466,
      "flow_matching_loss": 0.039716145396232604,
      "grad_norm": 0.11129608750343323,
      "learning_rate": 7.312083103738206e-05,
      "loss": 0.0437,
      "relation_loss": 0.004012429900467396,
      "selected_pairs": 26967.1,
      "state_query_norm": 43.6265453338623,
      "state_relation_loss": 0.0002619530030642636,
      "step": 11390,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.004012429900467396,
      "weighted_relation_loss": 0.004012429900467396
    },
    {
      "action_query_norm": 42.453539276123045,
      "action_relation_loss": 0.003939251997508109,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.664172123479888,
      "flow_matching_loss": 0.040512082353234294,
      "grad_norm": 0.12398877739906311,
      "learning_rate": 7.307194799595958e-05,
      "loss": 0.0447,
      "relation_loss": 0.004205524129793048,
      "selected_pairs": 26484.3,
      "state_query_norm": 43.62082099914551,
      "state_relation_loss": 0.0002662721206434071,
      "step": 11400,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.004205524129793048,
      "weighted_relation_loss": 0.004205524129793048
    },
    {
      "action_query_norm": 42.48221015930176,
      "action_relation_loss": 0.0036574746714904903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.673526660430309,
      "flow_matching_loss": 0.036918189749121665,
      "grad_norm": 0.13050469756126404,
      "learning_rate": 7.302303691994473e-05,
      "loss": 0.0408,
      "relation_loss": 0.003910388983786106,
      "selected_pairs": 26506.0,
      "state_query_norm": 43.63505172729492,
      "state_relation_loss": 0.0002529143341234885,
      "step": 11410,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.003910388983786106,
      "weighted_relation_loss": 0.003910388983786106
    },
    {
      "action_query_norm": 42.45124053955078,
      "action_relation_loss": 0.003552659787237644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.68288119738073,
      "flow_matching_loss": 0.03865865282714367,
      "grad_norm": 0.1342371255159378,
      "learning_rate": 7.29740978687691e-05,
      "loss": 0.0425,
      "relation_loss": 0.0038184088189154863,
      "selected_pairs": 26430.6,
      "state_query_norm": 43.62551994323731,
      "state_relation_loss": 0.00026574910443741826,
      "step": 11420,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0038184088189154863,
      "weighted_relation_loss": 0.0038184088189154863
    },
    {
      "action_query_norm": 42.46763687133789,
      "action_relation_loss": 0.0038007811177521943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.69223573433115,
      "flow_matching_loss": 0.03682890702039003,
      "grad_norm": 0.1277501881122589,
      "learning_rate": 7.292513090189825e-05,
      "loss": 0.0409,
      "relation_loss": 0.004055665736086666,
      "selected_pairs": 26664.2,
      "state_query_norm": 43.6168399810791,
      "state_relation_loss": 0.00025488453975412994,
      "step": 11430,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.004055665736086666,
      "weighted_relation_loss": 0.004055665736086666
    },
    {
      "action_query_norm": 42.448143005371094,
      "action_relation_loss": 0.00393029572442174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.701590271281571,
      "flow_matching_loss": 0.03842181414365768,
      "grad_norm": 0.10813024640083313,
      "learning_rate": 7.287613607883163e-05,
      "loss": 0.0426,
      "relation_loss": 0.0041753211291506885,
      "selected_pairs": 27060.1,
      "state_query_norm": 43.62413101196289,
      "state_relation_loss": 0.00024502541054971516,
      "step": 11440,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0041753211291506885,
      "weighted_relation_loss": 0.0041753211291506885
    },
    {
      "action_query_norm": 42.46218414306641,
      "action_relation_loss": 0.0037626934703439475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.710944808231993,
      "flow_matching_loss": 0.03653469141572714,
      "grad_norm": 0.1013704240322113,
      "learning_rate": 7.282711345910263e-05,
      "loss": 0.0406,
      "relation_loss": 0.004018176998943091,
      "selected_pairs": 27407.9,
      "state_query_norm": 43.62498931884765,
      "state_relation_loss": 0.0002554835213231854,
      "step": 11450,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.004018176998943091,
      "weighted_relation_loss": 0.004018176998943091
    },
    {
      "action_query_norm": 42.45627098083496,
      "action_relation_loss": 0.0037105883937329056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.720299345182413,
      "flow_matching_loss": 0.03732910193502903,
      "grad_norm": 0.13422533869743347,
      "learning_rate": 7.27780631022783e-05,
      "loss": 0.0413,
      "relation_loss": 0.00395360121037811,
      "selected_pairs": 27029.1,
      "state_query_norm": 43.624030685424806,
      "state_relation_loss": 0.00024301280500367285,
      "step": 11460,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.00395360121037811,
      "weighted_relation_loss": 0.00395360121037811
    },
    {
      "action_query_norm": 42.435899353027345,
      "action_relation_loss": 0.003734982805326581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.729653882132835,
      "flow_matching_loss": 0.03562651704996824,
      "grad_norm": 0.11972232908010483,
      "learning_rate": 7.272898506795948e-05,
      "loss": 0.0396,
      "relation_loss": 0.003975698770955205,
      "selected_pairs": 26784.8,
      "state_query_norm": 43.62402763366699,
      "state_relation_loss": 0.00024071599618764595,
      "step": 11470,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003975698770955205,
      "weighted_relation_loss": 0.003975698770955205
    },
    {
      "action_query_norm": 42.45033645629883,
      "action_relation_loss": 0.003587994142435491,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.739008419083255,
      "flow_matching_loss": 0.03708444088697434,
      "grad_norm": 0.08588443696498871,
      "learning_rate": 7.267987941578058e-05,
      "loss": 0.0409,
      "relation_loss": 0.003833629353903234,
      "selected_pairs": 26668.7,
      "state_query_norm": 43.61119270324707,
      "state_relation_loss": 0.00024563516926718877,
      "step": 11480,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003833629353903234,
      "weighted_relation_loss": 0.003833629353903234
    },
    {
      "action_query_norm": 42.4368408203125,
      "action_relation_loss": 0.0037312254775315524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.748362956033676,
      "flow_matching_loss": 0.03892152979969978,
      "grad_norm": 0.12516972422599792,
      "learning_rate": 7.263074620540962e-05,
      "loss": 0.0429,
      "relation_loss": 0.003987521003000438,
      "selected_pairs": 27226.3,
      "state_query_norm": 43.60736503601074,
      "state_relation_loss": 0.0002562955472967587,
      "step": 11490,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.003987521003000438,
      "weighted_relation_loss": 0.003987521003000438
    },
    {
      "action_query_norm": 42.45409812927246,
      "action_relation_loss": 0.0036492061568424105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.757717492984098,
      "flow_matching_loss": 0.038077342137694356,
      "grad_norm": 0.1559092402458191,
      "learning_rate": 7.25815854965481e-05,
      "loss": 0.042,
      "relation_loss": 0.0038806664990261195,
      "selected_pairs": 26736.3,
      "state_query_norm": 43.61596870422363,
      "state_relation_loss": 0.00023146030289353803,
      "step": 11500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0038806664990261195,
      "weighted_relation_loss": 0.0038806664990261195
    },
    {
      "action_query_norm": 42.43342323303223,
      "action_relation_loss": 0.0036085039377212525,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.767072029934518,
      "flow_matching_loss": 0.03774402663111687,
      "grad_norm": 0.09225007891654968,
      "learning_rate": 7.253239734893089e-05,
      "loss": 0.0416,
      "relation_loss": 0.0038455025060102344,
      "selected_pairs": 26833.8,
      "state_query_norm": 43.60246315002441,
      "state_relation_loss": 0.00023699855373706668,
      "step": 11510,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0038455025060102344,
      "weighted_relation_loss": 0.0038455025060102344
    },
    {
      "action_query_norm": 42.455580139160155,
      "action_relation_loss": 0.003732660529203713,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.77642656688494,
      "flow_matching_loss": 0.03883260823786259,
      "grad_norm": 0.09858469665050507,
      "learning_rate": 7.248318182232623e-05,
      "loss": 0.0428,
      "relation_loss": 0.0039659417001530525,
      "selected_pairs": 26438.2,
      "state_query_norm": 43.62210807800293,
      "state_relation_loss": 0.0002332811855012551,
      "step": 11520,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0039659417001530525,
      "weighted_relation_loss": 0.0039659417001530525
    },
    {
      "action_query_norm": 42.461940383911134,
      "action_relation_loss": 0.0037737198639661075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.78578110383536,
      "flow_matching_loss": 0.03833786658942699,
      "grad_norm": 0.1044459268450737,
      "learning_rate": 7.243393897653565e-05,
      "loss": 0.0424,
      "relation_loss": 0.004030094435438514,
      "selected_pairs": 26876.3,
      "state_query_norm": 43.60829811096191,
      "state_relation_loss": 0.00025637455692049114,
      "step": 11530,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.004030094435438514,
      "weighted_relation_loss": 0.004030094435438514
    },
    {
      "action_query_norm": 42.423621368408206,
      "action_relation_loss": 0.003728268900886178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.795135640785782,
      "flow_matching_loss": 0.037476501613855365,
      "grad_norm": 0.10854345560073853,
      "learning_rate": 7.23846688713938e-05,
      "loss": 0.0415,
      "relation_loss": 0.003983361390419304,
      "selected_pairs": 27482.0,
      "state_query_norm": 43.606336212158205,
      "state_relation_loss": 0.00025509249971946704,
      "step": 11540,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.003983361390419304,
      "weighted_relation_loss": 0.003983361390419304
    },
    {
      "action_query_norm": 42.427508163452146,
      "action_relation_loss": 0.003774117724969983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.804490177736202,
      "flow_matching_loss": 0.038198532164096834,
      "grad_norm": 0.12635554373264313,
      "learning_rate": 7.233537156676854e-05,
      "loss": 0.0422,
      "relation_loss": 0.0040302003733813764,
      "selected_pairs": 27172.1,
      "state_query_norm": 43.59918327331543,
      "state_relation_loss": 0.00025608264259062706,
      "step": 11550,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0040302003733813764,
      "weighted_relation_loss": 0.0040302003733813764
    },
    {
      "action_query_norm": 42.40202331542969,
      "action_relation_loss": 0.0036914914380759,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.813844714686622,
      "flow_matching_loss": 0.03748804926872253,
      "grad_norm": 0.15182484686374664,
      "learning_rate": 7.228604712256076e-05,
      "loss": 0.0414,
      "relation_loss": 0.003934590169228614,
      "selected_pairs": 26418.3,
      "state_query_norm": 43.60436019897461,
      "state_relation_loss": 0.00024309865548275412,
      "step": 11560,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.003934590169228614,
      "weighted_relation_loss": 0.003934590169228614
    },
    {
      "action_query_norm": 42.425008392333986,
      "action_relation_loss": 0.0037711341166868804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.823199251637044,
      "flow_matching_loss": 0.03739603161811829,
      "grad_norm": 0.1072160080075264,
      "learning_rate": 7.223669559870426e-05,
      "loss": 0.0414,
      "relation_loss": 0.0040149078471586105,
      "selected_pairs": 26900.5,
      "state_query_norm": 43.60955123901367,
      "state_relation_loss": 0.00024377376394113526,
      "step": 11570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0040149078471586105,
      "weighted_relation_loss": 0.0040149078471586105
    },
    {
      "action_query_norm": 42.41862869262695,
      "action_relation_loss": 0.0037719205021858214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.832553788587465,
      "flow_matching_loss": 0.03823546562343836,
      "grad_norm": 0.11370602995157242,
      "learning_rate": 7.218731705516585e-05,
      "loss": 0.0422,
      "relation_loss": 0.004009839799255133,
      "selected_pairs": 26733.7,
      "state_query_norm": 43.601902389526366,
      "state_relation_loss": 0.0002379192810622044,
      "step": 11580,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.004009839799255133,
      "weighted_relation_loss": 0.004009839799255133
    },
    {
      "action_query_norm": 42.428458786010744,
      "action_relation_loss": 0.0036130375694483517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.841908325537887,
      "flow_matching_loss": 0.03827218972146511,
      "grad_norm": 0.12780217826366425,
      "learning_rate": 7.21379115519451e-05,
      "loss": 0.0421,
      "relation_loss": 0.0038397091208025812,
      "selected_pairs": 26987.5,
      "state_query_norm": 43.5946159362793,
      "state_relation_loss": 0.0002266716110170819,
      "step": 11590,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0038397091208025812,
      "weighted_relation_loss": 0.0038397091208025812
    },
    {
      "action_query_norm": 42.41688461303711,
      "action_relation_loss": 0.0038119214354082943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.851262862488307,
      "flow_matching_loss": 0.0387168250977993,
      "grad_norm": 0.13210700452327728,
      "learning_rate": 7.20884791490743e-05,
      "loss": 0.0428,
      "relation_loss": 0.004065980180166661,
      "selected_pairs": 27298.1,
      "state_query_norm": 43.5938117980957,
      "state_relation_loss": 0.0002540587811381556,
      "step": 11600,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.004065980180166661,
      "weighted_relation_loss": 0.004065980180166661
    },
    {
      "action_query_norm": 42.405569839477536,
      "action_relation_loss": 0.0036175458692014216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.860617399438727,
      "flow_matching_loss": 0.03886929638683796,
      "grad_norm": 0.10244089365005493,
      "learning_rate": 7.203901990661857e-05,
      "loss": 0.0427,
      "relation_loss": 0.0038653263822197915,
      "selected_pairs": 26891.0,
      "state_query_norm": 43.58897933959961,
      "state_relation_loss": 0.0002477805232047103,
      "step": 11610,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0038653263822197915,
      "weighted_relation_loss": 0.0038653263822197915
    },
    {
      "action_query_norm": 42.413814544677734,
      "action_relation_loss": 0.003602752136066556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.869971936389149,
      "flow_matching_loss": 0.03777401968836784,
      "grad_norm": 0.1202302947640419,
      "learning_rate": 7.198953388467549e-05,
      "loss": 0.0416,
      "relation_loss": 0.003847617260180414,
      "selected_pairs": 26734.3,
      "state_query_norm": 43.591176223754886,
      "state_relation_loss": 0.0002448651473969221,
      "step": 11620,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.003847617260180414,
      "weighted_relation_loss": 0.003847617260180414
    },
    {
      "action_query_norm": 42.42245445251465,
      "action_relation_loss": 0.0036000187275931237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.87932647333957,
      "flow_matching_loss": 0.040098662301898,
      "grad_norm": 0.12460705637931824,
      "learning_rate": 7.194002114337526e-05,
      "loss": 0.0439,
      "relation_loss": 0.003842568211257458,
      "selected_pairs": 27008.5,
      "state_query_norm": 43.59437370300293,
      "state_relation_loss": 0.0002425494822091423,
      "step": 11630,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.003842568211257458,
      "weighted_relation_loss": 0.003842568211257458
    },
    {
      "action_query_norm": 42.40473976135254,
      "action_relation_loss": 0.0036828134441748263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.888681010289991,
      "flow_matching_loss": 0.038198215514421464,
      "grad_norm": 0.1294078826904297,
      "learning_rate": 7.189048174288054e-05,
      "loss": 0.0421,
      "relation_loss": 0.003917076950892806,
      "selected_pairs": 26854.2,
      "state_query_norm": 43.58635330200195,
      "state_relation_loss": 0.00023426348343491554,
      "step": 11640,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.003917076950892806,
      "weighted_relation_loss": 0.003917076950892806
    },
    {
      "action_query_norm": 42.40801315307617,
      "action_relation_loss": 0.0037796142511069774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.898035547240411,
      "flow_matching_loss": 0.03876293003559113,
      "grad_norm": 0.11468164622783661,
      "learning_rate": 7.184091574338636e-05,
      "loss": 0.0428,
      "relation_loss": 0.004024223005399108,
      "selected_pairs": 26297.4,
      "state_query_norm": 43.583551025390626,
      "state_relation_loss": 0.00024460874119540675,
      "step": 11650,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.004024223005399108,
      "weighted_relation_loss": 0.004024223005399108
    },
    {
      "action_query_norm": 42.42103309631348,
      "action_relation_loss": 0.003648995654657483,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.907390084190833,
      "flow_matching_loss": 0.03693828657269478,
      "grad_norm": 0.13006332516670227,
      "learning_rate": 7.179132320512009e-05,
      "loss": 0.0408,
      "relation_loss": 0.003875729185529053,
      "selected_pairs": 26598.3,
      "state_query_norm": 43.58899116516113,
      "state_relation_loss": 0.0002267334988573566,
      "step": 11660,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.003875729185529053,
      "weighted_relation_loss": 0.003875729185529053
    },
    {
      "action_query_norm": 42.40011787414551,
      "action_relation_loss": 0.003584091062657535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.916744621141254,
      "flow_matching_loss": 0.037109697982668874,
      "grad_norm": 0.11835719645023346,
      "learning_rate": 7.174170418834134e-05,
      "loss": 0.0409,
      "relation_loss": 0.003821089304983616,
      "selected_pairs": 27315.0,
      "state_query_norm": 43.57693519592285,
      "state_relation_loss": 0.00023699824669165537,
      "step": 11670,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.003821089304983616,
      "weighted_relation_loss": 0.003821089304983616
    },
    {
      "action_query_norm": 42.391740798950195,
      "action_relation_loss": 0.0035796999232843517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.926099158091674,
      "flow_matching_loss": 0.03876571655273438,
      "grad_norm": 0.12644948065280914,
      "learning_rate": 7.169205875334189e-05,
      "loss": 0.0426,
      "relation_loss": 0.0038082350278273226,
      "selected_pairs": 27383.0,
      "state_query_norm": 43.56834106445312,
      "state_relation_loss": 0.0002285351394675672,
      "step": 11680,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0038082350278273226,
      "weighted_relation_loss": 0.0038082350278273226
    },
    {
      "action_query_norm": 42.40126457214355,
      "action_relation_loss": 0.003619626141153276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.935453695042096,
      "flow_matching_loss": 0.03649444915354252,
      "grad_norm": 0.12171107530593872,
      "learning_rate": 7.164238696044562e-05,
      "loss": 0.0403,
      "relation_loss": 0.0038521720096468927,
      "selected_pairs": 27036.3,
      "state_query_norm": 43.57538833618164,
      "state_relation_loss": 0.00023254589614225553,
      "step": 11690,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0038521720096468927,
      "weighted_relation_loss": 0.0038521720096468927
    },
    {
      "action_query_norm": 42.38236694335937,
      "action_relation_loss": 0.0036372171016409993,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.944808231992516,
      "flow_matching_loss": 0.037063416093587875,
      "grad_norm": 0.1098167821764946,
      "learning_rate": 7.159268887000846e-05,
      "loss": 0.0409,
      "relation_loss": 0.0038627667352557184,
      "selected_pairs": 26338.3,
      "state_query_norm": 43.57545433044434,
      "state_relation_loss": 0.0002255496452562511,
      "step": 11700,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0038627667352557184,
      "weighted_relation_loss": 0.0038627667352557184
    },
    {
      "action_query_norm": 42.397587585449216,
      "action_relation_loss": 0.003604103927500546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.954162768942938,
      "flow_matching_loss": 0.03982544466853142,
      "grad_norm": 0.14919298887252808,
      "learning_rate": 7.154296454241827e-05,
      "loss": 0.0437,
      "relation_loss": 0.0038245956413447855,
      "selected_pairs": 27289.7,
      "state_query_norm": 43.57457313537598,
      "state_relation_loss": 0.00022049169056117534,
      "step": 11710,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0038245956413447855,
      "weighted_relation_loss": 0.0038245956413447855
    },
    {
      "action_query_norm": 42.36821174621582,
      "action_relation_loss": 0.0036532497499138115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.963517305893358,
      "flow_matching_loss": 0.036900711432099345,
      "grad_norm": 0.15877558290958405,
      "learning_rate": 7.149321403809479e-05,
      "loss": 0.0408,
      "relation_loss": 0.0038927322486415504,
      "selected_pairs": 27196.9,
      "state_query_norm": 43.56714057922363,
      "state_relation_loss": 0.00023948256130097435,
      "step": 11720,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0038927322486415504,
      "weighted_relation_loss": 0.0038927322486415504
    },
    {
      "action_query_norm": 42.37388725280762,
      "action_relation_loss": 0.003453964344225824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.97287184284378,
      "flow_matching_loss": 0.03781819604337215,
      "grad_norm": 0.12911392748355865,
      "learning_rate": 7.144343741748956e-05,
      "loss": 0.0415,
      "relation_loss": 0.0036862875567749143,
      "selected_pairs": 27101.6,
      "state_query_norm": 43.56731605529785,
      "state_relation_loss": 0.00023232314415508882,
      "step": 11730,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0036862875567749143,
      "weighted_relation_loss": 0.0036862875567749143
    },
    {
      "action_query_norm": 42.37904663085938,
      "action_relation_loss": 0.0037812442053109406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.9822263797942,
      "flow_matching_loss": 0.039108434692025185,
      "grad_norm": 0.09989620745182037,
      "learning_rate": 7.139363474108589e-05,
      "loss": 0.0431,
      "relation_loss": 0.004017701582051813,
      "selected_pairs": 26139.4,
      "state_query_norm": 43.569894027709964,
      "state_relation_loss": 0.000236457385472022,
      "step": 11740,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.004017701582051813,
      "weighted_relation_loss": 0.004017701582051813
    },
    {
      "action_query_norm": 42.36351089477539,
      "action_relation_loss": 0.0036596670746803285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 10.99158091674462,
      "flow_matching_loss": 0.03809768855571747,
      "grad_norm": 0.1452966034412384,
      "learning_rate": 7.134380606939874e-05,
      "loss": 0.042,
      "relation_loss": 0.003891059965826571,
      "selected_pairs": 26686.4,
      "state_query_norm": 43.57379989624023,
      "state_relation_loss": 0.00023139294498832896,
      "step": 11750,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.003891059965826571,
      "weighted_relation_loss": 0.003891059965826571
    },
    {
      "action_query_norm": 42.35508232116699,
      "action_relation_loss": 0.00365597337950021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.000935453695043,
      "flow_matching_loss": 0.038877416774630545,
      "grad_norm": 0.12224941700696945,
      "learning_rate": 7.12939514629746e-05,
      "loss": 0.0428,
      "relation_loss": 0.003872591583058238,
      "selected_pairs": 24618.5,
      "state_query_norm": 43.56531982421875,
      "state_relation_loss": 0.0002166182384826243,
      "step": 11760,
      "valid_action_samples": 215.4,
      "weighted_auxiliary_loss": 0.003872591583058238,
      "weighted_relation_loss": 0.003872591583058238
    },
    {
      "action_query_norm": 42.37021255493164,
      "action_relation_loss": 0.003633129014633596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.010289990645463,
      "flow_matching_loss": 0.03568105064332485,
      "grad_norm": 0.12540215253829956,
      "learning_rate": 7.124407098239155e-05,
      "loss": 0.0395,
      "relation_loss": 0.0038614805322140454,
      "selected_pairs": 26585.4,
      "state_query_norm": 43.56328811645508,
      "state_relation_loss": 0.00022835152340121566,
      "step": 11770,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0038614805322140454,
      "weighted_relation_loss": 0.0038614805322140454
    },
    {
      "action_query_norm": 42.360263442993165,
      "action_relation_loss": 0.003497185930609703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.019644527595885,
      "flow_matching_loss": 0.03663153536617756,
      "grad_norm": 0.12542320787906647,
      "learning_rate": 7.119416468825908e-05,
      "loss": 0.0404,
      "relation_loss": 0.00373678901232779,
      "selected_pairs": 26483.4,
      "state_query_norm": 43.55881690979004,
      "state_relation_loss": 0.00023960309335961937,
      "step": 11780,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.00373678901232779,
      "weighted_relation_loss": 0.00373678901232779
    },
    {
      "action_query_norm": 42.35309600830078,
      "action_relation_loss": 0.0034952991874888538,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.028999064546305,
      "flow_matching_loss": 0.03746380507946014,
      "grad_norm": 0.11284725368022919,
      "learning_rate": 7.114423264121804e-05,
      "loss": 0.0412,
      "relation_loss": 0.0037123818416148423,
      "selected_pairs": 26434.3,
      "state_query_norm": 43.54737396240235,
      "state_relation_loss": 0.0002170827123336494,
      "step": 11790,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0037123818416148423,
      "weighted_relation_loss": 0.0037123818416148423
    },
    {
      "action_query_norm": 42.34851989746094,
      "action_relation_loss": 0.0034641702892258762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.038353601496725,
      "flow_matching_loss": 0.03739466108381748,
      "grad_norm": 0.11381558328866959,
      "learning_rate": 7.109427490194056e-05,
      "loss": 0.0411,
      "relation_loss": 0.0036998297087848186,
      "selected_pairs": 26965.4,
      "state_query_norm": 43.55856857299805,
      "state_relation_loss": 0.00023565941664855928,
      "step": 11800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0036998297087848186,
      "weighted_relation_loss": 0.0036998297087848186
    },
    {
      "action_query_norm": 42.32049827575683,
      "action_relation_loss": 0.0037752619246020913,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.047708138447147,
      "flow_matching_loss": 0.03967537842690945,
      "grad_norm": 0.14739255607128143,
      "learning_rate": 7.104429153113001e-05,
      "loss": 0.0437,
      "relation_loss": 0.003999856323935092,
      "selected_pairs": 27017.2,
      "state_query_norm": 43.54858245849609,
      "state_relation_loss": 0.0002245944633614272,
      "step": 11810,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.003999856323935092,
      "weighted_relation_loss": 0.003999856323935092
    },
    {
      "action_query_norm": 42.34736022949219,
      "action_relation_loss": 0.0036123333731666206,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.057062675397567,
      "flow_matching_loss": 0.038926729932427406,
      "grad_norm": 0.13695771992206573,
      "learning_rate": 7.099428258952092e-05,
      "loss": 0.0428,
      "relation_loss": 0.003865779913030565,
      "selected_pairs": 26448.5,
      "state_query_norm": 43.552411270141604,
      "state_relation_loss": 0.0002534464889322408,
      "step": 11820,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.003865779913030565,
      "weighted_relation_loss": 0.003865779913030565
    },
    {
      "action_query_norm": 42.351097106933594,
      "action_relation_loss": 0.0035909790080040694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.06641721234799,
      "flow_matching_loss": 0.037048653699457645,
      "grad_norm": 0.12222053855657578,
      "learning_rate": 7.094424813787883e-05,
      "loss": 0.0409,
      "relation_loss": 0.0038276487495750187,
      "selected_pairs": 27315.3,
      "state_query_norm": 43.56328201293945,
      "state_relation_loss": 0.00023666967899771407,
      "step": 11830,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0038276487495750187,
      "weighted_relation_loss": 0.0038276487495750187
    },
    {
      "action_query_norm": 42.3194164276123,
      "action_relation_loss": 0.0035924048395827414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.07577174929841,
      "flow_matching_loss": 0.037527157180011275,
      "grad_norm": 0.12969203293323517,
      "learning_rate": 7.089418823700034e-05,
      "loss": 0.0414,
      "relation_loss": 0.00382664252538234,
      "selected_pairs": 27105.6,
      "state_query_norm": 43.55876121520996,
      "state_relation_loss": 0.00023423768870998174,
      "step": 11840,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.00382664252538234,
      "weighted_relation_loss": 0.00382664252538234
    },
    {
      "action_query_norm": 42.32690582275391,
      "action_relation_loss": 0.003547381982207298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.085126286248832,
      "flow_matching_loss": 0.03713646996766329,
      "grad_norm": 0.11730026453733444,
      "learning_rate": 7.084410294771298e-05,
      "loss": 0.0409,
      "relation_loss": 0.003778058080933988,
      "selected_pairs": 26826.0,
      "state_query_norm": 43.550790786743164,
      "state_relation_loss": 0.00023067608854034915,
      "step": 11850,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.003778058080933988,
      "weighted_relation_loss": 0.003778058080933988
    },
    {
      "action_query_norm": 42.34451599121094,
      "action_relation_loss": 0.0034550120821222664,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.094480823199252,
      "flow_matching_loss": 0.03644000105559826,
      "grad_norm": 0.11030587553977966,
      "learning_rate": 7.079399233087504e-05,
      "loss": 0.0401,
      "relation_loss": 0.0036741197574883698,
      "selected_pairs": 27038.3,
      "state_query_norm": 43.55036239624023,
      "state_relation_loss": 0.0002191077423049137,
      "step": 11860,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0036741197574883698,
      "weighted_relation_loss": 0.0036741197574883698
    },
    {
      "action_query_norm": 42.33675842285156,
      "action_relation_loss": 0.0035656235180795194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.103835360149672,
      "flow_matching_loss": 0.03746991772204637,
      "grad_norm": 0.11063242703676224,
      "learning_rate": 7.074385644737568e-05,
      "loss": 0.0413,
      "relation_loss": 0.003801486408337951,
      "selected_pairs": 26963.6,
      "state_query_norm": 43.54142723083496,
      "state_relation_loss": 0.00023586285824421792,
      "step": 11870,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.003801486408337951,
      "weighted_relation_loss": 0.003801486408337951
    },
    {
      "action_query_norm": 42.3243766784668,
      "action_relation_loss": 0.0035198388388380407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.113189897100094,
      "flow_matching_loss": 0.036405515670776364,
      "grad_norm": 0.13034045696258545,
      "learning_rate": 7.069369535813473e-05,
      "loss": 0.0401,
      "relation_loss": 0.0037391575751826167,
      "selected_pairs": 27737.2,
      "state_query_norm": 43.540317916870116,
      "state_relation_loss": 0.00021931870287517086,
      "step": 11880,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.0037391575751826167,
      "weighted_relation_loss": 0.0037391575751826167
    },
    {
      "action_query_norm": 42.33038215637207,
      "action_relation_loss": 0.0032293646596372127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.122544434050514,
      "flow_matching_loss": 0.03646541628986597,
      "grad_norm": 0.11326856911182404,
      "learning_rate": 7.06435091241026e-05,
      "loss": 0.0399,
      "relation_loss": 0.0034476053901016713,
      "selected_pairs": 26622.0,
      "state_query_norm": 43.53963356018066,
      "state_relation_loss": 0.00021824069990543648,
      "step": 11890,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0034476053901016713,
      "weighted_relation_loss": 0.0034476053901016713
    },
    {
      "action_query_norm": 42.305916213989256,
      "action_relation_loss": 0.003561692498624325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.131898971000936,
      "flow_matching_loss": 0.037929421104490754,
      "grad_norm": 0.1210503876209259,
      "learning_rate": 7.059329780626034e-05,
      "loss": 0.0417,
      "relation_loss": 0.0037738950457423927,
      "selected_pairs": 26446.0,
      "state_query_norm": 43.53893623352051,
      "state_relation_loss": 0.00021220257913228123,
      "step": 11900,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0037738950457423927,
      "weighted_relation_loss": 0.0037738950457423927
    },
    {
      "action_query_norm": 42.31995124816895,
      "action_relation_loss": 0.0034545114962384106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.141253507951356,
      "flow_matching_loss": 0.036911404505372045,
      "grad_norm": 0.10308606922626495,
      "learning_rate": 7.054306146561944e-05,
      "loss": 0.0406,
      "relation_loss": 0.0036713786190375685,
      "selected_pairs": 26214.8,
      "state_query_norm": 43.533585739135745,
      "state_relation_loss": 0.00021686708932975308,
      "step": 11910,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0036713786190375685,
      "weighted_relation_loss": 0.0036713786190375685
    },
    {
      "action_query_norm": 42.36518859863281,
      "action_relation_loss": 0.0032392114168033003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.150608044901777,
      "flow_matching_loss": 0.03838099539279938,
      "grad_norm": 0.1458093672990799,
      "learning_rate": 7.049280016322178e-05,
      "loss": 0.0418,
      "relation_loss": 0.003453323687426746,
      "selected_pairs": 26755.9,
      "state_query_norm": 43.53419380187988,
      "state_relation_loss": 0.00021411229245131834,
      "step": 11920,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.003453323687426746,
      "weighted_relation_loss": 0.003453323687426746
    },
    {
      "action_query_norm": 42.34226684570312,
      "action_relation_loss": 0.0033316632732748987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.159962581852199,
      "flow_matching_loss": 0.03503253236413002,
      "grad_norm": 0.11295643448829651,
      "learning_rate": 7.044251396013957e-05,
      "loss": 0.0386,
      "relation_loss": 0.003551654936745763,
      "selected_pairs": 27012.8,
      "state_query_norm": 43.524257278442384,
      "state_relation_loss": 0.0002199916634708643,
      "step": 11930,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.003551654936745763,
      "weighted_relation_loss": 0.003551654936745763
    },
    {
      "action_query_norm": 42.33619613647461,
      "action_relation_loss": 0.003348483215086162,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.169317118802619,
      "flow_matching_loss": 0.0362206606194377,
      "grad_norm": 0.13446305692195892,
      "learning_rate": 7.039220291747528e-05,
      "loss": 0.0398,
      "relation_loss": 0.0035720309242606163,
      "selected_pairs": 26807.0,
      "state_query_norm": 43.527140045166014,
      "state_relation_loss": 0.00022354768152581527,
      "step": 11940,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0035720309242606163,
      "weighted_relation_loss": 0.0035720309242606163
    },
    {
      "action_query_norm": 42.3299072265625,
      "action_relation_loss": 0.003405768959783018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.17867165575304,
      "flow_matching_loss": 0.03650190383195877,
      "grad_norm": 0.11119038611650467,
      "learning_rate": 7.034186709636159e-05,
      "loss": 0.0401,
      "relation_loss": 0.003628933313302696,
      "selected_pairs": 26144.8,
      "state_query_norm": 43.52467269897461,
      "state_relation_loss": 0.0002231643520644866,
      "step": 11950,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.003628933313302696,
      "weighted_relation_loss": 0.003628933313302696
    },
    {
      "action_query_norm": 42.287973403930664,
      "action_relation_loss": 0.0035743589978665113,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.188026192703461,
      "flow_matching_loss": 0.03805495277047157,
      "grad_norm": 0.13549154996871948,
      "learning_rate": 7.029150655796129e-05,
      "loss": 0.0419,
      "relation_loss": 0.0038005071924999355,
      "selected_pairs": 26782.3,
      "state_query_norm": 43.534489440917966,
      "state_relation_loss": 0.00022614823101321235,
      "step": 11960,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0038005071924999355,
      "weighted_relation_loss": 0.0038005071924999355
    },
    {
      "action_query_norm": 42.304917907714845,
      "action_relation_loss": 0.0034460857277736068,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.197380729653883,
      "flow_matching_loss": 0.036011919006705286,
      "grad_norm": 0.14135892689228058,
      "learning_rate": 7.024112136346712e-05,
      "loss": 0.0397,
      "relation_loss": 0.003662392194382846,
      "selected_pairs": 26667.8,
      "state_query_norm": 43.52930107116699,
      "state_relation_loss": 0.0002163064302294515,
      "step": 11970,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003662392194382846,
      "weighted_relation_loss": 0.003662392194382846
    },
    {
      "action_query_norm": 42.27889747619629,
      "action_relation_loss": 0.0036159967072308063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.206735266604303,
      "flow_matching_loss": 0.03816225733608007,
      "grad_norm": 0.13666009902954102,
      "learning_rate": 7.01907115741019e-05,
      "loss": 0.042,
      "relation_loss": 0.0038520455360412597,
      "selected_pairs": 26966.5,
      "state_query_norm": 43.52654838562012,
      "state_relation_loss": 0.00023604883608641103,
      "step": 11980,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0038520455360412597,
      "weighted_relation_loss": 0.0038520455360412597
    },
    {
      "action_query_norm": 42.25491638183594,
      "action_relation_loss": 0.0034213276579976083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.216089803554723,
      "flow_matching_loss": 0.03883342146873474,
      "grad_norm": 0.13527381420135498,
      "learning_rate": 7.014027725111825e-05,
      "loss": 0.0425,
      "relation_loss": 0.0036470713559538124,
      "selected_pairs": 26689.1,
      "state_query_norm": 43.52734222412109,
      "state_relation_loss": 0.00022574372560484334,
      "step": 11990,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0036470713559538124,
      "weighted_relation_loss": 0.0036470713559538124
    },
    {
      "action_query_norm": 42.24286422729492,
      "action_relation_loss": 0.0034426589962095024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.225444340505145,
      "flow_matching_loss": 0.03827880173921585,
      "grad_norm": 0.15436220169067383,
      "learning_rate": 7.008981845579864e-05,
      "loss": 0.042,
      "relation_loss": 0.003672029823064804,
      "selected_pairs": 26622.4,
      "state_query_norm": 43.51313819885254,
      "state_relation_loss": 0.0002293708283104934,
      "step": 12000,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.003672029823064804,
      "weighted_relation_loss": 0.003672029823064804
    },
    {
      "action_query_norm": 42.24184379577637,
      "action_relation_loss": 0.003416877263225615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.234798877455566,
      "flow_matching_loss": 0.035625309124588965,
      "grad_norm": 0.12083075940608978,
      "learning_rate": 7.003933524945528e-05,
      "loss": 0.0393,
      "relation_loss": 0.0036448305938392876,
      "selected_pairs": 26753.4,
      "state_query_norm": 43.51805534362793,
      "state_relation_loss": 0.00022795336553826928,
      "step": 12010,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0036448305938392876,
      "weighted_relation_loss": 0.0036448305938392876
    },
    {
      "action_query_norm": 42.282898330688475,
      "action_relation_loss": 0.0034751614090055226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.244153414405988,
      "flow_matching_loss": 0.03555261306464672,
      "grad_norm": 0.10374335944652557,
      "learning_rate": 6.998882769342998e-05,
      "loss": 0.0393,
      "relation_loss": 0.003708282788284123,
      "selected_pairs": 26365.0,
      "state_query_norm": 43.52282295227051,
      "state_relation_loss": 0.0002331213909201324,
      "step": 12020,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.003708282788284123,
      "weighted_relation_loss": 0.003708282788284123
    },
    {
      "action_query_norm": 42.27407875061035,
      "action_relation_loss": 0.003472167206928134,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.253507951356408,
      "flow_matching_loss": 0.03555713426321745,
      "grad_norm": 0.12096989899873734,
      "learning_rate": 6.993829584909423e-05,
      "loss": 0.0392,
      "relation_loss": 0.0036855238489806653,
      "selected_pairs": 27320.1,
      "state_query_norm": 43.51217346191406,
      "state_relation_loss": 0.00021335663623176516,
      "step": 12030,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0036855238489806653,
      "weighted_relation_loss": 0.0036855238489806653
    },
    {
      "action_query_norm": 42.26821517944336,
      "action_relation_loss": 0.003671755688264966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.262862488306828,
      "flow_matching_loss": 0.037407879345119,
      "grad_norm": 0.10966039448976517,
      "learning_rate": 6.988773977784895e-05,
      "loss": 0.0413,
      "relation_loss": 0.0038921623956412075,
      "selected_pairs": 26660.9,
      "state_query_norm": 43.513708877563474,
      "state_relation_loss": 0.0002204066753620282,
      "step": 12040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0038921623956412075,
      "weighted_relation_loss": 0.0038921623956412075
    },
    {
      "action_query_norm": 42.25817222595215,
      "action_relation_loss": 0.0036381388548761607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.27221702525725,
      "flow_matching_loss": 0.03824984394013882,
      "grad_norm": 0.11886917054653168,
      "learning_rate": 6.983715954112454e-05,
      "loss": 0.0421,
      "relation_loss": 0.0038498122245073318,
      "selected_pairs": 26876.1,
      "state_query_norm": 43.52042236328125,
      "state_relation_loss": 0.00021167334489291534,
      "step": 12050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0038498122245073318,
      "weighted_relation_loss": 0.0038498122245073318
    },
    {
      "action_query_norm": 42.248786544799806,
      "action_relation_loss": 0.003366266004741192,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.28157156220767,
      "flow_matching_loss": 0.034600745886564255,
      "grad_norm": 0.12573550641536713,
      "learning_rate": 6.978655520038079e-05,
      "loss": 0.0382,
      "relation_loss": 0.003586231544613838,
      "selected_pairs": 27069.4,
      "state_query_norm": 43.49939308166504,
      "state_relation_loss": 0.00021996557043166832,
      "step": 12060,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.003586231544613838,
      "weighted_relation_loss": 0.003586231544613838
    },
    {
      "action_query_norm": 42.23551483154297,
      "action_relation_loss": 0.0033538424409925937,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.290926099158092,
      "flow_matching_loss": 0.035318396240472796,
      "grad_norm": 0.10542241483926773,
      "learning_rate": 6.97359268171067e-05,
      "loss": 0.0389,
      "relation_loss": 0.003572947671636939,
      "selected_pairs": 27086.9,
      "state_query_norm": 43.500262069702146,
      "state_relation_loss": 0.00021910526265855879,
      "step": 12070,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.003572947671636939,
      "weighted_relation_loss": 0.003572947671636939
    },
    {
      "action_query_norm": 42.23811264038086,
      "action_relation_loss": 0.0034176427870988846,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.300280636108512,
      "flow_matching_loss": 0.03766727186739445,
      "grad_norm": 0.13184860348701477,
      "learning_rate": 6.968527445282056e-05,
      "loss": 0.0413,
      "relation_loss": 0.0036328897811472418,
      "selected_pairs": 26880.7,
      "state_query_norm": 43.51065940856934,
      "state_relation_loss": 0.00021524700423469767,
      "step": 12080,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0036328897811472418,
      "weighted_relation_loss": 0.0036328897811472418
    },
    {
      "action_query_norm": 42.27214012145996,
      "action_relation_loss": 0.003390810312703252,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.309635173058934,
      "flow_matching_loss": 0.037195426598191264,
      "grad_norm": 0.12563468515872955,
      "learning_rate": 6.963459816906977e-05,
      "loss": 0.0408,
      "relation_loss": 0.003600653843022883,
      "selected_pairs": 27314.1,
      "state_query_norm": 43.50993347167969,
      "state_relation_loss": 0.00020984351285733283,
      "step": 12090,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.003600653843022883,
      "weighted_relation_loss": 0.003600653843022883
    },
    {
      "action_query_norm": 42.2795955657959,
      "action_relation_loss": 0.0033652760786935685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.318989710009355,
      "flow_matching_loss": 0.035007215663790704,
      "grad_norm": 0.11614283174276352,
      "learning_rate": 6.958389802743077e-05,
      "loss": 0.0386,
      "relation_loss": 0.003585871332325041,
      "selected_pairs": 27246.4,
      "state_query_norm": 43.499834060668945,
      "state_relation_loss": 0.00022059529001126066,
      "step": 12100,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.003585871332325041,
      "weighted_relation_loss": 0.003585871332325041
    },
    {
      "action_query_norm": 42.24223251342774,
      "action_relation_loss": 0.003368758084252477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.328344246959775,
      "flow_matching_loss": 0.035058594308793545,
      "grad_norm": 0.1457776129245758,
      "learning_rate": 6.953317408950903e-05,
      "loss": 0.0387,
      "relation_loss": 0.0035915407352149487,
      "selected_pairs": 27183.5,
      "state_query_norm": 43.50501823425293,
      "state_relation_loss": 0.00022278266987996176,
      "step": 12110,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0035915407352149487,
      "weighted_relation_loss": 0.0035915407352149487
    },
    {
      "action_query_norm": 42.25391159057617,
      "action_relation_loss": 0.003306548297405243,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.337698783910197,
      "flow_matching_loss": 0.035430072247982024,
      "grad_norm": 0.09050969779491425,
      "learning_rate": 6.94824264169389e-05,
      "loss": 0.0389,
      "relation_loss": 0.003514466411434114,
      "selected_pairs": 26641.9,
      "state_query_norm": 43.498739624023436,
      "state_relation_loss": 0.00020791810966329648,
      "step": 12120,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.003514466411434114,
      "weighted_relation_loss": 0.003514466411434114
    },
    {
      "action_query_norm": 42.22072334289551,
      "action_relation_loss": 0.003423997899517417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.347053320860617,
      "flow_matching_loss": 0.03641915619373322,
      "grad_norm": 0.11177719384431839,
      "learning_rate": 6.94316550713836e-05,
      "loss": 0.0401,
      "relation_loss": 0.0036477734567597507,
      "selected_pairs": 27012.4,
      "state_query_norm": 43.493578338623045,
      "state_relation_loss": 0.0002237755703390576,
      "step": 12130,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0036477734567597507,
      "weighted_relation_loss": 0.0036477734567597507
    },
    {
      "action_query_norm": 42.24471244812012,
      "action_relation_loss": 0.003355250298045576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.356407857811039,
      "flow_matching_loss": 0.037905776873230934,
      "grad_norm": 0.1078900396823883,
      "learning_rate": 6.938086011453513e-05,
      "loss": 0.0415,
      "relation_loss": 0.0035804214887320994,
      "selected_pairs": 25660.6,
      "state_query_norm": 43.48552284240723,
      "state_relation_loss": 0.00022517118050018324,
      "step": 12140,
      "valid_action_samples": 227.0,
      "weighted_auxiliary_loss": 0.0035804214887320994,
      "weighted_relation_loss": 0.0035804214887320994
    },
    {
      "action_query_norm": 42.25706825256348,
      "action_relation_loss": 0.0032800512621179224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.36576239476146,
      "flow_matching_loss": 0.036764952912926675,
      "grad_norm": 0.14182786643505096,
      "learning_rate": 6.93300416081141e-05,
      "loss": 0.0403,
      "relation_loss": 0.0034995307214558125,
      "selected_pairs": 26953.9,
      "state_query_norm": 43.49894905090332,
      "state_relation_loss": 0.00021947946515865624,
      "step": 12150,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0034995307214558125,
      "weighted_relation_loss": 0.0034995307214558125
    },
    {
      "action_query_norm": 42.2324031829834,
      "action_relation_loss": 0.0033803251571953297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.37511693171188,
      "flow_matching_loss": 0.03748547099530697,
      "grad_norm": 0.1377258449792862,
      "learning_rate": 6.927919961386984e-05,
      "loss": 0.0411,
      "relation_loss": 0.003597947908565402,
      "selected_pairs": 26287.6,
      "state_query_norm": 43.489799118041994,
      "state_relation_loss": 0.00021762277174275368,
      "step": 12160,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.003597947908565402,
      "weighted_relation_loss": 0.003597947908565402
    },
    {
      "action_query_norm": 42.213935470581056,
      "action_relation_loss": 0.003558741579763591,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.384471468662301,
      "flow_matching_loss": 0.0376663438975811,
      "grad_norm": 0.13855285942554474,
      "learning_rate": 6.922833419358013e-05,
      "loss": 0.0415,
      "relation_loss": 0.003785207774490118,
      "selected_pairs": 26914.2,
      "state_query_norm": 43.48393783569336,
      "state_relation_loss": 0.00022646618017461151,
      "step": 12170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.003785207774490118,
      "weighted_relation_loss": 0.003785207774490118
    },
    {
      "action_query_norm": 42.233512496948244,
      "action_relation_loss": 0.003348171501420438,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.393826005612722,
      "flow_matching_loss": 0.034783700853586196,
      "grad_norm": 0.12188196182250977,
      "learning_rate": 6.917744540905124e-05,
      "loss": 0.0384,
      "relation_loss": 0.003573523135855794,
      "selected_pairs": 26782.5,
      "state_query_norm": 43.487797546386716,
      "state_relation_loss": 0.0002253515805932693,
      "step": 12180,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003573523135855794,
      "weighted_relation_loss": 0.003573523135855794
    },
    {
      "action_query_norm": 42.24207382202148,
      "action_relation_loss": 0.003244507382623851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.403180542563144,
      "flow_matching_loss": 0.037160313315689565,
      "grad_norm": 0.13451427221298218,
      "learning_rate": 6.912653332211787e-05,
      "loss": 0.0406,
      "relation_loss": 0.0034763125237077474,
      "selected_pairs": 26827.0,
      "state_query_norm": 43.490618515014646,
      "state_relation_loss": 0.00023180512507678941,
      "step": 12190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0034763125237077474,
      "weighted_relation_loss": 0.0034763125237077474
    },
    {
      "action_query_norm": 42.23967933654785,
      "action_relation_loss": 0.003185544302687049,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.412535079513564,
      "flow_matching_loss": 0.03762156888842583,
      "grad_norm": 0.17974194884300232,
      "learning_rate": 6.9075597994643e-05,
      "loss": 0.041,
      "relation_loss": 0.0034044865518808366,
      "selected_pairs": 27693.7,
      "state_query_norm": 43.48233413696289,
      "state_relation_loss": 0.00021894226374570281,
      "step": 12200,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.0034044865518808366,
      "weighted_relation_loss": 0.0034044865518808366
    },
    {
      "action_query_norm": 42.20636253356933,
      "action_relation_loss": 0.0033727270318195226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.421889616463986,
      "flow_matching_loss": 0.038810350745916364,
      "grad_norm": 0.11631488800048828,
      "learning_rate": 6.902463948851787e-05,
      "loss": 0.0424,
      "relation_loss": 0.003601385769434273,
      "selected_pairs": 26669.6,
      "state_query_norm": 43.47470054626465,
      "state_relation_loss": 0.0002286587332491763,
      "step": 12210,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003601385769434273,
      "weighted_relation_loss": 0.003601385769434273
    },
    {
      "action_query_norm": 42.2312686920166,
      "action_relation_loss": 0.0033509741770103574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.431244153414406,
      "flow_matching_loss": 0.03482390698045492,
      "grad_norm": 0.12551163136959076,
      "learning_rate": 6.897365786566184e-05,
      "loss": 0.0384,
      "relation_loss": 0.0035686072427779437,
      "selected_pairs": 26777.7,
      "state_query_norm": 43.4859619140625,
      "state_relation_loss": 0.00021763306722277775,
      "step": 12220,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0035686072427779437,
      "weighted_relation_loss": 0.0035686072427779437
    },
    {
      "action_query_norm": 42.23380699157715,
      "action_relation_loss": 0.003264257707633078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.440598690364826,
      "flow_matching_loss": 0.03465246427804232,
      "grad_norm": 0.11762004345655441,
      "learning_rate": 6.892265318802242e-05,
      "loss": 0.0381,
      "relation_loss": 0.0034862666158005594,
      "selected_pairs": 27125.6,
      "state_query_norm": 43.48010139465332,
      "state_relation_loss": 0.00022200888633960858,
      "step": 12230,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0034862666158005594,
      "weighted_relation_loss": 0.0034862666158005594
    },
    {
      "action_query_norm": 42.19971542358398,
      "action_relation_loss": 0.0034942379919812084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.449953227315248,
      "flow_matching_loss": 0.0343641372397542,
      "grad_norm": 0.11113609373569489,
      "learning_rate": 6.887162551757507e-05,
      "loss": 0.0381,
      "relation_loss": 0.003711665538139641,
      "selected_pairs": 26224.0,
      "state_query_norm": 43.483327865600586,
      "state_relation_loss": 0.00021742749522672965,
      "step": 12240,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.003711665538139641,
      "weighted_relation_loss": 0.003711665538139641
    },
    {
      "action_query_norm": 42.19910736083985,
      "action_relation_loss": 0.0035403405781835318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.459307764265668,
      "flow_matching_loss": 0.03532683998346329,
      "grad_norm": 0.11339490115642548,
      "learning_rate": 6.882057491632326e-05,
      "loss": 0.0391,
      "relation_loss": 0.0037671525264158846,
      "selected_pairs": 26712.1,
      "state_query_norm": 43.46441307067871,
      "state_relation_loss": 0.0002268119584186934,
      "step": 12250,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0037671525264158846,
      "weighted_relation_loss": 0.0037671525264158846
    },
    {
      "action_query_norm": 42.180677795410155,
      "action_relation_loss": 0.0034042609622702003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.46866230121609,
      "flow_matching_loss": 0.03748497497290373,
      "grad_norm": 0.10912596434354782,
      "learning_rate": 6.876950144629824e-05,
      "loss": 0.0411,
      "relation_loss": 0.0036230734549462795,
      "selected_pairs": 27131.5,
      "state_query_norm": 43.47650032043457,
      "state_relation_loss": 0.00021881246648263185,
      "step": 12260,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0036230734549462795,
      "weighted_relation_loss": 0.0036230734549462795
    },
    {
      "action_query_norm": 42.167978286743164,
      "action_relation_loss": 0.0032847695983946323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.47801683816651,
      "flow_matching_loss": 0.0369851129129529,
      "grad_norm": 0.11887135356664658,
      "learning_rate": 6.87184051695591e-05,
      "loss": 0.0405,
      "relation_loss": 0.0035010908497497438,
      "selected_pairs": 27336.5,
      "state_query_norm": 43.48338851928711,
      "state_relation_loss": 0.00021632125572068617,
      "step": 12270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0035010908497497438,
      "weighted_relation_loss": 0.0035010908497497438
    },
    {
      "action_query_norm": 42.18276252746582,
      "action_relation_loss": 0.0034068903420120477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.487371375116933,
      "flow_matching_loss": 0.037061866372823715,
      "grad_norm": 0.12417113035917282,
      "learning_rate": 6.866728614819268e-05,
      "loss": 0.0407,
      "relation_loss": 0.0036249352619051934,
      "selected_pairs": 26895.2,
      "state_query_norm": 43.47976722717285,
      "state_relation_loss": 0.0002180449228035286,
      "step": 12280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0036249352619051934,
      "weighted_relation_loss": 0.0036249352619051934
    },
    {
      "action_query_norm": 42.1812801361084,
      "action_relation_loss": 0.0032841098960489034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.496725912067353,
      "flow_matching_loss": 0.03636789582669735,
      "grad_norm": 0.1399625539779663,
      "learning_rate": 6.861614444431337e-05,
      "loss": 0.0399,
      "relation_loss": 0.0034910645335912704,
      "selected_pairs": 27101.7,
      "state_query_norm": 43.476810455322266,
      "state_relation_loss": 0.00020695462153526024,
      "step": 12290,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0034910645335912704,
      "weighted_relation_loss": 0.0034910645335912704
    },
    {
      "action_query_norm": 42.19993629455566,
      "action_relation_loss": 0.003112882887944579,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.506080449017773,
      "flow_matching_loss": 0.03534938730299473,
      "grad_norm": 0.1512395739555359,
      "learning_rate": 6.856498012006318e-05,
      "loss": 0.0387,
      "relation_loss": 0.0033248526975512504,
      "selected_pairs": 27151.1,
      "state_query_norm": 43.469760513305665,
      "state_relation_loss": 0.00021196980233071372,
      "step": 12300,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0033248526975512504,
      "weighted_relation_loss": 0.0033248526975512504
    },
    {
      "action_query_norm": 42.17736320495605,
      "action_relation_loss": 0.003239030879922211,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.515434985968195,
      "flow_matching_loss": 0.03666394148021936,
      "grad_norm": 0.1421726495027542,
      "learning_rate": 6.851379323761157e-05,
      "loss": 0.0401,
      "relation_loss": 0.0034507296280935408,
      "selected_pairs": 27295.4,
      "state_query_norm": 43.46566619873047,
      "state_relation_loss": 0.0002116987685440108,
      "step": 12310,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0034507296280935408,
      "weighted_relation_loss": 0.0034507296280935408
    },
    {
      "action_query_norm": 42.19910011291504,
      "action_relation_loss": 0.003235983825288713,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.524789522918615,
      "flow_matching_loss": 0.0348039485514164,
      "grad_norm": 0.14795462787151337,
      "learning_rate": 6.846258385915545e-05,
      "loss": 0.0382,
      "relation_loss": 0.003444705554284155,
      "selected_pairs": 27245.4,
      "state_query_norm": 43.461478042602536,
      "state_relation_loss": 0.00020872171880910174,
      "step": 12320,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.003444705554284155,
      "weighted_relation_loss": 0.003444705554284155
    },
    {
      "action_query_norm": 42.2327262878418,
      "action_relation_loss": 0.003251852537505329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.534144059869037,
      "flow_matching_loss": 0.036949438601732255,
      "grad_norm": 0.12003280967473984,
      "learning_rate": 6.841135204691902e-05,
      "loss": 0.0404,
      "relation_loss": 0.0034732329891994595,
      "selected_pairs": 26643.9,
      "state_query_norm": 43.46166458129883,
      "state_relation_loss": 0.0002213804458733648,
      "step": 12330,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0034732329891994595,
      "weighted_relation_loss": 0.0034732329891994595
    },
    {
      "action_query_norm": 42.1659797668457,
      "action_relation_loss": 0.0033567221835255623,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.543498596819457,
      "flow_matching_loss": 0.033931861817836764,
      "grad_norm": 0.13398581743240356,
      "learning_rate": 6.836009786315377e-05,
      "loss": 0.0375,
      "relation_loss": 0.003574499418027699,
      "selected_pairs": 26457.6,
      "state_query_norm": 43.489004135131836,
      "state_relation_loss": 0.0002177772272261791,
      "step": 12340,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.003574499418027699,
      "weighted_relation_loss": 0.003574499418027699
    },
    {
      "action_query_norm": 42.15822334289551,
      "action_relation_loss": 0.0032732214778661726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.552853133769878,
      "flow_matching_loss": 0.03678263761103153,
      "grad_norm": 0.13553957641124725,
      "learning_rate": 6.830882137013839e-05,
      "loss": 0.0403,
      "relation_loss": 0.0034794495208188892,
      "selected_pairs": 26918.0,
      "state_query_norm": 43.4560546875,
      "state_relation_loss": 0.0002062280007521622,
      "step": 12350,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0034794495208188892,
      "weighted_relation_loss": 0.0034794495208188892
    },
    {
      "action_query_norm": 42.17784385681152,
      "action_relation_loss": 0.0033801913727074862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.5622076707203,
      "flow_matching_loss": 0.03464647363871336,
      "grad_norm": 0.11690334975719452,
      "learning_rate": 6.825752263017863e-05,
      "loss": 0.0382,
      "relation_loss": 0.0035919335670769213,
      "selected_pairs": 26665.1,
      "state_query_norm": 43.459957885742185,
      "state_relation_loss": 0.00021174221474211662,
      "step": 12360,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0035919335670769213,
      "weighted_relation_loss": 0.0035919335670769213
    },
    {
      "action_query_norm": 42.20817909240723,
      "action_relation_loss": 0.0032825684174895287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.57156220767072,
      "flow_matching_loss": 0.03586600814014673,
      "grad_norm": 0.10910698026418686,
      "learning_rate": 6.820620170560731e-05,
      "loss": 0.0394,
      "relation_loss": 0.003508445667102933,
      "selected_pairs": 26782.0,
      "state_query_norm": 43.46208610534668,
      "state_relation_loss": 0.00022587722633033992,
      "step": 12370,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003508445667102933,
      "weighted_relation_loss": 0.003508445667102933
    },
    {
      "action_query_norm": 42.16415100097656,
      "action_relation_loss": 0.0032461652997881176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.580916744621142,
      "flow_matching_loss": 0.03573173247277737,
      "grad_norm": 0.1503521203994751,
      "learning_rate": 6.815485865878418e-05,
      "loss": 0.0392,
      "relation_loss": 0.0034614575328305365,
      "selected_pairs": 27293.4,
      "state_query_norm": 43.45157737731934,
      "state_relation_loss": 0.0002152922228560783,
      "step": 12380,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0034614575328305365,
      "weighted_relation_loss": 0.0034614575328305365
    },
    {
      "action_query_norm": 42.17634620666504,
      "action_relation_loss": 0.0032445505261421204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.590271281571562,
      "flow_matching_loss": 0.03368823491036892,
      "grad_norm": 0.1349669098854065,
      "learning_rate": 6.810349355209588e-05,
      "loss": 0.0371,
      "relation_loss": 0.0034366451669484376,
      "selected_pairs": 26920.8,
      "state_query_norm": 43.44979629516602,
      "state_relation_loss": 0.00019209464953746648,
      "step": 12390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0034366451669484376,
      "weighted_relation_loss": 0.0034366451669484376
    },
    {
      "action_query_norm": 42.1641170501709,
      "action_relation_loss": 0.003509943536482751,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.599625818521982,
      "flow_matching_loss": 0.03718550745397806,
      "grad_norm": 0.12181929498910904,
      "learning_rate": 6.805210644795588e-05,
      "loss": 0.0409,
      "relation_loss": 0.0037096495041623713,
      "selected_pairs": 27360.9,
      "state_query_norm": 43.452945709228516,
      "state_relation_loss": 0.00019970594439655542,
      "step": 12400,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0037096495041623713,
      "weighted_relation_loss": 0.0037096495041623713
    },
    {
      "action_query_norm": 42.1680305480957,
      "action_relation_loss": 0.0031360001070424915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.608980355472404,
      "flow_matching_loss": 0.035378886573016646,
      "grad_norm": 0.10737279802560806,
      "learning_rate": 6.800069740880433e-05,
      "loss": 0.0387,
      "relation_loss": 0.0033513715025037527,
      "selected_pairs": 27626.9,
      "state_query_norm": 43.451899719238284,
      "state_relation_loss": 0.0002153713838197291,
      "step": 12410,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0033513715025037527,
      "weighted_relation_loss": 0.0033513715025037527
    },
    {
      "action_query_norm": 42.153960418701175,
      "action_relation_loss": 0.003339984640479088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.618334892422824,
      "flow_matching_loss": 0.03643034845590591,
      "grad_norm": 0.17068736255168915,
      "learning_rate": 6.794926649710807e-05,
      "loss": 0.04,
      "relation_loss": 0.0035481836879625917,
      "selected_pairs": 26575.5,
      "state_query_norm": 43.454064559936526,
      "state_relation_loss": 0.00020819902565563097,
      "step": 12420,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0035481836879625917,
      "weighted_relation_loss": 0.0035481836879625917
    },
    {
      "action_query_norm": 42.162287902832034,
      "action_relation_loss": 0.003384105861186981,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.627689429373246,
      "flow_matching_loss": 0.03526771124452353,
      "grad_norm": 0.12669725716114044,
      "learning_rate": 6.789781377536052e-05,
      "loss": 0.0389,
      "relation_loss": 0.0035952350357547403,
      "selected_pairs": 26898.8,
      "state_query_norm": 43.451137161254884,
      "state_relation_loss": 0.0002111291862092912,
      "step": 12430,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0035952350357547403,
      "weighted_relation_loss": 0.0035952350357547403
    },
    {
      "action_query_norm": 42.15617790222168,
      "action_relation_loss": 0.0031788289081305264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.637043966323667,
      "flow_matching_loss": 0.036006140522658825,
      "grad_norm": 0.13722069561481476,
      "learning_rate": 6.784633930608158e-05,
      "loss": 0.0394,
      "relation_loss": 0.0033991184318438172,
      "selected_pairs": 26990.0,
      "state_query_norm": 43.44210433959961,
      "state_relation_loss": 0.000220289510616567,
      "step": 12440,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0033991184318438172,
      "weighted_relation_loss": 0.0033991184318438172
    },
    {
      "action_query_norm": 42.147071838378906,
      "action_relation_loss": 0.0036741125164553522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.646398503274089,
      "flow_matching_loss": 0.03687986768782139,
      "grad_norm": 0.14447014033794403,
      "learning_rate": 6.779484315181759e-05,
      "loss": 0.0408,
      "relation_loss": 0.00388653501868248,
      "selected_pairs": 26112.4,
      "state_query_norm": 43.447697448730466,
      "state_relation_loss": 0.00021242253278614954,
      "step": 12450,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.00388653501868248,
      "weighted_relation_loss": 0.00388653501868248
    },
    {
      "action_query_norm": 42.17148628234863,
      "action_relation_loss": 0.0033325975760817527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.655753040224509,
      "flow_matching_loss": 0.035694096609950064,
      "grad_norm": 0.09845362603664398,
      "learning_rate": 6.774332537514122e-05,
      "loss": 0.0392,
      "relation_loss": 0.003542292071506381,
      "selected_pairs": 27047.1,
      "state_query_norm": 43.435228729248045,
      "state_relation_loss": 0.0002096944605000317,
      "step": 12460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003542292071506381,
      "weighted_relation_loss": 0.003542292071506381
    },
    {
      "action_query_norm": 42.11215744018555,
      "action_relation_loss": 0.0036054196767508983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.665107577174929,
      "flow_matching_loss": 0.03608983419835567,
      "grad_norm": 0.12846402823925018,
      "learning_rate": 6.769178603865143e-05,
      "loss": 0.0399,
      "relation_loss": 0.0038288127398118377,
      "selected_pairs": 26392.8,
      "state_query_norm": 43.434486389160156,
      "state_relation_loss": 0.00022339302231557667,
      "step": 12470,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0038288127398118377,
      "weighted_relation_loss": 0.0038288127398118377
    },
    {
      "action_query_norm": 42.15213394165039,
      "action_relation_loss": 0.0031204638071358204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.674462114125351,
      "flow_matching_loss": 0.036912243254482745,
      "grad_norm": 0.09330590814352036,
      "learning_rate": 6.764022520497337e-05,
      "loss": 0.0403,
      "relation_loss": 0.0033380566630512474,
      "selected_pairs": 27253.1,
      "state_query_norm": 43.4385498046875,
      "state_relation_loss": 0.0002175928675569594,
      "step": 12480,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0033380566630512474,
      "weighted_relation_loss": 0.0033380566630512474
    },
    {
      "action_query_norm": 42.136787796020506,
      "action_relation_loss": 0.003384478925727308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.683816651075771,
      "flow_matching_loss": 0.03630836009979248,
      "grad_norm": 0.12670718133449554,
      "learning_rate": 6.758864293675833e-05,
      "loss": 0.0399,
      "relation_loss": 0.0036014342214912174,
      "selected_pairs": 26385.2,
      "state_query_norm": 43.446577072143555,
      "state_relation_loss": 0.00021695530886063353,
      "step": 12490,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0036014342214912174,
      "weighted_relation_loss": 0.0036014342214912174
    },
    {
      "action_query_norm": 42.13591537475586,
      "action_relation_loss": 0.003408984513953328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.693171188026193,
      "flow_matching_loss": 0.03601077198982239,
      "grad_norm": 0.10600659996271133,
      "learning_rate": 6.753703929668362e-05,
      "loss": 0.0396,
      "relation_loss": 0.0036164599703624846,
      "selected_pairs": 26874.4,
      "state_query_norm": 43.43670425415039,
      "state_relation_loss": 0.00020747546368511395,
      "step": 12500,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0036164599703624846,
      "weighted_relation_loss": 0.0036164599703624846
    },
    {
      "action_query_norm": 42.1272388458252,
      "action_relation_loss": 0.003271537250839174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.702525724976613,
      "flow_matching_loss": 0.03469538874924183,
      "grad_norm": 0.12563477456569672,
      "learning_rate": 6.748541434745253e-05,
      "loss": 0.0382,
      "relation_loss": 0.0034792701015248896,
      "selected_pairs": 26785.2,
      "state_query_norm": 43.42907485961914,
      "state_relation_loss": 0.0002077328899758868,
      "step": 12510,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0034792701015248896,
      "weighted_relation_loss": 0.0034792701015248896
    },
    {
      "action_query_norm": 42.13860397338867,
      "action_relation_loss": 0.0030992010608315466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.711880261927035,
      "flow_matching_loss": 0.03535322360694408,
      "grad_norm": 0.15090161561965942,
      "learning_rate": 6.743376815179424e-05,
      "loss": 0.0387,
      "relation_loss": 0.0033018662827089427,
      "selected_pairs": 27300.0,
      "state_query_norm": 43.434522247314455,
      "state_relation_loss": 0.00020266522478777914,
      "step": 12520,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0033018662827089427,
      "weighted_relation_loss": 0.0033018662827089427
    },
    {
      "action_query_norm": 42.13251953125,
      "action_relation_loss": 0.003163106762804091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.721234798877456,
      "flow_matching_loss": 0.03805111199617386,
      "grad_norm": 0.12468741089105606,
      "learning_rate": 6.738210077246376e-05,
      "loss": 0.0414,
      "relation_loss": 0.003385440004058182,
      "selected_pairs": 27669.9,
      "state_query_norm": 43.428938674926755,
      "state_relation_loss": 0.00022233326453715562,
      "step": 12530,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.003385440004058182,
      "weighted_relation_loss": 0.003385440004058182
    },
    {
      "action_query_norm": 42.128719329833984,
      "action_relation_loss": 0.0031965394271537663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.730589335827876,
      "flow_matching_loss": 0.036211360991001126,
      "grad_norm": 0.11946740746498108,
      "learning_rate": 6.733041227224181e-05,
      "loss": 0.0396,
      "relation_loss": 0.0034204736817628144,
      "selected_pairs": 26340.6,
      "state_query_norm": 43.41685180664062,
      "state_relation_loss": 0.0002239342444227077,
      "step": 12540,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0034204736817628144,
      "weighted_relation_loss": 0.0034204736817628144
    },
    {
      "action_query_norm": 42.139271926879886,
      "action_relation_loss": 0.003373961732722819,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.739943872778298,
      "flow_matching_loss": 0.03611021302640438,
      "grad_norm": 0.1084761992096901,
      "learning_rate": 6.72787027139348e-05,
      "loss": 0.0397,
      "relation_loss": 0.0035954131744802,
      "selected_pairs": 27527.7,
      "state_query_norm": 43.43634719848633,
      "state_relation_loss": 0.000221451444667764,
      "step": 12550,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0035954131744802,
      "weighted_relation_loss": 0.0035954131744802
    },
    {
      "action_query_norm": 42.099867248535155,
      "action_relation_loss": 0.0033215171191841365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.749298409728718,
      "flow_matching_loss": 0.03610061276704073,
      "grad_norm": 0.15541520714759827,
      "learning_rate": 6.72269721603747e-05,
      "loss": 0.0396,
      "relation_loss": 0.003530876338481903,
      "selected_pairs": 26340.6,
      "state_query_norm": 43.427049255371095,
      "state_relation_loss": 0.00020935925422236324,
      "step": 12560,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.003530876338481903,
      "weighted_relation_loss": 0.003530876338481903
    },
    {
      "action_query_norm": 42.13784561157227,
      "action_relation_loss": 0.0032907841727137566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.75865294667914,
      "flow_matching_loss": 0.0330638837069273,
      "grad_norm": 0.10872554779052734,
      "learning_rate": 6.717522067441904e-05,
      "loss": 0.0366,
      "relation_loss": 0.0035239846911281347,
      "selected_pairs": 27059.0,
      "state_query_norm": 43.4247241973877,
      "state_relation_loss": 0.00023320053733186797,
      "step": 12570,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0035239846911281347,
      "weighted_relation_loss": 0.0035239846911281347
    },
    {
      "action_query_norm": 42.14399452209473,
      "action_relation_loss": 0.0033002663170918823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.76800748362956,
      "flow_matching_loss": 0.0357510257512331,
      "grad_norm": 0.11451537907123566,
      "learning_rate": 6.712344831895074e-05,
      "loss": 0.0393,
      "relation_loss": 0.0035325733944773672,
      "selected_pairs": 27197.7,
      "state_query_norm": 43.43232955932617,
      "state_relation_loss": 0.000232307065743953,
      "step": 12580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0035325733944773672,
      "weighted_relation_loss": 0.0035325733944773672
    },
    {
      "action_query_norm": 42.12469902038574,
      "action_relation_loss": 0.003286406840197742,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.77736202057998,
      "flow_matching_loss": 0.035647550225257875,
      "grad_norm": 0.10018999129533768,
      "learning_rate": 6.707165515687811e-05,
      "loss": 0.0391,
      "relation_loss": 0.00350063000805676,
      "selected_pairs": 26733.8,
      "state_query_norm": 43.41707878112793,
      "state_relation_loss": 0.00021422317367978395,
      "step": 12590,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00350063000805676,
      "weighted_relation_loss": 0.00350063000805676
    },
    {
      "action_query_norm": 42.11401443481445,
      "action_relation_loss": 0.0032594076823443175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.786716557530402,
      "flow_matching_loss": 0.03258440177887678,
      "grad_norm": 0.14452065527439117,
      "learning_rate": 6.70198412511347e-05,
      "loss": 0.0361,
      "relation_loss": 0.003482065978460014,
      "selected_pairs": 26995.4,
      "state_query_norm": 43.417243576049806,
      "state_relation_loss": 0.0002226582946605049,
      "step": 12600,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003482065978460014,
      "weighted_relation_loss": 0.003482065978460014
    },
    {
      "action_query_norm": 42.089165115356444,
      "action_relation_loss": 0.0032584309112280605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.796071094480823,
      "flow_matching_loss": 0.035713966190814975,
      "grad_norm": 0.13447514176368713,
      "learning_rate": 6.69680066646793e-05,
      "loss": 0.0392,
      "relation_loss": 0.003467001160606742,
      "selected_pairs": 26547.1,
      "state_query_norm": 43.42060813903809,
      "state_relation_loss": 0.0002085702624754049,
      "step": 12610,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003467001160606742,
      "weighted_relation_loss": 0.003467001160606742
    },
    {
      "action_query_norm": 42.07159614562988,
      "action_relation_loss": 0.0034908968722447754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.805425631431245,
      "flow_matching_loss": 0.03516863565891981,
      "grad_norm": 0.18321366608142853,
      "learning_rate": 6.691615146049583e-05,
      "loss": 0.0389,
      "relation_loss": 0.003696297248825431,
      "selected_pairs": 27256.1,
      "state_query_norm": 43.423404693603516,
      "state_relation_loss": 0.00020540037512546406,
      "step": 12620,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.003696297248825431,
      "weighted_relation_loss": 0.003696297248825431
    },
    {
      "action_query_norm": 42.11720924377441,
      "action_relation_loss": 0.0032715510576963426,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.814780168381665,
      "flow_matching_loss": 0.03669705949723721,
      "grad_norm": 0.09841006994247437,
      "learning_rate": 6.686427570159324e-05,
      "loss": 0.0402,
      "relation_loss": 0.003482982935383916,
      "selected_pairs": 26552.8,
      "state_query_norm": 43.416984939575194,
      "state_relation_loss": 0.0002114318616804667,
      "step": 12630,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003482982935383916,
      "weighted_relation_loss": 0.003482982935383916
    },
    {
      "action_query_norm": 42.1185115814209,
      "action_relation_loss": 0.0034717173082754018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.824134705332087,
      "flow_matching_loss": 0.03486535605043173,
      "grad_norm": 0.14246077835559845,
      "learning_rate": 6.681237945100548e-05,
      "loss": 0.0385,
      "relation_loss": 0.003682415303774178,
      "selected_pairs": 26994.9,
      "state_query_norm": 43.40112190246582,
      "state_relation_loss": 0.0002106979678501375,
      "step": 12640,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003682415303774178,
      "weighted_relation_loss": 0.003682415303774178
    },
    {
      "action_query_norm": 42.06296234130859,
      "action_relation_loss": 0.00320038516074419,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.833489242282507,
      "flow_matching_loss": 0.03619080148637295,
      "grad_norm": 0.18118497729301453,
      "learning_rate": 6.676046277179139e-05,
      "loss": 0.0396,
      "relation_loss": 0.0033979439176619053,
      "selected_pairs": 26810.6,
      "state_query_norm": 43.41137161254883,
      "state_relation_loss": 0.00019755875400733202,
      "step": 12650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0033979439176619053,
      "weighted_relation_loss": 0.0033979439176619053
    },
    {
      "action_query_norm": 42.10098075866699,
      "action_relation_loss": 0.003268791502341628,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.842843779232927,
      "flow_matching_loss": 0.03744703866541386,
      "grad_norm": 0.11715467274188995,
      "learning_rate": 6.670852572703462e-05,
      "loss": 0.0409,
      "relation_loss": 0.003497178968973458,
      "selected_pairs": 27151.7,
      "state_query_norm": 43.402547454833986,
      "state_relation_loss": 0.0002283874637214467,
      "step": 12660,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.003497178968973458,
      "weighted_relation_loss": 0.003497178968973458
    },
    {
      "action_query_norm": 42.09425086975098,
      "action_relation_loss": 0.003417373262345791,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.85219831618335,
      "flow_matching_loss": 0.03763245977461338,
      "grad_norm": 0.15791797637939453,
      "learning_rate": 6.665656837984359e-05,
      "loss": 0.0413,
      "relation_loss": 0.0036306986585259438,
      "selected_pairs": 26317.5,
      "state_query_norm": 43.39705543518066,
      "state_relation_loss": 0.00021332540054572747,
      "step": 12670,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0036306986585259438,
      "weighted_relation_loss": 0.0036306986585259438
    },
    {
      "action_query_norm": 42.06493072509765,
      "action_relation_loss": 0.0033443977357819677,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.86155285313377,
      "flow_matching_loss": 0.035151008516550064,
      "grad_norm": 0.14945101737976074,
      "learning_rate": 6.660459079335135e-05,
      "loss": 0.0387,
      "relation_loss": 0.003568142093718052,
      "selected_pairs": 26664.8,
      "state_query_norm": 43.39111175537109,
      "state_relation_loss": 0.00022374436230165884,
      "step": 12680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.003568142093718052,
      "weighted_relation_loss": 0.003568142093718052
    },
    {
      "action_query_norm": 42.07370300292969,
      "action_relation_loss": 0.0035307564539834857,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.870907390084191,
      "flow_matching_loss": 0.03522318638861179,
      "grad_norm": 0.11899380385875702,
      "learning_rate": 6.655259303071558e-05,
      "loss": 0.039,
      "relation_loss": 0.0037455454235896467,
      "selected_pairs": 26989.2,
      "state_query_norm": 43.3939453125,
      "state_relation_loss": 0.00021478890848811716,
      "step": 12690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0037455454235896467,
      "weighted_relation_loss": 0.0037455454235896467
    },
    {
      "action_query_norm": 42.112248229980466,
      "action_relation_loss": 0.0031870581209659577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.880261927034612,
      "flow_matching_loss": 0.035374893248081206,
      "grad_norm": 0.14075924456119537,
      "learning_rate": 6.650057515511848e-05,
      "loss": 0.0388,
      "relation_loss": 0.003386860410682857,
      "selected_pairs": 27255.2,
      "state_query_norm": 43.407301330566405,
      "state_relation_loss": 0.00019980223587481304,
      "step": 12700,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.003386860410682857,
      "weighted_relation_loss": 0.003386860410682857
    },
    {
      "action_query_norm": 42.04957275390625,
      "action_relation_loss": 0.0033884773962199687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.889616463985032,
      "flow_matching_loss": 0.03720669746398926,
      "grad_norm": 0.11417224258184433,
      "learning_rate": 6.644853722976667e-05,
      "loss": 0.0408,
      "relation_loss": 0.0035873363725841046,
      "selected_pairs": 26874.4,
      "state_query_norm": 43.395621871948244,
      "state_relation_loss": 0.00019885896908817814,
      "step": 12710,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0035873363725841046,
      "weighted_relation_loss": 0.0035873363725841046
    },
    {
      "action_query_norm": 42.08283576965332,
      "action_relation_loss": 0.0032439684029668568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.898971000935454,
      "flow_matching_loss": 0.035357263311743735,
      "grad_norm": 0.12198728322982788,
      "learning_rate": 6.639647931789113e-05,
      "loss": 0.0388,
      "relation_loss": 0.0034452999709174035,
      "selected_pairs": 26895.6,
      "state_query_norm": 43.39553451538086,
      "state_relation_loss": 0.0002013315839576535,
      "step": 12720,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0034452999709174035,
      "weighted_relation_loss": 0.0034452999709174035
    },
    {
      "action_query_norm": 42.07964172363281,
      "action_relation_loss": 0.0033548516919836404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.908325537885874,
      "flow_matching_loss": 0.03628058359026909,
      "grad_norm": 0.10545831173658371,
      "learning_rate": 6.634440148274713e-05,
      "loss": 0.0398,
      "relation_loss": 0.003561158780939877,
      "selected_pairs": 27104.5,
      "state_query_norm": 43.38041801452637,
      "state_relation_loss": 0.00020630710205296054,
      "step": 12730,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.003561158780939877,
      "weighted_relation_loss": 0.003561158780939877
    },
    {
      "action_query_norm": 42.0982063293457,
      "action_relation_loss": 0.0032496941974386575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.917680074836296,
      "flow_matching_loss": 0.03626366425305605,
      "grad_norm": 0.10581057518720627,
      "learning_rate": 6.629230378761415e-05,
      "loss": 0.0397,
      "relation_loss": 0.0034594799857586623,
      "selected_pairs": 26848.5,
      "state_query_norm": 43.395045852661134,
      "state_relation_loss": 0.0002097857679473236,
      "step": 12740,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0034594799857586623,
      "weighted_relation_loss": 0.0034594799857586623
    },
    {
      "action_query_norm": 42.081613540649414,
      "action_relation_loss": 0.003233614703640342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.927034611786716,
      "flow_matching_loss": 0.03411154765635729,
      "grad_norm": 0.09740237891674042,
      "learning_rate": 6.624018629579581e-05,
      "loss": 0.0376,
      "relation_loss": 0.0034446246922016145,
      "selected_pairs": 26568.8,
      "state_query_norm": 43.38403511047363,
      "state_relation_loss": 0.00021101000456837937,
      "step": 12750,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0034446246922016145,
      "weighted_relation_loss": 0.0034446246922016145
    },
    {
      "action_query_norm": 42.09049263000488,
      "action_relation_loss": 0.003326917183585465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.936389148737138,
      "flow_matching_loss": 0.037549202516675,
      "grad_norm": 0.11423471570014954,
      "learning_rate": 6.618804907061976e-05,
      "loss": 0.0411,
      "relation_loss": 0.003524695825763047,
      "selected_pairs": 26992.3,
      "state_query_norm": 43.38533592224121,
      "state_relation_loss": 0.00019777865963988006,
      "step": 12760,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.003524695825763047,
      "weighted_relation_loss": 0.003524695825763047
    },
    {
      "action_query_norm": 42.0783390045166,
      "action_relation_loss": 0.0032315643038600683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.945743685687559,
      "flow_matching_loss": 0.034297749027609827,
      "grad_norm": 0.11897691339254379,
      "learning_rate": 6.613589217543766e-05,
      "loss": 0.0377,
      "relation_loss": 0.0034473280422389506,
      "selected_pairs": 26709.7,
      "state_query_norm": 43.38562774658203,
      "state_relation_loss": 0.0002157637500204146,
      "step": 12770,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0034473280422389506,
      "weighted_relation_loss": 0.0034473280422389506
    },
    {
      "action_query_norm": 42.093906784057616,
      "action_relation_loss": 0.003161029447801411,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.955098222637979,
      "flow_matching_loss": 0.034375982359051704,
      "grad_norm": 0.10936271399259567,
      "learning_rate": 6.608371567362504e-05,
      "loss": 0.0377,
      "relation_loss": 0.003364280262030661,
      "selected_pairs": 26477.0,
      "state_query_norm": 43.3805118560791,
      "state_relation_loss": 0.00020325079676695169,
      "step": 12780,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.003364280262030661,
      "weighted_relation_loss": 0.003364280262030661
    },
    {
      "action_query_norm": 42.077139663696286,
      "action_relation_loss": 0.00316439897287637,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.9644527595884,
      "flow_matching_loss": 0.03525025900453329,
      "grad_norm": 0.09012307226657867,
      "learning_rate": 6.60315196285813e-05,
      "loss": 0.0386,
      "relation_loss": 0.003355544013902545,
      "selected_pairs": 26526.1,
      "state_query_norm": 43.37594795227051,
      "state_relation_loss": 0.00019114501483272762,
      "step": 12790,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.003355544013902545,
      "weighted_relation_loss": 0.003355544013902545
    },
    {
      "action_query_norm": 42.061435317993165,
      "action_relation_loss": 0.0031819280935451387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.973807296538821,
      "flow_matching_loss": 0.03566908612847328,
      "grad_norm": 0.11817358434200287,
      "learning_rate": 6.597930410372954e-05,
      "loss": 0.0391,
      "relation_loss": 0.00338738108985126,
      "selected_pairs": 26178.8,
      "state_query_norm": 43.371619033813474,
      "state_relation_loss": 0.00020545297884382306,
      "step": 12800,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.00338738108985126,
      "weighted_relation_loss": 0.00338738108985126
    },
    {
      "action_query_norm": 42.066127395629884,
      "action_relation_loss": 0.00308361304923892,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.983161833489243,
      "flow_matching_loss": 0.0350887842476368,
      "grad_norm": 0.1320609748363495,
      "learning_rate": 6.592706916251653e-05,
      "loss": 0.0384,
      "relation_loss": 0.0032824303256347776,
      "selected_pairs": 26595.4,
      "state_query_norm": 43.38021125793457,
      "state_relation_loss": 0.0001988172691198997,
      "step": 12810,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0032824303256347776,
      "weighted_relation_loss": 0.0032824303256347776
    },
    {
      "action_query_norm": 42.08383445739746,
      "action_relation_loss": 0.003136694896966219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 11.992516370439663,
      "flow_matching_loss": 0.03580314517021179,
      "grad_norm": 0.14879296720027924,
      "learning_rate": 6.587481486841267e-05,
      "loss": 0.0391,
      "relation_loss": 0.0033356362488120793,
      "selected_pairs": 27595.8,
      "state_query_norm": 43.38139343261719,
      "state_relation_loss": 0.00019894134748028592,
      "step": 12820,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.0033356362488120793,
      "weighted_relation_loss": 0.0033356362488120793
    },
    {
      "action_query_norm": 42.04866256713867,
      "action_relation_loss": 0.0030959404539316893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.001870907390083,
      "flow_matching_loss": 0.03514289930462837,
      "grad_norm": 0.11402806639671326,
      "learning_rate": 6.582254128491184e-05,
      "loss": 0.0384,
      "relation_loss": 0.00330482660792768,
      "selected_pairs": 24547.5,
      "state_query_norm": 43.37799072265625,
      "state_relation_loss": 0.00020888611470581964,
      "step": 12830,
      "valid_action_samples": 215.1,
      "weighted_auxiliary_loss": 0.00330482660792768,
      "weighted_relation_loss": 0.00330482660792768
    },
    {
      "action_query_norm": 42.04716873168945,
      "action_relation_loss": 0.0031221042620018123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.011225444340505,
      "flow_matching_loss": 0.03295574858784676,
      "grad_norm": 0.13004425168037415,
      "learning_rate": 6.577024847553139e-05,
      "loss": 0.0363,
      "relation_loss": 0.0033320561749860646,
      "selected_pairs": 26880.0,
      "state_query_norm": 43.37228698730469,
      "state_relation_loss": 0.00020995188679080458,
      "step": 12840,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0033320561749860646,
      "weighted_relation_loss": 0.0033320561749860646
    },
    {
      "action_query_norm": 42.050313568115236,
      "action_relation_loss": 0.003190901433117688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.020579981290926,
      "flow_matching_loss": 0.03371278904378414,
      "grad_norm": 0.14309906959533691,
      "learning_rate": 6.571793650381201e-05,
      "loss": 0.0371,
      "relation_loss": 0.0033862671349197625,
      "selected_pairs": 27223.7,
      "state_query_norm": 43.37340774536133,
      "state_relation_loss": 0.00019536567851901053,
      "step": 12850,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0033862671349197625,
      "weighted_relation_loss": 0.0033862671349197625
    },
    {
      "action_query_norm": 42.01391067504883,
      "action_relation_loss": 0.0030988536309450865,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.029934518241348,
      "flow_matching_loss": 0.03574510347098112,
      "grad_norm": 0.14291833341121674,
      "learning_rate": 6.566560543331766e-05,
      "loss": 0.039,
      "relation_loss": 0.003300968883559108,
      "selected_pairs": 26802.0,
      "state_query_norm": 43.36585693359375,
      "state_relation_loss": 0.00020211523806210608,
      "step": 12860,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.003300968883559108,
      "weighted_relation_loss": 0.003300968883559108
    },
    {
      "action_query_norm": 42.083334350585936,
      "action_relation_loss": 0.002893327036872506,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.039289055191768,
      "flow_matching_loss": 0.033046936988830565,
      "grad_norm": 0.1001598984003067,
      "learning_rate": 6.561325532763554e-05,
      "loss": 0.0362,
      "relation_loss": 0.003114529582671821,
      "selected_pairs": 26877.0,
      "state_query_norm": 43.36498870849609,
      "state_relation_loss": 0.0002212025414337404,
      "step": 12870,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.003114529582671821,
      "weighted_relation_loss": 0.003114529582671821
    },
    {
      "action_query_norm": 42.03871803283691,
      "action_relation_loss": 0.002979818731546402,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.04864359214219,
      "flow_matching_loss": 0.03476782720535994,
      "grad_norm": 0.14181570708751678,
      "learning_rate": 6.556088625037597e-05,
      "loss": 0.038,
      "relation_loss": 0.0031822217395529153,
      "selected_pairs": 26921.9,
      "state_query_norm": 43.3635913848877,
      "state_relation_loss": 0.00020240300655132159,
      "step": 12880,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0031822217395529153,
      "weighted_relation_loss": 0.0031822217395529153
    },
    {
      "action_query_norm": 42.04298362731934,
      "action_relation_loss": 0.003023512125946581,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.05799812909261,
      "flow_matching_loss": 0.032402877882122996,
      "grad_norm": 0.14820265769958496,
      "learning_rate": 6.55084982651723e-05,
      "loss": 0.0356,
      "relation_loss": 0.0032207869691774247,
      "selected_pairs": 26688.2,
      "state_query_norm": 43.35374374389649,
      "state_relation_loss": 0.00019727482576854528,
      "step": 12890,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0032207869691774247,
      "weighted_relation_loss": 0.0032207869691774247
    },
    {
      "action_query_norm": 42.046190643310545,
      "action_relation_loss": 0.003085822588764131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.06735266604303,
      "flow_matching_loss": 0.034634676203131674,
      "grad_norm": 0.10619772225618362,
      "learning_rate": 6.54560914356809e-05,
      "loss": 0.0379,
      "relation_loss": 0.003276353795081377,
      "selected_pairs": 27220.9,
      "state_query_norm": 43.36613006591797,
      "state_relation_loss": 0.000190531209227629,
      "step": 12900,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.003276353795081377,
      "weighted_relation_loss": 0.003276353795081377
    },
    {
      "action_query_norm": 42.026324081420896,
      "action_relation_loss": 0.003068832284770906,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.076707202993452,
      "flow_matching_loss": 0.0336170980706811,
      "grad_norm": 0.12278197705745697,
      "learning_rate": 6.540366582558097e-05,
      "loss": 0.0369,
      "relation_loss": 0.0032732379622757435,
      "selected_pairs": 26476.0,
      "state_query_norm": 43.356804275512694,
      "state_relation_loss": 0.00020440566004253924,
      "step": 12910,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0032732379622757435,
      "weighted_relation_loss": 0.0032732379622757435
    },
    {
      "action_query_norm": 42.00962409973145,
      "action_relation_loss": 0.0032684650970622895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.086061739943872,
      "flow_matching_loss": 0.037249767780303956,
      "grad_norm": 0.14539669454097748,
      "learning_rate": 6.53512214985746e-05,
      "loss": 0.0407,
      "relation_loss": 0.003457650658674538,
      "selected_pairs": 26573.0,
      "state_query_norm": 43.36070518493652,
      "state_relation_loss": 0.00018918558344012127,
      "step": 12920,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.003457650658674538,
      "weighted_relation_loss": 0.003457650658674538
    },
    {
      "action_query_norm": 42.04133911132813,
      "action_relation_loss": 0.0029819006100296973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.095416276894294,
      "flow_matching_loss": 0.034303052723407744,
      "grad_norm": 0.1127200648188591,
      "learning_rate": 6.529875851838658e-05,
      "loss": 0.0375,
      "relation_loss": 0.003178106411360204,
      "selected_pairs": 27602.7,
      "state_query_norm": 43.350215911865234,
      "state_relation_loss": 0.00019620577513705938,
      "step": 12930,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.003178106411360204,
      "weighted_relation_loss": 0.003178106411360204
    },
    {
      "action_query_norm": 42.01273422241211,
      "action_relation_loss": 0.0031150704715400936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.104770813844715,
      "flow_matching_loss": 0.03480273336172104,
      "grad_norm": 0.13901492953300476,
      "learning_rate": 6.52462769487644e-05,
      "loss": 0.0381,
      "relation_loss": 0.003314757253974676,
      "selected_pairs": 27436.1,
      "state_query_norm": 43.34239501953125,
      "state_relation_loss": 0.0001996867489651777,
      "step": 12940,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.003314757253974676,
      "weighted_relation_loss": 0.003314757253974676
    },
    {
      "action_query_norm": 42.00590515136719,
      "action_relation_loss": 0.002969392924569547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.114125350795135,
      "flow_matching_loss": 0.03248721435666084,
      "grad_norm": 0.11887712776660919,
      "learning_rate": 6.519377685347807e-05,
      "loss": 0.0356,
      "relation_loss": 0.003160699619911611,
      "selected_pairs": 27340.1,
      "state_query_norm": 43.363122177124026,
      "state_relation_loss": 0.00019130669097648932,
      "step": 12950,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.003160699619911611,
      "weighted_relation_loss": 0.003160699619911611
    },
    {
      "action_query_norm": 41.96333312988281,
      "action_relation_loss": 0.003198953974060714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.123479887745557,
      "flow_matching_loss": 0.03252903632819652,
      "grad_norm": 0.10666616261005402,
      "learning_rate": 6.514125829632021e-05,
      "loss": 0.0359,
      "relation_loss": 0.003387809474952519,
      "selected_pairs": 26253.3,
      "state_query_norm": 43.351890563964844,
      "state_relation_loss": 0.00018885553436120972,
      "step": 12960,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.003387809474952519,
      "weighted_relation_loss": 0.003387809474952519
    },
    {
      "action_query_norm": 42.00170707702637,
      "action_relation_loss": 0.0028604473220184444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.132834424695977,
      "flow_matching_loss": 0.033934765681624415,
      "grad_norm": 0.12301512807607651,
      "learning_rate": 6.508872134110577e-05,
      "loss": 0.037,
      "relation_loss": 0.0030592439230531453,
      "selected_pairs": 26734.1,
      "state_query_norm": 43.352596282958984,
      "state_relation_loss": 0.00019879662286257372,
      "step": 12970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0030592439230531453,
      "weighted_relation_loss": 0.0030592439230531453
    },
    {
      "action_query_norm": 41.961161041259764,
      "action_relation_loss": 0.0030927300453186036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.142188961646399,
      "flow_matching_loss": 0.03441220819950104,
      "grad_norm": 0.10151702910661697,
      "learning_rate": 6.503616605167213e-05,
      "loss": 0.0377,
      "relation_loss": 0.003293701191432774,
      "selected_pairs": 27276.7,
      "state_query_norm": 43.344367599487306,
      "state_relation_loss": 0.00020097113301744684,
      "step": 12980,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.003293701191432774,
      "weighted_relation_loss": 0.003293701191432774
    },
    {
      "action_query_norm": 42.014617919921875,
      "action_relation_loss": 0.0029540361603721975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.15154349859682,
      "flow_matching_loss": 0.03293301034718752,
      "grad_norm": 0.10673095285892487,
      "learning_rate": 6.498359249187893e-05,
      "loss": 0.0361,
      "relation_loss": 0.0031520267017185687,
      "selected_pairs": 26964.9,
      "state_query_norm": 43.3379581451416,
      "state_relation_loss": 0.00019799052533926443,
      "step": 12990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0031520267017185687,
      "weighted_relation_loss": 0.0031520267017185687
    },
    {
      "action_query_norm": 42.0279052734375,
      "action_relation_loss": 0.002908687829039991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.160898035547241,
      "flow_matching_loss": 0.03397692572325468,
      "grad_norm": 0.10573019832372665,
      "learning_rate": 6.493100072560799e-05,
      "loss": 0.0371,
      "relation_loss": 0.0031020934926345944,
      "selected_pairs": 27036.2,
      "state_query_norm": 43.331865310668945,
      "state_relation_loss": 0.00019340563158039005,
      "step": 13000,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0031020934926345944,
      "weighted_relation_loss": 0.0031020934926345944
    },
    {
      "action_query_norm": 41.981890487670896,
      "action_relation_loss": 0.0029812086606398223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.170252572497661,
      "flow_matching_loss": 0.032801967300474646,
      "grad_norm": 0.11215309053659439,
      "learning_rate": 6.487839081676327e-05,
      "loss": 0.036,
      "relation_loss": 0.003185033146291971,
      "selected_pairs": 27311.5,
      "state_query_norm": 43.329850387573245,
      "state_relation_loss": 0.0002038244769209996,
      "step": 13010,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.003185033146291971,
      "weighted_relation_loss": 0.003185033146291971
    },
    {
      "action_query_norm": 42.01356086730957,
      "action_relation_loss": 0.003017530124634504,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.179607109448082,
      "flow_matching_loss": 0.03306339010596275,
      "grad_norm": 0.09973546117544174,
      "learning_rate": 6.482576282927076e-05,
      "loss": 0.0363,
      "relation_loss": 0.0032194812316447496,
      "selected_pairs": 26485.3,
      "state_query_norm": 43.33837127685547,
      "state_relation_loss": 0.00020195110846543685,
      "step": 13020,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0032194812316447496,
      "weighted_relation_loss": 0.0032194812316447496
    },
    {
      "action_query_norm": 42.00211181640625,
      "action_relation_loss": 0.0028858553851023316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.188961646398504,
      "flow_matching_loss": 0.032318707555532455,
      "grad_norm": 0.10658583790063858,
      "learning_rate": 6.477311682707845e-05,
      "loss": 0.0354,
      "relation_loss": 0.003073032689280808,
      "selected_pairs": 26778.8,
      "state_query_norm": 43.34107360839844,
      "state_relation_loss": 0.00018717730272328482,
      "step": 13030,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003073032689280808,
      "weighted_relation_loss": 0.003073032689280808
    },
    {
      "action_query_norm": 41.98744506835938,
      "action_relation_loss": 0.0029449273832142355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.198316183348924,
      "flow_matching_loss": 0.0343806080520153,
      "grad_norm": 0.13801437616348267,
      "learning_rate": 6.472045287415616e-05,
      "loss": 0.0375,
      "relation_loss": 0.0031410392140969632,
      "selected_pairs": 26755.5,
      "state_query_norm": 43.34403648376465,
      "state_relation_loss": 0.0001961118294275366,
      "step": 13040,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0031410392140969632,
      "weighted_relation_loss": 0.0031410392140969632
    },
    {
      "action_query_norm": 41.975955200195315,
      "action_relation_loss": 0.0028778354404494166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.207670720299346,
      "flow_matching_loss": 0.0319796497002244,
      "grad_norm": 0.1225477010011673,
      "learning_rate": 6.466777103449558e-05,
      "loss": 0.035,
      "relation_loss": 0.0030649506486952307,
      "selected_pairs": 27207.5,
      "state_query_norm": 43.327941513061525,
      "state_relation_loss": 0.00018711523152887822,
      "step": 13050,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0030649506486952307,
      "weighted_relation_loss": 0.0030649506486952307
    },
    {
      "action_query_norm": 42.007928466796876,
      "action_relation_loss": 0.002881701779551804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.217025257249766,
      "flow_matching_loss": 0.032753498293459415,
      "grad_norm": 0.10961809754371643,
      "learning_rate": 6.461507137211012e-05,
      "loss": 0.0358,
      "relation_loss": 0.003076604660600424,
      "selected_pairs": 27175.2,
      "state_query_norm": 43.33794937133789,
      "state_relation_loss": 0.0001949029217939824,
      "step": 13060,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.003076604660600424,
      "weighted_relation_loss": 0.003076604660600424
    },
    {
      "action_query_norm": 42.011256790161134,
      "action_relation_loss": 0.0029142208863049746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.226379794200188,
      "flow_matching_loss": 0.03407270666211844,
      "grad_norm": 0.13027186691761017,
      "learning_rate": 6.456235395103483e-05,
      "loss": 0.0372,
      "relation_loss": 0.0031133118318393826,
      "selected_pairs": 27244.8,
      "state_query_norm": 43.33836059570312,
      "state_relation_loss": 0.00019909094116883353,
      "step": 13070,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0031133118318393826,
      "weighted_relation_loss": 0.0031133118318393826
    },
    {
      "action_query_norm": 41.953257369995114,
      "action_relation_loss": 0.002952860062941909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.235734331150608,
      "flow_matching_loss": 0.03292595576494932,
      "grad_norm": 0.09938924759626389,
      "learning_rate": 6.450961883532634e-05,
      "loss": 0.0361,
      "relation_loss": 0.003153616189956665,
      "selected_pairs": 26650.8,
      "state_query_norm": 43.33512573242187,
      "state_relation_loss": 0.00020075615320820362,
      "step": 13080,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.003153616189956665,
      "weighted_relation_loss": 0.003153616189956665
    },
    {
      "action_query_norm": 41.954030227661136,
      "action_relation_loss": 0.0030038283206522466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.245088868101028,
      "flow_matching_loss": 0.03414360173046589,
      "grad_norm": 0.12156412750482559,
      "learning_rate": 6.445686608906283e-05,
      "loss": 0.0373,
      "relation_loss": 0.0032057388219982387,
      "selected_pairs": 26325.9,
      "state_query_norm": 43.32361831665039,
      "state_relation_loss": 0.00020191053481539712,
      "step": 13090,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0032057388219982387,
      "weighted_relation_loss": 0.0032057388219982387
    },
    {
      "action_query_norm": 41.96389541625977,
      "action_relation_loss": 0.00301289155613631,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.25444340505145,
      "flow_matching_loss": 0.03402236346155405,
      "grad_norm": 0.11320826411247253,
      "learning_rate": 6.44040957763438e-05,
      "loss": 0.0372,
      "relation_loss": 0.0031990136252716185,
      "selected_pairs": 27033.9,
      "state_query_norm": 43.32292366027832,
      "state_relation_loss": 0.00018612209387356416,
      "step": 13100,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0031990136252716185,
      "weighted_relation_loss": 0.0031990136252716185
    },
    {
      "action_query_norm": 41.99753837585449,
      "action_relation_loss": 0.0029820204479619862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.26379794200187,
      "flow_matching_loss": 0.03459729105234146,
      "grad_norm": 0.11916480958461761,
      "learning_rate": 6.435130796129018e-05,
      "loss": 0.0378,
      "relation_loss": 0.0031749474117532374,
      "selected_pairs": 26436.6,
      "state_query_norm": 43.3400318145752,
      "state_relation_loss": 0.0001929269390529953,
      "step": 13110,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0031749474117532374,
      "weighted_relation_loss": 0.0031749474117532374
    },
    {
      "action_query_norm": 41.980713272094725,
      "action_relation_loss": 0.0029620779678225517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.273152478952293,
      "flow_matching_loss": 0.03542806431651115,
      "grad_norm": 0.11734075099229813,
      "learning_rate": 6.429850270804416e-05,
      "loss": 0.0386,
      "relation_loss": 0.0031668080948293207,
      "selected_pairs": 26513.1,
      "state_query_norm": 43.321742630004884,
      "state_relation_loss": 0.00020473011973081156,
      "step": 13120,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0031668080948293207,
      "weighted_relation_loss": 0.0031668080948293207
    },
    {
      "action_query_norm": 41.99038848876953,
      "action_relation_loss": 0.0030551114585250618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.282507015902713,
      "flow_matching_loss": 0.03469827063381672,
      "grad_norm": 0.12794241309165955,
      "learning_rate": 6.424568008076909e-05,
      "loss": 0.038,
      "relation_loss": 0.0032577066449448466,
      "selected_pairs": 27102.7,
      "state_query_norm": 43.32226333618164,
      "state_relation_loss": 0.000202595216978807,
      "step": 13130,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0032577066449448466,
      "weighted_relation_loss": 0.0032577066449448466
    },
    {
      "action_query_norm": 41.970523071289065,
      "action_relation_loss": 0.0028120413888245823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.291861552853133,
      "flow_matching_loss": 0.03245031703263521,
      "grad_norm": 0.11805905401706696,
      "learning_rate": 6.419284014364943e-05,
      "loss": 0.0355,
      "relation_loss": 0.0030132144456729293,
      "selected_pairs": 26779.5,
      "state_query_norm": 43.31773834228515,
      "state_relation_loss": 0.0002011730946833268,
      "step": 13140,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0030132144456729293,
      "weighted_relation_loss": 0.0030132144456729293
    },
    {
      "action_query_norm": 41.968004608154295,
      "action_relation_loss": 0.0029705721652135255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.301216089803555,
      "flow_matching_loss": 0.03323113154619932,
      "grad_norm": 0.10220222920179367,
      "learning_rate": 6.41399829608907e-05,
      "loss": 0.0364,
      "relation_loss": 0.0031597331864759324,
      "selected_pairs": 26763.6,
      "state_query_norm": 43.3296012878418,
      "state_relation_loss": 0.00018916102417279034,
      "step": 13150,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0031597331864759324,
      "weighted_relation_loss": 0.0031597331864759324
    },
    {
      "action_query_norm": 41.93251533508301,
      "action_relation_loss": 0.003016302525065839,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.310570626753975,
      "flow_matching_loss": 0.034129302762448785,
      "grad_norm": 0.13460874557495117,
      "learning_rate": 6.408710859671938e-05,
      "loss": 0.0373,
      "relation_loss": 0.003206933080218732,
      "selected_pairs": 27382.1,
      "state_query_norm": 43.32910385131836,
      "state_relation_loss": 0.000190630518773105,
      "step": 13160,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.003206933080218732,
      "weighted_relation_loss": 0.003206933080218732
    },
    {
      "action_query_norm": 41.94967918395996,
      "action_relation_loss": 0.003145502251572907,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.319925163704397,
      "flow_matching_loss": 0.03533425536006689,
      "grad_norm": 0.13616247475147247,
      "learning_rate": 6.403421711538278e-05,
      "loss": 0.0387,
      "relation_loss": 0.003328917850740254,
      "selected_pairs": 27035.4,
      "state_query_norm": 43.31529083251953,
      "state_relation_loss": 0.00018341561954002828,
      "step": 13170,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003328917850740254,
      "weighted_relation_loss": 0.003328917850740254
    },
    {
      "action_query_norm": 41.94817886352539,
      "action_relation_loss": 0.0029600247740745544,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.329279700654817,
      "flow_matching_loss": 0.036843419820070264,
      "grad_norm": 0.1033196821808815,
      "learning_rate": 6.398130858114903e-05,
      "loss": 0.04,
      "relation_loss": 0.0031516788294538856,
      "selected_pairs": 26919.8,
      "state_query_norm": 43.31308288574219,
      "state_relation_loss": 0.0001916540917591192,
      "step": 13180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0031516788294538856,
      "weighted_relation_loss": 0.0031516788294538856
    },
    {
      "action_query_norm": 41.97570381164551,
      "action_relation_loss": 0.0028692334657534958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.33863423760524,
      "flow_matching_loss": 0.03450008239597082,
      "grad_norm": 0.11780206114053726,
      "learning_rate": 6.392838305830701e-05,
      "loss": 0.0376,
      "relation_loss": 0.00307121928781271,
      "selected_pairs": 26835.9,
      "state_query_norm": 43.304320907592775,
      "state_relation_loss": 0.00020198582496959715,
      "step": 13190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.00307121928781271,
      "weighted_relation_loss": 0.00307121928781271
    },
    {
      "action_query_norm": 41.92032585144043,
      "action_relation_loss": 0.0030541214160621165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.34798877455566,
      "flow_matching_loss": 0.03348560817539692,
      "grad_norm": 0.1025887206196785,
      "learning_rate": 6.387544061116621e-05,
      "loss": 0.0367,
      "relation_loss": 0.0032472912222146987,
      "selected_pairs": 26901.6,
      "state_query_norm": 43.31240577697754,
      "state_relation_loss": 0.0001931698323460296,
      "step": 13200,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0032472912222146987,
      "weighted_relation_loss": 0.0032472912222146987
    },
    {
      "action_query_norm": 41.89239234924317,
      "action_relation_loss": 0.00302297945600003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.35734331150608,
      "flow_matching_loss": 0.03267549406737089,
      "grad_norm": 0.12010189890861511,
      "learning_rate": 6.382248130405671e-05,
      "loss": 0.0359,
      "relation_loss": 0.003224602760747075,
      "selected_pairs": 26781.2,
      "state_query_norm": 43.295893859863284,
      "state_relation_loss": 0.00020162330765742807,
      "step": 13210,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003224602760747075,
      "weighted_relation_loss": 0.003224602760747075
    },
    {
      "action_query_norm": 41.91617965698242,
      "action_relation_loss": 0.003093109605833888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.366697848456502,
      "flow_matching_loss": 0.033379473350942136,
      "grad_norm": 0.10182957351207733,
      "learning_rate": 6.376950520132905e-05,
      "loss": 0.0367,
      "relation_loss": 0.0033016825793311,
      "selected_pairs": 26045.0,
      "state_query_norm": 43.306893539428714,
      "state_relation_loss": 0.0002085729662212543,
      "step": 13220,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0033016825793311,
      "weighted_relation_loss": 0.0033016825793311
    },
    {
      "action_query_norm": 41.898946380615236,
      "action_relation_loss": 0.002925509866327047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.376052385406922,
      "flow_matching_loss": 0.03360453099012375,
      "grad_norm": 0.11867193877696991,
      "learning_rate": 6.371651236735418e-05,
      "loss": 0.0367,
      "relation_loss": 0.003130483045242727,
      "selected_pairs": 27433.1,
      "state_query_norm": 43.29015274047852,
      "state_relation_loss": 0.00020497318328125403,
      "step": 13230,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.003130483045242727,
      "weighted_relation_loss": 0.003130483045242727
    },
    {
      "action_query_norm": 41.93071670532227,
      "action_relation_loss": 0.0029353083577007056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.385406922357344,
      "flow_matching_loss": 0.032769221998751166,
      "grad_norm": 0.10601264983415604,
      "learning_rate": 6.366350286652341e-05,
      "loss": 0.0359,
      "relation_loss": 0.0031400663545355203,
      "selected_pairs": 26117.2,
      "state_query_norm": 43.30243453979492,
      "state_relation_loss": 0.0002047579735517502,
      "step": 13240,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.0031400663545355203,
      "weighted_relation_loss": 0.0031400663545355203
    },
    {
      "action_query_norm": 41.9407241821289,
      "action_relation_loss": 0.0028477767948061226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.394761459307764,
      "flow_matching_loss": 0.0342065641656518,
      "grad_norm": 0.09010843932628632,
      "learning_rate": 6.361047676324827e-05,
      "loss": 0.0373,
      "relation_loss": 0.0030528426403179766,
      "selected_pairs": 26666.4,
      "state_query_norm": 43.30059928894043,
      "state_relation_loss": 0.0002050658586085774,
      "step": 13250,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030528426403179766,
      "weighted_relation_loss": 0.0030528426403179766
    },
    {
      "action_query_norm": 41.91566963195801,
      "action_relation_loss": 0.003065446624532342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.404115996258184,
      "flow_matching_loss": 0.03256406635046005,
      "grad_norm": 0.1041155755519867,
      "learning_rate": 6.355743412196047e-05,
      "loss": 0.0358,
      "relation_loss": 0.003254598123021424,
      "selected_pairs": 26467.0,
      "state_query_norm": 43.29671974182129,
      "state_relation_loss": 0.00018915150285465644,
      "step": 13260,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.003254598123021424,
      "weighted_relation_loss": 0.003254598123021424
    },
    {
      "action_query_norm": 41.92751045227051,
      "action_relation_loss": 0.002995756780728698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.413470533208606,
      "flow_matching_loss": 0.03344463277608156,
      "grad_norm": 0.10944859683513641,
      "learning_rate": 6.350437500711184e-05,
      "loss": 0.0366,
      "relation_loss": 0.0031917183427140118,
      "selected_pairs": 26553.9,
      "state_query_norm": 43.29373359680176,
      "state_relation_loss": 0.00019596156780608,
      "step": 13270,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0031917183427140118,
      "weighted_relation_loss": 0.0031917183427140118
    },
    {
      "action_query_norm": 41.90259056091308,
      "action_relation_loss": 0.003034013882279396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.422825070159027,
      "flow_matching_loss": 0.032311699353158474,
      "grad_norm": 0.09716824442148209,
      "learning_rate": 6.345129948317418e-05,
      "loss": 0.0355,
      "relation_loss": 0.0032311022980138658,
      "selected_pairs": 27041.1,
      "state_query_norm": 43.28889503479004,
      "state_relation_loss": 0.00019708845065906643,
      "step": 13280,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0032311022980138658,
      "weighted_relation_loss": 0.0032311022980138658
    },
    {
      "action_query_norm": 41.86978340148926,
      "action_relation_loss": 0.0030159737449139356,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.432179607109449,
      "flow_matching_loss": 0.032954678125679496,
      "grad_norm": 0.11297447234392166,
      "learning_rate": 6.33982076146393e-05,
      "loss": 0.0362,
      "relation_loss": 0.003209255775436759,
      "selected_pairs": 26554.9,
      "state_query_norm": 43.2890811920166,
      "state_relation_loss": 0.00019328202906763182,
      "step": 13290,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.003209255775436759,
      "weighted_relation_loss": 0.003209255775436759
    },
    {
      "action_query_norm": 41.866848373413085,
      "action_relation_loss": 0.003042169287800789,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.441534144059869,
      "flow_matching_loss": 0.03332643546164036,
      "grad_norm": 0.11318192631006241,
      "learning_rate": 6.334509946601879e-05,
      "loss": 0.0366,
      "relation_loss": 0.003238321514800191,
      "selected_pairs": 27034.5,
      "state_query_norm": 43.29000968933106,
      "state_relation_loss": 0.00019615225173765793,
      "step": 13300,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.003238321514800191,
      "weighted_relation_loss": 0.003238321514800191
    },
    {
      "action_query_norm": 41.849219131469724,
      "action_relation_loss": 0.0031022023409605025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.45088868101029,
      "flow_matching_loss": 0.0341860955581069,
      "grad_norm": 0.09534257650375366,
      "learning_rate": 6.329197510184406e-05,
      "loss": 0.0375,
      "relation_loss": 0.003282605530694127,
      "selected_pairs": 26698.5,
      "state_query_norm": 43.29604797363281,
      "state_relation_loss": 0.00018040317227132618,
      "step": 13310,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.003282605530694127,
      "weighted_relation_loss": 0.003282605530694127
    },
    {
      "action_query_norm": 41.91279602050781,
      "action_relation_loss": 0.002999415062367916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.460243217960711,
      "flow_matching_loss": 0.03491746559739113,
      "grad_norm": 0.11335205286741257,
      "learning_rate": 6.323883458666624e-05,
      "loss": 0.0381,
      "relation_loss": 0.0031960950000211596,
      "selected_pairs": 26439.3,
      "state_query_norm": 43.2944034576416,
      "state_relation_loss": 0.00019667996384669094,
      "step": 13320,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0031960950000211596,
      "weighted_relation_loss": 0.0031960950000211596
    },
    {
      "action_query_norm": 41.89822540283203,
      "action_relation_loss": 0.0028760334942489864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.469597754911131,
      "flow_matching_loss": 0.03269033823162317,
      "grad_norm": 0.09444337338209152,
      "learning_rate": 6.318567798505605e-05,
      "loss": 0.0358,
      "relation_loss": 0.00307386617641896,
      "selected_pairs": 26685.4,
      "state_query_norm": 43.29493293762207,
      "state_relation_loss": 0.0001978326865355484,
      "step": 13330,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00307386617641896,
      "weighted_relation_loss": 0.00307386617641896
    },
    {
      "action_query_norm": 41.92545509338379,
      "action_relation_loss": 0.002930720127187669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.478952291861553,
      "flow_matching_loss": 0.03582060281187296,
      "grad_norm": 0.12613148987293243,
      "learning_rate": 6.313250536160378e-05,
      "loss": 0.039,
      "relation_loss": 0.0031340774381533264,
      "selected_pairs": 26621.0,
      "state_query_norm": 43.28831787109375,
      "state_relation_loss": 0.00020335732406238095,
      "step": 13340,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0031340774381533264,
      "weighted_relation_loss": 0.0031340774381533264
    },
    {
      "action_query_norm": 41.90429267883301,
      "action_relation_loss": 0.0028215087717399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.488306828811973,
      "flow_matching_loss": 0.034691262245178225,
      "grad_norm": 0.12997838854789734,
      "learning_rate": 6.307931678091917e-05,
      "loss": 0.0377,
      "relation_loss": 0.003013942064717412,
      "selected_pairs": 27183.3,
      "state_query_norm": 43.277807998657224,
      "state_relation_loss": 0.0001924332813359797,
      "step": 13350,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.003013942064717412,
      "weighted_relation_loss": 0.003013942064717412
    },
    {
      "action_query_norm": 41.8583740234375,
      "action_relation_loss": 0.0029031962621957064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.497661365762395,
      "flow_matching_loss": 0.033163034729659556,
      "grad_norm": 0.10633807629346848,
      "learning_rate": 6.302611230763138e-05,
      "loss": 0.0363,
      "relation_loss": 0.0030976733658462765,
      "selected_pairs": 26871.9,
      "state_query_norm": 43.28503150939942,
      "state_relation_loss": 0.00019447711383691058,
      "step": 13360,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0030976733658462765,
      "weighted_relation_loss": 0.0030976733658462765
    },
    {
      "action_query_norm": 41.88643341064453,
      "action_relation_loss": 0.0029071632772684096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.507015902712816,
      "flow_matching_loss": 0.033706601709127426,
      "grad_norm": 0.10939639806747437,
      "learning_rate": 6.297289200638887e-05,
      "loss": 0.0368,
      "relation_loss": 0.0031117254169657827,
      "selected_pairs": 26944.6,
      "state_query_norm": 43.267642211914065,
      "state_relation_loss": 0.00020456212951103225,
      "step": 13370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0031117254169657827,
      "weighted_relation_loss": 0.0031117254169657827
    },
    {
      "action_query_norm": 41.871853256225585,
      "action_relation_loss": 0.002914038603194058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.516370439663238,
      "flow_matching_loss": 0.032556484639644626,
      "grad_norm": 0.11325214803218842,
      "learning_rate": 6.29196559418593e-05,
      "loss": 0.0357,
      "relation_loss": 0.003102351608686149,
      "selected_pairs": 27368.6,
      "state_query_norm": 43.26554946899414,
      "state_relation_loss": 0.00018831299530575052,
      "step": 13380,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.003102351608686149,
      "weighted_relation_loss": 0.003102351608686149
    },
    {
      "action_query_norm": 41.83406753540039,
      "action_relation_loss": 0.002988861803896725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.525724976613658,
      "flow_matching_loss": 0.03345633409917355,
      "grad_norm": 0.11330179870128632,
      "learning_rate": 6.286640417872951e-05,
      "loss": 0.0366,
      "relation_loss": 0.0031806466868147254,
      "selected_pairs": 26754.9,
      "state_query_norm": 43.27794494628906,
      "state_relation_loss": 0.00019178489892510698,
      "step": 13390,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0031806466868147254,
      "weighted_relation_loss": 0.0031806466868147254
    },
    {
      "action_query_norm": 41.86244659423828,
      "action_relation_loss": 0.0030071608023718,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.535079513564078,
      "flow_matching_loss": 0.03382675815373659,
      "grad_norm": 0.12354860454797745,
      "learning_rate": 6.281313678170542e-05,
      "loss": 0.037,
      "relation_loss": 0.0032045750878751276,
      "selected_pairs": 26960.9,
      "state_query_norm": 43.27913246154785,
      "state_relation_loss": 0.00019741428404813633,
      "step": 13400,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0032045750878751276,
      "weighted_relation_loss": 0.0032045750878751276
    },
    {
      "action_query_norm": 41.88579139709473,
      "action_relation_loss": 0.002933054510504007,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.5444340505145,
      "flow_matching_loss": 0.035057196766138075,
      "grad_norm": 0.13142099976539612,
      "learning_rate": 6.275985381551194e-05,
      "loss": 0.0382,
      "relation_loss": 0.0031481345416978002,
      "selected_pairs": 27107.3,
      "state_query_norm": 43.27128372192383,
      "state_relation_loss": 0.00021508005011128262,
      "step": 13410,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0031481345416978002,
      "weighted_relation_loss": 0.0031481345416978002
    },
    {
      "action_query_norm": 41.86007537841797,
      "action_relation_loss": 0.0028997532092034815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.55378858746492,
      "flow_matching_loss": 0.03312976080924272,
      "grad_norm": 0.1709231436252594,
      "learning_rate": 6.270655534489292e-05,
      "loss": 0.0362,
      "relation_loss": 0.003084767796099186,
      "selected_pairs": 26902.8,
      "state_query_norm": 43.27973327636719,
      "state_relation_loss": 0.0001850145752541721,
      "step": 13420,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.003084767796099186,
      "weighted_relation_loss": 0.003084767796099186
    },
    {
      "action_query_norm": 41.84692535400391,
      "action_relation_loss": 0.00292145034763962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.563143124415342,
      "flow_matching_loss": 0.03524723369628191,
      "grad_norm": 0.15679430961608887,
      "learning_rate": 6.265324143461098e-05,
      "loss": 0.0384,
      "relation_loss": 0.0031103172339499,
      "selected_pairs": 27383.8,
      "state_query_norm": 43.27015495300293,
      "state_relation_loss": 0.0001888669008621946,
      "step": 13430,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0031103172339499,
      "weighted_relation_loss": 0.0031103172339499
    },
    {
      "action_query_norm": 41.85115547180176,
      "action_relation_loss": 0.0028126528020948173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.572497661365762,
      "flow_matching_loss": 0.034038602374494076,
      "grad_norm": 0.103431336581707,
      "learning_rate": 6.259991214944758e-05,
      "loss": 0.037,
      "relation_loss": 0.0030042525148019196,
      "selected_pairs": 26664.5,
      "state_query_norm": 43.28014869689942,
      "state_relation_loss": 0.00019159968651365488,
      "step": 13440,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030042525148019196,
      "weighted_relation_loss": 0.0030042525148019196
    },
    {
      "action_query_norm": 41.82633056640625,
      "action_relation_loss": 0.003076864150352776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.581852198316183,
      "flow_matching_loss": 0.033009641058743,
      "grad_norm": 0.12960004806518555,
      "learning_rate": 6.254656755420283e-05,
      "loss": 0.0363,
      "relation_loss": 0.0032693895045667887,
      "selected_pairs": 26691.5,
      "state_query_norm": 43.262054443359375,
      "state_relation_loss": 0.00019252534984843807,
      "step": 13450,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0032693895045667887,
      "weighted_relation_loss": 0.0032693895045667887
    },
    {
      "action_query_norm": 41.870100784301755,
      "action_relation_loss": 0.0027395717334002256,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.591206735266605,
      "flow_matching_loss": 0.03509743772447109,
      "grad_norm": 0.13361085951328278,
      "learning_rate": 6.249320771369545e-05,
      "loss": 0.038,
      "relation_loss": 0.0029277759604156017,
      "selected_pairs": 26973.5,
      "state_query_norm": 43.267233276367186,
      "state_relation_loss": 0.00018820423865690828,
      "step": 13460,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0029277759604156017,
      "weighted_relation_loss": 0.0029277759604156017
    },
    {
      "action_query_norm": 41.86938629150391,
      "action_relation_loss": 0.00288407274056226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.600561272217025,
      "flow_matching_loss": 0.03322108201682568,
      "grad_norm": 0.10932095348834991,
      "learning_rate": 6.243983269276263e-05,
      "loss": 0.0363,
      "relation_loss": 0.0030708693666383624,
      "selected_pairs": 26165.8,
      "state_query_norm": 43.27266883850098,
      "state_relation_loss": 0.00018679661880014465,
      "step": 13470,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0030708693666383624,
      "weighted_relation_loss": 0.0030708693666383624
    },
    {
      "action_query_norm": 41.82278938293457,
      "action_relation_loss": 0.002899501868523657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.609915809167447,
      "flow_matching_loss": 0.0346441088244319,
      "grad_norm": 0.1065073236823082,
      "learning_rate": 6.238644255626012e-05,
      "loss": 0.0377,
      "relation_loss": 0.0030802174238488077,
      "selected_pairs": 26592.1,
      "state_query_norm": 43.27066535949707,
      "state_relation_loss": 0.0001807155378628522,
      "step": 13480,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0030802174238488077,
      "weighted_relation_loss": 0.0030802174238488077
    },
    {
      "action_query_norm": 41.861328125,
      "action_relation_loss": 0.0028849321650341155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.619270346117867,
      "flow_matching_loss": 0.03497606571763754,
      "grad_norm": 0.1592881977558136,
      "learning_rate": 6.233303736906193e-05,
      "loss": 0.038,
      "relation_loss": 0.0030730700120329855,
      "selected_pairs": 27061.4,
      "state_query_norm": 43.262345504760745,
      "state_relation_loss": 0.00018813784554367886,
      "step": 13490,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0030730700120329855,
      "weighted_relation_loss": 0.0030730700120329855
    },
    {
      "action_query_norm": 41.820254516601565,
      "action_relation_loss": 0.0029724041698500512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.628624883068287,
      "flow_matching_loss": 0.035082435235381126,
      "grad_norm": 0.10989081859588623,
      "learning_rate": 6.227961719606039e-05,
      "loss": 0.0382,
      "relation_loss": 0.003152004908770323,
      "selected_pairs": 27171.5,
      "state_query_norm": 43.26218872070312,
      "state_relation_loss": 0.00017960076074814425,
      "step": 13500,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.003152004908770323,
      "weighted_relation_loss": 0.003152004908770323
    },
    {
      "action_query_norm": 41.84276924133301,
      "action_relation_loss": 0.0029079233529046176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.63797942001871,
      "flow_matching_loss": 0.036152249574661253,
      "grad_norm": 0.12044554948806763,
      "learning_rate": 6.222618210216608e-05,
      "loss": 0.0392,
      "relation_loss": 0.0030934217851608992,
      "selected_pairs": 26555.0,
      "state_query_norm": 43.26091537475586,
      "state_relation_loss": 0.00018549843953223898,
      "step": 13510,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0030934217851608992,
      "weighted_relation_loss": 0.0030934217851608992
    },
    {
      "action_query_norm": 41.844107437133786,
      "action_relation_loss": 0.00284826192073524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.64733395696913,
      "flow_matching_loss": 0.03332687932997942,
      "grad_norm": 0.101325623691082,
      "learning_rate": 6.217273215230767e-05,
      "loss": 0.0364,
      "relation_loss": 0.0030334654729813337,
      "selected_pairs": 26808.8,
      "state_query_norm": 43.26084671020508,
      "state_relation_loss": 0.00018520351877668873,
      "step": 13520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0030334654729813337,
      "weighted_relation_loss": 0.0030334654729813337
    },
    {
      "action_query_norm": 41.81576042175293,
      "action_relation_loss": 0.002952558547258377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.656688493919551,
      "flow_matching_loss": 0.03372565116733313,
      "grad_norm": 0.11374499648809433,
      "learning_rate": 6.211926741143188e-05,
      "loss": 0.0369,
      "relation_loss": 0.003153354418464005,
      "selected_pairs": 27130.9,
      "state_query_norm": 43.2516544342041,
      "state_relation_loss": 0.00020079587848158552,
      "step": 13530,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.003153354418464005,
      "weighted_relation_loss": 0.003153354418464005
    },
    {
      "action_query_norm": 41.83631629943848,
      "action_relation_loss": 0.002956402930431068,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.666043030869972,
      "flow_matching_loss": 0.03488182481378317,
      "grad_norm": 0.1536562144756317,
      "learning_rate": 6.20657879445034e-05,
      "loss": 0.038,
      "relation_loss": 0.0031427535228431224,
      "selected_pairs": 26985.8,
      "state_query_norm": 43.25921211242676,
      "state_relation_loss": 0.00018635056476341562,
      "step": 13540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0031427535228431224,
      "weighted_relation_loss": 0.0031427535228431224
    },
    {
      "action_query_norm": 41.76440162658692,
      "action_relation_loss": 0.002902032248675823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.675397567820394,
      "flow_matching_loss": 0.033957007341086864,
      "grad_norm": 0.13751523196697235,
      "learning_rate": 6.201229381650485e-05,
      "loss": 0.0371,
      "relation_loss": 0.0030997324734926223,
      "selected_pairs": 26808.9,
      "state_query_norm": 43.25201759338379,
      "state_relation_loss": 0.00019770020444411784,
      "step": 13550,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0030997324734926223,
      "weighted_relation_loss": 0.0030997324734926223
    },
    {
      "action_query_norm": 41.80410957336426,
      "action_relation_loss": 0.0028138609137386083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.684752104770814,
      "flow_matching_loss": 0.032915026694536206,
      "grad_norm": 0.13491666316986084,
      "learning_rate": 6.195878509243661e-05,
      "loss": 0.0359,
      "relation_loss": 0.00300018263515085,
      "selected_pairs": 26989.8,
      "state_query_norm": 43.25928421020508,
      "state_relation_loss": 0.00018632173887453974,
      "step": 13560,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.00300018263515085,
      "weighted_relation_loss": 0.00300018263515085
    },
    {
      "action_query_norm": 41.79946327209473,
      "action_relation_loss": 0.002975516114383936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.694106641721234,
      "flow_matching_loss": 0.0356860026717186,
      "grad_norm": 0.12024487555027008,
      "learning_rate": 6.190526183731685e-05,
      "loss": 0.0388,
      "relation_loss": 0.0031617589527741076,
      "selected_pairs": 27036.0,
      "state_query_norm": 43.25304107666015,
      "state_relation_loss": 0.0001862428500317037,
      "step": 13570,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0031617589527741076,
      "weighted_relation_loss": 0.0031617589527741076
    },
    {
      "action_query_norm": 41.75775680541992,
      "action_relation_loss": 0.003053256147541106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.703461178671656,
      "flow_matching_loss": 0.03303366955369711,
      "grad_norm": 0.11125186085700989,
      "learning_rate": 6.185172411618138e-05,
      "loss": 0.0363,
      "relation_loss": 0.003236802527680993,
      "selected_pairs": 27200.5,
      "state_query_norm": 43.25198173522949,
      "state_relation_loss": 0.00018354642670601607,
      "step": 13580,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.003236802527680993,
      "weighted_relation_loss": 0.003236802527680993
    },
    {
      "action_query_norm": 41.80714149475098,
      "action_relation_loss": 0.0028601220343261956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.712815715622076,
      "flow_matching_loss": 0.032706938497722146,
      "grad_norm": 0.10008835792541504,
      "learning_rate": 6.179817199408355e-05,
      "loss": 0.0357,
      "relation_loss": 0.0030402620090171696,
      "selected_pairs": 26666.7,
      "state_query_norm": 43.2433780670166,
      "state_relation_loss": 0.0001801399834221229,
      "step": 13590,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030402620090171696,
      "weighted_relation_loss": 0.0030402620090171696
    },
    {
      "action_query_norm": 41.726399993896486,
      "action_relation_loss": 0.0031355346320196984,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.722170252572498,
      "flow_matching_loss": 0.0356296231970191,
      "grad_norm": 0.12275291234254837,
      "learning_rate": 6.174460553609426e-05,
      "loss": 0.039,
      "relation_loss": 0.003325915685854852,
      "selected_pairs": 26779.6,
      "state_query_norm": 43.232056045532225,
      "state_relation_loss": 0.00019038103200728074,
      "step": 13600,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.003325915685854852,
      "weighted_relation_loss": 0.003325915685854852
    },
    {
      "action_query_norm": 41.786921691894534,
      "action_relation_loss": 0.002803436736576259,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.731524789522918,
      "flow_matching_loss": 0.03181898333132267,
      "grad_norm": 0.11635058373212814,
      "learning_rate": 6.16910248073018e-05,
      "loss": 0.0348,
      "relation_loss": 0.002978592598810792,
      "selected_pairs": 27351.9,
      "state_query_norm": 43.23227157592773,
      "state_relation_loss": 0.0001751558796968311,
      "step": 13610,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.002978592598810792,
      "weighted_relation_loss": 0.002978592598810792
    },
    {
      "action_query_norm": 41.78145980834961,
      "action_relation_loss": 0.00284370391163975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.74087932647334,
      "flow_matching_loss": 0.032880335487425326,
      "grad_norm": 0.11177191883325577,
      "learning_rate": 6.16374298728118e-05,
      "loss": 0.0359,
      "relation_loss": 0.003030073409900069,
      "selected_pairs": 27082.4,
      "state_query_norm": 43.23311157226563,
      "state_relation_loss": 0.00018636949098436162,
      "step": 13620,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.003030073409900069,
      "weighted_relation_loss": 0.003030073409900069
    },
    {
      "action_query_norm": 41.79292259216309,
      "action_relation_loss": 0.002946065901778638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.75023386342376,
      "flow_matching_loss": 0.034875626862049106,
      "grad_norm": 0.12276195734739304,
      "learning_rate": 6.158382079774716e-05,
      "loss": 0.038,
      "relation_loss": 0.003129570372402668,
      "selected_pairs": 26715.2,
      "state_query_norm": 43.24094161987305,
      "state_relation_loss": 0.00018350445752730594,
      "step": 13630,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.003129570372402668,
      "weighted_relation_loss": 0.003129570372402668
    },
    {
      "action_query_norm": 41.803466796875,
      "action_relation_loss": 0.002827559458091855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.75958840037418,
      "flow_matching_loss": 0.03228934854269028,
      "grad_norm": 0.09399166703224182,
      "learning_rate": 6.153019764724799e-05,
      "loss": 0.0353,
      "relation_loss": 0.0030101365875452755,
      "selected_pairs": 26780.5,
      "state_query_norm": 43.23357810974121,
      "state_relation_loss": 0.0001825771396397613,
      "step": 13640,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0030101365875452755,
      "weighted_relation_loss": 0.0030101365875452755
    },
    {
      "action_query_norm": 41.818054580688475,
      "action_relation_loss": 0.0027198756812140345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.768942937324603,
      "flow_matching_loss": 0.03297201506793499,
      "grad_norm": 0.12140477448701859,
      "learning_rate": 6.147656048647144e-05,
      "loss": 0.0359,
      "relation_loss": 0.0028987676836550237,
      "selected_pairs": 27152.0,
      "state_query_norm": 43.23104820251465,
      "state_relation_loss": 0.00017889200244098903,
      "step": 13650,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0028987676836550237,
      "weighted_relation_loss": 0.0028987676836550237
    },
    {
      "action_query_norm": 41.7976676940918,
      "action_relation_loss": 0.0029378564329817893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.778297474275023,
      "flow_matching_loss": 0.03355851750820875,
      "grad_norm": 0.12996293604373932,
      "learning_rate": 6.142290938059173e-05,
      "loss": 0.0367,
      "relation_loss": 0.0031126665649935602,
      "selected_pairs": 26666.9,
      "state_query_norm": 43.24694900512695,
      "state_relation_loss": 0.00017481013492215425,
      "step": 13660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0031126665649935602,
      "weighted_relation_loss": 0.0031126665649935602
    },
    {
      "action_query_norm": 41.75788879394531,
      "action_relation_loss": 0.002793299313634634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.787652011225445,
      "flow_matching_loss": 0.03232938479632139,
      "grad_norm": 0.1182917132973671,
      "learning_rate": 6.136924439480001e-05,
      "loss": 0.0353,
      "relation_loss": 0.0029846960213035347,
      "selected_pairs": 26531.1,
      "state_query_norm": 43.214606857299806,
      "state_relation_loss": 0.00019139676151098685,
      "step": 13670,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0029846960213035347,
      "weighted_relation_loss": 0.0029846960213035347
    },
    {
      "action_query_norm": 41.75819320678711,
      "action_relation_loss": 0.0028785782866179942,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.797006548175865,
      "flow_matching_loss": 0.032946360483765605,
      "grad_norm": 0.11825943738222122,
      "learning_rate": 6.13155655943043e-05,
      "loss": 0.036,
      "relation_loss": 0.0030594338197261093,
      "selected_pairs": 26679.4,
      "state_query_norm": 43.23197937011719,
      "state_relation_loss": 0.00018085552001139148,
      "step": 13680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030594338197261093,
      "weighted_relation_loss": 0.0030594338197261093
    },
    {
      "action_query_norm": 41.78599548339844,
      "action_relation_loss": 0.002868602494709194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.806361085126285,
      "flow_matching_loss": 0.031242551654577254,
      "grad_norm": 0.12157600373029709,
      "learning_rate": 6.126187304432941e-05,
      "loss": 0.0343,
      "relation_loss": 0.0030549932969734074,
      "selected_pairs": 26899.8,
      "state_query_norm": 43.22645301818848,
      "state_relation_loss": 0.00018639084737515077,
      "step": 13690,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0030549932969734074,
      "weighted_relation_loss": 0.0030549932969734074
    },
    {
      "action_query_norm": 41.802193832397464,
      "action_relation_loss": 0.002765170717611909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.815715622076707,
      "flow_matching_loss": 0.03408365193754435,
      "grad_norm": 0.12915067374706268,
      "learning_rate": 6.120816681011684e-05,
      "loss": 0.037,
      "relation_loss": 0.00295600404497236,
      "selected_pairs": 26870.5,
      "state_query_norm": 43.22727966308594,
      "state_relation_loss": 0.00019083331135334448,
      "step": 13700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00295600404497236,
      "weighted_relation_loss": 0.00295600404497236
    },
    {
      "action_query_norm": 41.772864532470706,
      "action_relation_loss": 0.0027947906870394944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.825070159027128,
      "flow_matching_loss": 0.03233078327029944,
      "grad_norm": 0.10690485686063766,
      "learning_rate": 6.115444695692474e-05,
      "loss": 0.0353,
      "relation_loss": 0.0029815097572281958,
      "selected_pairs": 26735.2,
      "state_query_norm": 43.22619476318359,
      "state_relation_loss": 0.00018671903235372155,
      "step": 13710,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0029815097572281958,
      "weighted_relation_loss": 0.0029815097572281958
    },
    {
      "action_query_norm": 41.78284912109375,
      "action_relation_loss": 0.00285655555780977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.83442469597755,
      "flow_matching_loss": 0.03445448372513056,
      "grad_norm": 0.1286262720823288,
      "learning_rate": 6.110071355002779e-05,
      "loss": 0.0375,
      "relation_loss": 0.0030462655005976556,
      "selected_pairs": 27173.6,
      "state_query_norm": 43.212781524658205,
      "state_relation_loss": 0.00018970996607095003,
      "step": 13720,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0030462655005976556,
      "weighted_relation_loss": 0.0030462655005976556
    },
    {
      "action_query_norm": 41.7567253112793,
      "action_relation_loss": 0.0030363045632839205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.84377923292797,
      "flow_matching_loss": 0.03405278790742159,
      "grad_norm": 0.13851739466190338,
      "learning_rate": 6.104696665471714e-05,
      "loss": 0.0373,
      "relation_loss": 0.0032134921988472343,
      "selected_pairs": 27318.0,
      "state_query_norm": 43.211961364746095,
      "state_relation_loss": 0.00017718764866003767,
      "step": 13730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0032134921988472343,
      "weighted_relation_loss": 0.0032134921988472343
    },
    {
      "action_query_norm": 41.75568161010742,
      "action_relation_loss": 0.0028671011561527847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.853133769878392,
      "flow_matching_loss": 0.03396223820745945,
      "grad_norm": 0.09743815660476685,
      "learning_rate": 6.099320633630036e-05,
      "loss": 0.037,
      "relation_loss": 0.0030454858904704453,
      "selected_pairs": 26666.7,
      "state_query_norm": 43.21788024902344,
      "state_relation_loss": 0.00017838472849689423,
      "step": 13740,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0030454858904704453,
      "weighted_relation_loss": 0.0030454858904704453
    },
    {
      "action_query_norm": 41.74366226196289,
      "action_relation_loss": 0.0029378441628068685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.862488306828812,
      "flow_matching_loss": 0.03271427284926176,
      "grad_norm": 0.09179746359586716,
      "learning_rate": 6.093943266010128e-05,
      "loss": 0.0358,
      "relation_loss": 0.003111955476924777,
      "selected_pairs": 26918.8,
      "state_query_norm": 43.21309547424316,
      "state_relation_loss": 0.00017411129083484412,
      "step": 13750,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.003111955476924777,
      "weighted_relation_loss": 0.003111955476924777
    },
    {
      "action_query_norm": 41.76531715393067,
      "action_relation_loss": 0.002841423498466611,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.871842843779232,
      "flow_matching_loss": 0.03185529727488756,
      "grad_norm": 0.0995904877781868,
      "learning_rate": 6.088564569146e-05,
      "loss": 0.0349,
      "relation_loss": 0.0030244153924286364,
      "selected_pairs": 26756.2,
      "state_query_norm": 43.20577392578125,
      "state_relation_loss": 0.00018299190123798325,
      "step": 13760,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0030244153924286364,
      "weighted_relation_loss": 0.0030244153924286364
    },
    {
      "action_query_norm": 41.76587753295898,
      "action_relation_loss": 0.002847148641012609,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.881197380729654,
      "flow_matching_loss": 0.033505094796419145,
      "grad_norm": 0.10749692469835281,
      "learning_rate": 6.083184549573274e-05,
      "loss": 0.0365,
      "relation_loss": 0.003035008767619729,
      "selected_pairs": 26640.2,
      "state_query_norm": 43.211767959594724,
      "state_relation_loss": 0.00018786015134537594,
      "step": 13770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.003035008767619729,
      "weighted_relation_loss": 0.003035008767619729
    },
    {
      "action_query_norm": 41.748711013793944,
      "action_relation_loss": 0.002857093117199838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.890551917680074,
      "flow_matching_loss": 0.03278735894709826,
      "grad_norm": 0.09544503688812256,
      "learning_rate": 6.077803213829184e-05,
      "loss": 0.0358,
      "relation_loss": 0.00304497207980603,
      "selected_pairs": 26778.2,
      "state_query_norm": 43.20301780700684,
      "state_relation_loss": 0.00018787898297887296,
      "step": 13780,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.00304497207980603,
      "weighted_relation_loss": 0.00304497207980603
    },
    {
      "action_query_norm": 41.71900901794434,
      "action_relation_loss": 0.0028207293478772046,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.899906454630496,
      "flow_matching_loss": 0.033283406496047975,
      "grad_norm": 0.1381189376115799,
      "learning_rate": 6.0724205684525594e-05,
      "loss": 0.0363,
      "relation_loss": 0.0030118506168946625,
      "selected_pairs": 26803.8,
      "state_query_norm": 43.20479278564453,
      "state_relation_loss": 0.00019112127920379863,
      "step": 13790,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0030118506168946625,
      "weighted_relation_loss": 0.0030118506168946625
    },
    {
      "action_query_norm": 41.723612213134764,
      "action_relation_loss": 0.0028960290364921093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.909260991580917,
      "flow_matching_loss": 0.030318896286189556,
      "grad_norm": 0.14075545966625214,
      "learning_rate": 6.0670366199838215e-05,
      "loss": 0.0334,
      "relation_loss": 0.0030787125462666152,
      "selected_pairs": 26410.9,
      "state_query_norm": 43.216823196411134,
      "state_relation_loss": 0.00018268349813297391,
      "step": 13800,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0030787125462666152,
      "weighted_relation_loss": 0.0030787125462666152
    },
    {
      "action_query_norm": 41.722738647460936,
      "action_relation_loss": 0.0028504672925919296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.918615528531337,
      "flow_matching_loss": 0.03467826880514622,
      "grad_norm": 0.1361103653907776,
      "learning_rate": 6.061651374964974e-05,
      "loss": 0.0377,
      "relation_loss": 0.003029725165106356,
      "selected_pairs": 27181.3,
      "state_query_norm": 43.20559196472168,
      "state_relation_loss": 0.00017925788706634193,
      "step": 13810,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.003029725165106356,
      "weighted_relation_loss": 0.003029725165106356
    },
    {
      "action_query_norm": 41.727501678466794,
      "action_relation_loss": 0.002775061852298677,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.927970065481759,
      "flow_matching_loss": 0.03182090707123279,
      "grad_norm": 0.11982747912406921,
      "learning_rate": 6.0562648399396004e-05,
      "loss": 0.0348,
      "relation_loss": 0.0029524913057684898,
      "selected_pairs": 26455.7,
      "state_query_norm": 43.20328559875488,
      "state_relation_loss": 0.00017742942582117395,
      "step": 13820,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0029524913057684898,
      "weighted_relation_loss": 0.0029524913057684898
    },
    {
      "action_query_norm": 41.69329566955567,
      "action_relation_loss": 0.0027693157084286214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.937324602432179,
      "flow_matching_loss": 0.03397236857563257,
      "grad_norm": 0.12275692075490952,
      "learning_rate": 6.050877021452845e-05,
      "loss": 0.0369,
      "relation_loss": 0.0029438107740134,
      "selected_pairs": 26668.8,
      "state_query_norm": 43.2056266784668,
      "state_relation_loss": 0.00017449505976401268,
      "step": 13830,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0029438107740134,
      "weighted_relation_loss": 0.0029438107740134
    },
    {
      "action_query_norm": 41.732859420776364,
      "action_relation_loss": 0.002769337431527674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.946679139382601,
      "flow_matching_loss": 0.03266721982508898,
      "grad_norm": 0.10904867202043533,
      "learning_rate": 6.04548792605142e-05,
      "loss": 0.0356,
      "relation_loss": 0.0029505772748962046,
      "selected_pairs": 26260.4,
      "state_query_norm": 43.19528350830078,
      "state_relation_loss": 0.0001812398375477642,
      "step": 13840,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0029505772748962046,
      "weighted_relation_loss": 0.0029505772748962046
    },
    {
      "action_query_norm": 41.74603424072266,
      "action_relation_loss": 0.002678123861551285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.956033676333021,
      "flow_matching_loss": 0.032396864145994186,
      "grad_norm": 0.10478413850069046,
      "learning_rate": 6.04009756028358e-05,
      "loss": 0.0352,
      "relation_loss": 0.0028530661715194584,
      "selected_pairs": 27220.0,
      "state_query_norm": 43.19814453125,
      "state_relation_loss": 0.00017494234343757854,
      "step": 13850,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0028530661715194584,
      "weighted_relation_loss": 0.0028530661715194584
    },
    {
      "action_query_norm": 41.74194412231445,
      "action_relation_loss": 0.0028776242630556225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.965388213283443,
      "flow_matching_loss": 0.03406483791768551,
      "grad_norm": 0.1225118339061737,
      "learning_rate": 6.0347059306991294e-05,
      "loss": 0.0371,
      "relation_loss": 0.003068928653374314,
      "selected_pairs": 26502.8,
      "state_query_norm": 43.20522766113281,
      "state_relation_loss": 0.00019130441214656457,
      "step": 13860,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.003068928653374314,
      "weighted_relation_loss": 0.003068928653374314
    },
    {
      "action_query_norm": 41.74155502319336,
      "action_relation_loss": 0.0027719844365492463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.974742750233863,
      "flow_matching_loss": 0.0320581553503871,
      "grad_norm": 0.11630269885063171,
      "learning_rate": 6.029313043849407e-05,
      "loss": 0.035,
      "relation_loss": 0.002947625773958862,
      "selected_pairs": 27290.1,
      "state_query_norm": 43.19786491394043,
      "state_relation_loss": 0.00017564136651344598,
      "step": 13870,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.002947625773958862,
      "weighted_relation_loss": 0.002947625773958862
    },
    {
      "action_query_norm": 41.68821258544922,
      "action_relation_loss": 0.002677802345715463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.984097287184284,
      "flow_matching_loss": 0.031844287738204004,
      "grad_norm": 0.1759508103132248,
      "learning_rate": 6.023918906287279e-05,
      "loss": 0.0347,
      "relation_loss": 0.002881640614941716,
      "selected_pairs": 26736.5,
      "state_query_norm": 43.19316215515137,
      "state_relation_loss": 0.0002038382866885513,
      "step": 13880,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.002881640614941716,
      "weighted_relation_loss": 0.002881640614941716
    },
    {
      "action_query_norm": 41.73072395324707,
      "action_relation_loss": 0.002785103488713503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 12.993451824134706,
      "flow_matching_loss": 0.03270441275089979,
      "grad_norm": 0.11761412024497986,
      "learning_rate": 6.01852352456713e-05,
      "loss": 0.0357,
      "relation_loss": 0.0029695060104131697,
      "selected_pairs": 27106.3,
      "state_query_norm": 43.18999824523926,
      "state_relation_loss": 0.00018440249405102805,
      "step": 13890,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0029695060104131697,
      "weighted_relation_loss": 0.0029695060104131697
    },
    {
      "action_query_norm": 41.702904891967776,
      "action_relation_loss": 0.002707251301035285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.002806361085126,
      "flow_matching_loss": 0.03204893637448549,
      "grad_norm": 0.12872856855392456,
      "learning_rate": 6.013126905244858e-05,
      "loss": 0.035,
      "relation_loss": 0.0029017137130722404,
      "selected_pairs": 24235.0,
      "state_query_norm": 43.19430160522461,
      "state_relation_loss": 0.00019446241058176383,
      "step": 13900,
      "valid_action_samples": 213.9,
      "weighted_auxiliary_loss": 0.0029017137130722404,
      "weighted_relation_loss": 0.0029017137130722404
    },
    {
      "action_query_norm": 41.70450706481934,
      "action_relation_loss": 0.002763181016780436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.012160898035548,
      "flow_matching_loss": 0.03286756798624992,
      "grad_norm": 0.1342812478542328,
      "learning_rate": 6.007729054877864e-05,
      "loss": 0.0358,
      "relation_loss": 0.002959742024540901,
      "selected_pairs": 26963.8,
      "state_query_norm": 43.187689208984374,
      "state_relation_loss": 0.00019656098447740078,
      "step": 13910,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002959742024540901,
      "weighted_relation_loss": 0.002959742024540901
    },
    {
      "action_query_norm": 41.70336227416992,
      "action_relation_loss": 0.002721926895901561,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.021515434985968,
      "flow_matching_loss": 0.03400987386703491,
      "grad_norm": 0.10685000568628311,
      "learning_rate": 6.0023299800250466e-05,
      "loss": 0.0369,
      "relation_loss": 0.002914327010512352,
      "selected_pairs": 27438.0,
      "state_query_norm": 43.19084510803223,
      "state_relation_loss": 0.00019240005349274725,
      "step": 13920,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.002914327010512352,
      "weighted_relation_loss": 0.002914327010512352
    },
    {
      "action_query_norm": 41.69489097595215,
      "action_relation_loss": 0.002672272804193199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.030869971936388,
      "flow_matching_loss": 0.03266658466309309,
      "grad_norm": 0.1337006837129593,
      "learning_rate": 5.9969296872467894e-05,
      "loss": 0.0355,
      "relation_loss": 0.0028674774803221226,
      "selected_pairs": 27668.2,
      "state_query_norm": 43.19401931762695,
      "state_relation_loss": 0.00019520467321854085,
      "step": 13930,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0028674774803221226,
      "weighted_relation_loss": 0.0028674774803221226
    },
    {
      "action_query_norm": 41.68865737915039,
      "action_relation_loss": 0.002552533452399075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.04022450888681,
      "flow_matching_loss": 0.0318056071177125,
      "grad_norm": 0.11261630803346634,
      "learning_rate": 5.9915281831049585e-05,
      "loss": 0.0345,
      "relation_loss": 0.0027430607937276364,
      "selected_pairs": 27129.6,
      "state_query_norm": 43.184980392456055,
      "state_relation_loss": 0.00019052733114222065,
      "step": 13940,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0027430607937276364,
      "weighted_relation_loss": 0.0027430607937276364
    },
    {
      "action_query_norm": 41.68984680175781,
      "action_relation_loss": 0.0025600607739761473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.04957904583723,
      "flow_matching_loss": 0.03399017136543989,
      "grad_norm": 0.1380489617586136,
      "learning_rate": 5.98612547416289e-05,
      "loss": 0.0367,
      "relation_loss": 0.0027423167368397115,
      "selected_pairs": 27150.1,
      "state_query_norm": 43.185363388061525,
      "state_relation_loss": 0.00018225598760182037,
      "step": 13950,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0027423167368397115,
      "weighted_relation_loss": 0.0027423167368397115
    },
    {
      "action_query_norm": 41.71786193847656,
      "action_relation_loss": 0.0026525626191869376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.058933582787652,
      "flow_matching_loss": 0.0319567559286952,
      "grad_norm": 0.12415856868028641,
      "learning_rate": 5.9807215669853855e-05,
      "loss": 0.0348,
      "relation_loss": 0.0028356063878163694,
      "selected_pairs": 26945.5,
      "state_query_norm": 43.17909317016601,
      "state_relation_loss": 0.00018304375844309108,
      "step": 13960,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0028356063878163694,
      "weighted_relation_loss": 0.0028356063878163694
    },
    {
      "action_query_norm": 41.691812133789064,
      "action_relation_loss": 0.002650402998551726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.068288119738073,
      "flow_matching_loss": 0.03192612621933222,
      "grad_norm": 0.1192890927195549,
      "learning_rate": 5.9753164681387e-05,
      "loss": 0.0348,
      "relation_loss": 0.002839657990261912,
      "selected_pairs": 27126.6,
      "state_query_norm": 43.18425827026367,
      "state_relation_loss": 0.0001892549786134623,
      "step": 13970,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.002839657990261912,
      "weighted_relation_loss": 0.002839657990261912
    },
    {
      "action_query_norm": 41.68690299987793,
      "action_relation_loss": 0.0026694554602727295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.077642656688495,
      "flow_matching_loss": 0.032333646155893805,
      "grad_norm": 0.09840419888496399,
      "learning_rate": 5.969910184190539e-05,
      "loss": 0.0352,
      "relation_loss": 0.0028491339879110457,
      "selected_pairs": 26726.3,
      "state_query_norm": 43.17464408874512,
      "state_relation_loss": 0.00017967854073503987,
      "step": 13980,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0028491339879110457,
      "weighted_relation_loss": 0.0028491339879110457
    },
    {
      "action_query_norm": 41.6570629119873,
      "action_relation_loss": 0.0026819884777069094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.086997193638915,
      "flow_matching_loss": 0.03166373427957296,
      "grad_norm": 0.11261782050132751,
      "learning_rate": 5.9645027217100477e-05,
      "loss": 0.0345,
      "relation_loss": 0.0028659125557169316,
      "selected_pairs": 26406.5,
      "state_query_norm": 43.166016006469725,
      "state_relation_loss": 0.00018392408092040568,
      "step": 13990,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0028659125557169316,
      "weighted_relation_loss": 0.0028659125557169316
    },
    {
      "action_query_norm": 41.64132919311523,
      "action_relation_loss": 0.0027491327142342924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.096351730589335,
      "flow_matching_loss": 0.03245857711881399,
      "grad_norm": 0.12091436982154846,
      "learning_rate": 5.959094087267804e-05,
      "loss": 0.0354,
      "relation_loss": 0.002927917940542102,
      "selected_pairs": 26987.7,
      "state_query_norm": 43.16328353881836,
      "state_relation_loss": 0.00017878524668049067,
      "step": 14000,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002927917940542102,
      "weighted_relation_loss": 0.002927917940542102
    },
    {
      "action_query_norm": 41.669013595581056,
      "action_relation_loss": 0.0026606102474033833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.105706267539757,
      "flow_matching_loss": 0.03206285685300827,
      "grad_norm": 0.10973099619150162,
      "learning_rate": 5.9536842874358065e-05,
      "loss": 0.0349,
      "relation_loss": 0.002842228324152529,
      "selected_pairs": 26228.4,
      "state_query_norm": 43.17374725341797,
      "state_relation_loss": 0.00018161806219723075,
      "step": 14010,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.002842228324152529,
      "weighted_relation_loss": 0.002842228324152529
    },
    {
      "action_query_norm": 41.67884407043457,
      "action_relation_loss": 0.002540618390776217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.115060804490177,
      "flow_matching_loss": 0.03244306668639183,
      "grad_norm": 0.10864913463592529,
      "learning_rate": 5.9482733287874734e-05,
      "loss": 0.0352,
      "relation_loss": 0.00271275679115206,
      "selected_pairs": 27153.3,
      "state_query_norm": 43.16658821105957,
      "state_relation_loss": 0.00017213843966601416,
      "step": 14020,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.00271275679115206,
      "weighted_relation_loss": 0.00271275679115206
    },
    {
      "action_query_norm": 41.66899871826172,
      "action_relation_loss": 0.0025796992471441627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.1244153414406,
      "flow_matching_loss": 0.032845993898808955,
      "grad_norm": 0.10766033828258514,
      "learning_rate": 5.942861217897631e-05,
      "loss": 0.0356,
      "relation_loss": 0.0027524637756869198,
      "selected_pairs": 26967.4,
      "state_query_norm": 43.17017440795898,
      "state_relation_loss": 0.00017276453727390617,
      "step": 14030,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0027524637756869198,
      "weighted_relation_loss": 0.0027524637756869198
    },
    {
      "action_query_norm": 41.686364364624026,
      "action_relation_loss": 0.0027530423132702707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.13376987839102,
      "flow_matching_loss": 0.03260344248265028,
      "grad_norm": 0.13522326946258545,
      "learning_rate": 5.9374479613425e-05,
      "loss": 0.0355,
      "relation_loss": 0.0029407073743641376,
      "selected_pairs": 26530.9,
      "state_query_norm": 43.15836524963379,
      "state_relation_loss": 0.0001876650596386753,
      "step": 14040,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0029407073743641376,
      "weighted_relation_loss": 0.0029407073743641376
    },
    {
      "action_query_norm": 41.70727195739746,
      "action_relation_loss": 0.0025108333909884097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.14312441534144,
      "flow_matching_loss": 0.032720235735177995,
      "grad_norm": 0.13120585680007935,
      "learning_rate": 5.932033565699704e-05,
      "loss": 0.0354,
      "relation_loss": 0.002685792208649218,
      "selected_pairs": 26044.5,
      "state_query_norm": 43.15317230224609,
      "state_relation_loss": 0.00017495877982582898,
      "step": 14050,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.002685792208649218,
      "weighted_relation_loss": 0.002685792208649218
    },
    {
      "action_query_norm": 41.68590812683105,
      "action_relation_loss": 0.002716427552513778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.152478952291862,
      "flow_matching_loss": 0.03215080443769693,
      "grad_norm": 0.16555023193359375,
      "learning_rate": 5.926618037548237e-05,
      "loss": 0.035,
      "relation_loss": 0.0028932733228430154,
      "selected_pairs": 26323.2,
      "state_query_norm": 43.16379890441895,
      "state_relation_loss": 0.0001768457645084709,
      "step": 14060,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0028932733228430154,
      "weighted_relation_loss": 0.0028932733228430154
    },
    {
      "action_query_norm": 41.611274337768556,
      "action_relation_loss": 0.0026313949609175326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.161833489242282,
      "flow_matching_loss": 0.032565840147435667,
      "grad_norm": 0.1255250722169876,
      "learning_rate": 5.9212013834684824e-05,
      "loss": 0.0354,
      "relation_loss": 0.002814700873568654,
      "selected_pairs": 26480.9,
      "state_query_norm": 43.143499755859374,
      "state_relation_loss": 0.0001833058980992064,
      "step": 14070,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.002814700873568654,
      "weighted_relation_loss": 0.002814700873568654
    },
    {
      "action_query_norm": 41.61160430908203,
      "action_relation_loss": 0.002714718715287745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.171188026192704,
      "flow_matching_loss": 0.03193984404206276,
      "grad_norm": 0.08631519973278046,
      "learning_rate": 5.9157836100421824e-05,
      "loss": 0.0348,
      "relation_loss": 0.0029074109392240644,
      "selected_pairs": 26952.2,
      "state_query_norm": 43.15033187866211,
      "state_relation_loss": 0.00019269219919806346,
      "step": 14080,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0029074109392240644,
      "weighted_relation_loss": 0.0029074109392240644
    },
    {
      "action_query_norm": 41.62991638183594,
      "action_relation_loss": 0.0026153260143473743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.180542563143124,
      "flow_matching_loss": 0.03404443096369505,
      "grad_norm": 0.11948150396347046,
      "learning_rate": 5.910364723852444e-05,
      "loss": 0.0368,
      "relation_loss": 0.0027974866097792983,
      "selected_pairs": 26897.0,
      "state_query_norm": 43.153506469726565,
      "state_relation_loss": 0.0001821606027078815,
      "step": 14090,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0027974866097792983,
      "weighted_relation_loss": 0.0027974866097792983
    },
    {
      "action_query_norm": 41.6226203918457,
      "action_relation_loss": 0.0025443622143939137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.189897100093546,
      "flow_matching_loss": 0.03280029781162739,
      "grad_norm": 0.11562467366456985,
      "learning_rate": 5.904944731483723e-05,
      "loss": 0.0355,
      "relation_loss": 0.0027255305321887136,
      "selected_pairs": 26528.1,
      "state_query_norm": 43.15703582763672,
      "state_relation_loss": 0.00018116830906365067,
      "step": 14100,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0027255305321887136,
      "weighted_relation_loss": 0.0027255305321887136
    },
    {
      "action_query_norm": 41.64524917602539,
      "action_relation_loss": 0.0025211991975083946,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.199251637043966,
      "flow_matching_loss": 0.033271894603967664,
      "grad_norm": 0.09604164212942123,
      "learning_rate": 5.899523639521825e-05,
      "loss": 0.036,
      "relation_loss": 0.002709701401181519,
      "selected_pairs": 26693.0,
      "state_query_norm": 43.15737800598144,
      "state_relation_loss": 0.00018850220367312432,
      "step": 14110,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002709701401181519,
      "weighted_relation_loss": 0.002709701401181519
    },
    {
      "action_query_norm": 41.66941680908203,
      "action_relation_loss": 0.002675431175157428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.208606173994387,
      "flow_matching_loss": 0.03203981090337038,
      "grad_norm": 0.1189599484205246,
      "learning_rate": 5.894101454553883e-05,
      "loss": 0.0349,
      "relation_loss": 0.0028469118988141418,
      "selected_pairs": 26299.5,
      "state_query_norm": 43.14051055908203,
      "state_relation_loss": 0.00017148072511190547,
      "step": 14120,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0028469118988141418,
      "weighted_relation_loss": 0.0028469118988141418
    },
    {
      "action_query_norm": 41.66209106445312,
      "action_relation_loss": 0.0026262545958161353,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.217960710944809,
      "flow_matching_loss": 0.03217937089502811,
      "grad_norm": 0.10693933814764023,
      "learning_rate": 5.888678183168368e-05,
      "loss": 0.035,
      "relation_loss": 0.002823381358757615,
      "selected_pairs": 26503.4,
      "state_query_norm": 43.1401782989502,
      "state_relation_loss": 0.0001971267760382034,
      "step": 14130,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002823381358757615,
      "weighted_relation_loss": 0.002823381358757615
    },
    {
      "action_query_norm": 41.6488094329834,
      "action_relation_loss": 0.002610251773148775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.227315247895229,
      "flow_matching_loss": 0.030752303265035152,
      "grad_norm": 0.09965366125106812,
      "learning_rate": 5.883253831955061e-05,
      "loss": 0.0335,
      "relation_loss": 0.002790782111696899,
      "selected_pairs": 27409.9,
      "state_query_norm": 43.14330291748047,
      "state_relation_loss": 0.00018053033854812384,
      "step": 14140,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.002790782111696899,
      "weighted_relation_loss": 0.002790782111696899
    },
    {
      "action_query_norm": 41.63585319519043,
      "action_relation_loss": 0.002591842575930059,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.23666978484565,
      "flow_matching_loss": 0.031590795144438744,
      "grad_norm": 0.09829344600439072,
      "learning_rate": 5.8778284075050625e-05,
      "loss": 0.0344,
      "relation_loss": 0.0027653489261865614,
      "selected_pairs": 26997.3,
      "state_query_norm": 43.15084724426269,
      "state_relation_loss": 0.00017350633279420436,
      "step": 14150,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0027653489261865614,
      "weighted_relation_loss": 0.0027653489261865614
    },
    {
      "action_query_norm": 41.617574310302736,
      "action_relation_loss": 0.0028255618177354335,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.246024321796071,
      "flow_matching_loss": 0.032532406784594056,
      "grad_norm": 0.11094606667757034,
      "learning_rate": 5.872401916410777e-05,
      "loss": 0.0355,
      "relation_loss": 0.0030004497151821853,
      "selected_pairs": 27043.7,
      "state_query_norm": 43.126050567626955,
      "state_relation_loss": 0.00017488791927462445,
      "step": 14160,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0030004497151821853,
      "weighted_relation_loss": 0.0030004497151821853
    },
    {
      "action_query_norm": 41.63901214599609,
      "action_relation_loss": 0.0026507738744840025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.255378858746493,
      "flow_matching_loss": 0.03219408188015223,
      "grad_norm": 0.15603135526180267,
      "learning_rate": 5.8669743652659014e-05,
      "loss": 0.035,
      "relation_loss": 0.002830668422393501,
      "selected_pairs": 27269.7,
      "state_query_norm": 43.14154396057129,
      "state_relation_loss": 0.00017989456246141343,
      "step": 14170,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.002830668422393501,
      "weighted_relation_loss": 0.002830668422393501
    },
    {
      "action_query_norm": 41.63070373535156,
      "action_relation_loss": 0.0024279051693156363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.264733395696913,
      "flow_matching_loss": 0.03160767182707787,
      "grad_norm": 0.15465885400772095,
      "learning_rate": 5.86154576066542e-05,
      "loss": 0.0342,
      "relation_loss": 0.0026071428321301935,
      "selected_pairs": 26804.7,
      "state_query_norm": 43.14645233154297,
      "state_relation_loss": 0.00017923769191838801,
      "step": 14180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0026071428321301935,
      "weighted_relation_loss": 0.0026071428321301935
    },
    {
      "action_query_norm": 41.62923583984375,
      "action_relation_loss": 0.0025072369491681455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.274087932647333,
      "flow_matching_loss": 0.029203133285045625,
      "grad_norm": 0.11052427440881729,
      "learning_rate": 5.856116109205602e-05,
      "loss": 0.0319,
      "relation_loss": 0.00269852913916111,
      "selected_pairs": 26897.6,
      "state_query_norm": 43.141105270385744,
      "state_relation_loss": 0.00019129218999296426,
      "step": 14190,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.00269852913916111,
      "weighted_relation_loss": 0.00269852913916111
    },
    {
      "action_query_norm": 41.59983940124512,
      "action_relation_loss": 0.002612106129527092,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.283442469597755,
      "flow_matching_loss": 0.03107796683907509,
      "grad_norm": 0.11796662211418152,
      "learning_rate": 5.850685417483983e-05,
      "loss": 0.0339,
      "relation_loss": 0.002790641342289746,
      "selected_pairs": 27012.3,
      "state_query_norm": 43.14540328979492,
      "state_relation_loss": 0.00017853520694188774,
      "step": 14200,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.002790641342289746,
      "weighted_relation_loss": 0.002790641342289746
    },
    {
      "action_query_norm": 41.612545013427734,
      "action_relation_loss": 0.0025225640507414936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.292797006548176,
      "flow_matching_loss": 0.030765284784138203,
      "grad_norm": 0.12773945927619934,
      "learning_rate": 5.8452536920993684e-05,
      "loss": 0.0335,
      "relation_loss": 0.0026972639840096234,
      "selected_pairs": 26422.8,
      "state_query_norm": 43.1350772857666,
      "state_relation_loss": 0.00017469993326812983,
      "step": 14210,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0026972639840096234,
      "weighted_relation_loss": 0.0026972639840096234
    },
    {
      "action_query_norm": 41.635353088378906,
      "action_relation_loss": 0.0027629460208117964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.302151543498598,
      "flow_matching_loss": 0.032225031033158304,
      "grad_norm": 0.09094971418380737,
      "learning_rate": 5.839820939651817e-05,
      "loss": 0.0352,
      "relation_loss": 0.002940128091722727,
      "selected_pairs": 26709.5,
      "state_query_norm": 43.13942947387695,
      "state_relation_loss": 0.00017718208255246282,
      "step": 14220,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.002940128091722727,
      "weighted_relation_loss": 0.002940128091722727
    },
    {
      "action_query_norm": 41.61963081359863,
      "action_relation_loss": 0.002709254715591669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.311506080449018,
      "flow_matching_loss": 0.03198587968945503,
      "grad_norm": 0.10819926112890244,
      "learning_rate": 5.834387166742632e-05,
      "loss": 0.0349,
      "relation_loss": 0.002882750425487757,
      "selected_pairs": 26697.4,
      "state_query_norm": 43.1336612701416,
      "state_relation_loss": 0.00017349569097859784,
      "step": 14230,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002882750425487757,
      "weighted_relation_loss": 0.002882750425487757
    },
    {
      "action_query_norm": 41.6201774597168,
      "action_relation_loss": 0.0026246943743899465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.320860617399438,
      "flow_matching_loss": 0.030369664542376995,
      "grad_norm": 0.10850170999765396,
      "learning_rate": 5.828952379974364e-05,
      "loss": 0.0332,
      "relation_loss": 0.002814620709978044,
      "selected_pairs": 26657.5,
      "state_query_norm": 43.135972595214845,
      "state_relation_loss": 0.00018992632976733148,
      "step": 14240,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002814620709978044,
      "weighted_relation_loss": 0.002814620709978044
    },
    {
      "action_query_norm": 41.60916328430176,
      "action_relation_loss": 0.002726189745590091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.33021515434986,
      "flow_matching_loss": 0.03081124257296324,
      "grad_norm": 0.13400492072105408,
      "learning_rate": 5.8235165859507864e-05,
      "loss": 0.0337,
      "relation_loss": 0.0029073266312479975,
      "selected_pairs": 27042.5,
      "state_query_norm": 43.12780494689942,
      "state_relation_loss": 0.00018113688565790653,
      "step": 14250,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0029073266312479975,
      "weighted_relation_loss": 0.0029073266312479975
    },
    {
      "action_query_norm": 41.61706352233887,
      "action_relation_loss": 0.002610704000107944,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.33956969130028,
      "flow_matching_loss": 0.03104844819754362,
      "grad_norm": 0.11855347454547882,
      "learning_rate": 5.818079791276907e-05,
      "loss": 0.0338,
      "relation_loss": 0.002781881322152913,
      "selected_pairs": 26709.1,
      "state_query_norm": 43.13263092041016,
      "state_relation_loss": 0.00017117734823841603,
      "step": 14260,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.002781881322152913,
      "weighted_relation_loss": 0.002781881322152913
    },
    {
      "action_query_norm": 41.595425033569335,
      "action_relation_loss": 0.0025588863296434282,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.348924228250702,
      "flow_matching_loss": 0.031961573474109176,
      "grad_norm": 0.12361830472946167,
      "learning_rate": 5.8126420025589415e-05,
      "loss": 0.0347,
      "relation_loss": 0.002750219707377255,
      "selected_pairs": 26344.3,
      "state_query_norm": 43.11985092163086,
      "state_relation_loss": 0.00019133342866552993,
      "step": 14270,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002750219707377255,
      "weighted_relation_loss": 0.002750219707377255
    },
    {
      "action_query_norm": 41.58005332946777,
      "action_relation_loss": 0.0025161349214613436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.358278765201122,
      "flow_matching_loss": 0.02874375693500042,
      "grad_norm": 0.12498969584703445,
      "learning_rate": 5.807203226404313e-05,
      "loss": 0.0315,
      "relation_loss": 0.0027089463314041496,
      "selected_pairs": 26672.0,
      "state_query_norm": 43.11675338745117,
      "state_relation_loss": 0.00019281138811493291,
      "step": 14280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0027089463314041496,
      "weighted_relation_loss": 0.0027089463314041496
    },
    {
      "action_query_norm": 41.57794723510742,
      "action_relation_loss": 0.0025945211295038463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.367633302151544,
      "flow_matching_loss": 0.03170188628137112,
      "grad_norm": 0.09963826090097427,
      "learning_rate": 5.801763469421652e-05,
      "loss": 0.0345,
      "relation_loss": 0.0027922932524234055,
      "selected_pairs": 26849.9,
      "state_query_norm": 43.12815933227539,
      "state_relation_loss": 0.00019777214038185775,
      "step": 14290,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0027922932524234055,
      "weighted_relation_loss": 0.0027922932524234055
    },
    {
      "action_query_norm": 41.60943260192871,
      "action_relation_loss": 0.002762200403958559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.376987839101965,
      "flow_matching_loss": 0.03180357832461596,
      "grad_norm": 0.09113297611474991,
      "learning_rate": 5.796322738220774e-05,
      "loss": 0.0348,
      "relation_loss": 0.0029593384824693203,
      "selected_pairs": 27014.1,
      "state_query_norm": 43.11144332885742,
      "state_relation_loss": 0.00019713809160748497,
      "step": 14300,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0029593384824693203,
      "weighted_relation_loss": 0.0029593384824693203
    },
    {
      "action_query_norm": 41.579960632324216,
      "action_relation_loss": 0.002556518348865211,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.386342376052385,
      "flow_matching_loss": 0.03202438857406378,
      "grad_norm": 0.09602361917495728,
      "learning_rate": 5.79088103941268e-05,
      "loss": 0.0348,
      "relation_loss": 0.0027343570021912456,
      "selected_pairs": 26665.7,
      "state_query_norm": 43.124352645874026,
      "state_relation_loss": 0.00017783862131182104,
      "step": 14310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0027343570021912456,
      "weighted_relation_loss": 0.0027343570021912456
    },
    {
      "action_query_norm": 41.55911827087402,
      "action_relation_loss": 0.002879503509029746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.395696913002807,
      "flow_matching_loss": 0.03283192478120327,
      "grad_norm": 0.11396311968564987,
      "learning_rate": 5.785438379609549e-05,
      "loss": 0.0359,
      "relation_loss": 0.0030561289051547645,
      "selected_pairs": 26482.2,
      "state_query_norm": 43.107578659057616,
      "state_relation_loss": 0.00017662537429714576,
      "step": 14320,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0030561289051547645,
      "weighted_relation_loss": 0.0030561289051547645
    },
    {
      "action_query_norm": 41.601322174072266,
      "action_relation_loss": 0.0025853579631075264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.405051449953227,
      "flow_matching_loss": 0.033901031129062176,
      "grad_norm": 0.11046604812145233,
      "learning_rate": 5.779994765424724e-05,
      "loss": 0.0367,
      "relation_loss": 0.0027755515184253454,
      "selected_pairs": 26630.1,
      "state_query_norm": 43.1251579284668,
      "state_relation_loss": 0.00019019352475879714,
      "step": 14330,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0027755515184253454,
      "weighted_relation_loss": 0.0027755515184253454
    },
    {
      "action_query_norm": 41.587759017944336,
      "action_relation_loss": 0.0025602350709959865,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.414405986903649,
      "flow_matching_loss": 0.03150638677179814,
      "grad_norm": 0.09336242079734802,
      "learning_rate": 5.7745502034727114e-05,
      "loss": 0.0343,
      "relation_loss": 0.0027507016202434897,
      "selected_pairs": 26548.8,
      "state_query_norm": 43.11112403869629,
      "state_relation_loss": 0.0001904665507026948,
      "step": 14340,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0027507016202434897,
      "weighted_relation_loss": 0.0027507016202434897
    },
    {
      "action_query_norm": 41.54907417297363,
      "action_relation_loss": 0.0026665513636544348,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.42376052385407,
      "flow_matching_loss": 0.03355022724717856,
      "grad_norm": 0.12158552557229996,
      "learning_rate": 5.769104700369164e-05,
      "loss": 0.0364,
      "relation_loss": 0.0028510089498013256,
      "selected_pairs": 26848.3,
      "state_query_norm": 43.120751953125,
      "state_relation_loss": 0.00018445755122229457,
      "step": 14350,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0028510089498013256,
      "weighted_relation_loss": 0.0028510089498013256
    },
    {
      "action_query_norm": 41.599083709716794,
      "action_relation_loss": 0.002635275898501277,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.43311506080449,
      "flow_matching_loss": 0.033085273765027526,
      "grad_norm": 0.12766307592391968,
      "learning_rate": 5.7636582627308854e-05,
      "loss": 0.0359,
      "relation_loss": 0.0028141404734924437,
      "selected_pairs": 26644.8,
      "state_query_norm": 43.121237182617186,
      "state_relation_loss": 0.0001788645633496344,
      "step": 14360,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0028141404734924437,
      "weighted_relation_loss": 0.0028141404734924437
    },
    {
      "action_query_norm": 41.60310516357422,
      "action_relation_loss": 0.0026021060533821584,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.442469597754911,
      "flow_matching_loss": 0.032126374915242194,
      "grad_norm": 0.09969300776720047,
      "learning_rate": 5.7582108971758095e-05,
      "loss": 0.0349,
      "relation_loss": 0.0027804333716630937,
      "selected_pairs": 26184.2,
      "state_query_norm": 43.1164363861084,
      "state_relation_loss": 0.00017832727753557264,
      "step": 14370,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0027804333716630937,
      "weighted_relation_loss": 0.0027804333716630937
    },
    {
      "action_query_norm": 41.55727996826172,
      "action_relation_loss": 0.0026893892092630265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.451824134705332,
      "flow_matching_loss": 0.03261926397681236,
      "grad_norm": 0.10514278709888458,
      "learning_rate": 5.752762610322999e-05,
      "loss": 0.0355,
      "relation_loss": 0.002871254784986377,
      "selected_pairs": 26801.9,
      "state_query_norm": 43.101884841918945,
      "state_relation_loss": 0.00018186555535066874,
      "step": 14380,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002871254784986377,
      "weighted_relation_loss": 0.002871254784986377
    },
    {
      "action_query_norm": 41.52287673950195,
      "action_relation_loss": 0.002804232412017882,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.461178671655754,
      "flow_matching_loss": 0.03038496505469084,
      "grad_norm": 0.11735621839761734,
      "learning_rate": 5.747313408792636e-05,
      "loss": 0.0334,
      "relation_loss": 0.0029803122393786906,
      "selected_pairs": 26741.5,
      "state_query_norm": 43.10949363708496,
      "state_relation_loss": 0.0001760798302711919,
      "step": 14390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0029803122393786906,
      "weighted_relation_loss": 0.0029803122393786906
    },
    {
      "action_query_norm": 41.538991165161136,
      "action_relation_loss": 0.0026213494827970864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.470533208606174,
      "flow_matching_loss": 0.03153781909495592,
      "grad_norm": 0.11636405438184738,
      "learning_rate": 5.741863299206014e-05,
      "loss": 0.0343,
      "relation_loss": 0.002810982777737081,
      "selected_pairs": 26550.0,
      "state_query_norm": 43.10961685180664,
      "state_relation_loss": 0.00018963325419463217,
      "step": 14400,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.002810982777737081,
      "weighted_relation_loss": 0.002810982777737081
    },
    {
      "action_query_norm": 41.54084701538086,
      "action_relation_loss": 0.002730115503072739,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.479887745556596,
      "flow_matching_loss": 0.030419374257326125,
      "grad_norm": 0.11888687312602997,
      "learning_rate": 5.73641228818553e-05,
      "loss": 0.0333,
      "relation_loss": 0.002916621929034591,
      "selected_pairs": 26754.5,
      "state_query_norm": 43.1036678314209,
      "state_relation_loss": 0.0001865064084995538,
      "step": 14410,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.002916621929034591,
      "weighted_relation_loss": 0.002916621929034591
    },
    {
      "action_query_norm": 41.54358253479004,
      "action_relation_loss": 0.0025667853420600294,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.489242282507016,
      "flow_matching_loss": 0.03103316333144903,
      "grad_norm": 0.11106965690851212,
      "learning_rate": 5.730960382354676e-05,
      "loss": 0.0338,
      "relation_loss": 0.002748607564717531,
      "selected_pairs": 27454.1,
      "state_query_norm": 43.10766754150391,
      "state_relation_loss": 0.0001818222168367356,
      "step": 14420,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.002748607564717531,
      "weighted_relation_loss": 0.002748607564717531
    },
    {
      "action_query_norm": 41.577459716796874,
      "action_relation_loss": 0.002541982522234321,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.498596819457436,
      "flow_matching_loss": 0.03127783183008433,
      "grad_norm": 0.14554449915885925,
      "learning_rate": 5.725507588338035e-05,
      "loss": 0.034,
      "relation_loss": 0.002738217543810606,
      "selected_pairs": 27368.8,
      "state_query_norm": 43.097974395751955,
      "state_relation_loss": 0.0001962350361282006,
      "step": 14430,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.002738217543810606,
      "weighted_relation_loss": 0.002738217543810606
    },
    {
      "action_query_norm": 41.55042991638184,
      "action_relation_loss": 0.0026131930062547326,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.507951356407858,
      "flow_matching_loss": 0.030075134709477425,
      "grad_norm": 0.08739639073610306,
      "learning_rate": 5.7200539127612604e-05,
      "loss": 0.0329,
      "relation_loss": 0.002786484220996499,
      "selected_pairs": 27182.5,
      "state_query_norm": 43.10537757873535,
      "state_relation_loss": 0.0001732911536237225,
      "step": 14440,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002786484220996499,
      "weighted_relation_loss": 0.002786484220996499
    },
    {
      "action_query_norm": 41.56784591674805,
      "action_relation_loss": 0.0025913241552188993,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.517305893358278,
      "flow_matching_loss": 0.032423791475594044,
      "grad_norm": 0.10539252310991287,
      "learning_rate": 5.714599362251087e-05,
      "loss": 0.0352,
      "relation_loss": 0.0027659115148708224,
      "selected_pairs": 27060.6,
      "state_query_norm": 43.09254913330078,
      "state_relation_loss": 0.0001745873669278808,
      "step": 14450,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0027659115148708224,
      "weighted_relation_loss": 0.0027659115148708224
    },
    {
      "action_query_norm": 41.57716522216797,
      "action_relation_loss": 0.0025907386327162383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.5266604303087,
      "flow_matching_loss": 0.03254869226366282,
      "grad_norm": 0.11425591260194778,
      "learning_rate": 5.709143943435307e-05,
      "loss": 0.0353,
      "relation_loss": 0.002766868029721081,
      "selected_pairs": 27106.4,
      "state_query_norm": 43.101220703125,
      "state_relation_loss": 0.00017612938245292753,
      "step": 14460,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.002766868029721081,
      "weighted_relation_loss": 0.002766868029721081
    },
    {
      "action_query_norm": 41.54859924316406,
      "action_relation_loss": 0.0024516557808965445,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.53601496725912,
      "flow_matching_loss": 0.02969760075211525,
      "grad_norm": 0.12707296013832092,
      "learning_rate": 5.7036876629427646e-05,
      "loss": 0.0323,
      "relation_loss": 0.0026232034200802446,
      "selected_pairs": 26642.5,
      "state_query_norm": 43.10434532165527,
      "state_relation_loss": 0.00017154765082523228,
      "step": 14470,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0026232034200802446,
      "weighted_relation_loss": 0.0026232034200802446
    },
    {
      "action_query_norm": 41.55517768859863,
      "action_relation_loss": 0.002498177229426801,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.545369504209543,
      "flow_matching_loss": 0.030031597800552846,
      "grad_norm": 0.11438342183828354,
      "learning_rate": 5.698230527403361e-05,
      "loss": 0.0327,
      "relation_loss": 0.0026726187206804752,
      "selected_pairs": 26872.1,
      "state_query_norm": 43.08795738220215,
      "state_relation_loss": 0.0001744414897984825,
      "step": 14480,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0026726187206804752,
      "weighted_relation_loss": 0.0026726187206804752
    },
    {
      "action_query_norm": 41.513519287109375,
      "action_relation_loss": 0.0025644656037911774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.554724041159963,
      "flow_matching_loss": 0.03236502781510353,
      "grad_norm": 0.1080157533288002,
      "learning_rate": 5.6927725434480274e-05,
      "loss": 0.0351,
      "relation_loss": 0.002741624368354678,
      "selected_pairs": 26482.2,
      "state_query_norm": 43.08880615234375,
      "state_relation_loss": 0.00017715881112962962,
      "step": 14490,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.002741624368354678,
      "weighted_relation_loss": 0.002741624368354678
    },
    {
      "action_query_norm": 41.53564949035645,
      "action_relation_loss": 0.002610297198407352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.564078578110383,
      "flow_matching_loss": 0.030099164694547653,
      "grad_norm": 0.10319487750530243,
      "learning_rate": 5.687313717708728e-05,
      "loss": 0.0329,
      "relation_loss": 0.002772411680780351,
      "selected_pairs": 26522.0,
      "state_query_norm": 43.09632530212402,
      "state_relation_loss": 0.00016211444890359416,
      "step": 14500,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.002772411680780351,
      "weighted_relation_loss": 0.002772411680780351
    },
    {
      "action_query_norm": 41.54655418395996,
      "action_relation_loss": 0.002614312176592648,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.573433115060805,
      "flow_matching_loss": 0.030298429355025292,
      "grad_norm": 0.12187708169221878,
      "learning_rate": 5.681854056818453e-05,
      "loss": 0.0331,
      "relation_loss": 0.0027875787112861873,
      "selected_pairs": 26919.3,
      "state_query_norm": 43.096004486083984,
      "state_relation_loss": 0.00017326650850009174,
      "step": 14510,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0027875787112861873,
      "weighted_relation_loss": 0.0027875787112861873
    },
    {
      "action_query_norm": 41.54738426208496,
      "action_relation_loss": 0.0026154259918257593,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.582787652011225,
      "flow_matching_loss": 0.031031363271176816,
      "grad_norm": 0.12197976559400558,
      "learning_rate": 5.676393567411205e-05,
      "loss": 0.0338,
      "relation_loss": 0.002788043092004955,
      "selected_pairs": 27484.0,
      "state_query_norm": 43.086516571044925,
      "state_relation_loss": 0.00017261708417208864,
      "step": 14520,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.002788043092004955,
      "weighted_relation_loss": 0.002788043092004955
    },
    {
      "action_query_norm": 41.51504707336426,
      "action_relation_loss": 0.0024305609287694095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.592142188961647,
      "flow_matching_loss": 0.03137594982981682,
      "grad_norm": 0.13370145857334137,
      "learning_rate": 5.670932256121991e-05,
      "loss": 0.034,
      "relation_loss": 0.002590748178772628,
      "selected_pairs": 26943.1,
      "state_query_norm": 43.0914852142334,
      "state_relation_loss": 0.00016018724709283562,
      "step": 14530,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.002590748178772628,
      "weighted_relation_loss": 0.002590748178772628
    },
    {
      "action_query_norm": 41.52819328308105,
      "action_relation_loss": 0.0024808021262288094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.601496725912067,
      "flow_matching_loss": 0.031596947088837625,
      "grad_norm": 0.09483646601438522,
      "learning_rate": 5.6654701295868216e-05,
      "loss": 0.0343,
      "relation_loss": 0.002659336756914854,
      "selected_pairs": 26988.0,
      "state_query_norm": 43.07091636657715,
      "state_relation_loss": 0.00017853462632047012,
      "step": 14540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002659336756914854,
      "weighted_relation_loss": 0.002659336756914854
    },
    {
      "action_query_norm": 41.52048530578613,
      "action_relation_loss": 0.0023999512661248444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.610851262862488,
      "flow_matching_loss": 0.030465246364474296,
      "grad_norm": 0.096675805747509,
      "learning_rate": 5.660007194442697e-05,
      "loss": 0.033,
      "relation_loss": 0.0025625809794291854,
      "selected_pairs": 26924.0,
      "state_query_norm": 43.08404388427734,
      "state_relation_loss": 0.00016262972494587302,
      "step": 14550,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0025625809794291854,
      "weighted_relation_loss": 0.0025625809794291854
    },
    {
      "action_query_norm": 41.492685317993164,
      "action_relation_loss": 0.002594326972030103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.62020579981291,
      "flow_matching_loss": 0.03171697687357664,
      "grad_norm": 0.12541812658309937,
      "learning_rate": 5.6545434573275946e-05,
      "loss": 0.0345,
      "relation_loss": 0.0027690902119502426,
      "selected_pairs": 27361.7,
      "state_query_norm": 43.07289199829101,
      "state_relation_loss": 0.00017476321227150038,
      "step": 14560,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0027690902119502426,
      "weighted_relation_loss": 0.0027690902119502426
    },
    {
      "action_query_norm": 41.53953399658203,
      "action_relation_loss": 0.002443992206826806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.62956033676333,
      "flow_matching_loss": 0.031945617869496346,
      "grad_norm": 0.10666047781705856,
      "learning_rate": 5.649078924880472e-05,
      "loss": 0.0346,
      "relation_loss": 0.002629681257531047,
      "selected_pairs": 27062.5,
      "state_query_norm": 43.07681541442871,
      "state_relation_loss": 0.0001856890434282832,
      "step": 14570,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002629681257531047,
      "weighted_relation_loss": 0.002629681257531047
    },
    {
      "action_query_norm": 41.553195571899415,
      "action_relation_loss": 0.0025130916154012086,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.638914873713752,
      "flow_matching_loss": 0.029333980940282344,
      "grad_norm": 0.09966187924146652,
      "learning_rate": 5.643613603741252e-05,
      "loss": 0.032,
      "relation_loss": 0.002691612346097827,
      "selected_pairs": 26802.0,
      "state_query_norm": 43.083256530761716,
      "state_relation_loss": 0.00017852072924142703,
      "step": 14580,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002691612346097827,
      "weighted_relation_loss": 0.002691612346097827
    },
    {
      "action_query_norm": 41.544565963745114,
      "action_relation_loss": 0.0024424701929092406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.648269410664172,
      "flow_matching_loss": 0.03302385415881872,
      "grad_norm": 0.11893200129270554,
      "learning_rate": 5.6381475005508154e-05,
      "loss": 0.0356,
      "relation_loss": 0.0026261239079758527,
      "selected_pairs": 26687.4,
      "state_query_norm": 43.06902809143067,
      "state_relation_loss": 0.00018365368741797284,
      "step": 14590,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0026261239079758527,
      "weighted_relation_loss": 0.0026261239079758527
    },
    {
      "action_query_norm": 41.47646865844727,
      "action_relation_loss": 0.0027260372415184976,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.657623947614592,
      "flow_matching_loss": 0.030664677359163763,
      "grad_norm": 0.11527373641729355,
      "learning_rate": 5.632680621950992e-05,
      "loss": 0.0336,
      "relation_loss": 0.0028961734846234322,
      "selected_pairs": 26598.0,
      "state_query_norm": 43.06471252441406,
      "state_relation_loss": 0.00017013623437378556,
      "step": 14600,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0028961734846234322,
      "weighted_relation_loss": 0.0028961734846234322
    },
    {
      "action_query_norm": 41.49259300231934,
      "action_relation_loss": 0.0025471349945291877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.666978484565014,
      "flow_matching_loss": 0.030343037284910678,
      "grad_norm": 0.0948474183678627,
      "learning_rate": 5.627212974584555e-05,
      "loss": 0.0331,
      "relation_loss": 0.002719006151892245,
      "selected_pairs": 26434.8,
      "state_query_norm": 43.06929244995117,
      "state_relation_loss": 0.00017187113990075887,
      "step": 14610,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002719006151892245,
      "weighted_relation_loss": 0.002719006151892245
    },
    {
      "action_query_norm": 41.46306457519531,
      "action_relation_loss": 0.0025940496707335114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.676333021515434,
      "flow_matching_loss": 0.03004087321460247,
      "grad_norm": 0.10291855037212372,
      "learning_rate": 5.62174456509521e-05,
      "loss": 0.0328,
      "relation_loss": 0.0027703468920663,
      "selected_pairs": 26718.9,
      "state_query_norm": 43.05446548461914,
      "state_relation_loss": 0.00017629723733989522,
      "step": 14620,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0027703468920663,
      "weighted_relation_loss": 0.0027703468920663
    },
    {
      "action_query_norm": 41.47540130615234,
      "action_relation_loss": 0.002758757909759879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.685687558465856,
      "flow_matching_loss": 0.03149610999971628,
      "grad_norm": 0.08443517237901688,
      "learning_rate": 5.616275400127594e-05,
      "loss": 0.0344,
      "relation_loss": 0.0029303195187821984,
      "selected_pairs": 26794.1,
      "state_query_norm": 43.074913024902344,
      "state_relation_loss": 0.00017156159156002105,
      "step": 14630,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0029303195187821984,
      "weighted_relation_loss": 0.0029303195187821984
    },
    {
      "action_query_norm": 41.49036407470703,
      "action_relation_loss": 0.0025947797810658814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.695042095416277,
      "flow_matching_loss": 0.029769693315029145,
      "grad_norm": 0.10114588588476181,
      "learning_rate": 5.610805486327254e-05,
      "loss": 0.0325,
      "relation_loss": 0.002766895410604775,
      "selected_pairs": 27032.2,
      "state_query_norm": 43.0750015258789,
      "state_relation_loss": 0.00017211563827004283,
      "step": 14640,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.002766895410604775,
      "weighted_relation_loss": 0.002766895410604775
    },
    {
      "action_query_norm": 41.49861526489258,
      "action_relation_loss": 0.0025150899542495607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.704396632366699,
      "flow_matching_loss": 0.03335609678179026,
      "grad_norm": 0.11110978573560715,
      "learning_rate": 5.6053348303406536e-05,
      "loss": 0.036,
      "relation_loss": 0.0026886893436312677,
      "selected_pairs": 27504.2,
      "state_query_norm": 43.062017822265624,
      "state_relation_loss": 0.00017359937482979148,
      "step": 14650,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0026886893436312677,
      "weighted_relation_loss": 0.0026886893436312677
    },
    {
      "action_query_norm": 41.44022216796875,
      "action_relation_loss": 0.002571877813898027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.713751169317119,
      "flow_matching_loss": 0.031388874538242816,
      "grad_norm": 0.2235197126865387,
      "learning_rate": 5.599863438815156e-05,
      "loss": 0.0341,
      "relation_loss": 0.0027436845703050494,
      "selected_pairs": 26939.2,
      "state_query_norm": 43.04357719421387,
      "state_relation_loss": 0.0001718067520414479,
      "step": 14660,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0027436845703050494,
      "weighted_relation_loss": 0.0027436845703050494
    },
    {
      "action_query_norm": 41.501982879638675,
      "action_relation_loss": 0.0026335295755416156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.723105706267539,
      "flow_matching_loss": 0.029756096377968787,
      "grad_norm": 0.12130746245384216,
      "learning_rate": 5.594391318399017e-05,
      "loss": 0.0326,
      "relation_loss": 0.0028115458553656936,
      "selected_pairs": 26504.1,
      "state_query_norm": 43.059349822998044,
      "state_relation_loss": 0.0001780162871000357,
      "step": 14670,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0028115458553656936,
      "weighted_relation_loss": 0.0028115458553656936
    },
    {
      "action_query_norm": 41.495896911621095,
      "action_relation_loss": 0.0025483997305855155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.732460243217961,
      "flow_matching_loss": 0.03200769741088152,
      "grad_norm": 0.11639444530010223,
      "learning_rate": 5.588918475741377e-05,
      "loss": 0.0347,
      "relation_loss": 0.002709351549856365,
      "selected_pairs": 27316.3,
      "state_query_norm": 43.069763565063475,
      "state_relation_loss": 0.00016095184546429663,
      "step": 14680,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.002709351549856365,
      "weighted_relation_loss": 0.002709351549856365
    },
    {
      "action_query_norm": 41.473008728027345,
      "action_relation_loss": 0.002599632111378014,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.741814780168381,
      "flow_matching_loss": 0.031142511032521726,
      "grad_norm": 0.10240864753723145,
      "learning_rate": 5.5834449174922586e-05,
      "loss": 0.0339,
      "relation_loss": 0.002759612095542252,
      "selected_pairs": 26320.7,
      "state_query_norm": 43.053306579589844,
      "state_relation_loss": 0.00015997996379155666,
      "step": 14690,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.002759612095542252,
      "weighted_relation_loss": 0.002759612095542252
    },
    {
      "action_query_norm": 41.48016090393067,
      "action_relation_loss": 0.0024555532028898595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.751169317118803,
      "flow_matching_loss": 0.029847703874111176,
      "grad_norm": 0.13764862716197968,
      "learning_rate": 5.57797065030255e-05,
      "loss": 0.0325,
      "relation_loss": 0.0026203697780147195,
      "selected_pairs": 26871.1,
      "state_query_norm": 43.05418853759765,
      "state_relation_loss": 0.00016481654602102936,
      "step": 14700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0026203697780147195,
      "weighted_relation_loss": 0.0026203697780147195
    },
    {
      "action_query_norm": 41.441440200805665,
      "action_relation_loss": 0.002491088258102536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.760523854069223,
      "flow_matching_loss": 0.03245267905294895,
      "grad_norm": 0.1140647754073143,
      "learning_rate": 5.5724956808240016e-05,
      "loss": 0.0351,
      "relation_loss": 0.0026661224896088243,
      "selected_pairs": 27149.8,
      "state_query_norm": 43.05586395263672,
      "state_relation_loss": 0.00017503421258879825,
      "step": 14710,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0026661224896088243,
      "weighted_relation_loss": 0.0026661224896088243
    },
    {
      "action_query_norm": 41.46618843078613,
      "action_relation_loss": 0.0024533880408853293,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.769878391019645,
      "flow_matching_loss": 0.030690247006714345,
      "grad_norm": 0.09120304137468338,
      "learning_rate": 5.567020015709219e-05,
      "loss": 0.0333,
      "relation_loss": 0.0026316808303818106,
      "selected_pairs": 27228.4,
      "state_query_norm": 43.05352668762207,
      "state_relation_loss": 0.00017829279822763057,
      "step": 14720,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0026316808303818106,
      "weighted_relation_loss": 0.0026316808303818106
    },
    {
      "action_query_norm": 41.45666923522949,
      "action_relation_loss": 0.002556543331593275,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.779232927970066,
      "flow_matching_loss": 0.02993825227022171,
      "grad_norm": 0.09733173996210098,
      "learning_rate": 5.561543661611649e-05,
      "loss": 0.0327,
      "relation_loss": 0.0027231434592977165,
      "selected_pairs": 27686.0,
      "state_query_norm": 43.05091514587402,
      "state_relation_loss": 0.00016660010733176022,
      "step": 14730,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.0027231434592977165,
      "weighted_relation_loss": 0.0027231434592977165
    },
    {
      "action_query_norm": 41.44828910827637,
      "action_relation_loss": 0.002526281122118235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.788587464920486,
      "flow_matching_loss": 0.03168849088251591,
      "grad_norm": 0.10685215890407562,
      "learning_rate": 5.556066625185583e-05,
      "loss": 0.0344,
      "relation_loss": 0.002706135530024767,
      "selected_pairs": 27458.2,
      "state_query_norm": 43.04075584411621,
      "state_relation_loss": 0.0001798544530174695,
      "step": 14740,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.002706135530024767,
      "weighted_relation_loss": 0.002706135530024767
    },
    {
      "action_query_norm": 41.44497985839844,
      "action_relation_loss": 0.002477175882086158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.797942001870908,
      "flow_matching_loss": 0.03210132904350758,
      "grad_norm": 0.10214433819055557,
      "learning_rate": 5.550588913086131e-05,
      "loss": 0.0347,
      "relation_loss": 0.002647184953093529,
      "selected_pairs": 27069.3,
      "state_query_norm": 43.05849494934082,
      "state_relation_loss": 0.0001700090491794981,
      "step": 14750,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002647184953093529,
      "weighted_relation_loss": 0.002647184953093529
    },
    {
      "action_query_norm": 41.47888526916504,
      "action_relation_loss": 0.0025006029056385158,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.807296538821328,
      "flow_matching_loss": 0.02981748804450035,
      "grad_norm": 0.1048126146197319,
      "learning_rate": 5.545110531969234e-05,
      "loss": 0.0325,
      "relation_loss": 0.0026621858589351175,
      "selected_pairs": 27461.8,
      "state_query_norm": 43.057478713989255,
      "state_relation_loss": 0.0001615829052752815,
      "step": 14760,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0026621858589351175,
      "weighted_relation_loss": 0.0026621858589351175
    },
    {
      "action_query_norm": 41.43233871459961,
      "action_relation_loss": 0.0027220301097258927,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.81665107577175,
      "flow_matching_loss": 0.03164856806397438,
      "grad_norm": 0.14014533162117004,
      "learning_rate": 5.539631488491641e-05,
      "loss": 0.0345,
      "relation_loss": 0.0028922588797286155,
      "selected_pairs": 26993.7,
      "state_query_norm": 43.051165390014646,
      "state_relation_loss": 0.000170228753995616,
      "step": 14770,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0028922588797286155,
      "weighted_relation_loss": 0.0028922588797286155
    },
    {
      "action_query_norm": 41.426523208618164,
      "action_relation_loss": 0.00253072795458138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.82600561272217,
      "flow_matching_loss": 0.031360137835145,
      "grad_norm": 0.14955051243305206,
      "learning_rate": 5.534151789310904e-05,
      "loss": 0.0341,
      "relation_loss": 0.002701622014865279,
      "selected_pairs": 26510.3,
      "state_query_norm": 43.039812850952146,
      "state_relation_loss": 0.00017089408356696367,
      "step": 14780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002701622014865279,
      "weighted_relation_loss": 0.002701622014865279
    },
    {
      "action_query_norm": 41.455223846435544,
      "action_relation_loss": 0.002447523083537817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.83536014967259,
      "flow_matching_loss": 0.028095726296305655,
      "grad_norm": 0.09663213044404984,
      "learning_rate": 5.528671441085376e-05,
      "loss": 0.0307,
      "relation_loss": 0.0026098887901753187,
      "selected_pairs": 26941.6,
      "state_query_norm": 43.03794059753418,
      "state_relation_loss": 0.00016236571682384237,
      "step": 14790,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0026098887901753187,
      "weighted_relation_loss": 0.0026098887901753187
    },
    {
      "action_query_norm": 41.40674591064453,
      "action_relation_loss": 0.0025307912030257283,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.844714686623012,
      "flow_matching_loss": 0.031607126258313654,
      "grad_norm": 0.08500514924526215,
      "learning_rate": 5.523190450474197e-05,
      "loss": 0.0343,
      "relation_loss": 0.00269531833473593,
      "selected_pairs": 26224.5,
      "state_query_norm": 43.03401756286621,
      "state_relation_loss": 0.0001645271186134778,
      "step": 14800,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.00269531833473593,
      "weighted_relation_loss": 0.00269531833473593
    },
    {
      "action_query_norm": 41.4221736907959,
      "action_relation_loss": 0.0026530269300565124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.854069223573433,
      "flow_matching_loss": 0.03451523035764694,
      "grad_norm": 0.12073744088411331,
      "learning_rate": 5.5177088241372865e-05,
      "loss": 0.0373,
      "relation_loss": 0.002823133207857609,
      "selected_pairs": 27159.1,
      "state_query_norm": 43.02204132080078,
      "state_relation_loss": 0.00017010627925628796,
      "step": 14810,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.002823133207857609,
      "weighted_relation_loss": 0.002823133207857609
    },
    {
      "action_query_norm": 41.4488109588623,
      "action_relation_loss": 0.0023851168574765325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.863423760523855,
      "flow_matching_loss": 0.029001163132488728,
      "grad_norm": 0.0913018137216568,
      "learning_rate": 5.512226568735338e-05,
      "loss": 0.0316,
      "relation_loss": 0.0025489156832918525,
      "selected_pairs": 26596.8,
      "state_query_norm": 43.034615707397464,
      "state_relation_loss": 0.00016379884036723524,
      "step": 14820,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0025489156832918525,
      "weighted_relation_loss": 0.0025489156832918525
    },
    {
      "action_query_norm": 41.43511123657227,
      "action_relation_loss": 0.0025736759649589657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.872778297474275,
      "flow_matching_loss": 0.030836624652147294,
      "grad_norm": 0.10529076308012009,
      "learning_rate": 5.506743690929809e-05,
      "loss": 0.0336,
      "relation_loss": 0.0027407252229750155,
      "selected_pairs": 26601.5,
      "state_query_norm": 43.030456161499025,
      "state_relation_loss": 0.00016704926238162444,
      "step": 14830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0027407252229750155,
      "weighted_relation_loss": 0.0027407252229750155
    },
    {
      "action_query_norm": 41.4616081237793,
      "action_relation_loss": 0.0024968019919469953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.882132834424697,
      "flow_matching_loss": 0.030459694750607013,
      "grad_norm": 0.0975336804986,
      "learning_rate": 5.501260197382912e-05,
      "loss": 0.0331,
      "relation_loss": 0.002669067960232496,
      "selected_pairs": 26644.9,
      "state_query_norm": 43.02113838195801,
      "state_relation_loss": 0.0001722659741062671,
      "step": 14840,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002669067960232496,
      "weighted_relation_loss": 0.002669067960232496
    },
    {
      "action_query_norm": 41.46588287353516,
      "action_relation_loss": 0.0024258267832919954,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.891487371375117,
      "flow_matching_loss": 0.031055612303316594,
      "grad_norm": 0.11435312032699585,
      "learning_rate": 5.49577609475761e-05,
      "loss": 0.0336,
      "relation_loss": 0.0025918398518115283,
      "selected_pairs": 26811.4,
      "state_query_norm": 43.02663459777832,
      "state_relation_loss": 0.00016601305833319202,
      "step": 14850,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0025918398518115283,
      "weighted_relation_loss": 0.0025918398518115283
    },
    {
      "action_query_norm": 41.44488716125488,
      "action_relation_loss": 0.0025103897554799915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.900841908325537,
      "flow_matching_loss": 0.030993316508829593,
      "grad_norm": 0.12697209417819977,
      "learning_rate": 5.490291389717603e-05,
      "loss": 0.0337,
      "relation_loss": 0.0026756422361359,
      "selected_pairs": 26966.9,
      "state_query_norm": 43.0314079284668,
      "state_relation_loss": 0.00016525246610399336,
      "step": 14860,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0026756422361359,
      "weighted_relation_loss": 0.0026756422361359
    },
    {
      "action_query_norm": 41.46325035095215,
      "action_relation_loss": 0.002375242533162236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.91019644527596,
      "flow_matching_loss": 0.03159539420157671,
      "grad_norm": 0.09548774361610413,
      "learning_rate": 5.484806088927329e-05,
      "loss": 0.0341,
      "relation_loss": 0.0025431963847950103,
      "selected_pairs": 27011.7,
      "state_query_norm": 43.04026222229004,
      "state_relation_loss": 0.0001679538530879654,
      "step": 14870,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0025431963847950103,
      "weighted_relation_loss": 0.0025431963847950103
    },
    {
      "action_query_norm": 41.44061813354492,
      "action_relation_loss": 0.0023200966417789457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.91955098222638,
      "flow_matching_loss": 0.02975611463189125,
      "grad_norm": 0.08618982881307602,
      "learning_rate": 5.479320199051942e-05,
      "loss": 0.0322,
      "relation_loss": 0.002491886890493333,
      "selected_pairs": 26893.4,
      "state_query_norm": 43.021836853027345,
      "state_relation_loss": 0.0001717902792734094,
      "step": 14880,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.002491886890493333,
      "weighted_relation_loss": 0.002491886890493333
    },
    {
      "action_query_norm": 41.45627212524414,
      "action_relation_loss": 0.002342474670149386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.928905519176801,
      "flow_matching_loss": 0.030428218841552734,
      "grad_norm": 0.11981292814016342,
      "learning_rate": 5.473833726757314e-05,
      "loss": 0.0329,
      "relation_loss": 0.0025070411153137683,
      "selected_pairs": 27177.7,
      "state_query_norm": 43.01782875061035,
      "state_relation_loss": 0.00016456644952995704,
      "step": 14890,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0025070411153137683,
      "weighted_relation_loss": 0.0025070411153137683
    },
    {
      "action_query_norm": 41.394722366333006,
      "action_relation_loss": 0.002448116196319461,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.938260056127222,
      "flow_matching_loss": 0.031732188537716866,
      "grad_norm": 0.10827654600143433,
      "learning_rate": 5.4683466787100304e-05,
      "loss": 0.0343,
      "relation_loss": 0.0026177807012572885,
      "selected_pairs": 26752.6,
      "state_query_norm": 43.01240577697754,
      "state_relation_loss": 0.00016966454277280718,
      "step": 14900,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0026177807012572885,
      "weighted_relation_loss": 0.0026177807012572885
    },
    {
      "action_query_norm": 41.39116439819336,
      "action_relation_loss": 0.0024813609197735786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.947614593077642,
      "flow_matching_loss": 0.02982050608843565,
      "grad_norm": 0.10380230098962784,
      "learning_rate": 5.462859061577369e-05,
      "loss": 0.0325,
      "relation_loss": 0.002655232045799494,
      "selected_pairs": 26453.7,
      "state_query_norm": 43.01723403930664,
      "state_relation_loss": 0.00017387112602591516,
      "step": 14910,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002655232045799494,
      "weighted_relation_loss": 0.002655232045799494
    },
    {
      "action_query_norm": 41.415570449829104,
      "action_relation_loss": 0.002631340967491269,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.956969130028064,
      "flow_matching_loss": 0.031064379960298538,
      "grad_norm": 0.12101740390062332,
      "learning_rate": 5.4573708820273026e-05,
      "loss": 0.0339,
      "relation_loss": 0.0028055035043507816,
      "selected_pairs": 27289.4,
      "state_query_norm": 43.017372131347656,
      "state_relation_loss": 0.00017416254995623605,
      "step": 14920,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0028055035043507816,
      "weighted_relation_loss": 0.0028055035043507816
    },
    {
      "action_query_norm": 41.40769500732422,
      "action_relation_loss": 0.002440994721837342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.966323666978484,
      "flow_matching_loss": 0.030292125791311263,
      "grad_norm": 0.1110200360417366,
      "learning_rate": 5.451882146728488e-05,
      "loss": 0.0329,
      "relation_loss": 0.0026109387166798116,
      "selected_pairs": 27504.0,
      "state_query_norm": 43.002711868286134,
      "state_relation_loss": 0.00016994400939438493,
      "step": 14930,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0026109387166798116,
      "weighted_relation_loss": 0.0026109387166798116
    },
    {
      "action_query_norm": 41.43537902832031,
      "action_relation_loss": 0.002477333741262555,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.975678203928906,
      "flow_matching_loss": 0.029875137284398078,
      "grad_norm": 0.12444251775741577,
      "learning_rate": 5.446392862350255e-05,
      "loss": 0.0325,
      "relation_loss": 0.002638457203283906,
      "selected_pairs": 26686.5,
      "state_query_norm": 43.020616149902345,
      "state_relation_loss": 0.0001611234329175204,
      "step": 14940,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002638457203283906,
      "weighted_relation_loss": 0.002638457203283906
    },
    {
      "action_query_norm": 41.39361038208008,
      "action_relation_loss": 0.002434714022092521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.985032740879326,
      "flow_matching_loss": 0.029502338357269765,
      "grad_norm": 0.14725758135318756,
      "learning_rate": 5.440903035562603e-05,
      "loss": 0.0321,
      "relation_loss": 0.0025893294718116523,
      "selected_pairs": 27024.4,
      "state_query_norm": 43.02208023071289,
      "state_relation_loss": 0.00015461542643606662,
      "step": 14950,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0025893294718116523,
      "weighted_relation_loss": 0.0025893294718116523
    },
    {
      "action_query_norm": 41.3810977935791,
      "action_relation_loss": 0.0025359497172757982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 13.994387277829748,
      "flow_matching_loss": 0.029262656532227992,
      "grad_norm": 0.12187071144580841,
      "learning_rate": 5.435412673036188e-05,
      "loss": 0.032,
      "relation_loss": 0.0026966565987095238,
      "selected_pairs": 26136.1,
      "state_query_norm": 43.01531791687012,
      "state_relation_loss": 0.0001607068843441084,
      "step": 14960,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0026966565987095238,
      "weighted_relation_loss": 0.0026966565987095238
    },
    {
      "action_query_norm": 41.400464630126955,
      "action_relation_loss": 0.0023740659235045314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.003741814780168,
      "flow_matching_loss": 0.03366111032664776,
      "grad_norm": 0.11455069482326508,
      "learning_rate": 5.429921781442318e-05,
      "loss": 0.0362,
      "relation_loss": 0.0025305562652647496,
      "selected_pairs": 24327.9,
      "state_query_norm": 43.011047744750975,
      "state_relation_loss": 0.0001564903010148555,
      "step": 14970,
      "valid_action_samples": 214.3,
      "weighted_auxiliary_loss": 0.0025305562652647496,
      "weighted_relation_loss": 0.0025305562652647496
    },
    {
      "action_query_norm": 41.43313751220703,
      "action_relation_loss": 0.002328155026771128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.013096351730589,
      "flow_matching_loss": 0.030167602561414243,
      "grad_norm": 0.10839974135160446,
      "learning_rate": 5.424430367452945e-05,
      "loss": 0.0327,
      "relation_loss": 0.0025016583502292635,
      "selected_pairs": 27086.2,
      "state_query_norm": 42.995859909057614,
      "state_relation_loss": 0.00017350329726468773,
      "step": 14980,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0025016583502292635,
      "weighted_relation_loss": 0.0025016583502292635
    },
    {
      "action_query_norm": 41.42660408020019,
      "action_relation_loss": 0.002334487740881741,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.02245088868101,
      "flow_matching_loss": 0.030128300189971924,
      "grad_norm": 0.12775641679763794,
      "learning_rate": 5.418938437740655e-05,
      "loss": 0.0326,
      "relation_loss": 0.002492466731928289,
      "selected_pairs": 26500.3,
      "state_query_norm": 43.00900497436523,
      "state_relation_loss": 0.0001579789648531005,
      "step": 14990,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002492466731928289,
      "weighted_relation_loss": 0.002492466731928289
    },
    {
      "action_query_norm": 41.39483909606933,
      "action_relation_loss": 0.0023938441299833357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.03180542563143,
      "flow_matching_loss": 0.029964335821568967,
      "grad_norm": 0.1271829903125763,
      "learning_rate": 5.413445998978658e-05,
      "loss": 0.0325,
      "relation_loss": 0.0025631014490500093,
      "selected_pairs": 27195.5,
      "state_query_norm": 43.00990180969238,
      "state_relation_loss": 0.00016925730160437523,
      "step": 15000,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0025631014490500093,
      "weighted_relation_loss": 0.0025631014490500093
    },
    {
      "action_query_norm": 41.385322952270506,
      "action_relation_loss": 0.00227470847312361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.041159962581853,
      "flow_matching_loss": 0.029862739145755768,
      "grad_norm": 0.11671029776334763,
      "learning_rate": 5.407953057840789e-05,
      "loss": 0.0323,
      "relation_loss": 0.0024338956689462067,
      "selected_pairs": 26386.7,
      "state_query_norm": 43.00817184448242,
      "state_relation_loss": 0.00015918719727778808,
      "step": 15010,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0024338956689462067,
      "weighted_relation_loss": 0.0024338956689462067
    },
    {
      "action_query_norm": 41.36498832702637,
      "action_relation_loss": 0.0022649332648143174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.050514499532273,
      "flow_matching_loss": 0.029878922924399375,
      "grad_norm": 0.09601347893476486,
      "learning_rate": 5.4024596210014853e-05,
      "loss": 0.0323,
      "relation_loss": 0.002432255819439888,
      "selected_pairs": 27927.0,
      "state_query_norm": 42.97785835266113,
      "state_relation_loss": 0.00016732254880480468,
      "step": 15020,
      "valid_action_samples": 236.8,
      "weighted_auxiliary_loss": 0.002432255819439888,
      "weighted_relation_loss": 0.002432255819439888
    },
    {
      "action_query_norm": 41.33726959228515,
      "action_relation_loss": 0.0023712610593065618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.059869036482693,
      "flow_matching_loss": 0.029573783837258817,
      "grad_norm": 0.11648114770650864,
      "learning_rate": 5.3969656951357935e-05,
      "loss": 0.0321,
      "relation_loss": 0.002526789018884301,
      "selected_pairs": 26389.8,
      "state_query_norm": 42.99521751403809,
      "state_relation_loss": 0.0001555279508465901,
      "step": 15030,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.002526789018884301,
      "weighted_relation_loss": 0.002526789018884301
    },
    {
      "action_query_norm": 41.374392700195315,
      "action_relation_loss": 0.0023189156781882048,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.069223573433115,
      "flow_matching_loss": 0.02822901923209429,
      "grad_norm": 0.10423168540000916,
      "learning_rate": 5.3914712869193504e-05,
      "loss": 0.0307,
      "relation_loss": 0.002473626984283328,
      "selected_pairs": 26648.8,
      "state_query_norm": 43.01819839477539,
      "state_relation_loss": 0.0001547113119158894,
      "step": 15040,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002473626984283328,
      "weighted_relation_loss": 0.002473626984283328
    },
    {
      "action_query_norm": 41.3661994934082,
      "action_relation_loss": 0.0023147563450038435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.078578110383535,
      "flow_matching_loss": 0.029019657336175442,
      "grad_norm": 0.11339303106069565,
      "learning_rate": 5.385976403028381e-05,
      "loss": 0.0315,
      "relation_loss": 0.0024731318233534695,
      "selected_pairs": 26926.5,
      "state_query_norm": 42.98641014099121,
      "state_relation_loss": 0.00015837545652175322,
      "step": 15050,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024731318233534695,
      "weighted_relation_loss": 0.0024731318233534695
    },
    {
      "action_query_norm": 41.33174057006836,
      "action_relation_loss": 0.0023728379281237723,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.087932647333957,
      "flow_matching_loss": 0.02914845123887062,
      "grad_norm": 0.09396027028560638,
      "learning_rate": 5.3804810501396875e-05,
      "loss": 0.0317,
      "relation_loss": 0.002531159413047135,
      "selected_pairs": 26620.7,
      "state_query_norm": 42.98603744506836,
      "state_relation_loss": 0.00015832149365451186,
      "step": 15060,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002531159413047135,
      "weighted_relation_loss": 0.002531159413047135
    },
    {
      "action_query_norm": 41.34365463256836,
      "action_relation_loss": 0.0023764848592691123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.097287184284378,
      "flow_matching_loss": 0.029222624748945235,
      "grad_norm": 0.12291073799133301,
      "learning_rate": 5.374985234930639e-05,
      "loss": 0.0318,
      "relation_loss": 0.002543779998086393,
      "selected_pairs": 26509.2,
      "state_query_norm": 42.987158966064456,
      "state_relation_loss": 0.00016729514172766357,
      "step": 15070,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002543779998086393,
      "weighted_relation_loss": 0.002543779998086393
    },
    {
      "action_query_norm": 41.405693435668944,
      "action_relation_loss": 0.0021967538516037164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.1066417212348,
      "flow_matching_loss": 0.02951185591518879,
      "grad_norm": 0.1285068541765213,
      "learning_rate": 5.3694889640791724e-05,
      "loss": 0.0319,
      "relation_loss": 0.0023622594540938734,
      "selected_pairs": 26578.4,
      "state_query_norm": 42.995919418334964,
      "state_relation_loss": 0.0001655056097661145,
      "step": 15080,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0023622594540938734,
      "weighted_relation_loss": 0.0023622594540938734
    },
    {
      "action_query_norm": 41.389212417602536,
      "action_relation_loss": 0.0024938286282122136,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.11599625818522,
      "flow_matching_loss": 0.029266800731420517,
      "grad_norm": 0.11848286539316177,
      "learning_rate": 5.3639922442637736e-05,
      "loss": 0.0319,
      "relation_loss": 0.00265656930860132,
      "selected_pairs": 27047.6,
      "state_query_norm": 42.99563407897949,
      "state_relation_loss": 0.00016274070803774522,
      "step": 15090,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00265656930860132,
      "weighted_relation_loss": 0.00265656930860132
    },
    {
      "action_query_norm": 41.364824295043945,
      "action_relation_loss": 0.002307430072687566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.12535079513564,
      "flow_matching_loss": 0.03177986741065979,
      "grad_norm": 0.14906327426433563,
      "learning_rate": 5.358495082163476e-05,
      "loss": 0.0342,
      "relation_loss": 0.00246764849871397,
      "selected_pairs": 27266.9,
      "state_query_norm": 42.99322395324707,
      "state_relation_loss": 0.00016021847841329873,
      "step": 15100,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.00246764849871397,
      "weighted_relation_loss": 0.00246764849871397
    },
    {
      "action_query_norm": 41.363743209838866,
      "action_relation_loss": 0.002438644599169493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.134705332086062,
      "flow_matching_loss": 0.031204788573086263,
      "grad_norm": 0.11312316358089447,
      "learning_rate": 5.35299748445785e-05,
      "loss": 0.0338,
      "relation_loss": 0.0026089650811627505,
      "selected_pairs": 26875.8,
      "state_query_norm": 42.98782730102539,
      "state_relation_loss": 0.00017032049363479017,
      "step": 15110,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0026089650811627505,
      "weighted_relation_loss": 0.0026089650811627505
    },
    {
      "action_query_norm": 41.3474308013916,
      "action_relation_loss": 0.0022454436868429182,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.144059869036482,
      "flow_matching_loss": 0.031022540666162966,
      "grad_norm": 0.10600972175598145,
      "learning_rate": 5.3474994578269944e-05,
      "loss": 0.0334,
      "relation_loss": 0.0024128060322254896,
      "selected_pairs": 27082.0,
      "state_query_norm": 42.98575325012207,
      "state_relation_loss": 0.000167362364300061,
      "step": 15120,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0024128060322254896,
      "weighted_relation_loss": 0.0024128060322254896
    },
    {
      "action_query_norm": 41.320991897583006,
      "action_relation_loss": 0.002390649914741516,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.153414405986904,
      "flow_matching_loss": 0.029843921400606634,
      "grad_norm": 0.13345122337341309,
      "learning_rate": 5.34200100895153e-05,
      "loss": 0.0324,
      "relation_loss": 0.0025551713770255446,
      "selected_pairs": 27270.8,
      "state_query_norm": 42.98528633117676,
      "state_relation_loss": 0.00016452145937364548,
      "step": 15130,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0025551713770255446,
      "weighted_relation_loss": 0.0025551713770255446
    },
    {
      "action_query_norm": 41.31506881713867,
      "action_relation_loss": 0.0024175051832571624,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.162768942937324,
      "flow_matching_loss": 0.03079098407179117,
      "grad_norm": 0.09852202236652374,
      "learning_rate": 5.3365021445125916e-05,
      "loss": 0.0334,
      "relation_loss": 0.0025875585386529567,
      "selected_pairs": 26434.4,
      "state_query_norm": 42.97543640136719,
      "state_relation_loss": 0.00017005333938868716,
      "step": 15140,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0025875585386529567,
      "weighted_relation_loss": 0.0025875585386529567
    },
    {
      "action_query_norm": 41.29890060424805,
      "action_relation_loss": 0.0023341595428064466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.172123479887745,
      "flow_matching_loss": 0.02919173687696457,
      "grad_norm": 0.10294768959283829,
      "learning_rate": 5.331002871191817e-05,
      "loss": 0.0317,
      "relation_loss": 0.0025145099963992832,
      "selected_pairs": 27185.7,
      "state_query_norm": 42.96765365600586,
      "state_relation_loss": 0.00018035044049611316,
      "step": 15150,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0025145099963992832,
      "weighted_relation_loss": 0.0025145099963992832
    },
    {
      "action_query_norm": 41.33688049316406,
      "action_relation_loss": 0.0023040930973365903,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.181478016838167,
      "flow_matching_loss": 0.030480059050023555,
      "grad_norm": 0.13760046660900116,
      "learning_rate": 5.3255031956713443e-05,
      "loss": 0.033,
      "relation_loss": 0.0024775164900347592,
      "selected_pairs": 26733.3,
      "state_query_norm": 42.975398254394534,
      "state_relation_loss": 0.00017342338105663657,
      "step": 15160,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0024775164900347592,
      "weighted_relation_loss": 0.0024775164900347592
    },
    {
      "action_query_norm": 41.30478363037109,
      "action_relation_loss": 0.002227455098181963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.190832553788587,
      "flow_matching_loss": 0.030807141959667207,
      "grad_norm": 0.10029821842908859,
      "learning_rate": 5.320003124633795e-05,
      "loss": 0.0332,
      "relation_loss": 0.0024030163884162905,
      "selected_pairs": 26667.5,
      "state_query_norm": 42.97430419921875,
      "state_relation_loss": 0.0001755613018758595,
      "step": 15170,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0024030163884162905,
      "weighted_relation_loss": 0.0024030163884162905
    },
    {
      "action_query_norm": 41.313266372680665,
      "action_relation_loss": 0.0023926730267703534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.200187090739009,
      "flow_matching_loss": 0.02944243848323822,
      "grad_norm": 0.1001153290271759,
      "learning_rate": 5.314502664762274e-05,
      "loss": 0.032,
      "relation_loss": 0.0025616985745728015,
      "selected_pairs": 27406.4,
      "state_query_norm": 42.9716236114502,
      "state_relation_loss": 0.00016902553034015,
      "step": 15180,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0025616985745728015,
      "weighted_relation_loss": 0.0025616985745728015
    },
    {
      "action_query_norm": 41.316894149780275,
      "action_relation_loss": 0.0023624743102118374,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.209541627689429,
      "flow_matching_loss": 0.02944932635873556,
      "grad_norm": 0.11555277556180954,
      "learning_rate": 5.3090018227403604e-05,
      "loss": 0.032,
      "relation_loss": 0.0025283002760261297,
      "selected_pairs": 26570.3,
      "state_query_norm": 42.98064956665039,
      "state_relation_loss": 0.00016582594253122807,
      "step": 15190,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0025283002760261297,
      "weighted_relation_loss": 0.0025283002760261297
    },
    {
      "action_query_norm": 41.383517837524415,
      "action_relation_loss": 0.0023413832765072583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.218896164639851,
      "flow_matching_loss": 0.029655744694173335,
      "grad_norm": 0.09838804602622986,
      "learning_rate": 5.3035006052520955e-05,
      "loss": 0.0322,
      "relation_loss": 0.002521605580113828,
      "selected_pairs": 27736.8,
      "state_query_norm": 42.972055435180664,
      "state_relation_loss": 0.00018022231961367652,
      "step": 15200,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.002521605580113828,
      "weighted_relation_loss": 0.002521605580113828
    },
    {
      "action_query_norm": 41.33180313110351,
      "action_relation_loss": 0.002334106503985822,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.228250701590271,
      "flow_matching_loss": 0.029632638953626157,
      "grad_norm": 0.0830380991101265,
      "learning_rate": 5.297999018981976e-05,
      "loss": 0.0321,
      "relation_loss": 0.002500439225696027,
      "selected_pairs": 27057.3,
      "state_query_norm": 42.968130111694336,
      "state_relation_loss": 0.00016633272462058813,
      "step": 15210,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002500439225696027,
      "weighted_relation_loss": 0.002500439225696027
    },
    {
      "action_query_norm": 41.30115432739258,
      "action_relation_loss": 0.002397141442634165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.237605238540691,
      "flow_matching_loss": 0.03004903271794319,
      "grad_norm": 0.15903420746326447,
      "learning_rate": 5.2924970706149505e-05,
      "loss": 0.0326,
      "relation_loss": 0.0025670280447229743,
      "selected_pairs": 26092.1,
      "state_query_norm": 42.97954063415527,
      "state_relation_loss": 0.0001698866268270649,
      "step": 15220,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.0025670280447229743,
      "weighted_relation_loss": 0.0025670280447229743
    },
    {
      "action_query_norm": 41.335672760009764,
      "action_relation_loss": 0.0022543593309819697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.246959775491113,
      "flow_matching_loss": 0.027614871971309185,
      "grad_norm": 0.11064782738685608,
      "learning_rate": 5.2869947668364014e-05,
      "loss": 0.03,
      "relation_loss": 0.002427562396042049,
      "selected_pairs": 26551.4,
      "state_query_norm": 42.97384262084961,
      "state_relation_loss": 0.00017320306360488758,
      "step": 15230,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.002427562396042049,
      "weighted_relation_loss": 0.002427562396042049
    },
    {
      "action_query_norm": 41.326040267944336,
      "action_relation_loss": 0.0023113515693694354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.256314312441534,
      "flow_matching_loss": 0.030150520987808703,
      "grad_norm": 0.07532810419797897,
      "learning_rate": 5.2814921143321504e-05,
      "loss": 0.0326,
      "relation_loss": 0.002466349513269961,
      "selected_pairs": 26454.3,
      "state_query_norm": 42.96452941894531,
      "state_relation_loss": 0.00015499790606554598,
      "step": 15240,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002466349513269961,
      "weighted_relation_loss": 0.002466349513269961
    },
    {
      "action_query_norm": 41.29961891174317,
      "action_relation_loss": 0.002301961136981845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.265668849391956,
      "flow_matching_loss": 0.029624277353286745,
      "grad_norm": 0.11502344161272049,
      "learning_rate": 5.275989119788436e-05,
      "loss": 0.0321,
      "relation_loss": 0.002464156621135771,
      "selected_pairs": 26371.2,
      "state_query_norm": 42.95566024780273,
      "state_relation_loss": 0.00016219551616813987,
      "step": 15250,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.002464156621135771,
      "weighted_relation_loss": 0.002464156621135771
    },
    {
      "action_query_norm": 41.307661819458005,
      "action_relation_loss": 0.0022986354772001503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.275023386342376,
      "flow_matching_loss": 0.02852322570979595,
      "grad_norm": 0.15193675458431244,
      "learning_rate": 5.270485789891919e-05,
      "loss": 0.031,
      "relation_loss": 0.0024621783290058375,
      "selected_pairs": 26929.4,
      "state_query_norm": 42.96787528991699,
      "state_relation_loss": 0.0001635428605368361,
      "step": 15260,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024621783290058375,
      "weighted_relation_loss": 0.0024621783290058375
    },
    {
      "action_query_norm": 41.288163375854495,
      "action_relation_loss": 0.002253209997434169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.284377923292798,
      "flow_matching_loss": 0.030459246411919595,
      "grad_norm": 0.12682382762432098,
      "learning_rate": 5.264982131329661e-05,
      "loss": 0.0329,
      "relation_loss": 0.002417596406303346,
      "selected_pairs": 27337.0,
      "state_query_norm": 42.96237640380859,
      "state_relation_loss": 0.00016438642487628385,
      "step": 15270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.002417596406303346,
      "weighted_relation_loss": 0.002417596406303346
    },
    {
      "action_query_norm": 41.30880355834961,
      "action_relation_loss": 0.002312990138307214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.293732460243218,
      "flow_matching_loss": 0.02978830933570862,
      "grad_norm": 0.1146930381655693,
      "learning_rate": 5.259478150789128e-05,
      "loss": 0.0323,
      "relation_loss": 0.0024860937148332594,
      "selected_pairs": 26663.3,
      "state_query_norm": 42.968780517578125,
      "state_relation_loss": 0.00017310356488451362,
      "step": 15280,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0024860937148332594,
      "weighted_relation_loss": 0.0024860937148332594
    },
    {
      "action_query_norm": 41.319051361083986,
      "action_relation_loss": 0.002355214348062873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.303086997193638,
      "flow_matching_loss": 0.029456140100955965,
      "grad_norm": 0.12014800310134888,
      "learning_rate": 5.253973854958173e-05,
      "loss": 0.032,
      "relation_loss": 0.0025182035751640797,
      "selected_pairs": 27337.3,
      "state_query_norm": 42.96622314453125,
      "state_relation_loss": 0.00016298923001158982,
      "step": 15290,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0025182035751640797,
      "weighted_relation_loss": 0.0025182035751640797
    },
    {
      "action_query_norm": 41.30305213928223,
      "action_relation_loss": 0.002293435507453978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.31244153414406,
      "flow_matching_loss": 0.0295817282050848,
      "grad_norm": 0.09700033068656921,
      "learning_rate": 5.2484692505250376e-05,
      "loss": 0.032,
      "relation_loss": 0.002456707088276744,
      "selected_pairs": 26891.5,
      "state_query_norm": 42.96093215942383,
      "state_relation_loss": 0.00016327155753970146,
      "step": 15300,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.002456707088276744,
      "weighted_relation_loss": 0.002456707088276744
    },
    {
      "action_query_norm": 41.276973724365234,
      "action_relation_loss": 0.0023247188655659556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.32179607109448,
      "flow_matching_loss": 0.029097563214600088,
      "grad_norm": 0.13410809636116028,
      "learning_rate": 5.242964344178333e-05,
      "loss": 0.0316,
      "relation_loss": 0.0024886684492230414,
      "selected_pairs": 27132.1,
      "state_query_norm": 42.955693054199216,
      "state_relation_loss": 0.00016394958656746895,
      "step": 15310,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0024886684492230414,
      "weighted_relation_loss": 0.0024886684492230414
    },
    {
      "action_query_norm": 41.26673278808594,
      "action_relation_loss": 0.0022562826285138725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.331150608044902,
      "flow_matching_loss": 0.031178838945925234,
      "grad_norm": 0.11239799112081528,
      "learning_rate": 5.237459142607041e-05,
      "loss": 0.0336,
      "relation_loss": 0.0024142791284248234,
      "selected_pairs": 27429.6,
      "state_query_norm": 42.96146125793457,
      "state_relation_loss": 0.00015799652464920655,
      "step": 15320,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0024142791284248234,
      "weighted_relation_loss": 0.0024142791284248234
    },
    {
      "action_query_norm": 41.28113059997558,
      "action_relation_loss": 0.0022966293385252357,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.340505144995323,
      "flow_matching_loss": 0.02789749763906002,
      "grad_norm": 0.10419735312461853,
      "learning_rate": 5.231953652500498e-05,
      "loss": 0.0304,
      "relation_loss": 0.002459580171853304,
      "selected_pairs": 26352.8,
      "state_query_norm": 42.95825004577637,
      "state_relation_loss": 0.00016295083478325977,
      "step": 15330,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002459580171853304,
      "weighted_relation_loss": 0.002459580171853304
    },
    {
      "action_query_norm": 41.28244323730469,
      "action_relation_loss": 0.0023013010621070864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.349859681945743,
      "flow_matching_loss": 0.02899544332176447,
      "grad_norm": 0.10446260869503021,
      "learning_rate": 5.2264478805483976e-05,
      "loss": 0.0315,
      "relation_loss": 0.002455496578477323,
      "selected_pairs": 26985.5,
      "state_query_norm": 42.95330696105957,
      "state_relation_loss": 0.00015419547708006576,
      "step": 15340,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002455496578477323,
      "weighted_relation_loss": 0.002455496578477323
    },
    {
      "action_query_norm": 41.24791488647461,
      "action_relation_loss": 0.002300493489019573,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.359214218896165,
      "flow_matching_loss": 0.030901887640357017,
      "grad_norm": 0.12217115610837936,
      "learning_rate": 5.220941833440768e-05,
      "loss": 0.0334,
      "relation_loss": 0.00246115883346647,
      "selected_pairs": 26387.7,
      "state_query_norm": 42.96107864379883,
      "state_relation_loss": 0.0001606653706403449,
      "step": 15350,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.00246115883346647,
      "weighted_relation_loss": 0.00246115883346647
    },
    {
      "action_query_norm": 41.2919506072998,
      "action_relation_loss": 0.002295263670384884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.368568755846585,
      "flow_matching_loss": 0.02768896222114563,
      "grad_norm": 0.12218828499317169,
      "learning_rate": 5.215435517867977e-05,
      "loss": 0.0301,
      "relation_loss": 0.0024495176738128068,
      "selected_pairs": 26984.3,
      "state_query_norm": 42.95314750671387,
      "state_relation_loss": 0.00015425400633830577,
      "step": 15360,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0024495176738128068,
      "weighted_relation_loss": 0.0024495176738128068
    },
    {
      "action_query_norm": 41.254146575927734,
      "action_relation_loss": 0.002305971528403461,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.377923292797007,
      "flow_matching_loss": 0.029706747457385062,
      "grad_norm": 0.09387662261724472,
      "learning_rate": 5.209928940520719e-05,
      "loss": 0.0322,
      "relation_loss": 0.002465247921645641,
      "selected_pairs": 26897.5,
      "state_query_norm": 42.94864959716797,
      "state_relation_loss": 0.00015927641215967013,
      "step": 15370,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.002465247921645641,
      "weighted_relation_loss": 0.002465247921645641
    },
    {
      "action_query_norm": 41.25102081298828,
      "action_relation_loss": 0.0023944755783304574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.387277829747427,
      "flow_matching_loss": 0.029709304682910442,
      "grad_norm": 0.12437231838703156,
      "learning_rate": 5.204422108090004e-05,
      "loss": 0.0323,
      "relation_loss": 0.0025577859254553915,
      "selected_pairs": 26852.0,
      "state_query_norm": 42.934981155395505,
      "state_relation_loss": 0.00016331036749761552,
      "step": 15380,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0025577859254553915,
      "weighted_relation_loss": 0.0025577859254553915
    },
    {
      "action_query_norm": 41.280926895141604,
      "action_relation_loss": 0.002399228932335973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.39663236669785,
      "flow_matching_loss": 0.029347293823957444,
      "grad_norm": 0.13854829967021942,
      "learning_rate": 5.19891502726715e-05,
      "loss": 0.0319,
      "relation_loss": 0.002570281131193042,
      "selected_pairs": 27217.9,
      "state_query_norm": 42.94170913696289,
      "state_relation_loss": 0.0001710521915811114,
      "step": 15390,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.002570281131193042,
      "weighted_relation_loss": 0.002570281131193042
    },
    {
      "action_query_norm": 41.25807647705078,
      "action_relation_loss": 0.002309787180274725,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.40598690364827,
      "flow_matching_loss": 0.03066202439367771,
      "grad_norm": 0.09175414592027664,
      "learning_rate": 5.193407704743782e-05,
      "loss": 0.0331,
      "relation_loss": 0.0024731843499466778,
      "selected_pairs": 27246.0,
      "state_query_norm": 42.940686798095705,
      "state_relation_loss": 0.00016339713911293075,
      "step": 15400,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0024731843499466778,
      "weighted_relation_loss": 0.0024731843499466778
    },
    {
      "action_query_norm": 41.24236831665039,
      "action_relation_loss": 0.0023240454494953154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.41534144059869,
      "flow_matching_loss": 0.02876077815890312,
      "grad_norm": 0.10743241757154465,
      "learning_rate": 5.187900147211815e-05,
      "loss": 0.0312,
      "relation_loss": 0.0024804445682093503,
      "selected_pairs": 26906.2,
      "state_query_norm": 42.928009033203125,
      "state_relation_loss": 0.00015639910998288542,
      "step": 15410,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0024804445682093503,
      "weighted_relation_loss": 0.0024804445682093503
    },
    {
      "action_query_norm": 41.28053932189941,
      "action_relation_loss": 0.0023087776615284385,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.424695977549112,
      "flow_matching_loss": 0.02749369740486145,
      "grad_norm": 0.1373436450958252,
      "learning_rate": 5.182392361363453e-05,
      "loss": 0.03,
      "relation_loss": 0.0024642099160701036,
      "selected_pairs": 26757.4,
      "state_query_norm": 42.93359565734863,
      "state_relation_loss": 0.000155432298197411,
      "step": 15420,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0024642099160701036,
      "weighted_relation_loss": 0.0024642099160701036
    },
    {
      "action_query_norm": 41.2662052154541,
      "action_relation_loss": 0.002306127198971808,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.434050514499532,
      "flow_matching_loss": 0.029317865148186684,
      "grad_norm": 0.09855061024427414,
      "learning_rate": 5.176884353891172e-05,
      "loss": 0.0318,
      "relation_loss": 0.0024713594233617187,
      "selected_pairs": 26598.6,
      "state_query_norm": 42.93187408447265,
      "state_relation_loss": 0.00016523222875548528,
      "step": 15430,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0024713594233617187,
      "weighted_relation_loss": 0.0024713594233617187
    },
    {
      "action_query_norm": 41.26071548461914,
      "action_relation_loss": 0.00234024110250175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.443405051449954,
      "flow_matching_loss": 0.028586569987237454,
      "grad_norm": 0.13550549745559692,
      "learning_rate": 5.171376131487722e-05,
      "loss": 0.0311,
      "relation_loss": 0.0024877714226022364,
      "selected_pairs": 26835.8,
      "state_query_norm": 42.943216705322264,
      "state_relation_loss": 0.00014753031282452867,
      "step": 15440,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0024877714226022364,
      "weighted_relation_loss": 0.0024877714226022364
    },
    {
      "action_query_norm": 41.23624114990234,
      "action_relation_loss": 0.0024421366630122065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.452759588400374,
      "flow_matching_loss": 0.029823427088558674,
      "grad_norm": 0.10729727149009705,
      "learning_rate": 5.165867700846113e-05,
      "loss": 0.0324,
      "relation_loss": 0.0026014620438218118,
      "selected_pairs": 26554.4,
      "state_query_norm": 42.92885894775391,
      "state_relation_loss": 0.00015932540700305253,
      "step": 15450,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0026014620438218118,
      "weighted_relation_loss": 0.0026014620438218118
    },
    {
      "action_query_norm": 41.21988143920898,
      "action_relation_loss": 0.0023385802283883094,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.462114125350794,
      "flow_matching_loss": 0.02921423353254795,
      "grad_norm": 0.13239933550357819,
      "learning_rate": 5.160359068659606e-05,
      "loss": 0.0317,
      "relation_loss": 0.0024964520474895836,
      "selected_pairs": 27078.4,
      "state_query_norm": 42.937618255615234,
      "state_relation_loss": 0.0001578718321979977,
      "step": 15460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0024964520474895836,
      "weighted_relation_loss": 0.0024964520474895836
    },
    {
      "action_query_norm": 41.23991394042969,
      "action_relation_loss": 0.0021618218859657647,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.471468662301216,
      "flow_matching_loss": 0.03220948241651058,
      "grad_norm": 0.11335728317499161,
      "learning_rate": 5.154850241621712e-05,
      "loss": 0.0345,
      "relation_loss": 0.002322232723236084,
      "selected_pairs": 26806.7,
      "state_query_norm": 42.93914031982422,
      "state_relation_loss": 0.00016041085909819232,
      "step": 15470,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002322232723236084,
      "weighted_relation_loss": 0.002322232723236084
    },
    {
      "action_query_norm": 41.181892013549806,
      "action_relation_loss": 0.002315632323734462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.480823199251637,
      "flow_matching_loss": 0.030990398302674294,
      "grad_norm": 0.1472698301076889,
      "learning_rate": 5.149341226426172e-05,
      "loss": 0.0335,
      "relation_loss": 0.0024898885982111096,
      "selected_pairs": 26925.2,
      "state_query_norm": 42.92241020202637,
      "state_relation_loss": 0.00017425626720068976,
      "step": 15480,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024898885982111096,
      "weighted_relation_loss": 0.0024898885982111096
    },
    {
      "action_query_norm": 41.21837120056152,
      "action_relation_loss": 0.0022978171473369002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.490177736202059,
      "flow_matching_loss": 0.02839848678559065,
      "grad_norm": 0.13310392200946808,
      "learning_rate": 5.14383202976696e-05,
      "loss": 0.0309,
      "relation_loss": 0.002460515801794827,
      "selected_pairs": 27174.2,
      "state_query_norm": 42.91330795288086,
      "state_relation_loss": 0.00016269866027869285,
      "step": 15490,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.002460515801794827,
      "weighted_relation_loss": 0.002460515801794827
    },
    {
      "action_query_norm": 41.181410217285155,
      "action_relation_loss": 0.0023237363900989296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.499532273152479,
      "flow_matching_loss": 0.030303737334907054,
      "grad_norm": 0.13724219799041748,
      "learning_rate": 5.1383226583382684e-05,
      "loss": 0.0328,
      "relation_loss": 0.0024935051100328566,
      "selected_pairs": 26833.9,
      "state_query_norm": 42.926117706298825,
      "state_relation_loss": 0.00016976871265796944,
      "step": 15500,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0024935051100328566,
      "weighted_relation_loss": 0.0024935051100328566
    },
    {
      "action_query_norm": 41.220791244506835,
      "action_relation_loss": 0.0023261882830411195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.5088868101029,
      "flow_matching_loss": 0.02863693665713072,
      "grad_norm": 0.13315434753894806,
      "learning_rate": 5.132813118834504e-05,
      "loss": 0.0311,
      "relation_loss": 0.0024951423518359663,
      "selected_pairs": 26923.3,
      "state_query_norm": 42.92774620056152,
      "state_relation_loss": 0.00016895404551178216,
      "step": 15510,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024951423518359663,
      "weighted_relation_loss": 0.0024951423518359663
    },
    {
      "action_query_norm": 41.210488891601564,
      "action_relation_loss": 0.0023341116728261115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.518241347053321,
      "flow_matching_loss": 0.028749627619981767,
      "grad_norm": 0.12162408232688904,
      "learning_rate": 5.1273034179502777e-05,
      "loss": 0.0313,
      "relation_loss": 0.0025113811250776054,
      "selected_pairs": 26917.3,
      "state_query_norm": 42.90745620727539,
      "state_relation_loss": 0.0001772694566170685,
      "step": 15520,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0025113811250776054,
      "weighted_relation_loss": 0.0025113811250776054
    },
    {
      "action_query_norm": 41.17873878479004,
      "action_relation_loss": 0.0023740083444863556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.527595884003741,
      "flow_matching_loss": 0.030166414752602577,
      "grad_norm": 0.11947061866521835,
      "learning_rate": 5.1217935623803945e-05,
      "loss": 0.0327,
      "relation_loss": 0.0025340714724734425,
      "selected_pairs": 27210.1,
      "state_query_norm": 42.92861824035644,
      "state_relation_loss": 0.00016006309306249022,
      "step": 15530,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0025340714724734425,
      "weighted_relation_loss": 0.0025340714724734425
    },
    {
      "action_query_norm": 41.207230758666995,
      "action_relation_loss": 0.0022623013239353893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.536950420954163,
      "flow_matching_loss": 0.027661681920289994,
      "grad_norm": 0.15744057297706604,
      "learning_rate": 5.1162835588198476e-05,
      "loss": 0.0301,
      "relation_loss": 0.002426032768562436,
      "selected_pairs": 26991.9,
      "state_query_norm": 42.9315185546875,
      "state_relation_loss": 0.00016373144753742964,
      "step": 15540,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.002426032768562436,
      "weighted_relation_loss": 0.002426032768562436
    },
    {
      "action_query_norm": 41.22012367248535,
      "action_relation_loss": 0.002262650732882321,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.546304957904583,
      "flow_matching_loss": 0.02878310400992632,
      "grad_norm": 0.1175604984164238,
      "learning_rate": 5.110773413963813e-05,
      "loss": 0.0312,
      "relation_loss": 0.0024184233509004114,
      "selected_pairs": 26785.5,
      "state_query_norm": 42.91436767578125,
      "state_relation_loss": 0.0001557725903694518,
      "step": 15550,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0024184233509004114,
      "weighted_relation_loss": 0.0024184233509004114
    },
    {
      "action_query_norm": 41.19505157470703,
      "action_relation_loss": 0.0021939845173619688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.555659494855005,
      "flow_matching_loss": 0.030059413611888887,
      "grad_norm": 0.09946043789386749,
      "learning_rate": 5.105263134507636e-05,
      "loss": 0.0324,
      "relation_loss": 0.0023438441334292293,
      "selected_pairs": 26803.6,
      "state_query_norm": 42.91164207458496,
      "state_relation_loss": 0.00014985965535743161,
      "step": 15560,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0023438441334292293,
      "weighted_relation_loss": 0.0023438441334292293
    },
    {
      "action_query_norm": 41.19878044128418,
      "action_relation_loss": 0.002132594946306199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.565014031805426,
      "flow_matching_loss": 0.02883633729070425,
      "grad_norm": 0.09221082180738449,
      "learning_rate": 5.099752727146824e-05,
      "loss": 0.0311,
      "relation_loss": 0.002288232836872339,
      "selected_pairs": 26345.7,
      "state_query_norm": 42.90175895690918,
      "state_relation_loss": 0.00015563785709673538,
      "step": 15570,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002288232836872339,
      "weighted_relation_loss": 0.002288232836872339
    },
    {
      "action_query_norm": 41.21460952758789,
      "action_relation_loss": 0.002201484446413815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.574368568755848,
      "flow_matching_loss": 0.030397436767816543,
      "grad_norm": 0.125510111451149,
      "learning_rate": 5.094242198577042e-05,
      "loss": 0.0327,
      "relation_loss": 0.0023482832126319408,
      "selected_pairs": 26460.5,
      "state_query_norm": 42.897609329223634,
      "state_relation_loss": 0.000146798777859658,
      "step": 15580,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0023482832126319408,
      "weighted_relation_loss": 0.0023482832126319408
    },
    {
      "action_query_norm": 41.18054542541504,
      "action_relation_loss": 0.00215472390409559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.583723105706268,
      "flow_matching_loss": 0.02805983182042837,
      "grad_norm": 0.11677093803882599,
      "learning_rate": 5.088731555494102e-05,
      "loss": 0.0304,
      "relation_loss": 0.002301568165421486,
      "selected_pairs": 26974.0,
      "state_query_norm": 42.90964965820312,
      "state_relation_loss": 0.0001468442613258958,
      "step": 15590,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002301568165421486,
      "weighted_relation_loss": 0.002301568165421486
    },
    {
      "action_query_norm": 41.19911689758301,
      "action_relation_loss": 0.002347953338176012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.593077642656688,
      "flow_matching_loss": 0.03069338612258434,
      "grad_norm": 0.11429457366466522,
      "learning_rate": 5.0832208045939556e-05,
      "loss": 0.0332,
      "relation_loss": 0.002500655874609947,
      "selected_pairs": 25839.1,
      "state_query_norm": 42.89355888366699,
      "state_relation_loss": 0.00015270253061316907,
      "step": 15600,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.002500655874609947,
      "weighted_relation_loss": 0.002500655874609947
    },
    {
      "action_query_norm": 41.15908927917481,
      "action_relation_loss": 0.0023230952909216286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.60243217960711,
      "flow_matching_loss": 0.02893178295344114,
      "grad_norm": 0.14460091292858124,
      "learning_rate": 5.0777099525726844e-05,
      "loss": 0.0314,
      "relation_loss": 0.002473013871349394,
      "selected_pairs": 26663.7,
      "state_query_norm": 42.88891525268555,
      "state_relation_loss": 0.0001499185644206591,
      "step": 15610,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.002473013871349394,
      "weighted_relation_loss": 0.002473013871349394
    },
    {
      "action_query_norm": 41.19443473815918,
      "action_relation_loss": 0.002221853961236775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.61178671655753,
      "flow_matching_loss": 0.02907407619059086,
      "grad_norm": 0.13452951610088348,
      "learning_rate": 5.072199006126494e-05,
      "loss": 0.0314,
      "relation_loss": 0.0023695371113717557,
      "selected_pairs": 26808.0,
      "state_query_norm": 42.89912147521973,
      "state_relation_loss": 0.00014768313994864,
      "step": 15620,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0023695371113717557,
      "weighted_relation_loss": 0.0023695371113717557
    },
    {
      "action_query_norm": 41.20175552368164,
      "action_relation_loss": 0.0023245861404575406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.621141253507952,
      "flow_matching_loss": 0.029580073431134224,
      "grad_norm": 0.12514546513557434,
      "learning_rate": 5.0666879719517026e-05,
      "loss": 0.0321,
      "relation_loss": 0.0024855352705344556,
      "selected_pairs": 27228.9,
      "state_query_norm": 42.910438537597656,
      "state_relation_loss": 0.00016094915335997939,
      "step": 15630,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0024855352705344556,
      "weighted_relation_loss": 0.0024855352705344556
    },
    {
      "action_query_norm": 41.21912307739258,
      "action_relation_loss": 0.002142976922914386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.630495790458372,
      "flow_matching_loss": 0.030551102943718433,
      "grad_norm": 0.09641923755407333,
      "learning_rate": 5.061176856744737e-05,
      "loss": 0.0329,
      "relation_loss": 0.002309904946014285,
      "selected_pairs": 26619.6,
      "state_query_norm": 42.8992431640625,
      "state_relation_loss": 0.00016692803474143146,
      "step": 15640,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002309904946014285,
      "weighted_relation_loss": 0.002309904946014285
    },
    {
      "action_query_norm": 41.2063045501709,
      "action_relation_loss": 0.002262749255169183,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.639850327408793,
      "flow_matching_loss": 0.029917311482131483,
      "grad_norm": 0.1282559484243393,
      "learning_rate": 5.0556656672021205e-05,
      "loss": 0.0323,
      "relation_loss": 0.0024233873933553695,
      "selected_pairs": 27014.6,
      "state_query_norm": 42.901710510253906,
      "state_relation_loss": 0.00016063813673099502,
      "step": 15650,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0024233873933553695,
      "weighted_relation_loss": 0.0024233873933553695
    },
    {
      "action_query_norm": 41.19926109313965,
      "action_relation_loss": 0.002198524191044271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.649204864359215,
      "flow_matching_loss": 0.028605986014008523,
      "grad_norm": 0.13958969712257385,
      "learning_rate": 5.050154410020472e-05,
      "loss": 0.0309,
      "relation_loss": 0.0023427366046234964,
      "selected_pairs": 26594.3,
      "state_query_norm": 42.88812828063965,
      "state_relation_loss": 0.0001442123859305866,
      "step": 15660,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0023427366046234964,
      "weighted_relation_loss": 0.0023427366046234964
    },
    {
      "action_query_norm": 41.17228698730469,
      "action_relation_loss": 0.0021590088261291386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.658559401309635,
      "flow_matching_loss": 0.030183988995850085,
      "grad_norm": 0.09534753113985062,
      "learning_rate": 5.044643091896485e-05,
      "loss": 0.0325,
      "relation_loss": 0.0023110796930268405,
      "selected_pairs": 26709.2,
      "state_query_norm": 42.9040283203125,
      "state_relation_loss": 0.00015207089018076658,
      "step": 15670,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0023110796930268405,
      "weighted_relation_loss": 0.0023110796930268405
    },
    {
      "action_query_norm": 41.1991470336914,
      "action_relation_loss": 0.0022337819216772916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.667913938260057,
      "flow_matching_loss": 0.028031531162559985,
      "grad_norm": 0.08783598244190216,
      "learning_rate": 5.039131719526932e-05,
      "loss": 0.0304,
      "relation_loss": 0.0023889405187219382,
      "selected_pairs": 26670.7,
      "state_query_norm": 42.883618927001955,
      "state_relation_loss": 0.00015515860141022131,
      "step": 15680,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0023889405187219382,
      "weighted_relation_loss": 0.0023889405187219382
    },
    {
      "action_query_norm": 41.18369483947754,
      "action_relation_loss": 0.002107472450006753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.677268475210477,
      "flow_matching_loss": 0.027966130897402762,
      "grad_norm": 0.10790453106164932,
      "learning_rate": 5.03362029960865e-05,
      "loss": 0.0302,
      "relation_loss": 0.0022632612148299814,
      "selected_pairs": 26666.3,
      "state_query_norm": 42.88842887878418,
      "state_relation_loss": 0.00015578876627841964,
      "step": 15690,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0022632612148299814,
      "weighted_relation_loss": 0.0022632612148299814
    },
    {
      "action_query_norm": 41.16953582763672,
      "action_relation_loss": 0.0022153749014250934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.686623012160897,
      "flow_matching_loss": 0.03066640216857195,
      "grad_norm": 0.09890665858983994,
      "learning_rate": 5.028108838838533e-05,
      "loss": 0.033,
      "relation_loss": 0.002375659509561956,
      "selected_pairs": 27431.7,
      "state_query_norm": 42.88687362670898,
      "state_relation_loss": 0.00016028458921937271,
      "step": 15700,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.002375659509561956,
      "weighted_relation_loss": 0.002375659509561956
    },
    {
      "action_query_norm": 41.14384651184082,
      "action_relation_loss": 0.0022764010238461195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.69597754911132,
      "flow_matching_loss": 0.029065300710499287,
      "grad_norm": 0.10866302996873856,
      "learning_rate": 5.0225973439135276e-05,
      "loss": 0.0315,
      "relation_loss": 0.0024428106378763916,
      "selected_pairs": 26760.9,
      "state_query_norm": 42.88862953186035,
      "state_relation_loss": 0.00016640964022371919,
      "step": 15710,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0024428106378763916,
      "weighted_relation_loss": 0.0024428106378763916
    },
    {
      "action_query_norm": 41.172628021240236,
      "action_relation_loss": 0.0022673743893392383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.70533208606174,
      "flow_matching_loss": 0.028593887761235236,
      "grad_norm": 0.1111496090888977,
      "learning_rate": 5.017085821530617e-05,
      "loss": 0.031,
      "relation_loss": 0.002415730897337198,
      "selected_pairs": 27088.1,
      "state_query_norm": 42.88378372192383,
      "state_relation_loss": 0.00014835651236353443,
      "step": 15720,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002415730897337198,
      "weighted_relation_loss": 0.002415730897337198
    },
    {
      "action_query_norm": 41.11067008972168,
      "action_relation_loss": 0.002158288890495896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.714686623012161,
      "flow_matching_loss": 0.02764780502766371,
      "grad_norm": 0.1152871772646904,
      "learning_rate": 5.011574278386822e-05,
      "loss": 0.03,
      "relation_loss": 0.0023095238488167525,
      "selected_pairs": 26668.7,
      "state_query_norm": 42.883191680908205,
      "state_relation_loss": 0.00015123495977604763,
      "step": 15730,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0023095238488167525,
      "weighted_relation_loss": 0.0023095238488167525
    },
    {
      "action_query_norm": 41.17008819580078,
      "action_relation_loss": 0.002125572308432311,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.724041159962582,
      "flow_matching_loss": 0.029383177869021893,
      "grad_norm": 0.12585806846618652,
      "learning_rate": 5.006062721179189e-05,
      "loss": 0.0317,
      "relation_loss": 0.0022812108742073177,
      "selected_pairs": 26573.9,
      "state_query_norm": 42.866824340820315,
      "state_relation_loss": 0.00015563856577500702,
      "step": 15740,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0022812108742073177,
      "weighted_relation_loss": 0.0022812108742073177
    },
    {
      "action_query_norm": 41.13433723449707,
      "action_relation_loss": 0.002107020525727421,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.733395696913004,
      "flow_matching_loss": 0.028424312919378282,
      "grad_norm": 0.10632065683603287,
      "learning_rate": 5.000551156604777e-05,
      "loss": 0.0307,
      "relation_loss": 0.0022584334947168827,
      "selected_pairs": 27247.2,
      "state_query_norm": 42.878726196289065,
      "state_relation_loss": 0.0001514129835413769,
      "step": 15750,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0022584334947168827,
      "weighted_relation_loss": 0.0022584334947168827
    },
    {
      "action_query_norm": 41.07797966003418,
      "action_relation_loss": 0.002292513637803495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.742750233863424,
      "flow_matching_loss": 0.030169853009283543,
      "grad_norm": 0.14157845079898834,
      "learning_rate": 4.9950395913606594e-05,
      "loss": 0.0326,
      "relation_loss": 0.0024410647805780173,
      "selected_pairs": 26918.3,
      "state_query_norm": 42.87555198669433,
      "state_relation_loss": 0.0001485511282226071,
      "step": 15760,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024410647805780173,
      "weighted_relation_loss": 0.0024410647805780173
    },
    {
      "action_query_norm": 41.08955078125,
      "action_relation_loss": 0.002206139359623194,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.752104770813844,
      "flow_matching_loss": 0.03006757255643606,
      "grad_norm": 0.12955342233181,
      "learning_rate": 4.989528032143903e-05,
      "loss": 0.0324,
      "relation_loss": 0.002350840764120221,
      "selected_pairs": 26343.4,
      "state_query_norm": 42.864871978759766,
      "state_relation_loss": 0.0001447013986762613,
      "step": 15770,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.002350840764120221,
      "weighted_relation_loss": 0.002350840764120221
    },
    {
      "action_query_norm": 41.10755805969238,
      "action_relation_loss": 0.0023259193520061673,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.761459307764266,
      "flow_matching_loss": 0.027854040265083313,
      "grad_norm": 0.10744751244783401,
      "learning_rate": 4.984016485651578e-05,
      "loss": 0.0303,
      "relation_loss": 0.002469363622367382,
      "selected_pairs": 26507.2,
      "state_query_norm": 42.87683296203613,
      "state_relation_loss": 0.00014344427036121487,
      "step": 15780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.002469363622367382,
      "weighted_relation_loss": 0.002469363622367382
    },
    {
      "action_query_norm": 41.13846664428711,
      "action_relation_loss": 0.0021011097356677054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.770813844714686,
      "flow_matching_loss": 0.028626021184027194,
      "grad_norm": 0.2311457246541977,
      "learning_rate": 4.9785049585807274e-05,
      "loss": 0.0309,
      "relation_loss": 0.002253609336912632,
      "selected_pairs": 27568.6,
      "state_query_norm": 42.87758598327637,
      "state_relation_loss": 0.000152499572141096,
      "step": 15790,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.002253609336912632,
      "weighted_relation_loss": 0.002253609336912632
    },
    {
      "action_query_norm": 41.11678848266602,
      "action_relation_loss": 0.0022702287882566453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.780168381665108,
      "flow_matching_loss": 0.030506146140396595,
      "grad_norm": 0.14388540387153625,
      "learning_rate": 4.972993457628381e-05,
      "loss": 0.0329,
      "relation_loss": 0.0024239353137090803,
      "selected_pairs": 26918.4,
      "state_query_norm": 42.87963676452637,
      "state_relation_loss": 0.00015370653563877567,
      "step": 15800,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0024239353137090803,
      "weighted_relation_loss": 0.0024239353137090803
    },
    {
      "action_query_norm": 41.15001945495605,
      "action_relation_loss": 0.0021968963323161007,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.789522918615528,
      "flow_matching_loss": 0.028671226277947427,
      "grad_norm": 0.12637192010879517,
      "learning_rate": 4.9674819894915306e-05,
      "loss": 0.031,
      "relation_loss": 0.002344102691859007,
      "selected_pairs": 26868.7,
      "state_query_norm": 42.875156021118165,
      "state_relation_loss": 0.00014720632971148006,
      "step": 15810,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.002344102691859007,
      "weighted_relation_loss": 0.002344102691859007
    },
    {
      "action_query_norm": 41.16165428161621,
      "action_relation_loss": 0.002187264640815556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.79887745556595,
      "flow_matching_loss": 0.029625074565410615,
      "grad_norm": 0.10573838651180267,
      "learning_rate": 4.9619705608671265e-05,
      "loss": 0.032,
      "relation_loss": 0.002341575105674565,
      "selected_pairs": 26462.3,
      "state_query_norm": 42.86364860534668,
      "state_relation_loss": 0.00015431049396283925,
      "step": 15820,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002341575105674565,
      "weighted_relation_loss": 0.002341575105674565
    },
    {
      "action_query_norm": 41.15527267456055,
      "action_relation_loss": 0.0022650188766419888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.80823199251637,
      "flow_matching_loss": 0.02930098455399275,
      "grad_norm": 0.11095733195543289,
      "learning_rate": 4.956459178452079e-05,
      "loss": 0.0317,
      "relation_loss": 0.002418484282679856,
      "selected_pairs": 26648.6,
      "state_query_norm": 42.860735702514646,
      "state_relation_loss": 0.00015346539003076033,
      "step": 15830,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.002418484282679856,
      "weighted_relation_loss": 0.002418484282679856
    },
    {
      "action_query_norm": 41.109938049316405,
      "action_relation_loss": 0.002297734236344695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.81758652946679,
      "flow_matching_loss": 0.029774156957864763,
      "grad_norm": 0.09841873496770859,
      "learning_rate": 4.950947848943235e-05,
      "loss": 0.0322,
      "relation_loss": 0.0024474862962961196,
      "selected_pairs": 27200.2,
      "state_query_norm": 42.87239532470703,
      "state_relation_loss": 0.00014975207595853136,
      "step": 15840,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0024474862962961196,
      "weighted_relation_loss": 0.0024474862962961196
    },
    {
      "action_query_norm": 41.09158897399902,
      "action_relation_loss": 0.0022447236580774186,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.826941066417213,
      "flow_matching_loss": 0.028397209383547305,
      "grad_norm": 0.10006824135780334,
      "learning_rate": 4.9454365790373805e-05,
      "loss": 0.0308,
      "relation_loss": 0.0023994937771931292,
      "selected_pairs": 26966.3,
      "state_query_norm": 42.84906768798828,
      "state_relation_loss": 0.00015477011038456112,
      "step": 15850,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0023994937771931292,
      "weighted_relation_loss": 0.0023994937771931292
    },
    {
      "action_query_norm": 41.093177795410156,
      "action_relation_loss": 0.0024020302575081588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.836295603367633,
      "flow_matching_loss": 0.0304222097620368,
      "grad_norm": 0.1317145824432373,
      "learning_rate": 4.939925375431226e-05,
      "loss": 0.033,
      "relation_loss": 0.00255377646535635,
      "selected_pairs": 26754.3,
      "state_query_norm": 42.85650520324707,
      "state_relation_loss": 0.00015174626460066066,
      "step": 15860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.00255377646535635,
      "weighted_relation_loss": 0.00255377646535635
    },
    {
      "action_query_norm": 41.06215057373047,
      "action_relation_loss": 0.0022195776109583676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.845650140318055,
      "flow_matching_loss": 0.029073050245642662,
      "grad_norm": 0.12712633609771729,
      "learning_rate": 4.9344142448214046e-05,
      "loss": 0.0314,
      "relation_loss": 0.0023664762265980244,
      "selected_pairs": 27386.8,
      "state_query_norm": 42.8638614654541,
      "state_relation_loss": 0.0001468986345571466,
      "step": 15870,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0023664762265980244,
      "weighted_relation_loss": 0.0023664762265980244
    },
    {
      "action_query_norm": 41.08943328857422,
      "action_relation_loss": 0.002133160550147295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.855004677268475,
      "flow_matching_loss": 0.02814090084284544,
      "grad_norm": 0.12330040335655212,
      "learning_rate": 4.928903193904461e-05,
      "loss": 0.0304,
      "relation_loss": 0.002287826593965292,
      "selected_pairs": 27322.8,
      "state_query_norm": 42.853380584716795,
      "state_relation_loss": 0.00015466604963876307,
      "step": 15880,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.002287826593965292,
      "weighted_relation_loss": 0.002287826593965292
    },
    {
      "action_query_norm": 41.08796844482422,
      "action_relation_loss": 0.0021975640906021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.864359214218895,
      "flow_matching_loss": 0.028326277248561384,
      "grad_norm": 0.11669556796550751,
      "learning_rate": 4.923392229376842e-05,
      "loss": 0.0307,
      "relation_loss": 0.0023485523415729403,
      "selected_pairs": 27087.7,
      "state_query_norm": 42.845194244384764,
      "state_relation_loss": 0.00015098828298505397,
      "step": 15890,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0023485523415729403,
      "weighted_relation_loss": 0.0023485523415729403
    },
    {
      "action_query_norm": 41.114881134033205,
      "action_relation_loss": 0.0021295457263477147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.873713751169317,
      "flow_matching_loss": 0.028368075378239155,
      "grad_norm": 0.09664517641067505,
      "learning_rate": 4.9178813579348865e-05,
      "loss": 0.0306,
      "relation_loss": 0.00227675368078053,
      "selected_pairs": 26689.0,
      "state_query_norm": 42.851402282714844,
      "state_relation_loss": 0.00014720797480549663,
      "step": 15900,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.00227675368078053,
      "weighted_relation_loss": 0.00227675368078053
    },
    {
      "action_query_norm": 41.10379219055176,
      "action_relation_loss": 0.002120348415337503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.883068288119738,
      "flow_matching_loss": 0.02799167577177286,
      "grad_norm": 0.13639101386070251,
      "learning_rate": 4.9123705862748253e-05,
      "loss": 0.0303,
      "relation_loss": 0.0022741828579455614,
      "selected_pairs": 27201.5,
      "state_query_norm": 42.856943130493164,
      "state_relation_loss": 0.00015383440331788733,
      "step": 15910,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0022741828579455614,
      "weighted_relation_loss": 0.0022741828579455614
    },
    {
      "action_query_norm": 41.054822540283205,
      "action_relation_loss": 0.0021936249802820386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.89242282507016,
      "flow_matching_loss": 0.028553316369652748,
      "grad_norm": 0.13120009005069733,
      "learning_rate": 4.9068599210927627e-05,
      "loss": 0.0309,
      "relation_loss": 0.002332753175869584,
      "selected_pairs": 26781.0,
      "state_query_norm": 42.861341094970705,
      "state_relation_loss": 0.00013912820868426933,
      "step": 15920,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002332753175869584,
      "weighted_relation_loss": 0.002332753175869584
    },
    {
      "action_query_norm": 41.067440795898435,
      "action_relation_loss": 0.0023087688256055116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.90177736202058,
      "flow_matching_loss": 0.02951693646609783,
      "grad_norm": 0.13090002536773682,
      "learning_rate": 4.901349369084681e-05,
      "loss": 0.032,
      "relation_loss": 0.0024517751298844814,
      "selected_pairs": 26867.7,
      "state_query_norm": 42.85561027526855,
      "state_relation_loss": 0.0001430063261068426,
      "step": 15930,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0024517751298844814,
      "weighted_relation_loss": 0.0024517751298844814
    },
    {
      "action_query_norm": 41.07767181396484,
      "action_relation_loss": 0.0020832972368225455,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.911131898971,
      "flow_matching_loss": 0.028159316070377825,
      "grad_norm": 0.11566536873579025,
      "learning_rate": 4.8958389369464156e-05,
      "loss": 0.0304,
      "relation_loss": 0.002232567546889186,
      "selected_pairs": 27218.8,
      "state_query_norm": 42.843241119384764,
      "state_relation_loss": 0.0001492703187977895,
      "step": 15940,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.002232567546889186,
      "weighted_relation_loss": 0.002232567546889186
    },
    {
      "action_query_norm": 41.08364028930664,
      "action_relation_loss": 0.0021613382850773633,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.920486435921422,
      "flow_matching_loss": 0.030485366471111774,
      "grad_norm": 0.11742845922708511,
      "learning_rate": 4.890328631373666e-05,
      "loss": 0.0328,
      "relation_loss": 0.002306398283690214,
      "selected_pairs": 26825.5,
      "state_query_norm": 42.84826240539551,
      "state_relation_loss": 0.00014506004663417116,
      "step": 15950,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002306398283690214,
      "weighted_relation_loss": 0.002306398283690214
    },
    {
      "action_query_norm": 41.07111511230469,
      "action_relation_loss": 0.002301990007981658,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.929840972871842,
      "flow_matching_loss": 0.02900688573718071,
      "grad_norm": 0.09645962715148926,
      "learning_rate": 4.88481845906197e-05,
      "loss": 0.0315,
      "relation_loss": 0.0024458222556859254,
      "selected_pairs": 26560.2,
      "state_query_norm": 42.84935684204102,
      "state_relation_loss": 0.0001438322273315862,
      "step": 15960,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0024458222556859254,
      "weighted_relation_loss": 0.0024458222556859254
    },
    {
      "action_query_norm": 41.07747116088867,
      "action_relation_loss": 0.0021908000460825862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.939195509822264,
      "flow_matching_loss": 0.02920593526214361,
      "grad_norm": 0.10505601763725281,
      "learning_rate": 4.879308426706707e-05,
      "loss": 0.0316,
      "relation_loss": 0.0023443803191185,
      "selected_pairs": 26880.1,
      "state_query_norm": 42.832769775390624,
      "state_relation_loss": 0.00015358028758782893,
      "step": 15970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0023443803191185,
      "weighted_relation_loss": 0.0023443803191185
    },
    {
      "action_query_norm": 41.04440002441406,
      "action_relation_loss": 0.002331328624859452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.948550046772684,
      "flow_matching_loss": 0.029513135366141797,
      "grad_norm": 0.09168694168329239,
      "learning_rate": 4.873798541003084e-05,
      "loss": 0.032,
      "relation_loss": 0.0024742185836657883,
      "selected_pairs": 27573.8,
      "state_query_norm": 42.84764633178711,
      "state_relation_loss": 0.00014288995007518678,
      "step": 15980,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0024742185836657883,
      "weighted_relation_loss": 0.0024742185836657883
    },
    {
      "action_query_norm": 41.031226348876956,
      "action_relation_loss": 0.002135140518657863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.957904583723106,
      "flow_matching_loss": 0.030308564938604832,
      "grad_norm": 0.12437121570110321,
      "learning_rate": 4.868288808646136e-05,
      "loss": 0.0326,
      "relation_loss": 0.0022849942790344356,
      "selected_pairs": 26729.0,
      "state_query_norm": 42.836433792114256,
      "state_relation_loss": 0.0001498537589213811,
      "step": 15990,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0022849942790344356,
      "weighted_relation_loss": 0.0022849942790344356
    },
    {
      "action_query_norm": 41.068250274658205,
      "action_relation_loss": 0.0021692539798095823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.967259120673527,
      "flow_matching_loss": 0.028177258372306824,
      "grad_norm": 0.09460944682359695,
      "learning_rate": 4.862779236330705e-05,
      "loss": 0.0305,
      "relation_loss": 0.0023206590674817564,
      "selected_pairs": 26467.5,
      "state_query_norm": 42.85048713684082,
      "state_relation_loss": 0.0001514051007688977,
      "step": 16000,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0023206590674817564,
      "weighted_relation_loss": 0.0023206590674817564
    },
    {
      "action_query_norm": 41.03181686401367,
      "action_relation_loss": 0.002206917479634285,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.976613657623947,
      "flow_matching_loss": 0.03013059925287962,
      "grad_norm": 0.12251504510641098,
      "learning_rate": 4.8572698307514395e-05,
      "loss": 0.0325,
      "relation_loss": 0.002349380636587739,
      "selected_pairs": 26179.4,
      "state_query_norm": 42.828102111816406,
      "state_relation_loss": 0.00014246314676711336,
      "step": 16010,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.002349380636587739,
      "weighted_relation_loss": 0.002349380636587739
    },
    {
      "action_query_norm": 41.069379043579104,
      "action_relation_loss": 0.0021690512541681526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.985968194574369,
      "flow_matching_loss": 0.028664163872599603,
      "grad_norm": 0.08581145852804184,
      "learning_rate": 4.85176059860279e-05,
      "loss": 0.031,
      "relation_loss": 0.002316692774184048,
      "selected_pairs": 26940.3,
      "state_query_norm": 42.835967254638675,
      "state_relation_loss": 0.00014764152438146995,
      "step": 16020,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.002316692774184048,
      "weighted_relation_loss": 0.002316692774184048
    },
    {
      "action_query_norm": 41.048125457763675,
      "action_relation_loss": 0.002210878417827189,
      "auxiliary_loss_weight": 1.0,
      "epoch": 14.995322731524789,
      "flow_matching_loss": 0.02935880869626999,
      "grad_norm": 0.09846559166908264,
      "learning_rate": 4.846251546578989e-05,
      "loss": 0.0317,
      "relation_loss": 0.0023596495622768996,
      "selected_pairs": 26806.9,
      "state_query_norm": 42.83401527404785,
      "state_relation_loss": 0.00014877119538141413,
      "step": 16030,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0023596495622768996,
      "weighted_relation_loss": 0.0023596495622768996
    },
    {
      "action_query_norm": 41.02888107299805,
      "action_relation_loss": 0.00214656445896253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.004677268475211,
      "flow_matching_loss": 0.02888547386974096,
      "grad_norm": 0.09293181449174881,
      "learning_rate": 4.840742681374058e-05,
      "loss": 0.0312,
      "relation_loss": 0.0022818655241280796,
      "selected_pairs": 23953.5,
      "state_query_norm": 42.83917770385742,
      "state_relation_loss": 0.0001353010564344004,
      "step": 16040,
      "valid_action_samples": 212.3,
      "weighted_auxiliary_loss": 0.0022818655241280796,
      "weighted_relation_loss": 0.0022818655241280796
    },
    {
      "action_query_norm": 41.03338890075683,
      "action_relation_loss": 0.002164738904684782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.014031805425631,
      "flow_matching_loss": 0.027677072398364543,
      "grad_norm": 0.11970598995685577,
      "learning_rate": 4.835234009681787e-05,
      "loss": 0.03,
      "relation_loss": 0.0023032054770737885,
      "selected_pairs": 26436.1,
      "state_query_norm": 42.83241310119629,
      "state_relation_loss": 0.00013846656220266594,
      "step": 16050,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0023032054770737885,
      "weighted_relation_loss": 0.0023032054770737885
    },
    {
      "action_query_norm": 41.08233108520508,
      "action_relation_loss": 0.002014639275148511,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.023386342376053,
      "flow_matching_loss": 0.028322992473840715,
      "grad_norm": 0.09702375531196594,
      "learning_rate": 4.8297255381957296e-05,
      "loss": 0.0305,
      "relation_loss": 0.002158070122823119,
      "selected_pairs": 27079.7,
      "state_query_norm": 42.824921417236325,
      "state_relation_loss": 0.00014343085422297008,
      "step": 16060,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002158070122823119,
      "weighted_relation_loss": 0.002158070122823119
    },
    {
      "action_query_norm": 41.07698554992676,
      "action_relation_loss": 0.0020822710474021735,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.032740879326473,
      "flow_matching_loss": 0.02774695660918951,
      "grad_norm": 0.1184396892786026,
      "learning_rate": 4.824217273609198e-05,
      "loss": 0.03,
      "relation_loss": 0.0022348892409354447,
      "selected_pairs": 26614.3,
      "state_query_norm": 42.83445243835449,
      "state_relation_loss": 0.00015261823282344266,
      "step": 16070,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0022348892409354447,
      "weighted_relation_loss": 0.0022348892409354447
    },
    {
      "action_query_norm": 40.98125686645508,
      "action_relation_loss": 0.0021704044193029404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.042095416276894,
      "flow_matching_loss": 0.026474021188914777,
      "grad_norm": 0.10207819938659668,
      "learning_rate": 4.8187092226152555e-05,
      "loss": 0.0288,
      "relation_loss": 0.0023200594820082187,
      "selected_pairs": 26833.3,
      "state_query_norm": 42.81319236755371,
      "state_relation_loss": 0.00014965510490583256,
      "step": 16080,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0023200594820082187,
      "weighted_relation_loss": 0.0023200594820082187
    },
    {
      "action_query_norm": 41.05115432739258,
      "action_relation_loss": 0.0020442740758880975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.051449953227316,
      "flow_matching_loss": 0.0275607330724597,
      "grad_norm": 0.12237254530191422,
      "learning_rate": 4.8132013919067016e-05,
      "loss": 0.0298,
      "relation_loss": 0.002200639445800334,
      "selected_pairs": 26782.4,
      "state_query_norm": 42.835868453979494,
      "state_relation_loss": 0.0001563653495395556,
      "step": 16090,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002200639445800334,
      "weighted_relation_loss": 0.002200639445800334
    },
    {
      "action_query_norm": 41.011785888671874,
      "action_relation_loss": 0.002029648853931576,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.060804490177736,
      "flow_matching_loss": 0.027939522452652455,
      "grad_norm": 0.1499151885509491,
      "learning_rate": 4.807693788176071e-05,
      "loss": 0.0301,
      "relation_loss": 0.002176818554289639,
      "selected_pairs": 27039.5,
      "state_query_norm": 42.80950813293457,
      "state_relation_loss": 0.00014716971054440364,
      "step": 16100,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.002176818554289639,
      "weighted_relation_loss": 0.002176818554289639
    },
    {
      "action_query_norm": 41.02779579162598,
      "action_relation_loss": 0.00201478258240968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.070159027128158,
      "flow_matching_loss": 0.02856382541358471,
      "grad_norm": 0.12855331599712372,
      "learning_rate": 4.8021864181156214e-05,
      "loss": 0.0307,
      "relation_loss": 0.0021576676866970957,
      "selected_pairs": 26386.3,
      "state_query_norm": 42.812186813354494,
      "state_relation_loss": 0.0001428850897355005,
      "step": 16110,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0021576676866970957,
      "weighted_relation_loss": 0.0021576676866970957
    },
    {
      "action_query_norm": 41.00155067443848,
      "action_relation_loss": 0.0021416683914139865,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.079513564078578,
      "flow_matching_loss": 0.028627341799438,
      "grad_norm": 0.13056614995002747,
      "learning_rate": 4.796679288417326e-05,
      "loss": 0.0309,
      "relation_loss": 0.002284976583905518,
      "selected_pairs": 26623.9,
      "state_query_norm": 42.822353744506835,
      "state_relation_loss": 0.00014330818084999918,
      "step": 16120,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002284976583905518,
      "weighted_relation_loss": 0.002284976583905518
    },
    {
      "action_query_norm": 41.03328094482422,
      "action_relation_loss": 0.0021635891636833547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.088868101028998,
      "flow_matching_loss": 0.02713435757905245,
      "grad_norm": 0.11105908453464508,
      "learning_rate": 4.791172405772866e-05,
      "loss": 0.0295,
      "relation_loss": 0.002316073002293706,
      "selected_pairs": 26922.5,
      "state_query_norm": 42.81952743530273,
      "state_relation_loss": 0.00015248385025188327,
      "step": 16130,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.002316073002293706,
      "weighted_relation_loss": 0.002316073002293706
    },
    {
      "action_query_norm": 41.011827087402345,
      "action_relation_loss": 0.0021989672910422086,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.09822263797942,
      "flow_matching_loss": 0.027120698429644107,
      "grad_norm": 0.1091744601726532,
      "learning_rate": 4.785665776873625e-05,
      "loss": 0.0295,
      "relation_loss": 0.0023507441394031046,
      "selected_pairs": 26484.2,
      "state_query_norm": 42.823111343383786,
      "state_relation_loss": 0.00015177682071225717,
      "step": 16140,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0023507441394031046,
      "weighted_relation_loss": 0.0023507441394031046
    },
    {
      "action_query_norm": 40.98885650634766,
      "action_relation_loss": 0.002153595071285963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.10757717492984,
      "flow_matching_loss": 0.028969948180019855,
      "grad_norm": 0.11627569794654846,
      "learning_rate": 4.7801594084106763e-05,
      "loss": 0.0313,
      "relation_loss": 0.0023084106855094434,
      "selected_pairs": 26597.9,
      "state_query_norm": 42.823660659790036,
      "state_relation_loss": 0.00015481562440982087,
      "step": 16150,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0023084106855094434,
      "weighted_relation_loss": 0.0023084106855094434
    },
    {
      "action_query_norm": 41.04726791381836,
      "action_relation_loss": 0.0021382115548476575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.116931711880262,
      "flow_matching_loss": 0.02792995311319828,
      "grad_norm": 0.10917601734399796,
      "learning_rate": 4.774653307074775e-05,
      "loss": 0.0302,
      "relation_loss": 0.0022873108042404055,
      "selected_pairs": 27200.5,
      "state_query_norm": 42.82055244445801,
      "state_relation_loss": 0.00014909923338564114,
      "step": 16160,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0022873108042404055,
      "weighted_relation_loss": 0.0022873108042404055
    },
    {
      "action_query_norm": 41.07034111022949,
      "action_relation_loss": 0.0020896215573884547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.126286248830683,
      "flow_matching_loss": 0.026380470581352712,
      "grad_norm": 0.1115059033036232,
      "learning_rate": 4.769147479556355e-05,
      "loss": 0.0286,
      "relation_loss": 0.0022445676848292353,
      "selected_pairs": 26597.5,
      "state_query_norm": 42.816267013549805,
      "state_relation_loss": 0.00015494615799980239,
      "step": 16170,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0022445676848292353,
      "weighted_relation_loss": 0.0022445676848292353
    },
    {
      "action_query_norm": 41.025079345703126,
      "action_relation_loss": 0.0022504706983454527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.135640785781105,
      "flow_matching_loss": 0.02745053879916668,
      "grad_norm": 0.16018560528755188,
      "learning_rate": 4.763641932545515e-05,
      "loss": 0.0299,
      "relation_loss": 0.0024043960263952615,
      "selected_pairs": 26625.3,
      "state_query_norm": 42.79365615844726,
      "state_relation_loss": 0.0001539253134978935,
      "step": 16180,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0024043960263952615,
      "weighted_relation_loss": 0.0024043960263952615
    },
    {
      "action_query_norm": 40.997260665893556,
      "action_relation_loss": 0.002010684763081372,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.144995322731525,
      "flow_matching_loss": 0.02903745286166668,
      "grad_norm": 0.1179654449224472,
      "learning_rate": 4.758136672732013e-05,
      "loss": 0.0312,
      "relation_loss": 0.002177548932377249,
      "selected_pairs": 26684.1,
      "state_query_norm": 42.7912841796875,
      "state_relation_loss": 0.00016686417075106874,
      "step": 16190,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002177548932377249,
      "weighted_relation_loss": 0.002177548932377249
    },
    {
      "action_query_norm": 40.96093101501465,
      "action_relation_loss": 0.0021371171227656306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.154349859681945,
      "flow_matching_loss": 0.0275705236941576,
      "grad_norm": 0.0983961746096611,
      "learning_rate": 4.752631706805261e-05,
      "loss": 0.0299,
      "relation_loss": 0.0022898137103766203,
      "selected_pairs": 26397.4,
      "state_query_norm": 42.78581581115723,
      "state_relation_loss": 0.0001526965876109898,
      "step": 16200,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0022898137103766203,
      "weighted_relation_loss": 0.0022898137103766203
    },
    {
      "action_query_norm": 41.03452415466309,
      "action_relation_loss": 0.0022252584574744107,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.163704396632367,
      "flow_matching_loss": 0.028280074521899224,
      "grad_norm": 0.10270997136831284,
      "learning_rate": 4.7471270414543114e-05,
      "loss": 0.0307,
      "relation_loss": 0.00237428160617128,
      "selected_pairs": 26830.7,
      "state_query_norm": 42.80417404174805,
      "state_relation_loss": 0.0001490231603384018,
      "step": 16210,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.00237428160617128,
      "weighted_relation_loss": 0.00237428160617128
    },
    {
      "action_query_norm": 41.01541976928711,
      "action_relation_loss": 0.0020929634221829472,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.173058933582787,
      "flow_matching_loss": 0.02821652013808489,
      "grad_norm": 0.1177351176738739,
      "learning_rate": 4.741622683367849e-05,
      "loss": 0.0305,
      "relation_loss": 0.0022367998491972683,
      "selected_pairs": 26804.1,
      "state_query_norm": 42.81569213867188,
      "state_relation_loss": 0.00014383645466296003,
      "step": 16220,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0022367998491972683,
      "weighted_relation_loss": 0.0022367998491972683
    },
    {
      "action_query_norm": 41.019476318359374,
      "action_relation_loss": 0.002049147291108966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.18241347053321,
      "flow_matching_loss": 0.02857813984155655,
      "grad_norm": 0.11473020166158676,
      "learning_rate": 4.736118639234191e-05,
      "loss": 0.0308,
      "relation_loss": 0.0021956503624096513,
      "selected_pairs": 26479.2,
      "state_query_norm": 42.798047637939455,
      "state_relation_loss": 0.0001465030902181752,
      "step": 16230,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0021956503624096513,
      "weighted_relation_loss": 0.0021956503624096513
    },
    {
      "action_query_norm": 40.97909049987793,
      "action_relation_loss": 0.0021834300947375595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.19176800748363,
      "flow_matching_loss": 0.02472191918641329,
      "grad_norm": 0.09897954016923904,
      "learning_rate": 4.730614915741267e-05,
      "loss": 0.0271,
      "relation_loss": 0.0023367160698398947,
      "selected_pairs": 26667.5,
      "state_query_norm": 42.80068130493164,
      "state_relation_loss": 0.00015328596346080302,
      "step": 16240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0023367160698398947,
      "weighted_relation_loss": 0.0023367160698398947
    },
    {
      "action_query_norm": 41.02536811828613,
      "action_relation_loss": 0.0021955760079436005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.20112254443405,
      "flow_matching_loss": 0.028216761723160743,
      "grad_norm": 0.10472393780946732,
      "learning_rate": 4.7251115195766234e-05,
      "loss": 0.0306,
      "relation_loss": 0.002341091725975275,
      "selected_pairs": 27042.2,
      "state_query_norm": 42.79315032958984,
      "state_relation_loss": 0.00014551571221090853,
      "step": 16250,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.002341091725975275,
      "weighted_relation_loss": 0.002341091725975275
    },
    {
      "action_query_norm": 41.06179428100586,
      "action_relation_loss": 0.0020814026822336016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.210477081384472,
      "flow_matching_loss": 0.027139571495354176,
      "grad_norm": 0.0879424512386322,
      "learning_rate": 4.719608457427404e-05,
      "loss": 0.0294,
      "relation_loss": 0.0022376606473699214,
      "selected_pairs": 26597.8,
      "state_query_norm": 42.79279975891113,
      "state_relation_loss": 0.0001562579462188296,
      "step": 16260,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0022376606473699214,
      "weighted_relation_loss": 0.0022376606473699214
    },
    {
      "action_query_norm": 40.97507858276367,
      "action_relation_loss": 0.0021535109844990075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.219831618334892,
      "flow_matching_loss": 0.028839132376015188,
      "grad_norm": 0.1118057370185852,
      "learning_rate": 4.714105735980347e-05,
      "loss": 0.0311,
      "relation_loss": 0.002299526007845998,
      "selected_pairs": 26482.7,
      "state_query_norm": 42.78405914306641,
      "state_relation_loss": 0.00014601504954043777,
      "step": 16270,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.002299526007845998,
      "weighted_relation_loss": 0.002299526007845998
    },
    {
      "action_query_norm": 40.96659622192383,
      "action_relation_loss": 0.0021540889283642173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.229186155285314,
      "flow_matching_loss": 0.026393692009150982,
      "grad_norm": 0.10933369398117065,
      "learning_rate": 4.70860336192178e-05,
      "loss": 0.0287,
      "relation_loss": 0.0022994910832494497,
      "selected_pairs": 26388.8,
      "state_query_norm": 42.80427360534668,
      "state_relation_loss": 0.0001454021352401469,
      "step": 16280,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0022994910832494497,
      "weighted_relation_loss": 0.0022994910832494497
    },
    {
      "action_query_norm": 40.9973258972168,
      "action_relation_loss": 0.0019217163207940757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.238540692235734,
      "flow_matching_loss": 0.026297074928879737,
      "grad_norm": 0.08000627160072327,
      "learning_rate": 4.7031013419376034e-05,
      "loss": 0.0284,
      "relation_loss": 0.0020768610178492964,
      "selected_pairs": 26894.5,
      "state_query_norm": 42.789899826049805,
      "state_relation_loss": 0.00015514471015194432,
      "step": 16290,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0020768610178492964,
      "weighted_relation_loss": 0.0020768610178492964
    },
    {
      "action_query_norm": 40.99926338195801,
      "action_relation_loss": 0.002039563748985529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.247895229186156,
      "flow_matching_loss": 0.02688975427299738,
      "grad_norm": 0.1005922183394432,
      "learning_rate": 4.6975996827132926e-05,
      "loss": 0.0291,
      "relation_loss": 0.002183488616719842,
      "selected_pairs": 27060.2,
      "state_query_norm": 42.80442771911621,
      "state_relation_loss": 0.00014392484299605713,
      "step": 16300,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.002183488616719842,
      "weighted_relation_loss": 0.002183488616719842
    },
    {
      "action_query_norm": 41.008504486083986,
      "action_relation_loss": 0.0020250515663065016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.257249766136576,
      "flow_matching_loss": 0.028561417385935782,
      "grad_norm": 0.10256706178188324,
      "learning_rate": 4.692098390933883e-05,
      "loss": 0.0307,
      "relation_loss": 0.0021687654545530675,
      "selected_pairs": 26761.0,
      "state_query_norm": 42.79050216674805,
      "state_relation_loss": 0.0001437138686014805,
      "step": 16310,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0021687654545530675,
      "weighted_relation_loss": 0.0021687654545530675
    },
    {
      "action_query_norm": 40.977071380615236,
      "action_relation_loss": 0.002093877049628645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.266604303086996,
      "flow_matching_loss": 0.02901224046945572,
      "grad_norm": 0.11653516441583633,
      "learning_rate": 4.6865974732839615e-05,
      "loss": 0.0312,
      "relation_loss": 0.0022368249716237187,
      "selected_pairs": 27195.6,
      "state_query_norm": 42.7907527923584,
      "state_relation_loss": 0.00014294789434643462,
      "step": 16320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0022368249716237187,
      "weighted_relation_loss": 0.0022368249716237187
    },
    {
      "action_query_norm": 40.9813777923584,
      "action_relation_loss": 0.0021302154520526527,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.275958840037418,
      "flow_matching_loss": 0.030179780162870883,
      "grad_norm": 0.11360982060432434,
      "learning_rate": 4.6810969364476615e-05,
      "loss": 0.0325,
      "relation_loss": 0.0022800172911956905,
      "selected_pairs": 26384.0,
      "state_query_norm": 42.771401977539064,
      "state_relation_loss": 0.00014980182459112257,
      "step": 16330,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0022800172911956905,
      "weighted_relation_loss": 0.0022800172911956905
    },
    {
      "action_query_norm": 40.98911972045899,
      "action_relation_loss": 0.0022375155007466675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.285313376987839,
      "flow_matching_loss": 0.02909160330891609,
      "grad_norm": 0.13615381717681885,
      "learning_rate": 4.675596787108653e-05,
      "loss": 0.0315,
      "relation_loss": 0.002391233039088547,
      "selected_pairs": 27460.7,
      "state_query_norm": 42.783443450927734,
      "state_relation_loss": 0.0001537175354314968,
      "step": 16340,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.002391233039088547,
      "weighted_relation_loss": 0.002391233039088547
    },
    {
      "action_query_norm": 40.964114379882815,
      "action_relation_loss": 0.002073674986604601,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.29466791393826,
      "flow_matching_loss": 0.027847948111593725,
      "grad_norm": 0.19270311295986176,
      "learning_rate": 4.670097031950139e-05,
      "loss": 0.0301,
      "relation_loss": 0.00222333543933928,
      "selected_pairs": 27337.5,
      "state_query_norm": 42.77220687866211,
      "state_relation_loss": 0.00014966049493523315,
      "step": 16350,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.00222333543933928,
      "weighted_relation_loss": 0.00222333543933928
    },
    {
      "action_query_norm": 41.00874404907226,
      "action_relation_loss": 0.0020711675751954317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.30402245088868,
      "flow_matching_loss": 0.02747804597020149,
      "grad_norm": 0.1107771247625351,
      "learning_rate": 4.664597677654838e-05,
      "loss": 0.0297,
      "relation_loss": 0.0022074217442423104,
      "selected_pairs": 27567.8,
      "state_query_norm": 42.771583557128906,
      "state_relation_loss": 0.00013625419887830503,
      "step": 16360,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0022074217442423104,
      "weighted_relation_loss": 0.0022074217442423104
    },
    {
      "action_query_norm": 40.98386077880859,
      "action_relation_loss": 0.002065330045297742,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.313376987839103,
      "flow_matching_loss": 0.02886719275265932,
      "grad_norm": 0.1149100586771965,
      "learning_rate": 4.659098730904986e-05,
      "loss": 0.0311,
      "relation_loss": 0.002210734284017235,
      "selected_pairs": 27596.4,
      "state_query_norm": 42.79231719970703,
      "state_relation_loss": 0.00014540425909217447,
      "step": 16370,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.002210734284017235,
      "weighted_relation_loss": 0.002210734284017235
    },
    {
      "action_query_norm": 41.02111320495605,
      "action_relation_loss": 0.0020653596613556145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.322731524789523,
      "flow_matching_loss": 0.027482898160815238,
      "grad_norm": 0.11616958677768707,
      "learning_rate": 4.6536001983823204e-05,
      "loss": 0.0297,
      "relation_loss": 0.0022053457563742996,
      "selected_pairs": 26575.6,
      "state_query_norm": 42.79644622802734,
      "state_relation_loss": 0.00013998609356349334,
      "step": 16380,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0022053457563742996,
      "weighted_relation_loss": 0.0022053457563742996
    },
    {
      "action_query_norm": 40.93114128112793,
      "action_relation_loss": 0.002054950373712927,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.332086061739943,
      "flow_matching_loss": 0.02540675662457943,
      "grad_norm": 0.12468969821929932,
      "learning_rate": 4.648102086768078e-05,
      "loss": 0.0276,
      "relation_loss": 0.0022045932477340103,
      "selected_pairs": 26801.4,
      "state_query_norm": 42.77534866333008,
      "state_relation_loss": 0.0001496429104008712,
      "step": 16390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0022045932477340103,
      "weighted_relation_loss": 0.0022045932477340103
    },
    {
      "action_query_norm": 40.949832153320315,
      "action_relation_loss": 0.0020560981472954156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.341440598690365,
      "flow_matching_loss": 0.02735130414366722,
      "grad_norm": 0.12201401591300964,
      "learning_rate": 4.6426044027429786e-05,
      "loss": 0.0295,
      "relation_loss": 0.0021941553335636855,
      "selected_pairs": 26965.0,
      "state_query_norm": 42.77728424072266,
      "state_relation_loss": 0.00013805713242618368,
      "step": 16400,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0021941553335636855,
      "weighted_relation_loss": 0.0021941553335636855
    },
    {
      "action_query_norm": 40.94884834289551,
      "action_relation_loss": 0.0020659724716097115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.350795135640785,
      "flow_matching_loss": 0.025697795487940312,
      "grad_norm": 0.1033768504858017,
      "learning_rate": 4.637107152987234e-05,
      "loss": 0.0279,
      "relation_loss": 0.002210604806896299,
      "selected_pairs": 26121.3,
      "state_query_norm": 42.76231575012207,
      "state_relation_loss": 0.00014463233383139596,
      "step": 16410,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.002210604806896299,
      "weighted_relation_loss": 0.002210604806896299
    },
    {
      "action_query_norm": 40.91171112060547,
      "action_relation_loss": 0.00211567145306617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.360149672591207,
      "flow_matching_loss": 0.029074339009821416,
      "grad_norm": 0.14028239250183105,
      "learning_rate": 4.631610344180515e-05,
      "loss": 0.0313,
      "relation_loss": 0.0022649339167401194,
      "selected_pairs": 27503.0,
      "state_query_norm": 42.76072387695312,
      "state_relation_loss": 0.00014926246949471534,
      "step": 16420,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0022649339167401194,
      "weighted_relation_loss": 0.0022649339167401194
    },
    {
      "action_query_norm": 40.94252662658691,
      "action_relation_loss": 0.002003431541379541,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.369504209541628,
      "flow_matching_loss": 0.026825182139873505,
      "grad_norm": 0.12716642022132874,
      "learning_rate": 4.626113983001965e-05,
      "loss": 0.029,
      "relation_loss": 0.002154062339104712,
      "selected_pairs": 26832.9,
      "state_query_norm": 42.76688003540039,
      "state_relation_loss": 0.00015063075989019126,
      "step": 16430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002154062339104712,
      "weighted_relation_loss": 0.002154062339104712
    },
    {
      "action_query_norm": 40.94577331542969,
      "action_relation_loss": 0.0020094890613108875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.378858746492048,
      "flow_matching_loss": 0.02791549973189831,
      "grad_norm": 0.12483228743076324,
      "learning_rate": 4.620618076130182e-05,
      "loss": 0.0301,
      "relation_loss": 0.0021549421246163545,
      "selected_pairs": 26994.1,
      "state_query_norm": 42.76242523193359,
      "state_relation_loss": 0.00014545304729836063,
      "step": 16440,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0021549421246163545,
      "weighted_relation_loss": 0.0021549421246163545
    },
    {
      "action_query_norm": 40.92798614501953,
      "action_relation_loss": 0.001983945700339973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.38821328344247,
      "flow_matching_loss": 0.028759030066430567,
      "grad_norm": 0.10412780940532684,
      "learning_rate": 4.6151226302432074e-05,
      "loss": 0.0309,
      "relation_loss": 0.002125769713893533,
      "selected_pairs": 26627.0,
      "state_query_norm": 42.76208419799805,
      "state_relation_loss": 0.00014182401209836826,
      "step": 16450,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002125769713893533,
      "weighted_relation_loss": 0.002125769713893533
    },
    {
      "action_query_norm": 40.99007949829102,
      "action_relation_loss": 0.001972700213082135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.39756782039289,
      "flow_matching_loss": 0.029514403268694877,
      "grad_norm": 0.11363095045089722,
      "learning_rate": 4.6096276520185304e-05,
      "loss": 0.0316,
      "relation_loss": 0.0021153366658836605,
      "selected_pairs": 26763.7,
      "state_query_norm": 42.778917694091795,
      "state_relation_loss": 0.00014263644552556798,
      "step": 16460,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0021153366658836605,
      "weighted_relation_loss": 0.0021153366658836605
    },
    {
      "action_query_norm": 40.91130599975586,
      "action_relation_loss": 0.0021251269383355973,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.406922357343312,
      "flow_matching_loss": 0.027775408513844014,
      "grad_norm": 0.12314784526824951,
      "learning_rate": 4.604133148133066e-05,
      "loss": 0.03,
      "relation_loss": 0.002269022271502763,
      "selected_pairs": 27572.1,
      "state_query_norm": 42.76219177246094,
      "state_relation_loss": 0.00014389533607754855,
      "step": 16470,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.002269022271502763,
      "weighted_relation_loss": 0.002269022271502763
    },
    {
      "action_query_norm": 40.95126724243164,
      "action_relation_loss": 0.0019476166926324368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.416276894293732,
      "flow_matching_loss": 0.026236013695597648,
      "grad_norm": 0.14759591221809387,
      "learning_rate": 4.598639125263155e-05,
      "loss": 0.0283,
      "relation_loss": 0.002084830321837217,
      "selected_pairs": 26531.1,
      "state_query_norm": 42.7571834564209,
      "state_relation_loss": 0.00013721363502554594,
      "step": 16480,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.002084830321837217,
      "weighted_relation_loss": 0.002084830321837217
    },
    {
      "action_query_norm": 40.925366973876955,
      "action_relation_loss": 0.0021163703757338228,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.425631431244154,
      "flow_matching_loss": 0.02813834846019745,
      "grad_norm": 0.11854575574398041,
      "learning_rate": 4.593145590084553e-05,
      "loss": 0.0304,
      "relation_loss": 0.0022563271224498747,
      "selected_pairs": 26943.6,
      "state_query_norm": 42.76736221313477,
      "state_relation_loss": 0.00013995675326441414,
      "step": 16490,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0022563271224498747,
      "weighted_relation_loss": 0.0022563271224498747
    },
    {
      "action_query_norm": 40.925489807128905,
      "action_relation_loss": 0.002057074615731835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.434985968194574,
      "flow_matching_loss": 0.026541346125304698,
      "grad_norm": 0.10466737300157547,
      "learning_rate": 4.58765254927242e-05,
      "loss": 0.0287,
      "relation_loss": 0.0021998620592057704,
      "selected_pairs": 26988.8,
      "state_query_norm": 42.76473693847656,
      "state_relation_loss": 0.00014278747912612743,
      "step": 16500,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0021998620592057704,
      "weighted_relation_loss": 0.0021998620592057704
    },
    {
      "action_query_norm": 40.92564506530762,
      "action_relation_loss": 0.0019479075097478926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.444340505144995,
      "flow_matching_loss": 0.02647493202239275,
      "grad_norm": 0.09557298570871353,
      "learning_rate": 4.582160009501323e-05,
      "loss": 0.0286,
      "relation_loss": 0.0020891744643449783,
      "selected_pairs": 26527.6,
      "state_query_norm": 42.768849563598636,
      "state_relation_loss": 0.00014126694295555354,
      "step": 16510,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0020891744643449783,
      "weighted_relation_loss": 0.0020891744643449783
    },
    {
      "action_query_norm": 40.93024597167969,
      "action_relation_loss": 0.0020857203635387123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.453695042095417,
      "flow_matching_loss": 0.028110622242093087,
      "grad_norm": 0.10158578306436539,
      "learning_rate": 4.5766679774452143e-05,
      "loss": 0.0303,
      "relation_loss": 0.002235468360595405,
      "selected_pairs": 27430.7,
      "state_query_norm": 42.75343246459961,
      "state_relation_loss": 0.00014974799996707587,
      "step": 16520,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.002235468360595405,
      "weighted_relation_loss": 0.002235468360595405
    },
    {
      "action_query_norm": 40.91517868041992,
      "action_relation_loss": 0.0020421611377969386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.463049579045837,
      "flow_matching_loss": 0.027574794366955757,
      "grad_norm": 0.11107718199491501,
      "learning_rate": 4.571176459777431e-05,
      "loss": 0.0298,
      "relation_loss": 0.002176687028259039,
      "selected_pairs": 26241.3,
      "state_query_norm": 42.7506591796875,
      "state_relation_loss": 0.00013452592975227162,
      "step": 16530,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.002176687028259039,
      "weighted_relation_loss": 0.002176687028259039
    },
    {
      "action_query_norm": 40.92530059814453,
      "action_relation_loss": 0.0019962315796874463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.472404115996259,
      "flow_matching_loss": 0.027286410704255103,
      "grad_norm": 0.10154205560684204,
      "learning_rate": 4.5656854631706846e-05,
      "loss": 0.0294,
      "relation_loss": 0.0021471629734151065,
      "selected_pairs": 26967.1,
      "state_query_norm": 42.74353523254395,
      "state_relation_loss": 0.0001509313369751908,
      "step": 16540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0021471629734151065,
      "weighted_relation_loss": 0.0021471629734151065
    },
    {
      "action_query_norm": 40.89697532653808,
      "action_relation_loss": 0.002114909142255783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.481758652946679,
      "flow_matching_loss": 0.028729626350104807,
      "grad_norm": 0.09795338660478592,
      "learning_rate": 4.560194994297054e-05,
      "loss": 0.031,
      "relation_loss": 0.0022630168357864024,
      "selected_pairs": 27369.5,
      "state_query_norm": 42.748033142089845,
      "state_relation_loss": 0.00014810768625466152,
      "step": 16550,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0022630168357864024,
      "weighted_relation_loss": 0.0022630168357864024
    },
    {
      "action_query_norm": 40.92398109436035,
      "action_relation_loss": 0.0018881146446801723,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.4911131898971,
      "flow_matching_loss": 0.02543699350208044,
      "grad_norm": 0.10043177753686905,
      "learning_rate": 4.554705059827974e-05,
      "loss": 0.0275,
      "relation_loss": 0.0020324908080510797,
      "selected_pairs": 26457.3,
      "state_query_norm": 42.755423355102536,
      "state_relation_loss": 0.0001443761619157158,
      "step": 16560,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0020324908080510797,
      "weighted_relation_loss": 0.0020324908080510797
    },
    {
      "action_query_norm": 40.86375274658203,
      "action_relation_loss": 0.0021024051355198027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.500467726847521,
      "flow_matching_loss": 0.028177638165652753,
      "grad_norm": 0.10333403199911118,
      "learning_rate": 4.5492156664342365e-05,
      "loss": 0.0304,
      "relation_loss": 0.0022502023144625126,
      "selected_pairs": 26831.7,
      "state_query_norm": 42.73203048706055,
      "state_relation_loss": 0.00014779714401811362,
      "step": 16570,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0022502023144625126,
      "weighted_relation_loss": 0.0022502023144625126
    },
    {
      "action_query_norm": 40.86713409423828,
      "action_relation_loss": 0.002209953358396888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.509822263797941,
      "flow_matching_loss": 0.027792132645845413,
      "grad_norm": 0.1302383691072464,
      "learning_rate": 4.5437268207859695e-05,
      "loss": 0.0301,
      "relation_loss": 0.002353336662054062,
      "selected_pairs": 26573.1,
      "state_query_norm": 42.73824234008789,
      "state_relation_loss": 0.00014338334294734524,
      "step": 16580,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.002353336662054062,
      "weighted_relation_loss": 0.002353336662054062
    },
    {
      "action_query_norm": 40.87396354675293,
      "action_relation_loss": 0.0020914522116072475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.519176800748363,
      "flow_matching_loss": 0.027846642024815082,
      "grad_norm": 0.15905499458312988,
      "learning_rate": 4.538238529552641e-05,
      "loss": 0.0301,
      "relation_loss": 0.002235789503902197,
      "selected_pairs": 27269.3,
      "state_query_norm": 42.743110275268556,
      "state_relation_loss": 0.00014433727774303405,
      "step": 16590,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.002235789503902197,
      "weighted_relation_loss": 0.002235789503902197
    },
    {
      "action_query_norm": 40.93153190612793,
      "action_relation_loss": 0.0020020728348754346,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.528531337698784,
      "flow_matching_loss": 0.02709035240113735,
      "grad_norm": 0.09908812493085861,
      "learning_rate": 4.5327507994030394e-05,
      "loss": 0.0292,
      "relation_loss": 0.0021486095618456604,
      "selected_pairs": 27010.5,
      "state_query_norm": 42.73758430480957,
      "state_relation_loss": 0.00014653675025328995,
      "step": 16600,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0021486095618456604,
      "weighted_relation_loss": 0.0021486095618456604
    },
    {
      "action_query_norm": 40.94937438964844,
      "action_relation_loss": 0.002015581191517413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.537885874649206,
      "flow_matching_loss": 0.02673769071698189,
      "grad_norm": 0.10269197821617126,
      "learning_rate": 4.527263637005273e-05,
      "loss": 0.0289,
      "relation_loss": 0.0021597771905362607,
      "selected_pairs": 26805.8,
      "state_query_norm": 42.732795333862306,
      "state_relation_loss": 0.0001441959808289539,
      "step": 16610,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0021597771905362607,
      "weighted_relation_loss": 0.0021597771905362607
    },
    {
      "action_query_norm": 40.89741134643555,
      "action_relation_loss": 0.0020387296099215746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.547240411599626,
      "flow_matching_loss": 0.02672426085919142,
      "grad_norm": 0.12643785774707794,
      "learning_rate": 4.521777049026766e-05,
      "loss": 0.0289,
      "relation_loss": 0.002189595066010952,
      "selected_pairs": 26711.9,
      "state_query_norm": 42.73913078308105,
      "state_relation_loss": 0.000150865493924357,
      "step": 16620,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.002189595066010952,
      "weighted_relation_loss": 0.002189595066010952
    },
    {
      "action_query_norm": 40.9090072631836,
      "action_relation_loss": 0.0020084494026377797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.556594948550046,
      "flow_matching_loss": 0.026509815454483034,
      "grad_norm": 0.11097966879606247,
      "learning_rate": 4.5162910421342386e-05,
      "loss": 0.0287,
      "relation_loss": 0.002153423149138689,
      "selected_pairs": 26051.0,
      "state_query_norm": 42.748233795166016,
      "state_relation_loss": 0.00014497376250801607,
      "step": 16630,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.002153423149138689,
      "weighted_relation_loss": 0.002153423149138689
    },
    {
      "action_query_norm": 40.90680160522461,
      "action_relation_loss": 0.001928658550605178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.565949485500468,
      "flow_matching_loss": 0.027159360609948634,
      "grad_norm": 0.12776023149490356,
      "learning_rate": 4.510805622993706e-05,
      "loss": 0.0292,
      "relation_loss": 0.0020754562690854074,
      "selected_pairs": 26943.8,
      "state_query_norm": 42.74134750366211,
      "state_relation_loss": 0.00014679769810754806,
      "step": 16640,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0020754562690854074,
      "weighted_relation_loss": 0.0020754562690854074
    },
    {
      "action_query_norm": 40.865330123901366,
      "action_relation_loss": 0.0019905786029994488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.575304022450888,
      "flow_matching_loss": 0.028309167921543123,
      "grad_norm": 0.15407879650592804,
      "learning_rate": 4.505320798270467e-05,
      "loss": 0.0304,
      "relation_loss": 0.0021404770785011353,
      "selected_pairs": 26361.5,
      "state_query_norm": 42.727870559692384,
      "state_relation_loss": 0.00014989846240496263,
      "step": 16650,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0021404770785011353,
      "weighted_relation_loss": 0.0021404770785011353
    },
    {
      "action_query_norm": 40.863609313964844,
      "action_relation_loss": 0.0019800745183601977,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.58465855940131,
      "flow_matching_loss": 0.026500475220382215,
      "grad_norm": 0.08658173680305481,
      "learning_rate": 4.499836574629105e-05,
      "loss": 0.0286,
      "relation_loss": 0.0021196539863012733,
      "selected_pairs": 27128.7,
      "state_query_norm": 42.73131484985352,
      "state_relation_loss": 0.00013957947376184166,
      "step": 16660,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0021196539863012733,
      "weighted_relation_loss": 0.0021196539863012733
    },
    {
      "action_query_norm": 40.88425750732422,
      "action_relation_loss": 0.0020225887536071243,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.59401309635173,
      "flow_matching_loss": 0.027551488019526005,
      "grad_norm": 0.13865329325199127,
      "learning_rate": 4.494352958733465e-05,
      "loss": 0.0297,
      "relation_loss": 0.0021634416654706,
      "selected_pairs": 27134.4,
      "state_query_norm": 42.72283592224121,
      "state_relation_loss": 0.00014085291550145485,
      "step": 16670,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0021634416654706,
      "weighted_relation_loss": 0.0021634416654706
    },
    {
      "action_query_norm": 40.85566635131836,
      "action_relation_loss": 0.0019913715543225406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.603367633302152,
      "flow_matching_loss": 0.028105448000133038,
      "grad_norm": 0.08544439822435379,
      "learning_rate": 4.488869957246663e-05,
      "loss": 0.0302,
      "relation_loss": 0.002132419089321047,
      "selected_pairs": 26765.7,
      "state_query_norm": 42.72420082092285,
      "state_relation_loss": 0.00014104758884059265,
      "step": 16680,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.002132419089321047,
      "weighted_relation_loss": 0.002132419089321047
    },
    {
      "action_query_norm": 40.83726005554199,
      "action_relation_loss": 0.0019605086650699376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.612722170252573,
      "flow_matching_loss": 0.02588987462222576,
      "grad_norm": 0.11647754907608032,
      "learning_rate": 4.4833875768310575e-05,
      "loss": 0.028,
      "relation_loss": 0.0020936294691637158,
      "selected_pairs": 26068.2,
      "state_query_norm": 42.74477767944336,
      "state_relation_loss": 0.00013312081573531033,
      "step": 16690,
      "valid_action_samples": 228.8,
      "weighted_auxiliary_loss": 0.0020936294691637158,
      "weighted_relation_loss": 0.0020936294691637158
    },
    {
      "action_query_norm": 40.85797882080078,
      "action_relation_loss": 0.002056120289489627,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.622076707202993,
      "flow_matching_loss": 0.026977119408547877,
      "grad_norm": 0.13411056995391846,
      "learning_rate": 4.47790582414826e-05,
      "loss": 0.0292,
      "relation_loss": 0.00219878819771111,
      "selected_pairs": 26926.0,
      "state_query_norm": 42.726244354248045,
      "state_relation_loss": 0.00014266788857639768,
      "step": 16700,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00219878819771111,
      "weighted_relation_loss": 0.00219878819771111
    },
    {
      "action_query_norm": 40.88660354614258,
      "action_relation_loss": 0.001966200605966151,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.631431244153415,
      "flow_matching_loss": 0.02755787707865238,
      "grad_norm": 0.10058201849460602,
      "learning_rate": 4.472424705859114e-05,
      "loss": 0.0297,
      "relation_loss": 0.00211827103048563,
      "selected_pairs": 27500.1,
      "state_query_norm": 42.72389602661133,
      "state_relation_loss": 0.00015207038377411662,
      "step": 16710,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.00211827103048563,
      "weighted_relation_loss": 0.00211827103048563
    },
    {
      "action_query_norm": 40.85154800415039,
      "action_relation_loss": 0.002034054754767567,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.640785781103835,
      "flow_matching_loss": 0.02792899440973997,
      "grad_norm": 0.10692952573299408,
      "learning_rate": 4.466944228623701e-05,
      "loss": 0.0301,
      "relation_loss": 0.0021699097473174334,
      "selected_pairs": 26386.0,
      "state_query_norm": 42.72663803100586,
      "state_relation_loss": 0.00013585505512310193,
      "step": 16720,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0021699097473174334,
      "weighted_relation_loss": 0.0021699097473174334
    },
    {
      "action_query_norm": 40.87112693786621,
      "action_relation_loss": 0.0018797199125401675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.650140318054257,
      "flow_matching_loss": 0.02700063996016979,
      "grad_norm": 0.12229272723197937,
      "learning_rate": 4.461464399101312e-05,
      "loss": 0.029,
      "relation_loss": 0.0020268909749574958,
      "selected_pairs": 26295.9,
      "state_query_norm": 42.7281436920166,
      "state_relation_loss": 0.00014717104641022162,
      "step": 16730,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0020268909749574958,
      "weighted_relation_loss": 0.0020268909749574958
    },
    {
      "action_query_norm": 40.82900733947754,
      "action_relation_loss": 0.002111967117525637,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.659494855004677,
      "flow_matching_loss": 0.028782302513718605,
      "grad_norm": 0.12587705254554749,
      "learning_rate": 4.45598522395046e-05,
      "loss": 0.031,
      "relation_loss": 0.0022647979203611612,
      "selected_pairs": 27226.0,
      "state_query_norm": 42.72362060546875,
      "state_relation_loss": 0.0001528308028355241,
      "step": 16740,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0022647979203611612,
      "weighted_relation_loss": 0.0022647979203611612
    },
    {
      "action_query_norm": 40.80218315124512,
      "action_relation_loss": 0.0019331928226165474,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.668849391955098,
      "flow_matching_loss": 0.02860191334038973,
      "grad_norm": 0.10381591320037842,
      "learning_rate": 4.4505067098288577e-05,
      "loss": 0.0307,
      "relation_loss": 0.002080924203619361,
      "selected_pairs": 26974.2,
      "state_query_norm": 42.71801567077637,
      "state_relation_loss": 0.00014773138536838815,
      "step": 16750,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.002080924203619361,
      "weighted_relation_loss": 0.002080924203619361
    },
    {
      "action_query_norm": 40.80883941650391,
      "action_relation_loss": 0.0019416814669966699,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.67820392890552,
      "flow_matching_loss": 0.027655894681811333,
      "grad_norm": 0.12024302035570145,
      "learning_rate": 4.445028863393417e-05,
      "loss": 0.0297,
      "relation_loss": 0.002088646753691137,
      "selected_pairs": 26320.4,
      "state_query_norm": 42.72703094482422,
      "state_relation_loss": 0.0001469652823288925,
      "step": 16760,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.002088646753691137,
      "weighted_relation_loss": 0.002088646753691137
    },
    {
      "action_query_norm": 40.81868896484375,
      "action_relation_loss": 0.002123763260897249,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.68755846585594,
      "flow_matching_loss": 0.02586027290672064,
      "grad_norm": 0.09553369879722595,
      "learning_rate": 4.439551691300236e-05,
      "loss": 0.0281,
      "relation_loss": 0.0022688849130645394,
      "selected_pairs": 26663.1,
      "state_query_norm": 42.71280784606934,
      "state_relation_loss": 0.00014512164052575826,
      "step": 16770,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0022688849130645394,
      "weighted_relation_loss": 0.0022688849130645394
    },
    {
      "action_query_norm": 40.8027587890625,
      "action_relation_loss": 0.002103013650048524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.696913002806362,
      "flow_matching_loss": 0.025743403658270835,
      "grad_norm": 0.0962323546409607,
      "learning_rate": 4.434075200204599e-05,
      "loss": 0.028,
      "relation_loss": 0.002255876688286662,
      "selected_pairs": 27221.9,
      "state_query_norm": 42.706369400024414,
      "state_relation_loss": 0.0001528630527900532,
      "step": 16780,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.002255876688286662,
      "weighted_relation_loss": 0.002255876688286662
    },
    {
      "action_query_norm": 40.869148635864256,
      "action_relation_loss": 0.00185506462585181,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.706267539756782,
      "flow_matching_loss": 0.026207008771598338,
      "grad_norm": 0.09761800616979599,
      "learning_rate": 4.428599396760956e-05,
      "loss": 0.0282,
      "relation_loss": 0.0020019079674966633,
      "selected_pairs": 26919.6,
      "state_query_norm": 42.708264923095705,
      "state_relation_loss": 0.00014684332709293812,
      "step": 16790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0020019079674966633,
      "weighted_relation_loss": 0.0020019079674966633
    },
    {
      "action_query_norm": 40.87564315795898,
      "action_relation_loss": 0.0019782527000643314,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.715622076707202,
      "flow_matching_loss": 0.027886082231998444,
      "grad_norm": 0.10115855187177658,
      "learning_rate": 4.423124287622926e-05,
      "loss": 0.03,
      "relation_loss": 0.002125299780163914,
      "selected_pairs": 26828.6,
      "state_query_norm": 42.71410293579102,
      "state_relation_loss": 0.00014704705754411406,
      "step": 16800,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.002125299780163914,
      "weighted_relation_loss": 0.002125299780163914
    },
    {
      "action_query_norm": 40.86308135986328,
      "action_relation_loss": 0.00193051794776693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.724976613657624,
      "flow_matching_loss": 0.02766245845705271,
      "grad_norm": 0.13403691351413727,
      "learning_rate": 4.417649879443282e-05,
      "loss": 0.0297,
      "relation_loss": 0.0020659680827520787,
      "selected_pairs": 27083.9,
      "state_query_norm": 42.71773262023926,
      "state_relation_loss": 0.00013545012916438282,
      "step": 16810,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0020659680827520787,
      "weighted_relation_loss": 0.0020659680827520787
    },
    {
      "action_query_norm": 40.82848587036133,
      "action_relation_loss": 0.001971466024406254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.734331150608044,
      "flow_matching_loss": 0.026952263712882996,
      "grad_norm": 0.09087168425321579,
      "learning_rate": 4.4121761788739444e-05,
      "loss": 0.0291,
      "relation_loss": 0.0021209866739809514,
      "selected_pairs": 26829.6,
      "state_query_norm": 42.696458435058595,
      "state_relation_loss": 0.00014952064375393093,
      "step": 16820,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0021209866739809514,
      "weighted_relation_loss": 0.0021209866739809514
    },
    {
      "action_query_norm": 40.80686378479004,
      "action_relation_loss": 0.001936901663430035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.743685687558466,
      "flow_matching_loss": 0.027156095206737518,
      "grad_norm": 0.10468840599060059,
      "learning_rate": 4.4067031925659805e-05,
      "loss": 0.0292,
      "relation_loss": 0.002074716705828905,
      "selected_pairs": 27243.5,
      "state_query_norm": 42.69953918457031,
      "state_relation_loss": 0.00013781502057099714,
      "step": 16830,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.002074716705828905,
      "weighted_relation_loss": 0.002074716705828905
    },
    {
      "action_query_norm": 40.82547492980957,
      "action_relation_loss": 0.0020832615555264056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.753040224508887,
      "flow_matching_loss": 0.02609914857894182,
      "grad_norm": 0.10428952425718307,
      "learning_rate": 4.4012309271695816e-05,
      "loss": 0.0283,
      "relation_loss": 0.0022237035213038326,
      "selected_pairs": 27393.9,
      "state_query_norm": 42.70392189025879,
      "state_relation_loss": 0.00014044198833289555,
      "step": 16840,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0022237035213038326,
      "weighted_relation_loss": 0.0022237035213038326
    },
    {
      "action_query_norm": 40.83094673156738,
      "action_relation_loss": 0.0018806345295161008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.762394761459309,
      "flow_matching_loss": 0.026164165884256362,
      "grad_norm": 0.1066536009311676,
      "learning_rate": 4.395759389334067e-05,
      "loss": 0.0282,
      "relation_loss": 0.0020199096063151954,
      "selected_pairs": 27478.0,
      "state_query_norm": 42.70562171936035,
      "state_relation_loss": 0.00013927508916822262,
      "step": 16850,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0020199096063151954,
      "weighted_relation_loss": 0.0020199096063151954
    },
    {
      "action_query_norm": 40.86631813049316,
      "action_relation_loss": 0.0019511039485223592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.771749298409729,
      "flow_matching_loss": 0.025453356467187403,
      "grad_norm": 0.09991640597581863,
      "learning_rate": 4.3902885857078685e-05,
      "loss": 0.0275,
      "relation_loss": 0.002095183136407286,
      "selected_pairs": 27087.4,
      "state_query_norm": 42.69884490966797,
      "state_relation_loss": 0.00014407921262318268,
      "step": 16860,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.002095183136407286,
      "weighted_relation_loss": 0.002095183136407286
    },
    {
      "action_query_norm": 40.83555183410645,
      "action_relation_loss": 0.001927066594362259,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.781103835360149,
      "flow_matching_loss": 0.02643111739307642,
      "grad_norm": 0.11082499474287033,
      "learning_rate": 4.384818522938531e-05,
      "loss": 0.0285,
      "relation_loss": 0.0020687269396148624,
      "selected_pairs": 26917.1,
      "state_query_norm": 42.70012283325195,
      "state_relation_loss": 0.000141660327790305,
      "step": 16870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0020687269396148624,
      "weighted_relation_loss": 0.0020687269396148624
    },
    {
      "action_query_norm": 40.83154563903808,
      "action_relation_loss": 0.0019679683027788997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.790458372310571,
      "flow_matching_loss": 0.02831730879843235,
      "grad_norm": 0.11867258697748184,
      "learning_rate": 4.379349207672695e-05,
      "loss": 0.0304,
      "relation_loss": 0.0021066790097393094,
      "selected_pairs": 27385.7,
      "state_query_norm": 42.702477264404294,
      "state_relation_loss": 0.00013871071132598445,
      "step": 16880,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0021066790097393094,
      "weighted_relation_loss": 0.0021066790097393094
    },
    {
      "action_query_norm": 40.801293563842776,
      "action_relation_loss": 0.0019283370696939529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.799812909260991,
      "flow_matching_loss": 0.02673528026789427,
      "grad_norm": 0.12252090871334076,
      "learning_rate": 4.373880646556098e-05,
      "loss": 0.0288,
      "relation_loss": 0.0020611398154869675,
      "selected_pairs": 27061.1,
      "state_query_norm": 42.69198837280273,
      "state_relation_loss": 0.00013280275888973846,
      "step": 16890,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0020611398154869675,
      "weighted_relation_loss": 0.0020611398154869675
    },
    {
      "action_query_norm": 40.82889404296875,
      "action_relation_loss": 0.0019743890734389424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.809167446211413,
      "flow_matching_loss": 0.026926122047007083,
      "grad_norm": 0.12081946432590485,
      "learning_rate": 4.3684128462335533e-05,
      "loss": 0.029,
      "relation_loss": 0.002117121999617666,
      "selected_pairs": 26692.7,
      "state_query_norm": 42.70530891418457,
      "state_relation_loss": 0.00014273292035795748,
      "step": 16900,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.002117121999617666,
      "weighted_relation_loss": 0.002117121999617666
    },
    {
      "action_query_norm": 40.86318855285644,
      "action_relation_loss": 0.0018641909351572394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.818521983161833,
      "flow_matching_loss": 0.02565052080899477,
      "grad_norm": 0.09754162281751633,
      "learning_rate": 4.362945813348955e-05,
      "loss": 0.0277,
      "relation_loss": 0.002005358482711017,
      "selected_pairs": 26625.6,
      "state_query_norm": 42.71294174194336,
      "state_relation_loss": 0.00014116751844994724,
      "step": 16910,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.002005358482711017,
      "weighted_relation_loss": 0.002005358482711017
    },
    {
      "action_query_norm": 40.80831642150879,
      "action_relation_loss": 0.001924784085713327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.827876520112255,
      "flow_matching_loss": 0.027241169288754464,
      "grad_norm": 0.13926224410533905,
      "learning_rate": 4.357479554545262e-05,
      "loss": 0.0293,
      "relation_loss": 0.002070630295202136,
      "selected_pairs": 27200.8,
      "state_query_norm": 42.689219284057614,
      "state_relation_loss": 0.00014584618184017019,
      "step": 16920,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.002070630295202136,
      "weighted_relation_loss": 0.002070630295202136
    },
    {
      "action_query_norm": 40.77109146118164,
      "action_relation_loss": 0.0019116719835437835,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.837231057062676,
      "flow_matching_loss": 0.026392952911555766,
      "grad_norm": 0.11075583100318909,
      "learning_rate": 4.352014076464499e-05,
      "loss": 0.0284,
      "relation_loss": 0.0020421785302460194,
      "selected_pairs": 26894.3,
      "state_query_norm": 42.69969215393066,
      "state_relation_loss": 0.00013050655470578932,
      "step": 16930,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0020421785302460194,
      "weighted_relation_loss": 0.0020421785302460194
    },
    {
      "action_query_norm": 40.871009826660156,
      "action_relation_loss": 0.0019570098142139614,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.846585594013096,
      "flow_matching_loss": 0.02698018979281187,
      "grad_norm": 0.12356553971767426,
      "learning_rate": 4.346549385747734e-05,
      "loss": 0.0291,
      "relation_loss": 0.0021005662740208207,
      "selected_pairs": 27008.5,
      "state_query_norm": 42.686360931396486,
      "state_relation_loss": 0.0001435564467101358,
      "step": 16940,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0021005662740208207,
      "weighted_relation_loss": 0.0021005662740208207
    },
    {
      "action_query_norm": 40.803130340576175,
      "action_relation_loss": 0.0018239295342937112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.855940130963518,
      "flow_matching_loss": 0.027622847631573678,
      "grad_norm": 0.11117660254240036,
      "learning_rate": 4.34108548903508e-05,
      "loss": 0.0296,
      "relation_loss": 0.00196213168092072,
      "selected_pairs": 27034.0,
      "state_query_norm": 42.69294624328613,
      "state_relation_loss": 0.00013820213716826403,
      "step": 16950,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00196213168092072,
      "weighted_relation_loss": 0.00196213168092072
    },
    {
      "action_query_norm": 40.79911231994629,
      "action_relation_loss": 0.0018835463444702328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.865294667913938,
      "flow_matching_loss": 0.026761292293667794,
      "grad_norm": 0.14553631842136383,
      "learning_rate": 4.335622392965689e-05,
      "loss": 0.0288,
      "relation_loss": 0.0020164627465419473,
      "selected_pairs": 26692.1,
      "state_query_norm": 42.68783721923828,
      "state_relation_loss": 0.0001329164137132466,
      "step": 16960,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0020164627465419473,
      "weighted_relation_loss": 0.0020164627465419473
    },
    {
      "action_query_norm": 40.79980812072754,
      "action_relation_loss": 0.0018913926673121751,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.87464920486436,
      "flow_matching_loss": 0.02776844222098589,
      "grad_norm": 0.10574588924646378,
      "learning_rate": 4.330160104177738e-05,
      "loss": 0.0298,
      "relation_loss": 0.0020286589744500814,
      "selected_pairs": 27527.9,
      "state_query_norm": 42.69362525939941,
      "state_relation_loss": 0.0001372663042275235,
      "step": 16970,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0020286589744500814,
      "weighted_relation_loss": 0.0020286589744500814
    },
    {
      "action_query_norm": 40.78924789428711,
      "action_relation_loss": 0.0020060930750332774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.88400374181478,
      "flow_matching_loss": 0.025868116319179534,
      "grad_norm": 0.09435927867889404,
      "learning_rate": 4.32469862930842e-05,
      "loss": 0.028,
      "relation_loss": 0.002141034952364862,
      "selected_pairs": 26807.5,
      "state_query_norm": 42.695912551879886,
      "state_relation_loss": 0.00013494187587639316,
      "step": 16980,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.002141034952364862,
      "weighted_relation_loss": 0.002141034952364862
    },
    {
      "action_query_norm": 40.77744560241699,
      "action_relation_loss": 0.001964309997856617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.8933582787652,
      "flow_matching_loss": 0.028285220079123975,
      "grad_norm": 0.11479820311069489,
      "learning_rate": 4.319237974993946e-05,
      "loss": 0.0304,
      "relation_loss": 0.002099845535121858,
      "selected_pairs": 27216.5,
      "state_query_norm": 42.68127975463867,
      "state_relation_loss": 0.00013553555036196485,
      "step": 16990,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.002099845535121858,
      "weighted_relation_loss": 0.002099845535121858
    },
    {
      "action_query_norm": 40.74194488525391,
      "action_relation_loss": 0.0018555202637799083,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.902712815715622,
      "flow_matching_loss": 0.0276576803997159,
      "grad_norm": 0.12343906611204147,
      "learning_rate": 4.3137781478695236e-05,
      "loss": 0.0296,
      "relation_loss": 0.0019857609760947524,
      "selected_pairs": 26617.3,
      "state_query_norm": 42.68232421875,
      "state_relation_loss": 0.00013024067375226877,
      "step": 17000,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0019857609760947524,
      "weighted_relation_loss": 0.0019857609760947524
    },
    {
      "action_query_norm": 40.75917625427246,
      "action_relation_loss": 0.0018783038016408683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.912067352666043,
      "flow_matching_loss": 0.02806731462478638,
      "grad_norm": 0.10726005584001541,
      "learning_rate": 4.308319154569358e-05,
      "loss": 0.0301,
      "relation_loss": 0.0020148826646618546,
      "selected_pairs": 26991.5,
      "state_query_norm": 42.69268684387207,
      "state_relation_loss": 0.0001365788404655177,
      "step": 17010,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0020148826646618546,
      "weighted_relation_loss": 0.0020148826646618546
    },
    {
      "action_query_norm": 40.762451171875,
      "action_relation_loss": 0.0019571786164306105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.921421889616465,
      "flow_matching_loss": 0.02595021054148674,
      "grad_norm": 0.08095492422580719,
      "learning_rate": 4.302861001726641e-05,
      "loss": 0.028,
      "relation_loss": 0.00209515307797119,
      "selected_pairs": 27675.7,
      "state_query_norm": 42.676553344726564,
      "state_relation_loss": 0.00013797447245451623,
      "step": 17020,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.00209515307797119,
      "weighted_relation_loss": 0.00209515307797119
    },
    {
      "action_query_norm": 40.7734489440918,
      "action_relation_loss": 0.0018807410029694437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.930776426566885,
      "flow_matching_loss": 0.026971714198589326,
      "grad_norm": 0.0923846885561943,
      "learning_rate": 4.2974036959735415e-05,
      "loss": 0.029,
      "relation_loss": 0.0020253762137144803,
      "selected_pairs": 27014.2,
      "state_query_norm": 42.66766700744629,
      "state_relation_loss": 0.00014463518600678072,
      "step": 17030,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0020253762137144803,
      "weighted_relation_loss": 0.0020253762137144803
    },
    {
      "action_query_norm": 40.7999885559082,
      "action_relation_loss": 0.0019436059053987264,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.940130963517305,
      "flow_matching_loss": 0.02774844504892826,
      "grad_norm": 0.10907593369483948,
      "learning_rate": 4.2919472439412034e-05,
      "loss": 0.0298,
      "relation_loss": 0.0020756083657033743,
      "selected_pairs": 26272.7,
      "state_query_norm": 42.677529907226564,
      "state_relation_loss": 0.00013200244793551975,
      "step": 17040,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0020756083657033743,
      "weighted_relation_loss": 0.0020756083657033743
    },
    {
      "action_query_norm": 40.75759773254394,
      "action_relation_loss": 0.0020506847533397375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.949485500467727,
      "flow_matching_loss": 0.025539403408765794,
      "grad_norm": 0.11229484528303146,
      "learning_rate": 4.286491652259729e-05,
      "loss": 0.0277,
      "relation_loss": 0.002189264795742929,
      "selected_pairs": 26462.8,
      "state_query_norm": 42.65746879577637,
      "state_relation_loss": 0.00013858004822395742,
      "step": 17050,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.002189264795742929,
      "weighted_relation_loss": 0.002189264795742929
    },
    {
      "action_query_norm": 40.8019603729248,
      "action_relation_loss": 0.001879702229052782,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.958840037418147,
      "flow_matching_loss": 0.029412411153316498,
      "grad_norm": 0.1142219603061676,
      "learning_rate": 4.281036927558178e-05,
      "loss": 0.0314,
      "relation_loss": 0.002016670792363584,
      "selected_pairs": 27129.1,
      "state_query_norm": 42.67545433044434,
      "state_relation_loss": 0.00013696854439331218,
      "step": 17060,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.002016670792363584,
      "weighted_relation_loss": 0.002016670792363584
    },
    {
      "action_query_norm": 40.82140121459961,
      "action_relation_loss": 0.0019358330289833247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.96819457436857,
      "flow_matching_loss": 0.027278580516576768,
      "grad_norm": 0.1490449756383896,
      "learning_rate": 4.275583076464552e-05,
      "loss": 0.0294,
      "relation_loss": 0.002073230256792158,
      "selected_pairs": 26433.5,
      "state_query_norm": 42.677778244018555,
      "state_relation_loss": 0.00013739722926402463,
      "step": 17070,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.002073230256792158,
      "weighted_relation_loss": 0.002073230256792158
    },
    {
      "action_query_norm": 40.821974182128905,
      "action_relation_loss": 0.0018426778842695057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.97754911131899,
      "flow_matching_loss": 0.026571835204958916,
      "grad_norm": 0.10636678338050842,
      "learning_rate": 4.270130105605794e-05,
      "loss": 0.0286,
      "relation_loss": 0.0019798541208729146,
      "selected_pairs": 26940.6,
      "state_query_norm": 42.66800994873047,
      "state_relation_loss": 0.000137176218413515,
      "step": 17080,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0019798541208729146,
      "weighted_relation_loss": 0.0019798541208729146
    },
    {
      "action_query_norm": 40.78549499511719,
      "action_relation_loss": 0.001953652105294168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.986903648269411,
      "flow_matching_loss": 0.02711614891886711,
      "grad_norm": 0.13402383029460907,
      "learning_rate": 4.264678021607782e-05,
      "loss": 0.0292,
      "relation_loss": 0.002088268497027457,
      "selected_pairs": 26878.4,
      "state_query_norm": 42.66143646240234,
      "state_relation_loss": 0.00013461639173328876,
      "step": 17090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.002088268497027457,
      "weighted_relation_loss": 0.002088268497027457
    },
    {
      "action_query_norm": 40.79930305480957,
      "action_relation_loss": 0.0019526016898453236,
      "auxiliary_loss_weight": 1.0,
      "epoch": 15.996258185219832,
      "flow_matching_loss": 0.02750228624790907,
      "grad_norm": 0.09519395977258682,
      "learning_rate": 4.25922683109531e-05,
      "loss": 0.0296,
      "relation_loss": 0.002087301702704281,
      "selected_pairs": 26776.8,
      "state_query_norm": 42.656842422485354,
      "state_relation_loss": 0.00013470000267261638,
      "step": 17100,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.002087301702704281,
      "weighted_relation_loss": 0.002087301702704281
    },
    {
      "action_query_norm": 40.81581573486328,
      "action_relation_loss": 0.00180713803274557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.00561272217025,
      "flow_matching_loss": 0.025903908535838127,
      "grad_norm": 0.11093860864639282,
      "learning_rate": 4.25377654069209e-05,
      "loss": 0.0278,
      "relation_loss": 0.0019440531497821212,
      "selected_pairs": 24332.7,
      "state_query_norm": 42.6701602935791,
      "state_relation_loss": 0.00013691509811906145,
      "step": 17110,
      "valid_action_samples": 214.4,
      "weighted_auxiliary_loss": 0.0019440531497821212,
      "weighted_relation_loss": 0.0019440531497821212
    },
    {
      "action_query_norm": 40.74397659301758,
      "action_relation_loss": 0.0018916376866400242,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.014967259120674,
      "flow_matching_loss": 0.028847026452422143,
      "grad_norm": 0.1277337670326233,
      "learning_rate": 4.248327157020737e-05,
      "loss": 0.0309,
      "relation_loss": 0.0020220509613864124,
      "selected_pairs": 26858.2,
      "state_query_norm": 42.66154594421387,
      "state_relation_loss": 0.000130413293663878,
      "step": 17120,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0020220509613864124,
      "weighted_relation_loss": 0.0020220509613864124
    },
    {
      "action_query_norm": 40.76003875732422,
      "action_relation_loss": 0.0018554432666860522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.024321796071096,
      "flow_matching_loss": 0.02514024768024683,
      "grad_norm": 0.10158047080039978,
      "learning_rate": 4.242878686702763e-05,
      "loss": 0.0271,
      "relation_loss": 0.0019923710613511504,
      "selected_pairs": 27454.4,
      "state_query_norm": 42.65883293151855,
      "state_relation_loss": 0.00013692780121346004,
      "step": 17130,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0019923710613511504,
      "weighted_relation_loss": 0.0019923710613511504
    },
    {
      "action_query_norm": 40.7587947845459,
      "action_relation_loss": 0.0017801601206883788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.033676333021514,
      "flow_matching_loss": 0.024792742542922498,
      "grad_norm": 0.0847906693816185,
      "learning_rate": 4.23743113635858e-05,
      "loss": 0.0267,
      "relation_loss": 0.001915646460838616,
      "selected_pairs": 26645.1,
      "state_query_norm": 42.66081085205078,
      "state_relation_loss": 0.0001354863736196421,
      "step": 17140,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.001915646460838616,
      "weighted_relation_loss": 0.001915646460838616
    },
    {
      "action_query_norm": 40.77411155700683,
      "action_relation_loss": 0.0018399150343611837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.043030869971936,
      "flow_matching_loss": 0.026108612678945063,
      "grad_norm": 0.09550037980079651,
      "learning_rate": 4.2319845126074705e-05,
      "loss": 0.0281,
      "relation_loss": 0.0019703777972608803,
      "selected_pairs": 26788.6,
      "state_query_norm": 42.6715648651123,
      "state_relation_loss": 0.00013046280218986794,
      "step": 17150,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0019703777972608803,
      "weighted_relation_loss": 0.0019703777972608803
    },
    {
      "action_query_norm": 40.807373046875,
      "action_relation_loss": 0.0017737766378559173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.052385406922358,
      "flow_matching_loss": 0.024669346772134303,
      "grad_norm": 0.10267515480518341,
      "learning_rate": 4.226538822067598e-05,
      "loss": 0.0266,
      "relation_loss": 0.0019035270204767584,
      "selected_pairs": 26862.8,
      "state_query_norm": 42.679541397094724,
      "state_relation_loss": 0.00012975037097930908,
      "step": 17160,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0019035270204767584,
      "weighted_relation_loss": 0.0019035270204767584
    },
    {
      "action_query_norm": 40.76453170776367,
      "action_relation_loss": 0.0019198071211576461,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.061739943872777,
      "flow_matching_loss": 0.026700873486697675,
      "grad_norm": 0.13906538486480713,
      "learning_rate": 4.2210940713559896e-05,
      "loss": 0.0288,
      "relation_loss": 0.0020549864741042255,
      "selected_pairs": 26880.3,
      "state_query_norm": 42.64834785461426,
      "state_relation_loss": 0.00013517936677089892,
      "step": 17170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0020549864741042255,
      "weighted_relation_loss": 0.0020549864741042255
    },
    {
      "action_query_norm": 40.75447425842285,
      "action_relation_loss": 0.0018404596252366901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.0710944808232,
      "flow_matching_loss": 0.025627658888697624,
      "grad_norm": 0.1083664819598198,
      "learning_rate": 4.21565026708853e-05,
      "loss": 0.0276,
      "relation_loss": 0.001972839783411473,
      "selected_pairs": 26399.7,
      "state_query_norm": 42.66345291137695,
      "state_relation_loss": 0.00013238013852969743,
      "step": 17180,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001972839783411473,
      "weighted_relation_loss": 0.001972839783411473
    },
    {
      "action_query_norm": 40.7582088470459,
      "action_relation_loss": 0.0018321292358450592,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.08044901777362,
      "flow_matching_loss": 0.023634298145771025,
      "grad_norm": 0.10611126571893692,
      "learning_rate": 4.2102074158799535e-05,
      "loss": 0.0256,
      "relation_loss": 0.001958441885653883,
      "selected_pairs": 26803.0,
      "state_query_norm": 42.645639419555664,
      "state_relation_loss": 0.00012631266436073928,
      "step": 17190,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001958441885653883,
      "weighted_relation_loss": 0.001958441885653883
    },
    {
      "action_query_norm": 40.72117042541504,
      "action_relation_loss": 0.0018442760570906102,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.089803554724043,
      "flow_matching_loss": 0.02575718816369772,
      "grad_norm": 0.11539370566606522,
      "learning_rate": 4.204765524343841e-05,
      "loss": 0.0277,
      "relation_loss": 0.001969881425611675,
      "selected_pairs": 26847.7,
      "state_query_norm": 42.65271110534668,
      "state_relation_loss": 0.00012560534305521286,
      "step": 17200,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001969881425611675,
      "weighted_relation_loss": 0.001969881425611675
    },
    {
      "action_query_norm": 40.693914794921874,
      "action_relation_loss": 0.0018733582459390164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.09915809167446,
      "flow_matching_loss": 0.025759087316691874,
      "grad_norm": 0.11634327471256256,
      "learning_rate": 4.1993245990926036e-05,
      "loss": 0.0278,
      "relation_loss": 0.002000825188588351,
      "selected_pairs": 26280.7,
      "state_query_norm": 42.63823547363281,
      "state_relation_loss": 0.00012746693901135586,
      "step": 17210,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.002000825188588351,
      "weighted_relation_loss": 0.002000825188588351
    },
    {
      "action_query_norm": 40.70313568115235,
      "action_relation_loss": 0.0018351545906625689,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.108512628624883,
      "flow_matching_loss": 0.02668808214366436,
      "grad_norm": 0.14964066445827484,
      "learning_rate": 4.193884646737474e-05,
      "loss": 0.0287,
      "relation_loss": 0.0019643271691165863,
      "selected_pairs": 26422.3,
      "state_query_norm": 42.65370483398438,
      "state_relation_loss": 0.00012917258427478374,
      "step": 17220,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0019643271691165863,
      "weighted_relation_loss": 0.0019643271691165863
    },
    {
      "action_query_norm": 40.75181198120117,
      "action_relation_loss": 0.0018693094374611975,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.117867165575305,
      "flow_matching_loss": 0.026327809505164623,
      "grad_norm": 0.10865689814090729,
      "learning_rate": 4.1884456738885124e-05,
      "loss": 0.0283,
      "relation_loss": 0.0019996468676254154,
      "selected_pairs": 26233.6,
      "state_query_norm": 42.644618606567384,
      "state_relation_loss": 0.00013033743307460098,
      "step": 17230,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0019996468676254154,
      "weighted_relation_loss": 0.0019996468676254154
    },
    {
      "action_query_norm": 40.781900787353514,
      "action_relation_loss": 0.0018482917337678372,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.127221702525723,
      "flow_matching_loss": 0.02636253535747528,
      "grad_norm": 0.11990481615066528,
      "learning_rate": 4.1830076871545795e-05,
      "loss": 0.0283,
      "relation_loss": 0.0019808238954283296,
      "selected_pairs": 27080.3,
      "state_query_norm": 42.63668594360352,
      "state_relation_loss": 0.00013253217475721611,
      "step": 17240,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0019808238954283296,
      "weighted_relation_loss": 0.0019808238954283296
    },
    {
      "action_query_norm": 40.77028884887695,
      "action_relation_loss": 0.0017857235972769559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.136576239476145,
      "flow_matching_loss": 0.02683706898242235,
      "grad_norm": 0.21796171367168427,
      "learning_rate": 4.177570693143346e-05,
      "loss": 0.0288,
      "relation_loss": 0.0019137742929160595,
      "selected_pairs": 27015.2,
      "state_query_norm": 42.64758567810058,
      "state_relation_loss": 0.0001280506890907418,
      "step": 17250,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0019137742929160595,
      "weighted_relation_loss": 0.0019137742929160595
    },
    {
      "action_query_norm": 40.72089385986328,
      "action_relation_loss": 0.001920619630254805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.145930776426567,
      "flow_matching_loss": 0.027486360259354116,
      "grad_norm": 0.1219119280576706,
      "learning_rate": 4.172134698461271e-05,
      "loss": 0.0295,
      "relation_loss": 0.0020532395108602943,
      "selected_pairs": 26547.1,
      "state_query_norm": 42.62863731384277,
      "state_relation_loss": 0.0001326198980677873,
      "step": 17260,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0020532395108602943,
      "weighted_relation_loss": 0.0020532395108602943
    },
    {
      "action_query_norm": 40.72635383605957,
      "action_relation_loss": 0.0017968177562579513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.15528531337699,
      "flow_matching_loss": 0.02502346206456423,
      "grad_norm": 0.09645617008209229,
      "learning_rate": 4.166699709713599e-05,
      "loss": 0.027,
      "relation_loss": 0.0019367333501577378,
      "selected_pairs": 27318.2,
      "state_query_norm": 42.643378829956056,
      "state_relation_loss": 0.00013991557061672212,
      "step": 17270,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0019367333501577378,
      "weighted_relation_loss": 0.0019367333501577378
    },
    {
      "action_query_norm": 40.75348091125488,
      "action_relation_loss": 0.0017818336840718985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.164639850327408,
      "flow_matching_loss": 0.02747940644621849,
      "grad_norm": 0.10310487449169159,
      "learning_rate": 4.1612657335043555e-05,
      "loss": 0.0294,
      "relation_loss": 0.0019180352450348438,
      "selected_pairs": 26962.5,
      "state_query_norm": 42.643425369262694,
      "state_relation_loss": 0.0001362015653285198,
      "step": 17280,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0019180352450348438,
      "weighted_relation_loss": 0.0019180352450348438
    },
    {
      "action_query_norm": 40.72154197692871,
      "action_relation_loss": 0.001870677643455565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.17399438727783,
      "flow_matching_loss": 0.02669179029762745,
      "grad_norm": 0.15080223977565765,
      "learning_rate": 4.155832776436331e-05,
      "loss": 0.0287,
      "relation_loss": 0.0019987583742477,
      "selected_pairs": 26946.3,
      "state_query_norm": 42.63488235473633,
      "state_relation_loss": 0.00012808073588530534,
      "step": 17290,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0019987583742477,
      "weighted_relation_loss": 0.0019987583742477
    },
    {
      "action_query_norm": 40.72788276672363,
      "action_relation_loss": 0.001870160666294396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.18334892422825,
      "flow_matching_loss": 0.027831114083528518,
      "grad_norm": 0.1376296579837799,
      "learning_rate": 4.1504008451110845e-05,
      "loss": 0.0298,
      "relation_loss": 0.0020051887375302615,
      "selected_pairs": 26807.3,
      "state_query_norm": 42.62615776062012,
      "state_relation_loss": 0.0001350280683254823,
      "step": 17300,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0020051887375302615,
      "weighted_relation_loss": 0.0020051887375302615
    },
    {
      "action_query_norm": 40.71028060913086,
      "action_relation_loss": 0.0018267799750901759,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.19270346117867,
      "flow_matching_loss": 0.028216480277478695,
      "grad_norm": 0.12052169442176819,
      "learning_rate": 4.1449699461289225e-05,
      "loss": 0.0302,
      "relation_loss": 0.001958580664359033,
      "selected_pairs": 26669.4,
      "state_query_norm": 42.6287181854248,
      "state_relation_loss": 0.00013180069872760215,
      "step": 17310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001958580664359033,
      "weighted_relation_loss": 0.001958580664359033
    },
    {
      "action_query_norm": 40.69994468688965,
      "action_relation_loss": 0.0017628215835429728,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.202057998129092,
      "flow_matching_loss": 0.026709019020199776,
      "grad_norm": 0.09814555943012238,
      "learning_rate": 4.139540086088901e-05,
      "loss": 0.0286,
      "relation_loss": 0.0019029323826543986,
      "selected_pairs": 26893.1,
      "state_query_norm": 42.6218620300293,
      "state_relation_loss": 0.00014011080493219196,
      "step": 17320,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0019029323826543986,
      "weighted_relation_loss": 0.0019029323826543986
    },
    {
      "action_query_norm": 40.703287506103514,
      "action_relation_loss": 0.001928376266732812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.211412535079514,
      "flow_matching_loss": 0.02690907660871744,
      "grad_norm": 0.1028454601764679,
      "learning_rate": 4.1341112715888106e-05,
      "loss": 0.029,
      "relation_loss": 0.002063163777347654,
      "selected_pairs": 26734.6,
      "state_query_norm": 42.64504776000977,
      "state_relation_loss": 0.00013478753535309807,
      "step": 17330,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.002063163777347654,
      "weighted_relation_loss": 0.002063163777347654
    },
    {
      "action_query_norm": 40.68726119995117,
      "action_relation_loss": 0.0018679700791835786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.220767072029936,
      "flow_matching_loss": 0.02458479441702366,
      "grad_norm": 0.1208045482635498,
      "learning_rate": 4.128683509225172e-05,
      "loss": 0.0266,
      "relation_loss": 0.0020116939558647573,
      "selected_pairs": 26870.3,
      "state_query_norm": 42.618842697143556,
      "state_relation_loss": 0.0001437239014194347,
      "step": 17340,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0020116939558647573,
      "weighted_relation_loss": 0.0020116939558647573
    },
    {
      "action_query_norm": 40.75039825439453,
      "action_relation_loss": 0.0017828643671236933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.230121608980355,
      "flow_matching_loss": 0.027366535365581514,
      "grad_norm": 0.12318538874387741,
      "learning_rate": 4.123256805593231e-05,
      "loss": 0.0293,
      "relation_loss": 0.001916245510801673,
      "selected_pairs": 27598.8,
      "state_query_norm": 42.62968063354492,
      "state_relation_loss": 0.0001333811625954695,
      "step": 17350,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.001916245510801673,
      "weighted_relation_loss": 0.001916245510801673
    },
    {
      "action_query_norm": 40.69527969360352,
      "action_relation_loss": 0.001844295603223145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.239476145930777,
      "flow_matching_loss": 0.026301308907568453,
      "grad_norm": 0.09335934370756149,
      "learning_rate": 4.117831167286943e-05,
      "loss": 0.0283,
      "relation_loss": 0.0019784923875704407,
      "selected_pairs": 26575.5,
      "state_query_norm": 42.6118106842041,
      "state_relation_loss": 0.00013419675669865683,
      "step": 17360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0019784923875704407,
      "weighted_relation_loss": 0.0019784923875704407
    },
    {
      "action_query_norm": 40.72581481933594,
      "action_relation_loss": 0.0018541515921242536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.2488306828812,
      "flow_matching_loss": 0.02733362391591072,
      "grad_norm": 0.11290934681892395,
      "learning_rate": 4.1124066008989685e-05,
      "loss": 0.0293,
      "relation_loss": 0.0019880752079188824,
      "selected_pairs": 26571.2,
      "state_query_norm": 42.61150932312012,
      "state_relation_loss": 0.0001339236128842458,
      "step": 17370,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0019880752079188824,
      "weighted_relation_loss": 0.0019880752079188824
    },
    {
      "action_query_norm": 40.7040901184082,
      "action_relation_loss": 0.0018871884443797172,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.258185219831617,
      "flow_matching_loss": 0.027030954882502557,
      "grad_norm": 0.10267995297908783,
      "learning_rate": 4.106983113020669e-05,
      "loss": 0.029,
      "relation_loss": 0.0020168673363514245,
      "selected_pairs": 27059.1,
      "state_query_norm": 42.628781127929685,
      "state_relation_loss": 0.0001296788890613243,
      "step": 17380,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0020168673363514245,
      "weighted_relation_loss": 0.0020168673363514245
    },
    {
      "action_query_norm": 40.6922191619873,
      "action_relation_loss": 0.0017938841017894446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.26753975678204,
      "flow_matching_loss": 0.025638300739228726,
      "grad_norm": 0.12029701471328735,
      "learning_rate": 4.101560710242094e-05,
      "loss": 0.0276,
      "relation_loss": 0.0019170225714333355,
      "selected_pairs": 27393.7,
      "state_query_norm": 42.620682907104495,
      "state_relation_loss": 0.0001231384594575502,
      "step": 17390,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0019170225714333355,
      "weighted_relation_loss": 0.0019170225714333355
    },
    {
      "action_query_norm": 40.68042526245117,
      "action_relation_loss": 0.0018092570127919317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.27689429373246,
      "flow_matching_loss": 0.02785532046109438,
      "grad_norm": 0.1024281457066536,
      "learning_rate": 4.096139399151971e-05,
      "loss": 0.0298,
      "relation_loss": 0.0019397891242988407,
      "selected_pairs": 26617.4,
      "state_query_norm": 42.62798538208008,
      "state_relation_loss": 0.00013053211296210065,
      "step": 17400,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0019397891242988407,
      "weighted_relation_loss": 0.0019397891242988407
    },
    {
      "action_query_norm": 40.69242858886719,
      "action_relation_loss": 0.0017624772735871375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.28624883068288,
      "flow_matching_loss": 0.025767639838159086,
      "grad_norm": 0.0949728786945343,
      "learning_rate": 4.090719186337709e-05,
      "loss": 0.0277,
      "relation_loss": 0.001888103468809277,
      "selected_pairs": 26439.1,
      "state_query_norm": 42.6171199798584,
      "state_relation_loss": 0.00012562616393552162,
      "step": 17410,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001888103468809277,
      "weighted_relation_loss": 0.001888103468809277
    },
    {
      "action_query_norm": 40.69801139831543,
      "action_relation_loss": 0.00182360002072528,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.2956033676333,
      "flow_matching_loss": 0.02591343428939581,
      "grad_norm": 0.1097288578748703,
      "learning_rate": 4.0853000783853746e-05,
      "loss": 0.0279,
      "relation_loss": 0.00195313609438017,
      "selected_pairs": 27312.6,
      "state_query_norm": 42.61549110412598,
      "state_relation_loss": 0.00012953607583767734,
      "step": 17420,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.00195313609438017,
      "weighted_relation_loss": 0.00195313609438017
    },
    {
      "action_query_norm": 40.71030693054199,
      "action_relation_loss": 0.0018634321400895714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.304957904583723,
      "flow_matching_loss": 0.02545769102871418,
      "grad_norm": 0.15131966769695282,
      "learning_rate": 4.0798820818796956e-05,
      "loss": 0.0274,
      "relation_loss": 0.0019864510861225424,
      "selected_pairs": 26836.7,
      "state_query_norm": 42.62717056274414,
      "state_relation_loss": 0.00012301894748816266,
      "step": 17430,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0019864510861225424,
      "weighted_relation_loss": 0.0019864510861225424
    },
    {
      "action_query_norm": 40.68474578857422,
      "action_relation_loss": 0.0018804038991220295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.314312441534145,
      "flow_matching_loss": 0.024415168538689615,
      "grad_norm": 0.10267769545316696,
      "learning_rate": 4.074465203404048e-05,
      "loss": 0.0264,
      "relation_loss": 0.0020036242553032936,
      "selected_pairs": 26873.4,
      "state_query_norm": 42.61462745666504,
      "state_relation_loss": 0.0001232203510880936,
      "step": 17440,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0020036242553032936,
      "weighted_relation_loss": 0.0020036242553032936
    },
    {
      "action_query_norm": 40.686455154418944,
      "action_relation_loss": 0.0017794350278563797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.323666978484564,
      "flow_matching_loss": 0.027652410231530665,
      "grad_norm": 0.12940384447574615,
      "learning_rate": 4.0690494495404495e-05,
      "loss": 0.0296,
      "relation_loss": 0.0019029082264751195,
      "selected_pairs": 27151.7,
      "state_query_norm": 42.628513717651366,
      "state_relation_loss": 0.00012347319861873984,
      "step": 17450,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0019029082264751195,
      "weighted_relation_loss": 0.0019029082264751195
    },
    {
      "action_query_norm": 40.68826599121094,
      "action_relation_loss": 0.001807584590278566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.333021515434986,
      "flow_matching_loss": 0.024708744138479233,
      "grad_norm": 0.11535533517599106,
      "learning_rate": 4.0636348268695535e-05,
      "loss": 0.0266,
      "relation_loss": 0.0019393594819121062,
      "selected_pairs": 27220.3,
      "state_query_norm": 42.612158203125,
      "state_relation_loss": 0.00013177488726796583,
      "step": 17460,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0019393594819121062,
      "weighted_relation_loss": 0.0019393594819121062
    },
    {
      "action_query_norm": 40.63784217834473,
      "action_relation_loss": 0.0018131701741367578,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.342376052385408,
      "flow_matching_loss": 0.02605273462831974,
      "grad_norm": 0.10004603862762451,
      "learning_rate": 4.058221341970637e-05,
      "loss": 0.028,
      "relation_loss": 0.001942434418015182,
      "selected_pairs": 27046.9,
      "state_query_norm": 42.61768798828125,
      "state_relation_loss": 0.0001292642380576581,
      "step": 17470,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001942434418015182,
      "weighted_relation_loss": 0.001942434418015182
    },
    {
      "action_query_norm": 40.63657188415527,
      "action_relation_loss": 0.00178110315464437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.351730589335826,
      "flow_matching_loss": 0.02782837990671396,
      "grad_norm": 0.09461402893066406,
      "learning_rate": 4.0528090014215945e-05,
      "loss": 0.0297,
      "relation_loss": 0.001912857056595385,
      "selected_pairs": 26392.7,
      "state_query_norm": 42.61134414672851,
      "state_relation_loss": 0.00013175390486139805,
      "step": 17480,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001912857056595385,
      "weighted_relation_loss": 0.001912857056595385
    },
    {
      "action_query_norm": 40.67886466979981,
      "action_relation_loss": 0.0017805346753448248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.36108512628625,
      "flow_matching_loss": 0.026969429291784763,
      "grad_norm": 0.1404639184474945,
      "learning_rate": 4.047397811798929e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019124816870316863,
      "selected_pairs": 27994.1,
      "state_query_norm": 42.61885070800781,
      "state_relation_loss": 0.00013194701459724457,
      "step": 17490,
      "valid_action_samples": 237.1,
      "weighted_auxiliary_loss": 0.0019124816870316863,
      "weighted_relation_loss": 0.0019124816870316863
    },
    {
      "action_query_norm": 40.666053009033206,
      "action_relation_loss": 0.0018781849881634115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.37043966323667,
      "flow_matching_loss": 0.027193455956876278,
      "grad_norm": 0.10258226096630096,
      "learning_rate": 4.041987779677745e-05,
      "loss": 0.0292,
      "relation_loss": 0.002004617836792022,
      "selected_pairs": 27208.8,
      "state_query_norm": 42.6045425415039,
      "state_relation_loss": 0.00012643284935620613,
      "step": 17500,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.002004617836792022,
      "weighted_relation_loss": 0.002004617836792022
    },
    {
      "action_query_norm": 40.635643005371094,
      "action_relation_loss": 0.001703904801979661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.379794200187092,
      "flow_matching_loss": 0.02707658112049103,
      "grad_norm": 0.10054103285074234,
      "learning_rate": 4.036578911631745e-05,
      "loss": 0.0289,
      "relation_loss": 0.0018249995075166226,
      "selected_pairs": 27363.2,
      "state_query_norm": 42.60732154846191,
      "state_relation_loss": 0.00012109469098504632,
      "step": 17510,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0018249995075166226,
      "weighted_relation_loss": 0.0018249995075166226
    },
    {
      "action_query_norm": 40.626417541503905,
      "action_relation_loss": 0.001767936977557838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.38914873713751,
      "flow_matching_loss": 0.025879632122814654,
      "grad_norm": 0.11658492684364319,
      "learning_rate": 4.031171214233211e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018971327343024313,
      "selected_pairs": 26322.3,
      "state_query_norm": 42.60982780456543,
      "state_relation_loss": 0.000129195764020551,
      "step": 17520,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0018971327343024313,
      "weighted_relation_loss": 0.0018971327343024313
    },
    {
      "action_query_norm": 40.63085250854492,
      "action_relation_loss": 0.0018252401961944998,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.398503274087933,
      "flow_matching_loss": 0.024669544771313667,
      "grad_norm": 0.13900572061538696,
      "learning_rate": 4.025764694053008e-05,
      "loss": 0.0266,
      "relation_loss": 0.001959381252527237,
      "selected_pairs": 27383.6,
      "state_query_norm": 42.60694923400879,
      "state_relation_loss": 0.00013414108761935494,
      "step": 17530,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.001959381252527237,
      "weighted_relation_loss": 0.001959381252527237
    },
    {
      "action_query_norm": 40.63867301940918,
      "action_relation_loss": 0.001766411354765296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.407857811038355,
      "flow_matching_loss": 0.023701498098671437,
      "grad_norm": 0.10740111023187637,
      "learning_rate": 4.020359357660566e-05,
      "loss": 0.0256,
      "relation_loss": 0.0018920738948509097,
      "selected_pairs": 26552.7,
      "state_query_norm": 42.59540367126465,
      "state_relation_loss": 0.00012566254881676286,
      "step": 17540,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0018920738948509097,
      "weighted_relation_loss": 0.0018920738948509097
    },
    {
      "action_query_norm": 40.63281135559082,
      "action_relation_loss": 0.0018172496929764748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.417212347988773,
      "flow_matching_loss": 0.02692259829491377,
      "grad_norm": 0.09941437095403671,
      "learning_rate": 4.014955211623875e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019491205923259259,
      "selected_pairs": 27151.7,
      "state_query_norm": 42.59337310791015,
      "state_relation_loss": 0.00013187091171857902,
      "step": 17550,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0019491205923259259,
      "weighted_relation_loss": 0.0019491205923259259
    },
    {
      "action_query_norm": 40.671756744384766,
      "action_relation_loss": 0.0017392105306498705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.426566884939195,
      "flow_matching_loss": 0.02451225072145462,
      "grad_norm": 0.10674557834863663,
      "learning_rate": 4.0095522625094895e-05,
      "loss": 0.0264,
      "relation_loss": 0.0018752695061266423,
      "selected_pairs": 26967.4,
      "state_query_norm": 42.596804809570315,
      "state_relation_loss": 0.00013605896820081397,
      "step": 17560,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0018752695061266423,
      "weighted_relation_loss": 0.0018752695061266423
    },
    {
      "action_query_norm": 40.64858779907227,
      "action_relation_loss": 0.001868196704890579,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.435921421889617,
      "flow_matching_loss": 0.026577101089060307,
      "grad_norm": 0.09449659287929535,
      "learning_rate": 4.004150516882497e-05,
      "loss": 0.0286,
      "relation_loss": 0.0020026468206197023,
      "selected_pairs": 26732.0,
      "state_query_norm": 42.58402557373047,
      "state_relation_loss": 0.00013445012664305977,
      "step": 17570,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0020026468206197023,
      "weighted_relation_loss": 0.0020026468206197023
    },
    {
      "action_query_norm": 40.617605972290036,
      "action_relation_loss": 0.0018102944013662636,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.44527595884004,
      "flow_matching_loss": 0.02697613351047039,
      "grad_norm": 0.13521189987659454,
      "learning_rate": 3.99874998130653e-05,
      "loss": 0.0289,
      "relation_loss": 0.0019448757637292146,
      "selected_pairs": 26926.2,
      "state_query_norm": 42.598839569091794,
      "state_relation_loss": 0.0001345813296211418,
      "step": 17580,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0019448757637292146,
      "weighted_relation_loss": 0.0019448757637292146
    },
    {
      "action_query_norm": 40.62738189697266,
      "action_relation_loss": 0.0018529957858845591,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.454630495790457,
      "flow_matching_loss": 0.02674428317695856,
      "grad_norm": 0.22361460328102112,
      "learning_rate": 3.9933506623437455e-05,
      "loss": 0.0287,
      "relation_loss": 0.001989715767558664,
      "selected_pairs": 27055.6,
      "state_query_norm": 42.596557235717775,
      "state_relation_loss": 0.00013671995693584905,
      "step": 17590,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001989715767558664,
      "weighted_relation_loss": 0.001989715767558664
    },
    {
      "action_query_norm": 40.63205490112305,
      "action_relation_loss": 0.0018205477274022997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.46398503274088,
      "flow_matching_loss": 0.025239368714392186,
      "grad_norm": 0.09900077432394028,
      "learning_rate": 3.9879525665548276e-05,
      "loss": 0.0272,
      "relation_loss": 0.001945959811564535,
      "selected_pairs": 26026.3,
      "state_query_norm": 42.59253768920898,
      "state_relation_loss": 0.00012541206524474546,
      "step": 17600,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.001945959811564535,
      "weighted_relation_loss": 0.001945959811564535
    },
    {
      "action_query_norm": 40.64144706726074,
      "action_relation_loss": 0.0017762165749445557,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.4733395696913,
      "flow_matching_loss": 0.025911687314510344,
      "grad_norm": 0.1147601529955864,
      "learning_rate": 3.9825557004989704e-05,
      "loss": 0.0278,
      "relation_loss": 0.0019129416323266923,
      "selected_pairs": 27162.2,
      "state_query_norm": 42.59743232727051,
      "state_relation_loss": 0.0001367250435578171,
      "step": 17610,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0019129416323266923,
      "weighted_relation_loss": 0.0019129416323266923
    },
    {
      "action_query_norm": 40.692803955078126,
      "action_relation_loss": 0.0017392940819263459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.48269410664172,
      "flow_matching_loss": 0.02470816895365715,
      "grad_norm": 0.07869954407215118,
      "learning_rate": 3.977160070733877e-05,
      "loss": 0.0266,
      "relation_loss": 0.0018814846058376133,
      "selected_pairs": 26940.3,
      "state_query_norm": 42.60547790527344,
      "state_relation_loss": 0.00014219054210116156,
      "step": 17620,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0018814846058376133,
      "weighted_relation_loss": 0.0018814846058376133
    },
    {
      "action_query_norm": 40.673241806030276,
      "action_relation_loss": 0.0017374519142322243,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.492048643592142,
      "flow_matching_loss": 0.027811692468822,
      "grad_norm": 0.1120164766907692,
      "learning_rate": 3.971765683815746e-05,
      "loss": 0.0297,
      "relation_loss": 0.0018747000955045222,
      "selected_pairs": 27475.4,
      "state_query_norm": 42.60533256530762,
      "state_relation_loss": 0.00013724820237257518,
      "step": 17630,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0018747000955045222,
      "weighted_relation_loss": 0.0018747000955045222
    },
    {
      "action_query_norm": 40.5867115020752,
      "action_relation_loss": 0.0019116948707960546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.501403180542564,
      "flow_matching_loss": 0.026088960655033588,
      "grad_norm": 0.12165287137031555,
      "learning_rate": 3.966372546299265e-05,
      "loss": 0.0281,
      "relation_loss": 0.0020407467731274665,
      "selected_pairs": 26815.7,
      "state_query_norm": 42.599076461791995,
      "state_relation_loss": 0.00012905189796583728,
      "step": 17640,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0020407467731274665,
      "weighted_relation_loss": 0.0020407467731274665
    },
    {
      "action_query_norm": 40.58545188903808,
      "action_relation_loss": 0.001766798400785774,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.510757717492986,
      "flow_matching_loss": 0.025444324314594268,
      "grad_norm": 0.12962889671325684,
      "learning_rate": 3.960980664737604e-05,
      "loss": 0.0273,
      "relation_loss": 0.0018994502257555722,
      "selected_pairs": 26658.6,
      "state_query_norm": 42.59017829895019,
      "state_relation_loss": 0.00013265182496979833,
      "step": 17650,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0018994502257555722,
      "weighted_relation_loss": 0.0018994502257555722
    },
    {
      "action_query_norm": 40.6553035736084,
      "action_relation_loss": 0.001749727176502347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.520112254443404,
      "flow_matching_loss": 0.02685244306921959,
      "grad_norm": 0.14040401577949524,
      "learning_rate": 3.955590045682408e-05,
      "loss": 0.0287,
      "relation_loss": 0.0018798828357830643,
      "selected_pairs": 26832.6,
      "state_query_norm": 42.59449996948242,
      "state_relation_loss": 0.00013015567456022836,
      "step": 17660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0018798828357830643,
      "weighted_relation_loss": 0.0018798828357830643
    },
    {
      "action_query_norm": 40.615499877929686,
      "action_relation_loss": 0.0018484602565877139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.529466791393826,
      "flow_matching_loss": 0.025282379984855653,
      "grad_norm": 0.09067708998918533,
      "learning_rate": 3.950200695683788e-05,
      "loss": 0.0273,
      "relation_loss": 0.001979864644818008,
      "selected_pairs": 27194.1,
      "state_query_norm": 42.5643367767334,
      "state_relation_loss": 0.00013140439623384736,
      "step": 17670,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001979864644818008,
      "weighted_relation_loss": 0.001979864644818008
    },
    {
      "action_query_norm": 40.59937858581543,
      "action_relation_loss": 0.0016916391672566534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.53882132834425,
      "flow_matching_loss": 0.025375019013881683,
      "grad_norm": 0.10892056673765182,
      "learning_rate": 3.944812621290314e-05,
      "loss": 0.0272,
      "relation_loss": 0.0018217244185507297,
      "selected_pairs": 27108.0,
      "state_query_norm": 42.57243690490723,
      "state_relation_loss": 0.00013008525420445948,
      "step": 17680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0018217244185507297,
      "weighted_relation_loss": 0.0018217244185507297
    },
    {
      "action_query_norm": 40.61254043579102,
      "action_relation_loss": 0.0018055640859529376,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.548175865294667,
      "flow_matching_loss": 0.02613874115049839,
      "grad_norm": 0.11954675614833832,
      "learning_rate": 3.939425829049002e-05,
      "loss": 0.0281,
      "relation_loss": 0.0019288414740003645,
      "selected_pairs": 27337.5,
      "state_query_norm": 42.574167251586914,
      "state_relation_loss": 0.00012327740259934216,
      "step": 17690,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0019288414740003645,
      "weighted_relation_loss": 0.0019288414740003645
    },
    {
      "action_query_norm": 40.61252288818359,
      "action_relation_loss": 0.0018213173025287688,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.55753040224509,
      "flow_matching_loss": 0.02651602290570736,
      "grad_norm": 0.10617955029010773,
      "learning_rate": 3.9340403255053135e-05,
      "loss": 0.0285,
      "relation_loss": 0.0019490658771246672,
      "selected_pairs": 26460.6,
      "state_query_norm": 42.59159469604492,
      "state_relation_loss": 0.00012774856077157892,
      "step": 17700,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0019490658771246672,
      "weighted_relation_loss": 0.0019490658771246672
    },
    {
      "action_query_norm": 40.62786521911621,
      "action_relation_loss": 0.0018199323676526546,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.56688493919551,
      "flow_matching_loss": 0.024656709656119345,
      "grad_norm": 0.1036410927772522,
      "learning_rate": 3.9286561172031407e-05,
      "loss": 0.0266,
      "relation_loss": 0.0019533700891770423,
      "selected_pairs": 26971.5,
      "state_query_norm": 42.56758499145508,
      "state_relation_loss": 0.00013343772006919609,
      "step": 17710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0019533700891770423,
      "weighted_relation_loss": 0.0019533700891770423
    },
    {
      "action_query_norm": 40.55849609375,
      "action_relation_loss": 0.0017271886463277042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.57623947614593,
      "flow_matching_loss": 0.02482542544603348,
      "grad_norm": 0.10336490720510483,
      "learning_rate": 3.923273210684809e-05,
      "loss": 0.0267,
      "relation_loss": 0.001858498645015061,
      "selected_pairs": 26762.8,
      "state_query_norm": 42.572953414916995,
      "state_relation_loss": 0.0001313099950493779,
      "step": 17720,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001858498645015061,
      "weighted_relation_loss": 0.001858498645015061
    },
    {
      "action_query_norm": 40.569880676269534,
      "action_relation_loss": 0.0018164845183491706,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.58559401309635,
      "flow_matching_loss": 0.024873358942568303,
      "grad_norm": 0.1281186044216156,
      "learning_rate": 3.9178916124910555e-05,
      "loss": 0.0268,
      "relation_loss": 0.0019436718546785414,
      "selected_pairs": 26547.8,
      "state_query_norm": 42.5846248626709,
      "state_relation_loss": 0.00012718731522909366,
      "step": 17730,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0019436718546785414,
      "weighted_relation_loss": 0.0019436718546785414
    },
    {
      "action_query_norm": 40.625687789916995,
      "action_relation_loss": 0.0017322175786830484,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.594948550046773,
      "flow_matching_loss": 0.025005542486906052,
      "grad_norm": 0.13362422585487366,
      "learning_rate": 3.9125113291610274e-05,
      "loss": 0.0269,
      "relation_loss": 0.0018651409773156047,
      "selected_pairs": 27244.7,
      "state_query_norm": 42.57665481567383,
      "state_relation_loss": 0.00013292340881889685,
      "step": 17740,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0018651409773156047,
      "weighted_relation_loss": 0.0018651409773156047
    },
    {
      "action_query_norm": 40.58743324279785,
      "action_relation_loss": 0.0017918065190315247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.604303086997195,
      "flow_matching_loss": 0.026229899562895297,
      "grad_norm": 0.14658017456531525,
      "learning_rate": 3.907132367232278e-05,
      "loss": 0.0281,
      "relation_loss": 0.0019171888125129044,
      "selected_pairs": 26736.8,
      "state_query_norm": 42.59070167541504,
      "state_relation_loss": 0.00012538228620542213,
      "step": 17750,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0019171888125129044,
      "weighted_relation_loss": 0.0019171888125129044
    },
    {
      "action_query_norm": 40.56834716796875,
      "action_relation_loss": 0.0018571226275525988,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.613657623947613,
      "flow_matching_loss": 0.025010989233851432,
      "grad_norm": 0.12571896612644196,
      "learning_rate": 3.901754733240752e-05,
      "loss": 0.027,
      "relation_loss": 0.0019948158063925804,
      "selected_pairs": 26618.2,
      "state_query_norm": 42.556850814819335,
      "state_relation_loss": 0.0001376931802951731,
      "step": 17760,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0019948158063925804,
      "weighted_relation_loss": 0.0019948158063925804
    },
    {
      "action_query_norm": 40.58712501525879,
      "action_relation_loss": 0.001819784042891115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.623012160898035,
      "flow_matching_loss": 0.024140207469463347,
      "grad_norm": 0.10411649942398071,
      "learning_rate": 3.896378433720785e-05,
      "loss": 0.0261,
      "relation_loss": 0.0019472162588499487,
      "selected_pairs": 26640.1,
      "state_query_norm": 42.55934219360351,
      "state_relation_loss": 0.0001274322195968125,
      "step": 17770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0019472162588499487,
      "weighted_relation_loss": 0.0019472162588499487
    },
    {
      "action_query_norm": 40.53862075805664,
      "action_relation_loss": 0.0016972818993963301,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.632366697848457,
      "flow_matching_loss": 0.02527778297662735,
      "grad_norm": 0.10942737758159637,
      "learning_rate": 3.891003475205086e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018185360124334693,
      "selected_pairs": 27084.0,
      "state_query_norm": 42.572014236450194,
      "state_relation_loss": 0.00012125409339205361,
      "step": 17780,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0018185360124334693,
      "weighted_relation_loss": 0.0018185360124334693
    },
    {
      "action_query_norm": 40.56444549560547,
      "action_relation_loss": 0.0017551713506691157,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.641721234798876,
      "flow_matching_loss": 0.02662219200283289,
      "grad_norm": 0.12642334401607513,
      "learning_rate": 3.885629864224736e-05,
      "loss": 0.0285,
      "relation_loss": 0.001880991505458951,
      "selected_pairs": 26903.4,
      "state_query_norm": 42.57239608764648,
      "state_relation_loss": 0.000125820124958409,
      "step": 17790,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001880991505458951,
      "weighted_relation_loss": 0.001880991505458951
    },
    {
      "action_query_norm": 40.615687942504884,
      "action_relation_loss": 0.001837798859924078,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.651075771749298,
      "flow_matching_loss": 0.025558273866772653,
      "grad_norm": 0.11638392508029938,
      "learning_rate": 3.880257607309178e-05,
      "loss": 0.0275,
      "relation_loss": 0.0019623399479314686,
      "selected_pairs": 26876.4,
      "state_query_norm": 42.56235466003418,
      "state_relation_loss": 0.00012454110183170996,
      "step": 17800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0019623399479314686,
      "weighted_relation_loss": 0.0019623399479314686
    },
    {
      "action_query_norm": 40.59383277893066,
      "action_relation_loss": 0.001770999061409384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.66043030869972,
      "flow_matching_loss": 0.02630599867552519,
      "grad_norm": 0.11358799040317535,
      "learning_rate": 3.874886710986213e-05,
      "loss": 0.0282,
      "relation_loss": 0.001898539869580418,
      "selected_pairs": 27157.5,
      "state_query_norm": 42.5736759185791,
      "state_relation_loss": 0.00012754081108141692,
      "step": 17810,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001898539869580418,
      "weighted_relation_loss": 0.001898539869580418
    },
    {
      "action_query_norm": 40.58760871887207,
      "action_relation_loss": 0.001738909282721579,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.669784845650142,
      "flow_matching_loss": 0.025516052916646003,
      "grad_norm": 0.09530673176050186,
      "learning_rate": 3.8695171817819834e-05,
      "loss": 0.0274,
      "relation_loss": 0.0018699781503528357,
      "selected_pairs": 26962.6,
      "state_query_norm": 42.542175674438475,
      "state_relation_loss": 0.00013106886617606506,
      "step": 17820,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0018699781503528357,
      "weighted_relation_loss": 0.0018699781503528357
    },
    {
      "action_query_norm": 40.559745407104494,
      "action_relation_loss": 0.0018314286833629012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.67913938260056,
      "flow_matching_loss": 0.025843478552997113,
      "grad_norm": 0.11305912584066391,
      "learning_rate": 3.8641490262209774e-05,
      "loss": 0.0278,
      "relation_loss": 0.001954438490793109,
      "selected_pairs": 26708.2,
      "state_query_norm": 42.56268081665039,
      "state_relation_loss": 0.00012300979578867556,
      "step": 17830,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001954438490793109,
      "weighted_relation_loss": 0.001954438490793109
    },
    {
      "action_query_norm": 40.56613006591797,
      "action_relation_loss": 0.0017002585576847196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.688493919550982,
      "flow_matching_loss": 0.024203314445912838,
      "grad_norm": 0.10151951014995575,
      "learning_rate": 3.8587822508260084e-05,
      "loss": 0.026,
      "relation_loss": 0.0018323188181966544,
      "selected_pairs": 26896.2,
      "state_query_norm": 42.56162986755371,
      "state_relation_loss": 0.00013206026633270085,
      "step": 17840,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0018323188181966544,
      "weighted_relation_loss": 0.0018323188181966544
    },
    {
      "action_query_norm": 40.56663360595703,
      "action_relation_loss": 0.0017526327748782933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.697848456501404,
      "flow_matching_loss": 0.024729752726852895,
      "grad_norm": 0.07454615086317062,
      "learning_rate": 3.853416862118214e-05,
      "loss": 0.0266,
      "relation_loss": 0.0018815597402863204,
      "selected_pairs": 26569.2,
      "state_query_norm": 42.55274887084961,
      "state_relation_loss": 0.000128926953766495,
      "step": 17850,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0018815597402863204,
      "weighted_relation_loss": 0.0018815597402863204
    },
    {
      "action_query_norm": 40.557661056518555,
      "action_relation_loss": 0.0017338998848572373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.707202993451823,
      "flow_matching_loss": 0.025514931976795198,
      "grad_norm": 0.13129135966300964,
      "learning_rate": 3.848052866617049e-05,
      "loss": 0.0274,
      "relation_loss": 0.0018606386031024158,
      "selected_pairs": 26688.0,
      "state_query_norm": 42.55484580993652,
      "state_relation_loss": 0.0001267387291591149,
      "step": 17860,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0018606386031024158,
      "weighted_relation_loss": 0.0018606386031024158
    },
    {
      "action_query_norm": 40.58356056213379,
      "action_relation_loss": 0.0017603779560886323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.716557530402245,
      "flow_matching_loss": 0.026445203460752964,
      "grad_norm": 0.12724092602729797,
      "learning_rate": 3.84269027084027e-05,
      "loss": 0.0283,
      "relation_loss": 0.0018864958779886364,
      "selected_pairs": 27124.5,
      "state_query_norm": 42.567068099975586,
      "state_relation_loss": 0.0001261179095308762,
      "step": 17870,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0018864958779886364,
      "weighted_relation_loss": 0.0018864958779886364
    },
    {
      "action_query_norm": 40.54547958374023,
      "action_relation_loss": 0.0018026711186394095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.725912067352667,
      "flow_matching_loss": 0.025053781270980836,
      "grad_norm": 0.0869232788681984,
      "learning_rate": 3.83732908130394e-05,
      "loss": 0.027,
      "relation_loss": 0.0019206549623049796,
      "selected_pairs": 26163.1,
      "state_query_norm": 42.54927024841309,
      "state_relation_loss": 0.00011798384075518697,
      "step": 17880,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0019206549623049796,
      "weighted_relation_loss": 0.0019206549623049796
    },
    {
      "action_query_norm": 40.55177879333496,
      "action_relation_loss": 0.0017275993595831096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.73526660430309,
      "flow_matching_loss": 0.02581271268427372,
      "grad_norm": 0.10865333676338196,
      "learning_rate": 3.831969304522409e-05,
      "loss": 0.0277,
      "relation_loss": 0.0018545539700426162,
      "selected_pairs": 27135.7,
      "state_query_norm": 42.556168746948245,
      "state_relation_loss": 0.00012695461482508107,
      "step": 17890,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0018545539700426162,
      "weighted_relation_loss": 0.0018545539700426162
    },
    {
      "action_query_norm": 40.58291015625,
      "action_relation_loss": 0.0017618792480789125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.744621141253507,
      "flow_matching_loss": 0.024582883715629576,
      "grad_norm": 0.11304422467947006,
      "learning_rate": 3.826610947008312e-05,
      "loss": 0.0265,
      "relation_loss": 0.001888881262857467,
      "selected_pairs": 26800.8,
      "state_query_norm": 42.54016876220703,
      "state_relation_loss": 0.00012700200895778834,
      "step": 17900,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001888881262857467,
      "weighted_relation_loss": 0.001888881262857467
    },
    {
      "action_query_norm": 40.506249237060544,
      "action_relation_loss": 0.0017591245123185217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.75397567820393,
      "flow_matching_loss": 0.024468003213405608,
      "grad_norm": 0.08516764640808105,
      "learning_rate": 3.821254015272559e-05,
      "loss": 0.0263,
      "relation_loss": 0.0018810189329087734,
      "selected_pairs": 27391.1,
      "state_query_norm": 42.544090270996094,
      "state_relation_loss": 0.00012189441040391102,
      "step": 17910,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0018810189329087734,
      "weighted_relation_loss": 0.0018810189329087734
    },
    {
      "action_query_norm": 40.56230583190918,
      "action_relation_loss": 0.0017224146984517574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.76333021515435,
      "flow_matching_loss": 0.02314553651958704,
      "grad_norm": 0.08968993276357651,
      "learning_rate": 3.8158985158243214e-05,
      "loss": 0.025,
      "relation_loss": 0.0018442228436470031,
      "selected_pairs": 27115.2,
      "state_query_norm": 42.54474220275879,
      "state_relation_loss": 0.00012180815101601184,
      "step": 17920,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0018442228436470031,
      "weighted_relation_loss": 0.0018442228436470031
    },
    {
      "action_query_norm": 40.532881546020505,
      "action_relation_loss": 0.0018428961047902703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.77268475210477,
      "flow_matching_loss": 0.025664561614394187,
      "grad_norm": 0.15458281338214874,
      "learning_rate": 3.810544455171044e-05,
      "loss": 0.0276,
      "relation_loss": 0.001966932392679155,
      "selected_pairs": 26901.2,
      "state_query_norm": 42.54485893249512,
      "state_relation_loss": 0.00012403629589243791,
      "step": 17930,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001966932392679155,
      "weighted_relation_loss": 0.001966932392679155
    },
    {
      "action_query_norm": 40.537329483032224,
      "action_relation_loss": 0.0017535959021188318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.78203928905519,
      "flow_matching_loss": 0.02588652465492487,
      "grad_norm": 0.11237142235040665,
      "learning_rate": 3.8051918398184114e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018878414179198445,
      "selected_pairs": 26666.4,
      "state_query_norm": 42.532758712768555,
      "state_relation_loss": 0.00013424551289062948,
      "step": 17940,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0018878414179198445,
      "weighted_relation_loss": 0.0018878414179198445
    },
    {
      "action_query_norm": 40.50659942626953,
      "action_relation_loss": 0.0017665299703367055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.791393826005613,
      "flow_matching_loss": 0.025182736106216908,
      "grad_norm": 0.11079178750514984,
      "learning_rate": 3.799840676270356e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018892427557148038,
      "selected_pairs": 26915.6,
      "state_query_norm": 42.54638862609863,
      "state_relation_loss": 0.0001227127831953112,
      "step": 17950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0018892427557148038,
      "weighted_relation_loss": 0.0018892427557148038
    },
    {
      "action_query_norm": 40.52074508666992,
      "action_relation_loss": 0.0017124072066508234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.800748362956035,
      "flow_matching_loss": 0.024996674619615078,
      "grad_norm": 0.12050929665565491,
      "learning_rate": 3.794490971029048e-05,
      "loss": 0.0268,
      "relation_loss": 0.0018410701886750758,
      "selected_pairs": 26718.6,
      "state_query_norm": 42.536929321289065,
      "state_relation_loss": 0.00012866299657616765,
      "step": 17960,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0018410701886750758,
      "weighted_relation_loss": 0.0018410701886750758
    },
    {
      "action_query_norm": 40.54872131347656,
      "action_relation_loss": 0.0017495753476396204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.810102899906454,
      "flow_matching_loss": 0.023205647058784962,
      "grad_norm": 0.08883368968963623,
      "learning_rate": 3.789142730594881e-05,
      "loss": 0.0251,
      "relation_loss": 0.0018748013069853186,
      "selected_pairs": 27102.6,
      "state_query_norm": 42.53559532165527,
      "state_relation_loss": 0.00012522596734925172,
      "step": 17970,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0018748013069853186,
      "weighted_relation_loss": 0.0018748013069853186
    },
    {
      "action_query_norm": 40.54382591247558,
      "action_relation_loss": 0.0017483991920016705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.819457436856876,
      "flow_matching_loss": 0.026226937770843506,
      "grad_norm": 0.12827304005622864,
      "learning_rate": 3.783795961466471e-05,
      "loss": 0.0281,
      "relation_loss": 0.0018747918307781219,
      "selected_pairs": 27081.8,
      "state_query_norm": 42.542626953125,
      "state_relation_loss": 0.00012639263804885558,
      "step": 17980,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0018747918307781219,
      "weighted_relation_loss": 0.0018747918307781219
    },
    {
      "action_query_norm": 40.53337287902832,
      "action_relation_loss": 0.0017421663738787173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.828811973807298,
      "flow_matching_loss": 0.02517158053815365,
      "grad_norm": 0.13279150426387787,
      "learning_rate": 3.778450670140651e-05,
      "loss": 0.027,
      "relation_loss": 0.0018683158443309367,
      "selected_pairs": 27055.1,
      "state_query_norm": 42.5441276550293,
      "state_relation_loss": 0.0001261494544451125,
      "step": 17990,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0018683158443309367,
      "weighted_relation_loss": 0.0018683158443309367
    },
    {
      "action_query_norm": 40.54199104309082,
      "action_relation_loss": 0.0016943220281973482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.838166510757716,
      "flow_matching_loss": 0.02356722895056009,
      "grad_norm": 0.11048024892807007,
      "learning_rate": 3.773106863112451e-05,
      "loss": 0.0254,
      "relation_loss": 0.0018167173955589532,
      "selected_pairs": 26365.2,
      "state_query_norm": 42.54266166687012,
      "state_relation_loss": 0.00012239537027198822,
      "step": 18000,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0018167173955589532,
      "weighted_relation_loss": 0.0018167173955589532
    },
    {
      "action_query_norm": 40.51105003356933,
      "action_relation_loss": 0.001789759227540344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.84752104770814,
      "flow_matching_loss": 0.027155145071446896,
      "grad_norm": 0.12840457260608673,
      "learning_rate": 3.7677645468751e-05,
      "loss": 0.0291,
      "relation_loss": 0.0019213020452298224,
      "selected_pairs": 27034.3,
      "state_query_norm": 42.515198135375975,
      "state_relation_loss": 0.00013154280895832926,
      "step": 18010,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0019213020452298224,
      "weighted_relation_loss": 0.0019213020452298224
    },
    {
      "action_query_norm": 40.51807823181152,
      "action_relation_loss": 0.001683620666153729,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.85687558465856,
      "flow_matching_loss": 0.027039753273129463,
      "grad_norm": 0.10932253301143646,
      "learning_rate": 3.762423727920018e-05,
      "loss": 0.0288,
      "relation_loss": 0.0018099514069035648,
      "selected_pairs": 26987.7,
      "state_query_norm": 42.518368530273435,
      "state_relation_loss": 0.00012633074875338933,
      "step": 18020,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0018099514069035648,
      "weighted_relation_loss": 0.0018099514069035648
    },
    {
      "action_query_norm": 40.52861976623535,
      "action_relation_loss": 0.0017824303009547292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.86623012160898,
      "flow_matching_loss": 0.024356781132519247,
      "grad_norm": 0.10704363137483597,
      "learning_rate": 3.7570844127367995e-05,
      "loss": 0.0263,
      "relation_loss": 0.0019048533169552684,
      "selected_pairs": 26320.8,
      "state_query_norm": 42.53125648498535,
      "state_relation_loss": 0.0001224229999934323,
      "step": 18030,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0019048533169552684,
      "weighted_relation_loss": 0.0019048533169552684
    },
    {
      "action_query_norm": 40.54550361633301,
      "action_relation_loss": 0.0017940360936336218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.8755846585594,
      "flow_matching_loss": 0.026439938135445117,
      "grad_norm": 0.11264976114034653,
      "learning_rate": 3.751746607813221e-05,
      "loss": 0.0284,
      "relation_loss": 0.0019143006415106357,
      "selected_pairs": 26644.7,
      "state_query_norm": 42.54155235290527,
      "state_relation_loss": 0.0001202645456942264,
      "step": 18040,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0019143006415106357,
      "weighted_relation_loss": 0.0019143006415106357
    },
    {
      "action_query_norm": 40.46122779846191,
      "action_relation_loss": 0.0017765124212019145,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.884939195509823,
      "flow_matching_loss": 0.024458420649170875,
      "grad_norm": 0.10199755430221558,
      "learning_rate": 3.746410319635217e-05,
      "loss": 0.0264,
      "relation_loss": 0.001896750903688371,
      "selected_pairs": 26136.2,
      "state_query_norm": 42.52267646789551,
      "state_relation_loss": 0.00012023848175886087,
      "step": 18050,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.001896750903688371,
      "weighted_relation_loss": 0.001896750903688371
    },
    {
      "action_query_norm": 40.48844871520996,
      "action_relation_loss": 0.001703765115235001,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.894293732460245,
      "flow_matching_loss": 0.024252705089747906,
      "grad_norm": 0.12339642643928528,
      "learning_rate": 3.7410755546868806e-05,
      "loss": 0.0261,
      "relation_loss": 0.0018250786000862718,
      "selected_pairs": 26662.5,
      "state_query_norm": 42.52112045288086,
      "state_relation_loss": 0.00012131349358242005,
      "step": 18060,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0018250786000862718,
      "weighted_relation_loss": 0.0018250786000862718
    },
    {
      "action_query_norm": 40.49062919616699,
      "action_relation_loss": 0.0016517684678547084,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.903648269410663,
      "flow_matching_loss": 0.02497087512165308,
      "grad_norm": 0.10360877960920334,
      "learning_rate": 3.735742319450454e-05,
      "loss": 0.0267,
      "relation_loss": 0.0017676913179457189,
      "selected_pairs": 26920.2,
      "state_query_norm": 42.5315601348877,
      "state_relation_loss": 0.00011592284572543576,
      "step": 18070,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0017676913179457189,
      "weighted_relation_loss": 0.0017676913179457189
    },
    {
      "action_query_norm": 40.50242233276367,
      "action_relation_loss": 0.0016523112426511943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.913002806361085,
      "flow_matching_loss": 0.026133438944816588,
      "grad_norm": 0.12110842019319534,
      "learning_rate": 3.7304106204063184e-05,
      "loss": 0.0279,
      "relation_loss": 0.0017722368822433055,
      "selected_pairs": 26907.8,
      "state_query_norm": 42.529232025146484,
      "state_relation_loss": 0.00011992564832326024,
      "step": 18080,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0017722368822433055,
      "weighted_relation_loss": 0.0017722368822433055
    },
    {
      "action_query_norm": 40.50707359313965,
      "action_relation_loss": 0.0018333986052311956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.922357343311507,
      "flow_matching_loss": 0.025860339775681494,
      "grad_norm": 0.13882769644260406,
      "learning_rate": 3.725080464032996e-05,
      "loss": 0.0278,
      "relation_loss": 0.0019603006425313652,
      "selected_pairs": 27315.1,
      "state_query_norm": 42.52743644714356,
      "state_relation_loss": 0.0001269020554900635,
      "step": 18090,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0019603006425313652,
      "weighted_relation_loss": 0.0019603006425313652
    },
    {
      "action_query_norm": 40.48587417602539,
      "action_relation_loss": 0.0017635681433603168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.931711880261926,
      "flow_matching_loss": 0.023974728398025037,
      "grad_norm": 0.10415706038475037,
      "learning_rate": 3.719751856807125e-05,
      "loss": 0.0259,
      "relation_loss": 0.001893733791075647,
      "selected_pairs": 26604.6,
      "state_query_norm": 42.50837783813476,
      "state_relation_loss": 0.00013016566008445808,
      "step": 18100,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001893733791075647,
      "weighted_relation_loss": 0.001893733791075647
    },
    {
      "action_query_norm": 40.50198516845703,
      "action_relation_loss": 0.0016523241298273205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.941066417212348,
      "flow_matching_loss": 0.025380656123161316,
      "grad_norm": 0.13628093898296356,
      "learning_rate": 3.71442480520347e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017767579294741154,
      "selected_pairs": 26166.6,
      "state_query_norm": 42.51624641418457,
      "state_relation_loss": 0.00012443377927411347,
      "step": 18110,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0017767579294741154,
      "weighted_relation_loss": 0.0017767579294741154
    },
    {
      "action_query_norm": 40.497310638427734,
      "action_relation_loss": 0.0017225050134584308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.95042095416277,
      "flow_matching_loss": 0.02575152590870857,
      "grad_norm": 0.09756455570459366,
      "learning_rate": 3.709099315694897e-05,
      "loss": 0.0276,
      "relation_loss": 0.0018506721826270223,
      "selected_pairs": 26676.7,
      "state_query_norm": 42.51352043151856,
      "state_relation_loss": 0.00012816718008252793,
      "step": 18120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0018506721826270223,
      "weighted_relation_loss": 0.0018506721826270223
    },
    {
      "action_query_norm": 40.52469444274902,
      "action_relation_loss": 0.001657581701874733,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.95977549111319,
      "flow_matching_loss": 0.02618345320224762,
      "grad_norm": 0.1116032674908638,
      "learning_rate": 3.703775394752378e-05,
      "loss": 0.028,
      "relation_loss": 0.0017818360356613994,
      "selected_pairs": 26386.2,
      "state_query_norm": 42.5283920288086,
      "state_relation_loss": 0.0001242543185071554,
      "step": 18130,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0017818360356613994,
      "weighted_relation_loss": 0.0017818360356613994
    },
    {
      "action_query_norm": 40.503651809692386,
      "action_relation_loss": 0.001716601091902703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.96913002806361,
      "flow_matching_loss": 0.023826068639755248,
      "grad_norm": 0.08788808435201645,
      "learning_rate": 3.698453048844983e-05,
      "loss": 0.0257,
      "relation_loss": 0.0018388916389085353,
      "selected_pairs": 26548.9,
      "state_query_norm": 42.506378173828125,
      "state_relation_loss": 0.00012229053900227881,
      "step": 18140,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0018388916389085353,
      "weighted_relation_loss": 0.0018388916389085353
    },
    {
      "action_query_norm": 40.53319664001465,
      "action_relation_loss": 0.0017282768385484815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.978484565014032,
      "flow_matching_loss": 0.025619247183203697,
      "grad_norm": 0.11284439265727997,
      "learning_rate": 3.693132284439861e-05,
      "loss": 0.0275,
      "relation_loss": 0.0018517246120609344,
      "selected_pairs": 26621.1,
      "state_query_norm": 42.521562576293945,
      "state_relation_loss": 0.00012344776478130371,
      "step": 18150,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0018517246120609344,
      "weighted_relation_loss": 0.0018517246120609344
    },
    {
      "action_query_norm": 40.525870132446286,
      "action_relation_loss": 0.0017576929065398871,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.987839101964454,
      "flow_matching_loss": 0.025222779624164103,
      "grad_norm": 0.1423344910144806,
      "learning_rate": 3.6878131080022414e-05,
      "loss": 0.0271,
      "relation_loss": 0.0018863720586523413,
      "selected_pairs": 27038.6,
      "state_query_norm": 42.494117736816406,
      "state_relation_loss": 0.0001286791513848584,
      "step": 18160,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0018863720586523413,
      "weighted_relation_loss": 0.0018863720586523413
    },
    {
      "action_query_norm": 40.48413925170898,
      "action_relation_loss": 0.0016581577947363257,
      "auxiliary_loss_weight": 1.0,
      "epoch": 16.997193638914872,
      "flow_matching_loss": 0.02554538380354643,
      "grad_norm": 0.13976174592971802,
      "learning_rate": 3.682495525995429e-05,
      "loss": 0.0273,
      "relation_loss": 0.00178134988527745,
      "selected_pairs": 27161.7,
      "state_query_norm": 42.51411247253418,
      "state_relation_loss": 0.00012319208690314554,
      "step": 18170,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.00178134988527745,
      "weighted_relation_loss": 0.00178134988527745
    },
    {
      "action_query_norm": 40.486441421508786,
      "action_relation_loss": 0.0017705217236652971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.006548175865294,
      "flow_matching_loss": 0.026608773320913316,
      "grad_norm": 0.1412728726863861,
      "learning_rate": 3.6771795448807846e-05,
      "loss": 0.0285,
      "relation_loss": 0.0018993415171280504,
      "selected_pairs": 24369.6,
      "state_query_norm": 42.52509803771973,
      "state_relation_loss": 0.00012881981674581767,
      "step": 18180,
      "valid_action_samples": 214.5,
      "weighted_auxiliary_loss": 0.0018993415171280504,
      "weighted_relation_loss": 0.0018993415171280504
    },
    {
      "action_query_norm": 40.529920959472655,
      "action_relation_loss": 0.0016602335614152252,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.015902712815716,
      "flow_matching_loss": 0.022867765463888644,
      "grad_norm": 0.0934407189488411,
      "learning_rate": 3.671865171117724e-05,
      "loss": 0.0247,
      "relation_loss": 0.001783473661635071,
      "selected_pairs": 27020.0,
      "state_query_norm": 42.510163116455075,
      "state_relation_loss": 0.00012324008712312208,
      "step": 18190,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001783473661635071,
      "weighted_relation_loss": 0.001783473661635071
    },
    {
      "action_query_norm": 40.48175163269043,
      "action_relation_loss": 0.0016935419873334468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.02525724976614,
      "flow_matching_loss": 0.02510246876627207,
      "grad_norm": 0.09758669883012772,
      "learning_rate": 3.666552411163718e-05,
      "loss": 0.0269,
      "relation_loss": 0.0018143972847610712,
      "selected_pairs": 26274.7,
      "state_query_norm": 42.50874061584473,
      "state_relation_loss": 0.00012085529961041175,
      "step": 18200,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0018143972847610712,
      "weighted_relation_loss": 0.0018143972847610712
    },
    {
      "action_query_norm": 40.567924118041994,
      "action_relation_loss": 0.001620455796364695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.034611786716557,
      "flow_matching_loss": 0.024962171353399752,
      "grad_norm": 0.11650774627923965,
      "learning_rate": 3.661241271474269e-05,
      "loss": 0.0267,
      "relation_loss": 0.0017393025103956462,
      "selected_pairs": 27469.4,
      "state_query_norm": 42.53266143798828,
      "state_relation_loss": 0.00011884670238941908,
      "step": 18210,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0017393025103956462,
      "weighted_relation_loss": 0.0017393025103956462
    },
    {
      "action_query_norm": 40.484642791748044,
      "action_relation_loss": 0.0016501731472089887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.04396632366698,
      "flow_matching_loss": 0.023137154616415502,
      "grad_norm": 0.12742242217063904,
      "learning_rate": 3.6559317585029116e-05,
      "loss": 0.0249,
      "relation_loss": 0.0017709581996314227,
      "selected_pairs": 27208.8,
      "state_query_norm": 42.50942840576172,
      "state_relation_loss": 0.00012078503059456124,
      "step": 18220,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0017709581996314227,
      "weighted_relation_loss": 0.0017709581996314227
    },
    {
      "action_query_norm": 40.481672286987305,
      "action_relation_loss": 0.001695373747497797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.0533208606174,
      "flow_matching_loss": 0.023809762299060823,
      "grad_norm": 0.09543650597333908,
      "learning_rate": 3.650623878701204e-05,
      "loss": 0.0256,
      "relation_loss": 0.0018172165961004793,
      "selected_pairs": 27229.6,
      "state_query_norm": 42.49374656677246,
      "state_relation_loss": 0.00012184282240923494,
      "step": 18230,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0018172165961004793,
      "weighted_relation_loss": 0.0018172165961004793
    },
    {
      "action_query_norm": 40.52188949584961,
      "action_relation_loss": 0.0015853977529332042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.06267539756782,
      "flow_matching_loss": 0.024578678235411645,
      "grad_norm": 0.10122368484735489,
      "learning_rate": 3.645317638518721e-05,
      "loss": 0.0263,
      "relation_loss": 0.0017134624067693949,
      "selected_pairs": 26947.0,
      "state_query_norm": 42.50536613464355,
      "state_relation_loss": 0.00012806466838810593,
      "step": 18240,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017134624067693949,
      "weighted_relation_loss": 0.0017134624067693949
    },
    {
      "action_query_norm": 40.48731689453125,
      "action_relation_loss": 0.0016525265527889132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.07202993451824,
      "flow_matching_loss": 0.024784772656857967,
      "grad_norm": 0.11922954022884369,
      "learning_rate": 3.640013044403046e-05,
      "loss": 0.0266,
      "relation_loss": 0.001770479104015976,
      "selected_pairs": 26506.6,
      "state_query_norm": 42.494528198242186,
      "state_relation_loss": 0.00011795255486504174,
      "step": 18250,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.001770479104015976,
      "weighted_relation_loss": 0.001770479104015976
    },
    {
      "action_query_norm": 40.51647872924805,
      "action_relation_loss": 0.0015798600274138153,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.081384471468663,
      "flow_matching_loss": 0.02505367323756218,
      "grad_norm": 0.10319089889526367,
      "learning_rate": 3.634710102799761e-05,
      "loss": 0.0268,
      "relation_loss": 0.001703980949241668,
      "selected_pairs": 26735.2,
      "state_query_norm": 42.5004280090332,
      "state_relation_loss": 0.0001241209276486188,
      "step": 18260,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001703980949241668,
      "weighted_relation_loss": 0.001703980949241668
    },
    {
      "action_query_norm": 40.50779495239258,
      "action_relation_loss": 0.0015181029215455056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.09073900841908,
      "flow_matching_loss": 0.023865779489278795,
      "grad_norm": 0.11724922060966492,
      "learning_rate": 3.6294088201524394e-05,
      "loss": 0.0255,
      "relation_loss": 0.001638985867612064,
      "selected_pairs": 26966.8,
      "state_query_norm": 42.49766883850098,
      "state_relation_loss": 0.00012088294315617531,
      "step": 18270,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001638985867612064,
      "weighted_relation_loss": 0.001638985867612064
    },
    {
      "action_query_norm": 40.487453842163085,
      "action_relation_loss": 0.001609377155546099,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.100093545369504,
      "flow_matching_loss": 0.025441464595496656,
      "grad_norm": 0.11731721460819244,
      "learning_rate": 3.62410920290264e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017336920485831798,
      "selected_pairs": 26620.9,
      "state_query_norm": 42.50129737854004,
      "state_relation_loss": 0.00012431487848516554,
      "step": 18280,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0017336920485831798,
      "weighted_relation_loss": 0.0017336920485831798
    },
    {
      "action_query_norm": 40.483689880371095,
      "action_relation_loss": 0.00162256034091115,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.109448082319926,
      "flow_matching_loss": 0.024362882226705553,
      "grad_norm": 0.11749549210071564,
      "learning_rate": 3.6188112574898954e-05,
      "loss": 0.0261,
      "relation_loss": 0.0017549775424413384,
      "selected_pairs": 26571.9,
      "state_query_norm": 42.49360389709473,
      "state_relation_loss": 0.00013241719861980527,
      "step": 18290,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0017549775424413384,
      "weighted_relation_loss": 0.0017549775424413384
    },
    {
      "action_query_norm": 40.46379508972168,
      "action_relation_loss": 0.0017508437391370534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.118802619270348,
      "flow_matching_loss": 0.025880923308432103,
      "grad_norm": 0.1111496090888977,
      "learning_rate": 3.6135149903517115e-05,
      "loss": 0.0278,
      "relation_loss": 0.0018823671038262547,
      "selected_pairs": 26864.4,
      "state_query_norm": 42.49766883850098,
      "state_relation_loss": 0.00013152336250641384,
      "step": 18300,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0018823671038262547,
      "weighted_relation_loss": 0.0018823671038262547
    },
    {
      "action_query_norm": 40.50947570800781,
      "action_relation_loss": 0.001625297823920846,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.128157156220766,
      "flow_matching_loss": 0.0260204428806901,
      "grad_norm": 0.1475355327129364,
      "learning_rate": 3.608220407923552e-05,
      "loss": 0.0278,
      "relation_loss": 0.0017470911610871554,
      "selected_pairs": 26574.4,
      "state_query_norm": 42.48075866699219,
      "state_relation_loss": 0.00012179335753899067,
      "step": 18310,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0017470911610871554,
      "weighted_relation_loss": 0.0017470911610871554
    },
    {
      "action_query_norm": 40.49874420166016,
      "action_relation_loss": 0.001543927297461778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.137511693171188,
      "flow_matching_loss": 0.024918897077441216,
      "grad_norm": 0.11277732998132706,
      "learning_rate": 3.602927516638833e-05,
      "loss": 0.0266,
      "relation_loss": 0.0016652503865770996,
      "selected_pairs": 27247.5,
      "state_query_norm": 42.5033878326416,
      "state_relation_loss": 0.00012132309202570469,
      "step": 18320,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0016652503865770996,
      "weighted_relation_loss": 0.0016652503865770996
    },
    {
      "action_query_norm": 40.47541885375976,
      "action_relation_loss": 0.0017104538623243571,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.14686623012161,
      "flow_matching_loss": 0.022486979700624944,
      "grad_norm": 0.11089383810758591,
      "learning_rate": 3.5976363229289165e-05,
      "loss": 0.0243,
      "relation_loss": 0.0018336323439143599,
      "selected_pairs": 27034.1,
      "state_query_norm": 42.48265495300293,
      "state_relation_loss": 0.00012317847285885363,
      "step": 18330,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0018336323439143599,
      "weighted_relation_loss": 0.0018336323439143599
    },
    {
      "action_query_norm": 40.47759742736817,
      "action_relation_loss": 0.0016490682261064648,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.15622076707203,
      "flow_matching_loss": 0.02376182470470667,
      "grad_norm": 0.1266685128211975,
      "learning_rate": 3.5923468332231e-05,
      "loss": 0.0255,
      "relation_loss": 0.0017640728037804365,
      "selected_pairs": 26848.3,
      "state_query_norm": 42.49208068847656,
      "state_relation_loss": 0.00011500458858790808,
      "step": 18340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0017640728037804365,
      "weighted_relation_loss": 0.0017640728037804365
    },
    {
      "action_query_norm": 40.48099708557129,
      "action_relation_loss": 0.0016066954820416867,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.16557530402245,
      "flow_matching_loss": 0.024479449726641178,
      "grad_norm": 0.10287836194038391,
      "learning_rate": 3.587059053948616e-05,
      "loss": 0.0262,
      "relation_loss": 0.0017331744660623372,
      "selected_pairs": 27015.3,
      "state_query_norm": 42.47647132873535,
      "state_relation_loss": 0.0001264789614651818,
      "step": 18350,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0017331744660623372,
      "weighted_relation_loss": 0.0017331744660623372
    },
    {
      "action_query_norm": 40.45982360839844,
      "action_relation_loss": 0.0015863032662309707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.174929840972872,
      "flow_matching_loss": 0.025996856018900873,
      "grad_norm": 0.10655108094215393,
      "learning_rate": 3.5817729915306134e-05,
      "loss": 0.0277,
      "relation_loss": 0.0017068804008886218,
      "selected_pairs": 26570.7,
      "state_query_norm": 42.483707427978516,
      "state_relation_loss": 0.00012057714702677913,
      "step": 18360,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0017068804008886218,
      "weighted_relation_loss": 0.0017068804008886218
    },
    {
      "action_query_norm": 40.459665298461914,
      "action_relation_loss": 0.001689647464081645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.184284377923294,
      "flow_matching_loss": 0.02290422860532999,
      "grad_norm": 0.11559829115867615,
      "learning_rate": 3.5764886523921566e-05,
      "loss": 0.0247,
      "relation_loss": 0.001809407293330878,
      "selected_pairs": 26692.0,
      "state_query_norm": 42.49685478210449,
      "state_relation_loss": 0.00011975982488365844,
      "step": 18370,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.001809407293330878,
      "weighted_relation_loss": 0.001809407293330878
    },
    {
      "action_query_norm": 40.45332794189453,
      "action_relation_loss": 0.0016896325978450477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.193638914873713,
      "flow_matching_loss": 0.025361070595681667,
      "grad_norm": 0.1097194105386734,
      "learning_rate": 3.571206042954214e-05,
      "loss": 0.0272,
      "relation_loss": 0.0018069295678287745,
      "selected_pairs": 27058.9,
      "state_query_norm": 42.475168228149414,
      "state_relation_loss": 0.00011729694306268356,
      "step": 18380,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0018069295678287745,
      "weighted_relation_loss": 0.0018069295678287745
    },
    {
      "action_query_norm": 40.43770523071289,
      "action_relation_loss": 0.0016293583437800407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.202993451824135,
      "flow_matching_loss": 0.023693242855370045,
      "grad_norm": 0.12005852907896042,
      "learning_rate": 3.5659251696356566e-05,
      "loss": 0.0254,
      "relation_loss": 0.00174243951914832,
      "selected_pairs": 26925.7,
      "state_query_norm": 42.494439697265626,
      "state_relation_loss": 0.00011308117464068346,
      "step": 18390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.00174243951914832,
      "weighted_relation_loss": 0.00174243951914832
    },
    {
      "action_query_norm": 40.49277610778809,
      "action_relation_loss": 0.0015555876307189465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.212347988774557,
      "flow_matching_loss": 0.023301079869270325,
      "grad_norm": 0.08601748198270798,
      "learning_rate": 3.56064603885324e-05,
      "loss": 0.025,
      "relation_loss": 0.001678171439561993,
      "selected_pairs": 26805.2,
      "state_query_norm": 42.48070030212402,
      "state_relation_loss": 0.00012258380302228035,
      "step": 18400,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001678171439561993,
      "weighted_relation_loss": 0.001678171439561993
    },
    {
      "action_query_norm": 40.43797225952149,
      "action_relation_loss": 0.0015969278756529093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.221702525724975,
      "flow_matching_loss": 0.023644811287522317,
      "grad_norm": 0.1212480291724205,
      "learning_rate": 3.5553686570216115e-05,
      "loss": 0.0254,
      "relation_loss": 0.0017214712454006076,
      "selected_pairs": 25846.2,
      "state_query_norm": 42.468309783935545,
      "state_relation_loss": 0.00012454336756491103,
      "step": 18410,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.0017214712454006076,
      "weighted_relation_loss": 0.0017214712454006076
    },
    {
      "action_query_norm": 40.43898735046387,
      "action_relation_loss": 0.0016694544348865748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.231057062675397,
      "flow_matching_loss": 0.025391456112265585,
      "grad_norm": 0.08279528468847275,
      "learning_rate": 3.550093030553284e-05,
      "loss": 0.0272,
      "relation_loss": 0.0017950162058696152,
      "selected_pairs": 26986.4,
      "state_query_norm": 42.478803634643555,
      "state_relation_loss": 0.00012556178189697674,
      "step": 18420,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0017950162058696152,
      "weighted_relation_loss": 0.0017950162058696152
    },
    {
      "action_query_norm": 40.46444320678711,
      "action_relation_loss": 0.001588511560112238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.24041159962582,
      "flow_matching_loss": 0.024498171359300613,
      "grad_norm": 0.08234690874814987,
      "learning_rate": 3.544819165858642e-05,
      "loss": 0.0262,
      "relation_loss": 0.0017114644986577332,
      "selected_pairs": 26486.2,
      "state_query_norm": 42.481344604492186,
      "state_relation_loss": 0.00012295295382500627,
      "step": 18430,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0017114644986577332,
      "weighted_relation_loss": 0.0017114644986577332
    },
    {
      "action_query_norm": 40.48366928100586,
      "action_relation_loss": 0.001547913404647261,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.24976613657624,
      "flow_matching_loss": 0.02501243185251951,
      "grad_norm": 0.10232463479042053,
      "learning_rate": 3.539547069345926e-05,
      "loss": 0.0267,
      "relation_loss": 0.0016681751701980829,
      "selected_pairs": 27104.0,
      "state_query_norm": 42.48874397277832,
      "state_relation_loss": 0.0001202617640956305,
      "step": 18440,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0016681751701980829,
      "weighted_relation_loss": 0.0016681751701980829
    },
    {
      "action_query_norm": 40.44849166870117,
      "action_relation_loss": 0.0016503558028489352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.25912067352666,
      "flow_matching_loss": 0.023950277641415597,
      "grad_norm": 0.09576813131570816,
      "learning_rate": 3.5342767474212344e-05,
      "loss": 0.0257,
      "relation_loss": 0.0017686138860881328,
      "selected_pairs": 27639.2,
      "state_query_norm": 42.47244148254394,
      "state_relation_loss": 0.00011825806504930369,
      "step": 18450,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0017686138860881328,
      "weighted_relation_loss": 0.0017686138860881328
    },
    {
      "action_query_norm": 40.41444320678711,
      "action_relation_loss": 0.0016138553619384766,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.26847521047708,
      "flow_matching_loss": 0.02394891865551472,
      "grad_norm": 0.11469753086566925,
      "learning_rate": 3.529008206488502e-05,
      "loss": 0.0257,
      "relation_loss": 0.0017360964324325324,
      "selected_pairs": 27262.2,
      "state_query_norm": 42.47697105407715,
      "state_relation_loss": 0.00012224109086673708,
      "step": 18460,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0017360964324325324,
      "weighted_relation_loss": 0.0017360964324325324
    },
    {
      "action_query_norm": 40.385922241210935,
      "action_relation_loss": 0.0016104597714729607,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.277829747427504,
      "flow_matching_loss": 0.023754271492362023,
      "grad_norm": 0.10844732820987701,
      "learning_rate": 3.5237414529495055e-05,
      "loss": 0.0255,
      "relation_loss": 0.001730529498308897,
      "selected_pairs": 26790.1,
      "state_query_norm": 42.45723114013672,
      "state_relation_loss": 0.00012006972829112784,
      "step": 18470,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001730529498308897,
      "weighted_relation_loss": 0.001730529498308897
    },
    {
      "action_query_norm": 40.405975341796875,
      "action_relation_loss": 0.00167800314957276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.287184284377922,
      "flow_matching_loss": 0.021805070526897907,
      "grad_norm": 0.11318787187337875,
      "learning_rate": 3.5184764932038454e-05,
      "loss": 0.0236,
      "relation_loss": 0.0017968939384445548,
      "selected_pairs": 26943.2,
      "state_query_norm": 42.47040710449219,
      "state_relation_loss": 0.00011889079469256103,
      "step": 18480,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017968939384445548,
      "weighted_relation_loss": 0.0017968939384445548
    },
    {
      "action_query_norm": 40.433636856079104,
      "action_relation_loss": 0.0015892748604528606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.296538821328344,
      "flow_matching_loss": 0.02526837233453989,
      "grad_norm": 0.09652428328990936,
      "learning_rate": 3.513213333648945e-05,
      "loss": 0.027,
      "relation_loss": 0.0017083571176044643,
      "selected_pairs": 26824.5,
      "state_query_norm": 42.46520805358887,
      "state_relation_loss": 0.00011908224914805032,
      "step": 18490,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0017083571176044643,
      "weighted_relation_loss": 0.0017083571176044643
    },
    {
      "action_query_norm": 40.4916934967041,
      "action_relation_loss": 0.001586211286485195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.305893358278766,
      "flow_matching_loss": 0.02299812566488981,
      "grad_norm": 0.11504559963941574,
      "learning_rate": 3.5079519806800374e-05,
      "loss": 0.0247,
      "relation_loss": 0.0017045745975337923,
      "selected_pairs": 26961.5,
      "state_query_norm": 42.47166366577149,
      "state_relation_loss": 0.00011836330668302253,
      "step": 18500,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0017045745975337923,
      "weighted_relation_loss": 0.0017045745975337923
    },
    {
      "action_query_norm": 40.44262199401855,
      "action_relation_loss": 0.001568022696301341,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.315247895229184,
      "flow_matching_loss": 0.025163339264690877,
      "grad_norm": 0.13999277353286743,
      "learning_rate": 3.502692440690165e-05,
      "loss": 0.0268,
      "relation_loss": 0.0016848577070049942,
      "selected_pairs": 26711.1,
      "state_query_norm": 42.46247482299805,
      "state_relation_loss": 0.00011683500342769548,
      "step": 18510,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0016848577070049942,
      "weighted_relation_loss": 0.0016848577070049942
    },
    {
      "action_query_norm": 40.45154914855957,
      "action_relation_loss": 0.0015704550431109964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.324602432179606,
      "flow_matching_loss": 0.02481807377189398,
      "grad_norm": 0.1025475338101387,
      "learning_rate": 3.497434720070165e-05,
      "loss": 0.0265,
      "relation_loss": 0.0016906521399505437,
      "selected_pairs": 27124.8,
      "state_query_norm": 42.45951538085937,
      "state_relation_loss": 0.0001201970866532065,
      "step": 18520,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0016906521399505437,
      "weighted_relation_loss": 0.0016906521399505437
    },
    {
      "action_query_norm": 40.442508697509766,
      "action_relation_loss": 0.0015953590744175016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.33395696913003,
      "flow_matching_loss": 0.02457695472985506,
      "grad_norm": 0.097411148250103,
      "learning_rate": 3.4921788252086616e-05,
      "loss": 0.0263,
      "relation_loss": 0.0017146166879683732,
      "selected_pairs": 26673.5,
      "state_query_norm": 42.46067733764649,
      "state_relation_loss": 0.00011925762955797837,
      "step": 18530,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0017146166879683732,
      "weighted_relation_loss": 0.0017146166879683732
    },
    {
      "action_query_norm": 40.43304557800293,
      "action_relation_loss": 0.0015803691232576967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.34331150608045,
      "flow_matching_loss": 0.024309279024600984,
      "grad_norm": 0.09920907020568848,
      "learning_rate": 3.486924762492065e-05,
      "loss": 0.026,
      "relation_loss": 0.001691051316447556,
      "selected_pairs": 27406.1,
      "state_query_norm": 42.47876319885254,
      "state_relation_loss": 0.00011068218518630601,
      "step": 18540,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001691051316447556,
      "weighted_relation_loss": 0.001691051316447556
    },
    {
      "action_query_norm": 40.45908126831055,
      "action_relation_loss": 0.0015907708439044654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.35266604303087,
      "flow_matching_loss": 0.024013804271817207,
      "grad_norm": 0.09794753789901733,
      "learning_rate": 3.4816725383045534e-05,
      "loss": 0.0257,
      "relation_loss": 0.0017055312055163085,
      "selected_pairs": 27017.8,
      "state_query_norm": 42.46637077331543,
      "state_relation_loss": 0.0001147603521530982,
      "step": 18550,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0017055312055163085,
      "weighted_relation_loss": 0.0017055312055163085
    },
    {
      "action_query_norm": 40.41128387451172,
      "action_relation_loss": 0.001599963754415512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.36202057998129,
      "flow_matching_loss": 0.024706477671861647,
      "grad_norm": 0.1044294685125351,
      "learning_rate": 3.476422159028079e-05,
      "loss": 0.0264,
      "relation_loss": 0.001724233035929501,
      "selected_pairs": 26832.8,
      "state_query_norm": 42.45659637451172,
      "state_relation_loss": 0.00012426928806235081,
      "step": 18560,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001724233035929501,
      "weighted_relation_loss": 0.001724233035929501
    },
    {
      "action_query_norm": 40.44753532409668,
      "action_relation_loss": 0.0015528055955655873,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.371375116931713,
      "flow_matching_loss": 0.023506592586636545,
      "grad_norm": 0.12870633602142334,
      "learning_rate": 3.471173631042345e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016756168683059514,
      "selected_pairs": 27248.1,
      "state_query_norm": 42.460577392578124,
      "state_relation_loss": 0.00012281127856113018,
      "step": 18570,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0016756168683059514,
      "weighted_relation_loss": 0.0016756168683059514
    },
    {
      "action_query_norm": 40.41986351013183,
      "action_relation_loss": 0.0016799465753138064,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.38072965388213,
      "flow_matching_loss": 0.024291733279824257,
      "grad_norm": 0.13808754086494446,
      "learning_rate": 3.465926960724808e-05,
      "loss": 0.0261,
      "relation_loss": 0.0017987402039580046,
      "selected_pairs": 26944.1,
      "state_query_norm": 42.457132720947264,
      "state_relation_loss": 0.00011879361263709143,
      "step": 18580,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017987402039580046,
      "weighted_relation_loss": 0.0017987402039580046
    },
    {
      "action_query_norm": 40.37553443908691,
      "action_relation_loss": 0.0015692711691372096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.390084190832553,
      "flow_matching_loss": 0.02288933489471674,
      "grad_norm": 0.09852629154920578,
      "learning_rate": 3.460682154450666e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016840771771967412,
      "selected_pairs": 26937.2,
      "state_query_norm": 42.453189086914065,
      "state_relation_loss": 0.00011480600587674417,
      "step": 18590,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016840771771967412,
      "weighted_relation_loss": 0.0016840771771967412
    },
    {
      "action_query_norm": 40.420708084106444,
      "action_relation_loss": 0.0016288113896735013,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.399438727782975,
      "flow_matching_loss": 0.0248952466994524,
      "grad_norm": 0.1293834149837494,
      "learning_rate": 3.4554392185928564e-05,
      "loss": 0.0266,
      "relation_loss": 0.0017421136260963976,
      "selected_pairs": 26597.6,
      "state_query_norm": 42.45585899353027,
      "state_relation_loss": 0.00011330221677781082,
      "step": 18600,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0017421136260963976,
      "weighted_relation_loss": 0.0017421136260963976
    },
    {
      "action_query_norm": 40.364127731323244,
      "action_relation_loss": 0.0016261338139884173,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.408793264733397,
      "flow_matching_loss": 0.02523853313177824,
      "grad_norm": 0.13300971686840057,
      "learning_rate": 3.450198159522037e-05,
      "loss": 0.027,
      "relation_loss": 0.001748342183418572,
      "selected_pairs": 26620.9,
      "state_query_norm": 42.45091857910156,
      "state_relation_loss": 0.0001222083708853461,
      "step": 18610,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001748342183418572,
      "weighted_relation_loss": 0.001748342183418572
    },
    {
      "action_query_norm": 40.379634094238284,
      "action_relation_loss": 0.001589604129549116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.418147801683816,
      "flow_matching_loss": 0.02301280777901411,
      "grad_norm": 0.1255439966917038,
      "learning_rate": 3.444958983606592e-05,
      "loss": 0.0247,
      "relation_loss": 0.0017087597865611316,
      "selected_pairs": 27264.3,
      "state_query_norm": 42.444388961791994,
      "state_relation_loss": 0.0001191556584672071,
      "step": 18620,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0017087597865611316,
      "weighted_relation_loss": 0.0017087597865611316
    },
    {
      "action_query_norm": 40.36450386047363,
      "action_relation_loss": 0.0015542337205260992,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.427502338634238,
      "flow_matching_loss": 0.024069728888571264,
      "grad_norm": 0.1473388969898224,
      "learning_rate": 3.439721697212612e-05,
      "loss": 0.0257,
      "relation_loss": 0.0016622619819827377,
      "selected_pairs": 26484.2,
      "state_query_norm": 42.4505558013916,
      "state_relation_loss": 0.00010802825781865976,
      "step": 18630,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0016622619819827377,
      "weighted_relation_loss": 0.0016622619819827377
    },
    {
      "action_query_norm": 40.34805679321289,
      "action_relation_loss": 0.0016172958770766855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.43685687558466,
      "flow_matching_loss": 0.022911952435970308,
      "grad_norm": 0.08533389121294022,
      "learning_rate": 3.4344863067038954e-05,
      "loss": 0.0246,
      "relation_loss": 0.0017337617464363575,
      "selected_pairs": 27013.4,
      "state_query_norm": 42.45719528198242,
      "state_relation_loss": 0.00011646587736322545,
      "step": 18640,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0017337617464363575,
      "weighted_relation_loss": 0.0017337617464363575
    },
    {
      "action_query_norm": 40.388597869873045,
      "action_relation_loss": 0.00156362367561087,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.446211412535078,
      "flow_matching_loss": 0.02441699169576168,
      "grad_norm": 0.10101891309022903,
      "learning_rate": 3.429252818441935e-05,
      "loss": 0.0261,
      "relation_loss": 0.0016829847358167172,
      "selected_pairs": 27164.1,
      "state_query_norm": 42.44886360168457,
      "state_relation_loss": 0.00011936104347114452,
      "step": 18650,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0016829847358167172,
      "weighted_relation_loss": 0.0016829847358167172
    },
    {
      "action_query_norm": 40.37818298339844,
      "action_relation_loss": 0.0015452981577254831,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.4555659494855,
      "flow_matching_loss": 0.02444200646132231,
      "grad_norm": 0.09400040656328201,
      "learning_rate": 3.4240212387859094e-05,
      "loss": 0.0261,
      "relation_loss": 0.001665373332798481,
      "selected_pairs": 26848.0,
      "state_query_norm": 42.439854049682616,
      "state_relation_loss": 0.00012007517361780628,
      "step": 18660,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001665373332798481,
      "weighted_relation_loss": 0.001665373332798481
    },
    {
      "action_query_norm": 40.447827911376955,
      "action_relation_loss": 0.0015371736022643744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.464920486435922,
      "flow_matching_loss": 0.025710353814065456,
      "grad_norm": 0.08183760195970535,
      "learning_rate": 3.418791574092686e-05,
      "loss": 0.0274,
      "relation_loss": 0.0016556105809286236,
      "selected_pairs": 27369.4,
      "state_query_norm": 42.44631233215332,
      "state_relation_loss": 0.00011843696120195091,
      "step": 18670,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0016556105809286236,
      "weighted_relation_loss": 0.0016556105809286236
    },
    {
      "action_query_norm": 40.41254997253418,
      "action_relation_loss": 0.0015506939613260328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.474275023386344,
      "flow_matching_loss": 0.0239560229703784,
      "grad_norm": 0.12035845220088959,
      "learning_rate": 3.413563830716796e-05,
      "loss": 0.0256,
      "relation_loss": 0.0016665791394189,
      "selected_pairs": 26993.8,
      "state_query_norm": 42.442787170410156,
      "state_relation_loss": 0.00011588519992073998,
      "step": 18680,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0016665791394189,
      "weighted_relation_loss": 0.0016665791394189
    },
    {
      "action_query_norm": 40.33022003173828,
      "action_relation_loss": 0.0016492470516823233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.483629560336762,
      "flow_matching_loss": 0.025069624558091162,
      "grad_norm": 0.1031174287199974,
      "learning_rate": 3.408338015010445e-05,
      "loss": 0.0268,
      "relation_loss": 0.001763138989917934,
      "selected_pairs": 26410.1,
      "state_query_norm": 42.42908782958985,
      "state_relation_loss": 0.00011389195351512172,
      "step": 18690,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001763138989917934,
      "weighted_relation_loss": 0.001763138989917934
    },
    {
      "action_query_norm": 40.39941520690918,
      "action_relation_loss": 0.0015479662572033704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.492984097287184,
      "flow_matching_loss": 0.024793009273707867,
      "grad_norm": 0.08329135924577713,
      "learning_rate": 3.40311413332349e-05,
      "loss": 0.0264,
      "relation_loss": 0.0016565996455028653,
      "selected_pairs": 27367.3,
      "state_query_norm": 42.447780227661134,
      "state_relation_loss": 0.00010863340212381445,
      "step": 18700,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0016565996455028653,
      "weighted_relation_loss": 0.0016565996455028653
    },
    {
      "action_query_norm": 40.35207824707031,
      "action_relation_loss": 0.001656209712382406,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.502338634237606,
      "flow_matching_loss": 0.023554310388863086,
      "grad_norm": 0.09129375964403152,
      "learning_rate": 3.3978921920034365e-05,
      "loss": 0.0253,
      "relation_loss": 0.0017637064796872438,
      "selected_pairs": 26653.3,
      "state_query_norm": 42.44151268005371,
      "state_relation_loss": 0.00010749675493570976,
      "step": 18710,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0017637064796872438,
      "weighted_relation_loss": 0.0017637064796872438
    },
    {
      "action_query_norm": 40.3886043548584,
      "action_relation_loss": 0.0016391026554629206,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.511693171188025,
      "flow_matching_loss": 0.022446645237505437,
      "grad_norm": 0.11787080764770508,
      "learning_rate": 3.392672197395441e-05,
      "loss": 0.0242,
      "relation_loss": 0.001751025568228215,
      "selected_pairs": 26851.6,
      "state_query_norm": 42.445582962036134,
      "state_relation_loss": 0.00011192291713086888,
      "step": 18720,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001751025568228215,
      "weighted_relation_loss": 0.001751025568228215
    },
    {
      "action_query_norm": 40.330055236816406,
      "action_relation_loss": 0.0016895493376068772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.521047708138447,
      "flow_matching_loss": 0.023028735257685184,
      "grad_norm": 0.15814511477947235,
      "learning_rate": 3.387454155842287e-05,
      "loss": 0.0248,
      "relation_loss": 0.0018058740766718983,
      "selected_pairs": 26643.4,
      "state_query_norm": 42.4318042755127,
      "state_relation_loss": 0.00011632472014753149,
      "step": 18730,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0018058740766718983,
      "weighted_relation_loss": 0.0018058740766718983
    },
    {
      "action_query_norm": 40.31231155395508,
      "action_relation_loss": 0.0015680119744502009,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.53040224508887,
      "flow_matching_loss": 0.023568779230117798,
      "grad_norm": 0.10253424197435379,
      "learning_rate": 3.382238073684386e-05,
      "loss": 0.0253,
      "relation_loss": 0.0016874399618245662,
      "selected_pairs": 26969.5,
      "state_query_norm": 42.43158378601074,
      "state_relation_loss": 0.00011942799974349328,
      "step": 18740,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016874399618245662,
      "weighted_relation_loss": 0.0016874399618245662
    },
    {
      "action_query_norm": 40.388261032104495,
      "action_relation_loss": 0.0015556482248939574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.53975678203929,
      "flow_matching_loss": 0.023186166398227216,
      "grad_norm": 0.09975621849298477,
      "learning_rate": 3.377023957259771e-05,
      "loss": 0.0249,
      "relation_loss": 0.0016687142313458025,
      "selected_pairs": 27088.0,
      "state_query_norm": 42.435120010375975,
      "state_relation_loss": 0.00011306601954856888,
      "step": 18750,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0016687142313458025,
      "weighted_relation_loss": 0.0016687142313458025
    },
    {
      "action_query_norm": 40.39969596862793,
      "action_relation_loss": 0.0015254130470566452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.54911131898971,
      "flow_matching_loss": 0.02362998928874731,
      "grad_norm": 0.10135731846094131,
      "learning_rate": 3.3718118129040835e-05,
      "loss": 0.0253,
      "relation_loss": 0.0016404670546762644,
      "selected_pairs": 27180.2,
      "state_query_norm": 42.43324165344238,
      "state_relation_loss": 0.00011505400252644903,
      "step": 18760,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0016404670546762644,
      "weighted_relation_loss": 0.0016404670546762644
    },
    {
      "action_query_norm": 40.384230041503905,
      "action_relation_loss": 0.0016181834740564226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.55846585594013,
      "flow_matching_loss": 0.02218532133847475,
      "grad_norm": 0.08722779899835587,
      "learning_rate": 3.3666016469505724e-05,
      "loss": 0.0239,
      "relation_loss": 0.0017359668505378067,
      "selected_pairs": 26943.6,
      "state_query_norm": 42.43015060424805,
      "state_relation_loss": 0.00011778337939176709,
      "step": 18770,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017359668505378067,
      "weighted_relation_loss": 0.0017359668505378067
    },
    {
      "action_query_norm": 40.397400283813475,
      "action_relation_loss": 0.001548596960492432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.567820392890553,
      "flow_matching_loss": 0.022575392946600915,
      "grad_norm": 0.08495713025331497,
      "learning_rate": 3.361393465730079e-05,
      "loss": 0.0242,
      "relation_loss": 0.0016604706062935293,
      "selected_pairs": 26875.8,
      "state_query_norm": 42.44436874389648,
      "state_relation_loss": 0.00011187365598743781,
      "step": 18780,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0016604706062935293,
      "weighted_relation_loss": 0.0016604706062935293
    },
    {
      "action_query_norm": 40.3536792755127,
      "action_relation_loss": 0.0016623802599497139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.57717492984097,
      "flow_matching_loss": 0.025910444930195807,
      "grad_norm": 0.14062926173210144,
      "learning_rate": 3.3561872755710366e-05,
      "loss": 0.0277,
      "relation_loss": 0.0017786088166758419,
      "selected_pairs": 26922.0,
      "state_query_norm": 42.446724319458006,
      "state_relation_loss": 0.00011622854799497873,
      "step": 18790,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0017786088166758419,
      "weighted_relation_loss": 0.0017786088166758419
    },
    {
      "action_query_norm": 40.355686569213866,
      "action_relation_loss": 0.0015462051494978368,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.586529466791394,
      "flow_matching_loss": 0.02404873538762331,
      "grad_norm": 0.08607503026723862,
      "learning_rate": 3.350983082799456e-05,
      "loss": 0.0257,
      "relation_loss": 0.001669105002656579,
      "selected_pairs": 26780.2,
      "state_query_norm": 42.425391006469724,
      "state_relation_loss": 0.00012289984442759305,
      "step": 18800,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.001669105002656579,
      "weighted_relation_loss": 0.001669105002656579
    },
    {
      "action_query_norm": 40.31562271118164,
      "action_relation_loss": 0.001627753011416644,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.595884003741816,
      "flow_matching_loss": 0.022445878200232982,
      "grad_norm": 0.09426531940698624,
      "learning_rate": 3.34578089373892e-05,
      "loss": 0.0242,
      "relation_loss": 0.0017425749567337335,
      "selected_pairs": 26756.7,
      "state_query_norm": 42.425223159790036,
      "state_relation_loss": 0.00011482195914140903,
      "step": 18810,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0017425749567337335,
      "weighted_relation_loss": 0.0017425749567337335
    },
    {
      "action_query_norm": 40.38239936828613,
      "action_relation_loss": 0.0015396129572764039,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.605238540692234,
      "flow_matching_loss": 0.021916012465953826,
      "grad_norm": 0.14619629085063934,
      "learning_rate": 3.340580714710581e-05,
      "loss": 0.0236,
      "relation_loss": 0.0016568695777095855,
      "selected_pairs": 26800.5,
      "state_query_norm": 42.432919311523435,
      "state_relation_loss": 0.00011725660442607477,
      "step": 18820,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0016568695777095855,
      "weighted_relation_loss": 0.0016568695777095855
    },
    {
      "action_query_norm": 40.404794692993164,
      "action_relation_loss": 0.0015204150113277137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.614593077642656,
      "flow_matching_loss": 0.02359054945409298,
      "grad_norm": 0.08718792349100113,
      "learning_rate": 3.3353825520331466e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016367596108466387,
      "selected_pairs": 27627.5,
      "state_query_norm": 42.41774063110351,
      "state_relation_loss": 0.00011634460825007408,
      "step": 18830,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0016367596108466387,
      "weighted_relation_loss": 0.0016367596108466387
    },
    {
      "action_query_norm": 40.31706504821777,
      "action_relation_loss": 0.0015754476189613342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.623947614593078,
      "flow_matching_loss": 0.022952424921095373,
      "grad_norm": 0.08608204126358032,
      "learning_rate": 3.330186412022876e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016919389367103578,
      "selected_pairs": 26158.0,
      "state_query_norm": 42.42121887207031,
      "state_relation_loss": 0.00011649134030449204,
      "step": 18840,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0016919389367103578,
      "weighted_relation_loss": 0.0016919389367103578
    },
    {
      "action_query_norm": 40.344783782958984,
      "action_relation_loss": 0.0015479050227440894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.6333021515435,
      "flow_matching_loss": 0.02138214763253927,
      "grad_norm": 0.14606650173664093,
      "learning_rate": 3.324992300993568e-05,
      "loss": 0.023,
      "relation_loss": 0.001664687437005341,
      "selected_pairs": 26622.7,
      "state_query_norm": 42.40609283447266,
      "state_relation_loss": 0.00011678242444759234,
      "step": 18850,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001664687437005341,
      "weighted_relation_loss": 0.001664687437005341
    },
    {
      "action_query_norm": 40.310794067382815,
      "action_relation_loss": 0.0015940899029374123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.64265668849392,
      "flow_matching_loss": 0.022889823652803898,
      "grad_norm": 0.10012976825237274,
      "learning_rate": 3.319800225256556e-05,
      "loss": 0.0246,
      "relation_loss": 0.0017049534479156136,
      "selected_pairs": 26768.6,
      "state_query_norm": 42.418688201904295,
      "state_relation_loss": 0.00011086355079896748,
      "step": 18860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0017049534479156136,
      "weighted_relation_loss": 0.0017049534479156136
    },
    {
      "action_query_norm": 40.33033294677735,
      "action_relation_loss": 0.0015158918569795788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.65201122544434,
      "flow_matching_loss": 0.023427916504442692,
      "grad_norm": 0.11405982077121735,
      "learning_rate": 3.314610191120702e-05,
      "loss": 0.0251,
      "relation_loss": 0.001626386446878314,
      "selected_pairs": 26481.6,
      "state_query_norm": 42.41748313903808,
      "state_relation_loss": 0.00011049457680201158,
      "step": 18870,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001626386446878314,
      "weighted_relation_loss": 0.001626386446878314
    },
    {
      "action_query_norm": 40.327871322631836,
      "action_relation_loss": 0.0015297221951186657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.661365762394762,
      "flow_matching_loss": 0.023801141418516636,
      "grad_norm": 0.12186632305383682,
      "learning_rate": 3.30942220489239e-05,
      "loss": 0.0254,
      "relation_loss": 0.0016443599481135607,
      "selected_pairs": 26968.2,
      "state_query_norm": 42.414516830444335,
      "state_relation_loss": 0.00011463773698778823,
      "step": 18880,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016443599481135607,
      "weighted_relation_loss": 0.0016443599481135607
    },
    {
      "action_query_norm": 40.33102531433106,
      "action_relation_loss": 0.0015310813556425273,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.67072029934518,
      "flow_matching_loss": 0.02390899509191513,
      "grad_norm": 0.08918683230876923,
      "learning_rate": 3.3042362728755086e-05,
      "loss": 0.0256,
      "relation_loss": 0.0016562373260967434,
      "selected_pairs": 26341.2,
      "state_query_norm": 42.39475936889649,
      "state_relation_loss": 0.00012515597627498208,
      "step": 18890,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0016562373260967434,
      "weighted_relation_loss": 0.0016562373260967434
    },
    {
      "action_query_norm": 40.30926322937012,
      "action_relation_loss": 0.001534457434900105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.680074836295603,
      "flow_matching_loss": 0.0236514613032341,
      "grad_norm": 0.08420528471469879,
      "learning_rate": 3.299052401371456e-05,
      "loss": 0.0253,
      "relation_loss": 0.001647778193000704,
      "selected_pairs": 27047.0,
      "state_query_norm": 42.430219650268555,
      "state_relation_loss": 0.00011332076537655667,
      "step": 18900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001647778193000704,
      "weighted_relation_loss": 0.001647778193000704
    },
    {
      "action_query_norm": 40.33830871582031,
      "action_relation_loss": 0.0014944328228011726,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.689429373246025,
      "flow_matching_loss": 0.022574184648692607,
      "grad_norm": 0.10669031739234924,
      "learning_rate": 3.293870596679125e-05,
      "loss": 0.0242,
      "relation_loss": 0.0016123245004564524,
      "selected_pairs": 27290.8,
      "state_query_norm": 42.41087455749512,
      "state_relation_loss": 0.0001178916747448966,
      "step": 18910,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0016123245004564524,
      "weighted_relation_loss": 0.0016123245004564524
    },
    {
      "action_query_norm": 40.337266540527345,
      "action_relation_loss": 0.0016607390833087265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.698783910196447,
      "flow_matching_loss": 0.02465057000517845,
      "grad_norm": 0.12261343002319336,
      "learning_rate": 3.288690865094895e-05,
      "loss": 0.0264,
      "relation_loss": 0.001777370204217732,
      "selected_pairs": 27160.7,
      "state_query_norm": 42.41219520568848,
      "state_relation_loss": 0.00011663111945381388,
      "step": 18920,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001777370204217732,
      "weighted_relation_loss": 0.001777370204217732
    },
    {
      "action_query_norm": 40.31446990966797,
      "action_relation_loss": 0.0016700868145562708,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.708138447146865,
      "flow_matching_loss": 0.024441079422831535,
      "grad_norm": 0.09917941689491272,
      "learning_rate": 3.283513212912632e-05,
      "loss": 0.0262,
      "relation_loss": 0.0017848656629212202,
      "selected_pairs": 26593.2,
      "state_query_norm": 42.417297744750975,
      "state_relation_loss": 0.00011477886218926869,
      "step": 18930,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0017848656629212202,
      "weighted_relation_loss": 0.0017848656629212202
    },
    {
      "action_query_norm": 40.3358699798584,
      "action_relation_loss": 0.0015347125707194209,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.717492984097287,
      "flow_matching_loss": 0.02457629833370447,
      "grad_norm": 0.087371826171875,
      "learning_rate": 3.2783376464236684e-05,
      "loss": 0.0262,
      "relation_loss": 0.001648886175826192,
      "selected_pairs": 26803.4,
      "state_query_norm": 42.39642486572266,
      "state_relation_loss": 0.00011417359710321761,
      "step": 18940,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001648886175826192,
      "weighted_relation_loss": 0.001648886175826192
    },
    {
      "action_query_norm": 40.322392272949216,
      "action_relation_loss": 0.0016401289496570826,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.72684752104771,
      "flow_matching_loss": 0.022722463682293893,
      "grad_norm": 0.12822848558425903,
      "learning_rate": 3.273164171916806e-05,
      "loss": 0.0245,
      "relation_loss": 0.0017561507644131779,
      "selected_pairs": 26943.3,
      "state_query_norm": 42.406669235229494,
      "state_relation_loss": 0.00011602180529735051,
      "step": 18950,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0017561507644131779,
      "weighted_relation_loss": 0.0017561507644131779
    },
    {
      "action_query_norm": 40.31393241882324,
      "action_relation_loss": 0.0015449414728209375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.736202057998128,
      "flow_matching_loss": 0.02395316883921623,
      "grad_norm": 0.09862974286079407,
      "learning_rate": 3.267992795678306e-05,
      "loss": 0.0256,
      "relation_loss": 0.0016596996574662626,
      "selected_pairs": 26634.4,
      "state_query_norm": 42.415846252441405,
      "state_relation_loss": 0.00011475819046609103,
      "step": 18960,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0016596996574662626,
      "weighted_relation_loss": 0.0016596996574662626
    },
    {
      "action_query_norm": 40.29828796386719,
      "action_relation_loss": 0.0015664419042877853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.74555659494855,
      "flow_matching_loss": 0.02340172491967678,
      "grad_norm": 0.08869294822216034,
      "learning_rate": 3.2628235239918744e-05,
      "loss": 0.0251,
      "relation_loss": 0.0016770151909440757,
      "selected_pairs": 26881.1,
      "state_query_norm": 42.397592544555664,
      "state_relation_loss": 0.00011057328738388606,
      "step": 18970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0016770151909440757,
      "weighted_relation_loss": 0.0016770151909440757
    },
    {
      "action_query_norm": 40.31925048828125,
      "action_relation_loss": 0.0015707908431068062,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.75491113189897,
      "flow_matching_loss": 0.023666886799037456,
      "grad_norm": 0.12527380883693695,
      "learning_rate": 3.2576563631386695e-05,
      "loss": 0.0253,
      "relation_loss": 0.0016820003977045417,
      "selected_pairs": 26942.5,
      "state_query_norm": 42.40212287902832,
      "state_relation_loss": 0.00011120954077341594,
      "step": 18980,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0016820003977045417,
      "weighted_relation_loss": 0.0016820003977045417
    },
    {
      "action_query_norm": 40.33901023864746,
      "action_relation_loss": 0.0015155689674429596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.764265668849394,
      "flow_matching_loss": 0.022802388295531273,
      "grad_norm": 0.12086664885282516,
      "learning_rate": 3.252491319397275e-05,
      "loss": 0.0244,
      "relation_loss": 0.0016305933124385775,
      "selected_pairs": 26335.8,
      "state_query_norm": 42.40420684814453,
      "state_relation_loss": 0.00011502433626446873,
      "step": 18990,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0016305933124385775,
      "weighted_relation_loss": 0.0016305933124385775
    },
    {
      "action_query_norm": 40.312736129760744,
      "action_relation_loss": 0.0015671896398998797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.773620205799812,
      "flow_matching_loss": 0.023548244312405587,
      "grad_norm": 0.15273793041706085,
      "learning_rate": 3.247328399043706e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016976692830212414,
      "selected_pairs": 26663.1,
      "state_query_norm": 42.39028663635254,
      "state_relation_loss": 0.00013047964457655326,
      "step": 19000,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0016976692830212414,
      "weighted_relation_loss": 0.0016976692830212414
    },
    {
      "action_query_norm": 40.34020652770996,
      "action_relation_loss": 0.0015586311928927898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.782974742750234,
      "flow_matching_loss": 0.023451091907918453,
      "grad_norm": 0.12925849854946136,
      "learning_rate": 3.242167608351399e-05,
      "loss": 0.0251,
      "relation_loss": 0.0016714613535441457,
      "selected_pairs": 27358.8,
      "state_query_norm": 42.41097755432129,
      "state_relation_loss": 0.00011283015701337717,
      "step": 19010,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0016714613535441457,
      "weighted_relation_loss": 0.0016714613535441457
    },
    {
      "action_query_norm": 40.3553524017334,
      "action_relation_loss": 0.001612967881374061,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.792329279700656,
      "flow_matching_loss": 0.023783283680677413,
      "grad_norm": 0.09369141608476639,
      "learning_rate": 3.2370089535911986e-05,
      "loss": 0.0255,
      "relation_loss": 0.0017332268878817558,
      "selected_pairs": 27173.5,
      "state_query_norm": 42.406258773803714,
      "state_relation_loss": 0.0001202590465254616,
      "step": 19020,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0017332268878817558,
      "weighted_relation_loss": 0.0017332268878817558
    },
    {
      "action_query_norm": 40.302772521972656,
      "action_relation_loss": 0.0017236973624676466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.801683816651074,
      "flow_matching_loss": 0.024373833276331424,
      "grad_norm": 0.1308932900428772,
      "learning_rate": 3.23185244103136e-05,
      "loss": 0.0262,
      "relation_loss": 0.0018327504745684563,
      "selected_pairs": 26572.0,
      "state_query_norm": 42.413261795043944,
      "state_relation_loss": 0.00010905310264206492,
      "step": 19030,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0018327504745684563,
      "weighted_relation_loss": 0.0018327504745684563
    },
    {
      "action_query_norm": 40.29361343383789,
      "action_relation_loss": 0.0015224563074298203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.811038353601496,
      "flow_matching_loss": 0.02413275055587292,
      "grad_norm": 0.12619055807590485,
      "learning_rate": 3.22669807693753e-05,
      "loss": 0.0258,
      "relation_loss": 0.001638163416646421,
      "selected_pairs": 26384.6,
      "state_query_norm": 42.39773788452148,
      "state_relation_loss": 0.00011570710412343033,
      "step": 19040,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001638163416646421,
      "weighted_relation_loss": 0.001638163416646421
    },
    {
      "action_query_norm": 40.39188652038574,
      "action_relation_loss": 0.0015524853835813702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.82039289055192,
      "flow_matching_loss": 0.02570802643895149,
      "grad_norm": 0.09693402051925659,
      "learning_rate": 3.2215458675727495e-05,
      "loss": 0.0274,
      "relation_loss": 0.0016638830536976456,
      "selected_pairs": 27579.6,
      "state_query_norm": 42.408098983764646,
      "state_relation_loss": 0.00011139767229906283,
      "step": 19050,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.0016638830536976456,
      "weighted_relation_loss": 0.0016638830536976456
    },
    {
      "action_query_norm": 40.34636001586914,
      "action_relation_loss": 0.0015726217883639038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.82974742750234,
      "flow_matching_loss": 0.023614867031574248,
      "grad_norm": 0.11452675610780716,
      "learning_rate": 3.216395819197438e-05,
      "loss": 0.0253,
      "relation_loss": 0.001690834341570735,
      "selected_pairs": 26638.7,
      "state_query_norm": 42.374121475219724,
      "state_relation_loss": 0.00011821256339317188,
      "step": 19060,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.001690834341570735,
      "weighted_relation_loss": 0.001690834341570735
    },
    {
      "action_query_norm": 40.31137046813965,
      "action_relation_loss": 0.0015908197732642294,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.83910196445276,
      "flow_matching_loss": 0.023583568818867206,
      "grad_norm": 0.12394092231988907,
      "learning_rate": 3.211247938069387e-05,
      "loss": 0.0253,
      "relation_loss": 0.001706920051947236,
      "selected_pairs": 27025.6,
      "state_query_norm": 42.384707641601565,
      "state_relation_loss": 0.00011610029469011351,
      "step": 19070,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001706920051947236,
      "weighted_relation_loss": 0.001706920051947236
    },
    {
      "action_query_norm": 40.338779830932616,
      "action_relation_loss": 0.001515414915047586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.84845650140318,
      "flow_matching_loss": 0.023907279409468175,
      "grad_norm": 0.11681164056062698,
      "learning_rate": 3.206102230443759e-05,
      "loss": 0.0255,
      "relation_loss": 0.001638883154373616,
      "selected_pairs": 26436.9,
      "state_query_norm": 42.384181594848634,
      "state_relation_loss": 0.00012346824441920035,
      "step": 19080,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001638883154373616,
      "weighted_relation_loss": 0.001638883154373616
    },
    {
      "action_query_norm": 40.34005584716797,
      "action_relation_loss": 0.0015599303413182497,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.857811038353603,
      "flow_matching_loss": 0.023567167297005653,
      "grad_norm": 0.1467694491147995,
      "learning_rate": 3.2009587025730764e-05,
      "loss": 0.0252,
      "relation_loss": 0.0016745531931519509,
      "selected_pairs": 26113.4,
      "state_query_norm": 42.38487510681152,
      "state_relation_loss": 0.00011462287147878669,
      "step": 19090,
      "valid_action_samples": 229.0,
      "weighted_auxiliary_loss": 0.0016745531931519509,
      "weighted_relation_loss": 0.0016745531931519509
    },
    {
      "action_query_norm": 40.349117279052734,
      "action_relation_loss": 0.0015504470909945666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.86716557530402,
      "flow_matching_loss": 0.02327827513217926,
      "grad_norm": 0.11260269582271576,
      "learning_rate": 3.195817360707207e-05,
      "loss": 0.0249,
      "relation_loss": 0.0016666434006765486,
      "selected_pairs": 27247.5,
      "state_query_norm": 42.37470054626465,
      "state_relation_loss": 0.00011619630313361995,
      "step": 19100,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0016666434006765486,
      "weighted_relation_loss": 0.0016666434006765486
    },
    {
      "action_query_norm": 40.33173179626465,
      "action_relation_loss": 0.001530449918936938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.876520112254443,
      "flow_matching_loss": 0.021355006471276285,
      "grad_norm": 0.08886592835187912,
      "learning_rate": 3.19067821109337e-05,
      "loss": 0.023,
      "relation_loss": 0.001649209822062403,
      "selected_pairs": 26526.9,
      "state_query_norm": 42.38904914855957,
      "state_relation_loss": 0.00011875988711835816,
      "step": 19110,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001649209822062403,
      "weighted_relation_loss": 0.001649209822062403
    },
    {
      "action_query_norm": 40.293570327758786,
      "action_relation_loss": 0.001630134356673807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.885874649204865,
      "flow_matching_loss": 0.02356332242488861,
      "grad_norm": 0.09997481107711792,
      "learning_rate": 3.185541259976114e-05,
      "loss": 0.0253,
      "relation_loss": 0.0017476257286034524,
      "selected_pairs": 26915.6,
      "state_query_norm": 42.380771255493165,
      "state_relation_loss": 0.00011749138429877349,
      "step": 19120,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0017476257286034524,
      "weighted_relation_loss": 0.0017476257286034524
    },
    {
      "action_query_norm": 40.297529983520505,
      "action_relation_loss": 0.0015830685617402196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.895229186155284,
      "flow_matching_loss": 0.02271660976111889,
      "grad_norm": 0.09237917512655258,
      "learning_rate": 3.180406513597316e-05,
      "loss": 0.0244,
      "relation_loss": 0.0016990436939522623,
      "selected_pairs": 27181.6,
      "state_query_norm": 42.39005661010742,
      "state_relation_loss": 0.00011597512784646823,
      "step": 19130,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0016990436939522623,
      "weighted_relation_loss": 0.0016990436939522623
    },
    {
      "action_query_norm": 40.27458190917969,
      "action_relation_loss": 0.001600894145667553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.904583723105706,
      "flow_matching_loss": 0.02448238655924797,
      "grad_norm": 0.10132019966840744,
      "learning_rate": 3.1752739781961835e-05,
      "loss": 0.0262,
      "relation_loss": 0.001714857609476894,
      "selected_pairs": 27081.5,
      "state_query_norm": 42.386785888671874,
      "state_relation_loss": 0.00011396345726097934,
      "step": 19140,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001714857609476894,
      "weighted_relation_loss": 0.001714857609476894
    },
    {
      "action_query_norm": 40.25190467834473,
      "action_relation_loss": 0.0014916391810402274,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.913938260056128,
      "flow_matching_loss": 0.024730159156024455,
      "grad_norm": 0.1113506481051445,
      "learning_rate": 3.170143660009228e-05,
      "loss": 0.0263,
      "relation_loss": 0.0016011244035325945,
      "selected_pairs": 26319.1,
      "state_query_norm": 42.39087753295898,
      "state_relation_loss": 0.00010948521012323908,
      "step": 19150,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0016011244035325945,
      "weighted_relation_loss": 0.0016011244035325945
    },
    {
      "action_query_norm": 40.25099182128906,
      "action_relation_loss": 0.0015528284129686654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.92329279700655,
      "flow_matching_loss": 0.022995823249220847,
      "grad_norm": 0.0937487930059433,
      "learning_rate": 3.16501556527027e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016645066672936082,
      "selected_pairs": 26629.6,
      "state_query_norm": 42.37118415832519,
      "state_relation_loss": 0.00011167825723532588,
      "step": 19160,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0016645066672936082,
      "weighted_relation_loss": 0.0016645066672936082
    },
    {
      "action_query_norm": 40.246087646484376,
      "action_relation_loss": 0.0015837531187571584,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.932647333956968,
      "flow_matching_loss": 0.023336624167859553,
      "grad_norm": 0.0720672458410263,
      "learning_rate": 3.1598897002104265e-05,
      "loss": 0.025,
      "relation_loss": 0.001695376296993345,
      "selected_pairs": 27526.3,
      "state_query_norm": 42.38106575012207,
      "state_relation_loss": 0.00011162316950503736,
      "step": 19170,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.001695376296993345,
      "weighted_relation_loss": 0.001695376296993345
    },
    {
      "action_query_norm": 40.27793235778809,
      "action_relation_loss": 0.0015632073627784848,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.94200187090739,
      "flow_matching_loss": 0.022970392368733883,
      "grad_norm": 0.10778006911277771,
      "learning_rate": 3.154766071058108e-05,
      "loss": 0.0246,
      "relation_loss": 0.0016779077239334583,
      "selected_pairs": 26483.2,
      "state_query_norm": 42.37139549255371,
      "state_relation_loss": 0.00011470035460661165,
      "step": 19180,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0016779077239334583,
      "weighted_relation_loss": 0.0016779077239334583
    },
    {
      "action_query_norm": 40.2419620513916,
      "action_relation_loss": 0.0015433477237820625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.951356407857812,
      "flow_matching_loss": 0.023081721365451814,
      "grad_norm": 0.09132503718137741,
      "learning_rate": 3.149644684039008e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016560149844735861,
      "selected_pairs": 26896.1,
      "state_query_norm": 42.36880111694336,
      "state_relation_loss": 0.00011266727087786422,
      "step": 19190,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0016560149844735861,
      "weighted_relation_loss": 0.0016560149844735861
    },
    {
      "action_query_norm": 40.25193557739258,
      "action_relation_loss": 0.0016125555615872146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.96071094480823,
      "flow_matching_loss": 0.023730276711285116,
      "grad_norm": 0.10700925439596176,
      "learning_rate": 3.144525545376095e-05,
      "loss": 0.0255,
      "relation_loss": 0.001720377302262932,
      "selected_pairs": 26225.4,
      "state_query_norm": 42.3676872253418,
      "state_relation_loss": 0.00010782172248582355,
      "step": 19200,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.001720377302262932,
      "weighted_relation_loss": 0.001720377302262932
    },
    {
      "action_query_norm": 40.27106399536133,
      "action_relation_loss": 0.0014976613339968026,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.970065481758652,
      "flow_matching_loss": 0.022540049441158772,
      "grad_norm": 0.08484331518411636,
      "learning_rate": 3.139408661289603e-05,
      "loss": 0.0242,
      "relation_loss": 0.001610082236584276,
      "selected_pairs": 26388.2,
      "state_query_norm": 42.36729888916015,
      "state_relation_loss": 0.00011242089749430306,
      "step": 19210,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001610082236584276,
      "weighted_relation_loss": 0.001610082236584276
    },
    {
      "action_query_norm": 40.26975288391113,
      "action_relation_loss": 0.0014739245059899985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.979420018709074,
      "flow_matching_loss": 0.022467042692005633,
      "grad_norm": 0.09929193556308746,
      "learning_rate": 3.134294037997032e-05,
      "loss": 0.0241,
      "relation_loss": 0.0015849710558541118,
      "selected_pairs": 26577.9,
      "state_query_norm": 42.38604164123535,
      "state_relation_loss": 0.00011104656223324128,
      "step": 19220,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015849710558541118,
      "weighted_relation_loss": 0.0015849710558541118
    },
    {
      "action_query_norm": 40.25622444152832,
      "action_relation_loss": 0.0015152983251027763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.988774555659496,
      "flow_matching_loss": 0.022356640361249448,
      "grad_norm": 0.10355456173419952,
      "learning_rate": 3.129181681713127e-05,
      "loss": 0.024,
      "relation_loss": 0.001625392714049667,
      "selected_pairs": 27057.2,
      "state_query_norm": 42.36984405517578,
      "state_relation_loss": 0.00011009438458131626,
      "step": 19230,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001625392714049667,
      "weighted_relation_loss": 0.001625392714049667
    },
    {
      "action_query_norm": 40.230978393554686,
      "action_relation_loss": 0.001542399567551911,
      "auxiliary_loss_weight": 1.0,
      "epoch": 17.998129092609915,
      "flow_matching_loss": 0.024097164161503315,
      "grad_norm": 0.14284147322177887,
      "learning_rate": 3.1240715986498855e-05,
      "loss": 0.0258,
      "relation_loss": 0.0016532306093722583,
      "selected_pairs": 26733.5,
      "state_query_norm": 42.37766532897949,
      "state_relation_loss": 0.00011083103308919817,
      "step": 19240,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0016532306093722583,
      "weighted_relation_loss": 0.0016532306093722583
    },
    {
      "action_query_norm": 40.28232917785645,
      "action_relation_loss": 0.0015223558293655515,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.007483629560337,
      "flow_matching_loss": 0.02347663827240467,
      "grad_norm": 0.12196467816829681,
      "learning_rate": 3.1189637950165394e-05,
      "loss": 0.0251,
      "relation_loss": 0.001633422903250903,
      "selected_pairs": 24955.9,
      "state_query_norm": 42.36165161132813,
      "state_relation_loss": 0.00011106705569545739,
      "step": 19250,
      "valid_action_samples": 217.0,
      "weighted_auxiliary_loss": 0.001633422903250903,
      "weighted_relation_loss": 0.001633422903250903
    },
    {
      "action_query_norm": 40.274153518676755,
      "action_relation_loss": 0.0015397528768517077,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.01683816651076,
      "flow_matching_loss": 0.02305841390043497,
      "grad_norm": 0.11476986855268478,
      "learning_rate": 3.1138582770195544e-05,
      "loss": 0.0247,
      "relation_loss": 0.0016563733224757016,
      "selected_pairs": 27433.0,
      "state_query_norm": 42.37289047241211,
      "state_relation_loss": 0.0001166204565379303,
      "step": 19260,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0016563733224757016,
      "weighted_relation_loss": 0.0016563733224757016
    },
    {
      "action_query_norm": 40.24690208435059,
      "action_relation_loss": 0.001532242854591459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.026192703461177,
      "flow_matching_loss": 0.02412547692656517,
      "grad_norm": 0.09916029125452042,
      "learning_rate": 3.108755050862615e-05,
      "loss": 0.0258,
      "relation_loss": 0.0016369298100471497,
      "selected_pairs": 26895.0,
      "state_query_norm": 42.359401321411134,
      "state_relation_loss": 0.00010468697146279738,
      "step": 19270,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0016369298100471497,
      "weighted_relation_loss": 0.0016369298100471497
    },
    {
      "action_query_norm": 40.293347930908205,
      "action_relation_loss": 0.0014868366066366433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.0355472404116,
      "flow_matching_loss": 0.024838621728122234,
      "grad_norm": 0.2109898328781128,
      "learning_rate": 3.1036541227466204e-05,
      "loss": 0.0264,
      "relation_loss": 0.0015950081986375153,
      "selected_pairs": 27379.5,
      "state_query_norm": 42.36170539855957,
      "state_relation_loss": 0.00010817158836289309,
      "step": 19280,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0015950081986375153,
      "weighted_relation_loss": 0.0015950081986375153
    },
    {
      "action_query_norm": 40.28658828735352,
      "action_relation_loss": 0.001473670091945678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.04490177736202,
      "flow_matching_loss": 0.023456555604934693,
      "grad_norm": 0.13753682374954224,
      "learning_rate": 3.0985554988696786e-05,
      "loss": 0.025,
      "relation_loss": 0.0015871202107518912,
      "selected_pairs": 26622.2,
      "state_query_norm": 42.37668952941895,
      "state_relation_loss": 0.00011345012462697923,
      "step": 19290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015871202107518912,
      "weighted_relation_loss": 0.0015871202107518912
    },
    {
      "action_query_norm": 40.269130325317384,
      "action_relation_loss": 0.0014535770053043962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.054256314312443,
      "flow_matching_loss": 0.023311647772789,
      "grad_norm": 0.11053916066884995,
      "learning_rate": 3.093459185427102e-05,
      "loss": 0.0249,
      "relation_loss": 0.001563346036709845,
      "selected_pairs": 27150.4,
      "state_query_norm": 42.35911598205566,
      "state_relation_loss": 0.00010976903286064044,
      "step": 19300,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001563346036709845,
      "weighted_relation_loss": 0.001563346036709845
    },
    {
      "action_query_norm": 40.249338912963864,
      "action_relation_loss": 0.0014278933173045515,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.06361085126286,
      "flow_matching_loss": 0.022204075008630753,
      "grad_norm": 0.0952620729804039,
      "learning_rate": 3.0883651886113905e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015396466013044118,
      "selected_pairs": 26582.5,
      "state_query_norm": 42.36290969848633,
      "state_relation_loss": 0.00011175328399986029,
      "step": 19310,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015396466013044118,
      "weighted_relation_loss": 0.0015396466013044118
    },
    {
      "action_query_norm": 40.25054359436035,
      "action_relation_loss": 0.0014726275810971856,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.072965388213284,
      "flow_matching_loss": 0.02396849785000086,
      "grad_norm": 0.10490258783102036,
      "learning_rate": 3.0832735146122294e-05,
      "loss": 0.0256,
      "relation_loss": 0.0015844904235564172,
      "selected_pairs": 26733.4,
      "state_query_norm": 42.3456470489502,
      "state_relation_loss": 0.00011186284391442313,
      "step": 19320,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0015844904235564172,
      "weighted_relation_loss": 0.0015844904235564172
    },
    {
      "action_query_norm": 40.277078247070314,
      "action_relation_loss": 0.0014119160710833966,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.082319925163706,
      "flow_matching_loss": 0.023570615611970425,
      "grad_norm": 0.14463014900684357,
      "learning_rate": 3.078184169616485e-05,
      "loss": 0.0251,
      "relation_loss": 0.0015222923830151558,
      "selected_pairs": 27105.8,
      "state_query_norm": 42.361908721923825,
      "state_relation_loss": 0.00011037629301426932,
      "step": 19330,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0015222923830151558,
      "weighted_relation_loss": 0.0015222923830151558
    },
    {
      "action_query_norm": 40.2452033996582,
      "action_relation_loss": 0.0013640694203786552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.091674462114124,
      "flow_matching_loss": 0.02343042064458132,
      "grad_norm": 0.1158759668469429,
      "learning_rate": 3.073097159808187e-05,
      "loss": 0.0249,
      "relation_loss": 0.0014812576700933278,
      "selected_pairs": 26442.9,
      "state_query_norm": 42.341729736328126,
      "state_relation_loss": 0.00011718824316631071,
      "step": 19340,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0014812576700933278,
      "weighted_relation_loss": 0.0014812576700933278
    },
    {
      "action_query_norm": 40.24891471862793,
      "action_relation_loss": 0.0014469833229668438,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.101028999064546,
      "flow_matching_loss": 0.02172483243048191,
      "grad_norm": 0.10164192318916321,
      "learning_rate": 3.068012491368537e-05,
      "loss": 0.0233,
      "relation_loss": 0.0015610081842169165,
      "selected_pairs": 26190.4,
      "state_query_norm": 42.34959526062012,
      "state_relation_loss": 0.00011402486197766847,
      "step": 19350,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0015610081842169165,
      "weighted_relation_loss": 0.0015610081842169165
    },
    {
      "action_query_norm": 40.233786010742186,
      "action_relation_loss": 0.0015178884263150395,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.110383536014968,
      "flow_matching_loss": 0.023220168612897395,
      "grad_norm": 0.09176146239042282,
      "learning_rate": 3.062930170475885e-05,
      "loss": 0.0249,
      "relation_loss": 0.0016344699542969465,
      "selected_pairs": 26709.4,
      "state_query_norm": 42.341840744018555,
      "state_relation_loss": 0.0001165815170679707,
      "step": 19360,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0016344699542969465,
      "weighted_relation_loss": 0.0016344699542969465
    },
    {
      "action_query_norm": 40.3165168762207,
      "action_relation_loss": 0.0015276295132935047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.119738072965387,
      "flow_matching_loss": 0.022684945724904537,
      "grad_norm": 0.13886411488056183,
      "learning_rate": 3.057850203305729e-05,
      "loss": 0.0243,
      "relation_loss": 0.0016434514662250876,
      "selected_pairs": 27129.6,
      "state_query_norm": 42.37299156188965,
      "state_relation_loss": 0.00011582195438677445,
      "step": 19370,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0016434514662250876,
      "weighted_relation_loss": 0.0016434514662250876
    },
    {
      "action_query_norm": 40.275469589233396,
      "action_relation_loss": 0.0014465504791587591,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.12909260991581,
      "flow_matching_loss": 0.02314288802444935,
      "grad_norm": 0.10454373806715012,
      "learning_rate": 3.052772596030708e-05,
      "loss": 0.0247,
      "relation_loss": 0.0015635355841368437,
      "selected_pairs": 26898.7,
      "state_query_norm": 42.35540504455567,
      "state_relation_loss": 0.00011698510788846762,
      "step": 19380,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0015635355841368437,
      "weighted_relation_loss": 0.0015635355841368437
    },
    {
      "action_query_norm": 40.264213180541994,
      "action_relation_loss": 0.001432302256580442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.13844714686623,
      "flow_matching_loss": 0.022325002029538153,
      "grad_norm": 0.0949106365442276,
      "learning_rate": 3.0476973548205943e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015362553531304002,
      "selected_pairs": 26962.1,
      "state_query_norm": 42.369388580322266,
      "state_relation_loss": 0.00010395310528110713,
      "step": 19390,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015362553531304002,
      "weighted_relation_loss": 0.0015362553531304002
    },
    {
      "action_query_norm": 40.267761611938475,
      "action_relation_loss": 0.0014038923312909902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.147801683816652,
      "flow_matching_loss": 0.024194084480404852,
      "grad_norm": 0.09955249726772308,
      "learning_rate": 3.0426244858422847e-05,
      "loss": 0.0257,
      "relation_loss": 0.0015193963306955993,
      "selected_pairs": 26778.1,
      "state_query_norm": 42.33617172241211,
      "state_relation_loss": 0.00011550399285624735,
      "step": 19400,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0015193963306955993,
      "weighted_relation_loss": 0.0015193963306955993
    },
    {
      "action_query_norm": 40.240016555786134,
      "action_relation_loss": 0.0014733352698385715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.15715622076707,
      "flow_matching_loss": 0.022912683337926863,
      "grad_norm": 0.09776927530765533,
      "learning_rate": 3.0375539952597943e-05,
      "loss": 0.0245,
      "relation_loss": 0.00159282092936337,
      "selected_pairs": 26666.9,
      "state_query_norm": 42.33177299499512,
      "state_relation_loss": 0.00011948566534556449,
      "step": 19410,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.00159282092936337,
      "weighted_relation_loss": 0.00159282092936337
    },
    {
      "action_query_norm": 40.230908966064455,
      "action_relation_loss": 0.0014178153360262513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.166510757717493,
      "flow_matching_loss": 0.02096203062683344,
      "grad_norm": 0.1256042718887329,
      "learning_rate": 3.0324858892342468e-05,
      "loss": 0.0225,
      "relation_loss": 0.001526252750772983,
      "selected_pairs": 26387.5,
      "state_query_norm": 42.356920623779295,
      "state_relation_loss": 0.00010843740892596543,
      "step": 19420,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.001526252750772983,
      "weighted_relation_loss": 0.001526252750772983
    },
    {
      "action_query_norm": 40.20894203186035,
      "action_relation_loss": 0.0014794223243370652,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.175865294667915,
      "flow_matching_loss": 0.022594355046749115,
      "grad_norm": 0.09299609065055847,
      "learning_rate": 3.0274201739238672e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015904761618003248,
      "selected_pairs": 26388.4,
      "state_query_norm": 42.32626037597656,
      "state_relation_loss": 0.0001110538447392173,
      "step": 19430,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0015904761618003248,
      "weighted_relation_loss": 0.0015904761618003248
    },
    {
      "action_query_norm": 40.25325088500976,
      "action_relation_loss": 0.0014482675236649812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.185219831618333,
      "flow_matching_loss": 0.021947108767926694,
      "grad_norm": 0.12303607910871506,
      "learning_rate": 3.0223568554839786e-05,
      "loss": 0.0235,
      "relation_loss": 0.001557916891761124,
      "selected_pairs": 27230.7,
      "state_query_norm": 42.3289291381836,
      "state_relation_loss": 0.0001096493775548879,
      "step": 19440,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.001557916891761124,
      "weighted_relation_loss": 0.001557916891761124
    },
    {
      "action_query_norm": 40.20291404724121,
      "action_relation_loss": 0.0014629698474891485,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.194574368568755,
      "flow_matching_loss": 0.021465571969747542,
      "grad_norm": 0.09051072597503662,
      "learning_rate": 3.0172959400669886e-05,
      "loss": 0.023,
      "relation_loss": 0.0015642420388758183,
      "selected_pairs": 26812.1,
      "state_query_norm": 42.34098320007324,
      "state_relation_loss": 0.00010127218483830802,
      "step": 19450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015642420388758183,
      "weighted_relation_loss": 0.0015642420388758183
    },
    {
      "action_query_norm": 40.237128829956056,
      "action_relation_loss": 0.0014184828265570103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.203928905519177,
      "flow_matching_loss": 0.022079170681536198,
      "grad_norm": 0.14473888278007507,
      "learning_rate": 3.0122374338223902e-05,
      "loss": 0.0236,
      "relation_loss": 0.00152448097942397,
      "selected_pairs": 26732.2,
      "state_query_norm": 42.337080764770505,
      "state_relation_loss": 0.00010599815359455534,
      "step": 19460,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.00152448097942397,
      "weighted_relation_loss": 0.00152448097942397
    },
    {
      "action_query_norm": 40.247808837890624,
      "action_relation_loss": 0.0014663128997199238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.2132834424696,
      "flow_matching_loss": 0.023109050653874873,
      "grad_norm": 0.07957242429256439,
      "learning_rate": 3.0071813428967427e-05,
      "loss": 0.0247,
      "relation_loss": 0.001576538966037333,
      "selected_pairs": 26601.3,
      "state_query_norm": 42.32837181091308,
      "state_relation_loss": 0.0001102260357583873,
      "step": 19470,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001576538966037333,
      "weighted_relation_loss": 0.001576538966037333
    },
    {
      "action_query_norm": 40.26794548034668,
      "action_relation_loss": 0.00140790450386703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.222637979420018,
      "flow_matching_loss": 0.0225267942994833,
      "grad_norm": 0.09902271628379822,
      "learning_rate": 3.0021276734336746e-05,
      "loss": 0.024,
      "relation_loss": 0.0015184772899374365,
      "selected_pairs": 26733.7,
      "state_query_norm": 42.35343856811524,
      "state_relation_loss": 0.00011057278097723611,
      "step": 19480,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0015184772899374365,
      "weighted_relation_loss": 0.0015184772899374365
    },
    {
      "action_query_norm": 40.23476066589355,
      "action_relation_loss": 0.0014758514706045389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.23199251637044,
      "flow_matching_loss": 0.021115784533321857,
      "grad_norm": 0.15602998435497284,
      "learning_rate": 2.997076431573871e-05,
      "loss": 0.0227,
      "relation_loss": 0.001585292851086706,
      "selected_pairs": 26814.7,
      "state_query_norm": 42.332675552368165,
      "state_relation_loss": 0.00010944140376523137,
      "step": 19490,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001585292851086706,
      "weighted_relation_loss": 0.001585292851086706
    },
    {
      "action_query_norm": 40.255774307250974,
      "action_relation_loss": 0.0014363498659804464,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.24134705332086,
      "flow_matching_loss": 0.022644514217972755,
      "grad_norm": 0.14498579502105713,
      "learning_rate": 2.9920276234550636e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015479318681173027,
      "selected_pairs": 26666.1,
      "state_query_norm": 42.32988510131836,
      "state_relation_loss": 0.00011158200723002665,
      "step": 19500,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0015479318681173027,
      "weighted_relation_loss": 0.0015479318681173027
    },
    {
      "action_query_norm": 40.23950386047363,
      "action_relation_loss": 0.0014581423834897577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.25070159027128,
      "flow_matching_loss": 0.02404415588825941,
      "grad_norm": 0.11277942359447479,
      "learning_rate": 2.986981255212035e-05,
      "loss": 0.0256,
      "relation_loss": 0.0015647441148757934,
      "selected_pairs": 26322.2,
      "state_query_norm": 42.33323860168457,
      "state_relation_loss": 0.0001066017102857586,
      "step": 19510,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0015647441148757934,
      "weighted_relation_loss": 0.0015647441148757934
    },
    {
      "action_query_norm": 40.196021270751956,
      "action_relation_loss": 0.0014291121857240795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.260056127221702,
      "flow_matching_loss": 0.02474020514637232,
      "grad_norm": 0.11088576912879944,
      "learning_rate": 2.9819373329765977e-05,
      "loss": 0.0263,
      "relation_loss": 0.0015424682176671922,
      "selected_pairs": 26897.4,
      "state_query_norm": 42.32055168151855,
      "state_relation_loss": 0.00011335602102917619,
      "step": 19520,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0015424682176671922,
      "weighted_relation_loss": 0.0015424682176671922
    },
    {
      "action_query_norm": 40.214207077026366,
      "action_relation_loss": 0.0014125767978839577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.269410664172124,
      "flow_matching_loss": 0.02151386011391878,
      "grad_norm": 0.10188472270965576,
      "learning_rate": 2.9768958628775902e-05,
      "loss": 0.023,
      "relation_loss": 0.0015212291502393783,
      "selected_pairs": 26965.3,
      "state_query_norm": 42.33750686645508,
      "state_relation_loss": 0.00010865237054531462,
      "step": 19530,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015212291502393783,
      "weighted_relation_loss": 0.0015212291502393783
    },
    {
      "action_query_norm": 40.23536148071289,
      "action_relation_loss": 0.0014363945461809634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.278765201122546,
      "flow_matching_loss": 0.02100010383874178,
      "grad_norm": 0.11713354289531708,
      "learning_rate": 2.9718568510408762e-05,
      "loss": 0.0225,
      "relation_loss": 0.0015478679561056197,
      "selected_pairs": 26598.2,
      "state_query_norm": 42.33158912658691,
      "state_relation_loss": 0.00011147342083859258,
      "step": 19540,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0015478679561056197,
      "weighted_relation_loss": 0.0015478679561056197
    },
    {
      "action_query_norm": 40.19166717529297,
      "action_relation_loss": 0.0014185336185619235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.288119738072965,
      "flow_matching_loss": 0.023680161871016025,
      "grad_norm": 0.10666441172361374,
      "learning_rate": 2.9668203035893272e-05,
      "loss": 0.0252,
      "relation_loss": 0.0015280532534234226,
      "selected_pairs": 27171.5,
      "state_query_norm": 42.339242935180664,
      "state_relation_loss": 0.00010951963704428636,
      "step": 19550,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0015280532534234226,
      "weighted_relation_loss": 0.0015280532534234226
    },
    {
      "action_query_norm": 40.23141441345215,
      "action_relation_loss": 0.0014733318123035133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.297474275023387,
      "flow_matching_loss": 0.023804616555571555,
      "grad_norm": 0.09996224194765091,
      "learning_rate": 2.9617862266428287e-05,
      "loss": 0.0254,
      "relation_loss": 0.0015845580259338022,
      "selected_pairs": 26642.5,
      "state_query_norm": 42.322950744628905,
      "state_relation_loss": 0.00011122621872345917,
      "step": 19560,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0015845580259338022,
      "weighted_relation_loss": 0.0015845580259338022
    },
    {
      "action_query_norm": 40.23401412963867,
      "action_relation_loss": 0.001493869093246758,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.30682881197381,
      "flow_matching_loss": 0.023483110032975672,
      "grad_norm": 0.11853593587875366,
      "learning_rate": 2.9567546263182556e-05,
      "loss": 0.0251,
      "relation_loss": 0.0016031450824812055,
      "selected_pairs": 27784.2,
      "state_query_norm": 42.317544555664064,
      "state_relation_loss": 0.00010927598559646867,
      "step": 19570,
      "valid_action_samples": 236.2,
      "weighted_auxiliary_loss": 0.0016031450824812055,
      "weighted_relation_loss": 0.0016031450824812055
    },
    {
      "action_query_norm": 40.215043258666995,
      "action_relation_loss": 0.0015319745987653732,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.316183348924227,
      "flow_matching_loss": 0.025013532489538193,
      "grad_norm": 0.11102929711341858,
      "learning_rate": 2.951725508729476e-05,
      "loss": 0.0267,
      "relation_loss": 0.0016499199671670794,
      "selected_pairs": 26463.2,
      "state_query_norm": 42.31522674560547,
      "state_relation_loss": 0.00011794536912930198,
      "step": 19580,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0016499199671670794,
      "weighted_relation_loss": 0.0016499199671670794
    },
    {
      "action_query_norm": 40.18159523010254,
      "action_relation_loss": 0.0015982456272467971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.32553788587465,
      "flow_matching_loss": 0.02247483693063259,
      "grad_norm": 0.09866933524608612,
      "learning_rate": 2.946698879987344e-05,
      "loss": 0.0242,
      "relation_loss": 0.0017114434158429503,
      "selected_pairs": 26761.1,
      "state_query_norm": 42.32282676696777,
      "state_relation_loss": 0.00011319779951008968,
      "step": 19590,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0017114434158429503,
      "weighted_relation_loss": 0.0017114434158429503
    },
    {
      "action_query_norm": 40.274857711791995,
      "action_relation_loss": 0.0014795066905207931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.33489242282507,
      "flow_matching_loss": 0.023293718323111536,
      "grad_norm": 0.10526841133832932,
      "learning_rate": 2.9416747461996853e-05,
      "loss": 0.0249,
      "relation_loss": 0.0015858553582802416,
      "selected_pairs": 26625.5,
      "state_query_norm": 42.33456954956055,
      "state_relation_loss": 0.00010634867358021438,
      "step": 19600,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015858553582802416,
      "weighted_relation_loss": 0.0015858553582802416
    },
    {
      "action_query_norm": 40.24706192016602,
      "action_relation_loss": 0.0014440864673815668,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.34424695977549,
      "flow_matching_loss": 0.022596369311213494,
      "grad_norm": 0.1144082322716713,
      "learning_rate": 2.9366531134712973e-05,
      "loss": 0.0242,
      "relation_loss": 0.0015553115983493627,
      "selected_pairs": 26521.6,
      "state_query_norm": 42.3235668182373,
      "state_relation_loss": 0.0001112251338781789,
      "step": 19610,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0015553115983493627,
      "weighted_relation_loss": 0.0015553115983493627
    },
    {
      "action_query_norm": 40.23445625305176,
      "action_relation_loss": 0.0015150239458307623,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.35360149672591,
      "flow_matching_loss": 0.022065840475261212,
      "grad_norm": 0.09990905225276947,
      "learning_rate": 2.9316339879039367e-05,
      "loss": 0.0237,
      "relation_loss": 0.0016211131820455193,
      "selected_pairs": 26971.1,
      "state_query_norm": 42.32612609863281,
      "state_relation_loss": 0.00010608922893879935,
      "step": 19620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0016211131820455193,
      "weighted_relation_loss": 0.0016211131820455193
    },
    {
      "action_query_norm": 40.1869441986084,
      "action_relation_loss": 0.0015064889099448918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.362956033676333,
      "flow_matching_loss": 0.021745460852980614,
      "grad_norm": 0.10023084282875061,
      "learning_rate": 2.9266173755963167e-05,
      "loss": 0.0234,
      "relation_loss": 0.001617384352721274,
      "selected_pairs": 27152.3,
      "state_query_norm": 42.319195938110354,
      "state_relation_loss": 0.00011089543550042435,
      "step": 19630,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001617384352721274,
      "weighted_relation_loss": 0.001617384352721274
    },
    {
      "action_query_norm": 40.173514938354494,
      "action_relation_loss": 0.001392880198545754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.372310570626755,
      "flow_matching_loss": 0.02172622103244066,
      "grad_norm": 0.07306059449911118,
      "learning_rate": 2.9216032826440926e-05,
      "loss": 0.0232,
      "relation_loss": 0.0014995740610174834,
      "selected_pairs": 26963.2,
      "state_query_norm": 42.31258201599121,
      "state_relation_loss": 0.00010669388793758117,
      "step": 19640,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014995740610174834,
      "weighted_relation_loss": 0.0014995740610174834
    },
    {
      "action_query_norm": 40.25970802307129,
      "action_relation_loss": 0.0014545738929882645,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.381665107577174,
      "flow_matching_loss": 0.02212550602853298,
      "grad_norm": 0.12018819153308868,
      "learning_rate": 2.9165917151398592e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015743713243864476,
      "selected_pairs": 26805.9,
      "state_query_norm": 42.32535362243652,
      "state_relation_loss": 0.00011979743867414072,
      "step": 19650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0015743713243864476,
      "weighted_relation_loss": 0.0015743713243864476
    },
    {
      "action_query_norm": 40.25158920288086,
      "action_relation_loss": 0.0014152468531392515,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.391019644527596,
      "flow_matching_loss": 0.022616379521787168,
      "grad_norm": 0.10915777087211609,
      "learning_rate": 2.9115826791731428e-05,
      "loss": 0.0241,
      "relation_loss": 0.0015280604013241828,
      "selected_pairs": 27093.6,
      "state_query_norm": 42.318664932250975,
      "state_relation_loss": 0.0001128135365433991,
      "step": 19660,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0015280604013241828,
      "weighted_relation_loss": 0.0015280604013241828
    },
    {
      "action_query_norm": 40.23040580749512,
      "action_relation_loss": 0.001447921944782138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.400374181478018,
      "flow_matching_loss": 0.02281076703220606,
      "grad_norm": 0.11435408145189285,
      "learning_rate": 2.906576180830398e-05,
      "loss": 0.0244,
      "relation_loss": 0.001555180933792144,
      "selected_pairs": 26831.8,
      "state_query_norm": 42.32902870178223,
      "state_relation_loss": 0.00010725898464443162,
      "step": 19670,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001555180933792144,
      "weighted_relation_loss": 0.001555180933792144
    },
    {
      "action_query_norm": 40.19349136352539,
      "action_relation_loss": 0.0014446111745201051,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.409728718428436,
      "flow_matching_loss": 0.021688717231154443,
      "grad_norm": 0.13231347501277924,
      "learning_rate": 2.9015722261949917e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015558964922092855,
      "selected_pairs": 26712.5,
      "state_query_norm": 42.29497947692871,
      "state_relation_loss": 0.00011128529658890329,
      "step": 19680,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0015558964922092855,
      "weighted_relation_loss": 0.0015558964922092855
    },
    {
      "action_query_norm": 40.15006790161133,
      "action_relation_loss": 0.001492735731881112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.419083255378858,
      "flow_matching_loss": 0.02293106559664011,
      "grad_norm": 0.11238273978233337,
      "learning_rate": 2.8965708213471986e-05,
      "loss": 0.0245,
      "relation_loss": 0.0015968322637490927,
      "selected_pairs": 26984.7,
      "state_query_norm": 42.31180839538574,
      "state_relation_loss": 0.00010409651513327844,
      "step": 19690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0015968322637490927,
      "weighted_relation_loss": 0.0015968322637490927
    },
    {
      "action_query_norm": 40.219357299804685,
      "action_relation_loss": 0.001491038640961051,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.42843779232928,
      "flow_matching_loss": 0.022171657904982565,
      "grad_norm": 0.10534101724624634,
      "learning_rate": 2.8915719723641975e-05,
      "loss": 0.0238,
      "relation_loss": 0.0016012364299967884,
      "selected_pairs": 26802.2,
      "state_query_norm": 42.298892211914065,
      "state_relation_loss": 0.00011019779267371633,
      "step": 19700,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0016012364299967884,
      "weighted_relation_loss": 0.0016012364299967884
    },
    {
      "action_query_norm": 40.24220161437988,
      "action_relation_loss": 0.0014627736527472734,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.437792329279702,
      "flow_matching_loss": 0.021983214560896157,
      "grad_norm": 0.1044926568865776,
      "learning_rate": 2.8865756853200604e-05,
      "loss": 0.0236,
      "relation_loss": 0.0015699570882134139,
      "selected_pairs": 26972.9,
      "state_query_norm": 42.32890663146973,
      "state_relation_loss": 0.00010718342164182105,
      "step": 19710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015699570882134139,
      "weighted_relation_loss": 0.0015699570882134139
    },
    {
      "action_query_norm": 40.214341735839845,
      "action_relation_loss": 0.0014227888779714704,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.44714686623012,
      "flow_matching_loss": 0.021524359285831452,
      "grad_norm": 0.11158021539449692,
      "learning_rate": 2.8815819662857503e-05,
      "loss": 0.0231,
      "relation_loss": 0.001536216272506863,
      "selected_pairs": 26343.1,
      "state_query_norm": 42.3193962097168,
      "state_relation_loss": 0.00011342738944222219,
      "step": 19720,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.001536216272506863,
      "weighted_relation_loss": 0.001536216272506863
    },
    {
      "action_query_norm": 40.190970993041994,
      "action_relation_loss": 0.0014741163584403693,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.456501403180543,
      "flow_matching_loss": 0.02400400061160326,
      "grad_norm": 0.18459899723529816,
      "learning_rate": 2.876590821329105e-05,
      "loss": 0.0256,
      "relation_loss": 0.001578533835709095,
      "selected_pairs": 27195.8,
      "state_query_norm": 42.329497528076175,
      "state_relation_loss": 0.00010441746562719345,
      "step": 19730,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001578533835709095,
      "weighted_relation_loss": 0.001578533835709095
    },
    {
      "action_query_norm": 40.193468475341795,
      "action_relation_loss": 0.0014551465632393957,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.465855940130965,
      "flow_matching_loss": 0.02299117371439934,
      "grad_norm": 0.11361082643270493,
      "learning_rate": 2.871602256514836e-05,
      "loss": 0.0245,
      "relation_loss": 0.0015582478721626104,
      "selected_pairs": 26966.2,
      "state_query_norm": 42.32767868041992,
      "state_relation_loss": 0.00010310128200217151,
      "step": 19740,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0015582478721626104,
      "weighted_relation_loss": 0.0015582478721626104
    },
    {
      "action_query_norm": 40.171755981445315,
      "action_relation_loss": 0.001482090970966965,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.475210477081383,
      "flow_matching_loss": 0.021045322343707085,
      "grad_norm": 0.13597995042800903,
      "learning_rate": 2.8666162779045203e-05,
      "loss": 0.0226,
      "relation_loss": 0.0015961339580826462,
      "selected_pairs": 27036.5,
      "state_query_norm": 42.320352172851564,
      "state_relation_loss": 0.00011404298493289389,
      "step": 19750,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0015961339580826462,
      "weighted_relation_loss": 0.0015961339580826462
    },
    {
      "action_query_norm": 40.14235992431641,
      "action_relation_loss": 0.001441888464614749,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.484565014031805,
      "flow_matching_loss": 0.02237282693386078,
      "grad_norm": 0.10715832561254501,
      "learning_rate": 2.8616328915565904e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015450815204530953,
      "selected_pairs": 27267.6,
      "state_query_norm": 42.3073371887207,
      "state_relation_loss": 0.0001031930383760482,
      "step": 19760,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0015450815204530953,
      "weighted_relation_loss": 0.0015450815204530953
    },
    {
      "action_query_norm": 40.21789894104004,
      "action_relation_loss": 0.001397505064960569,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.493919550982227,
      "flow_matching_loss": 0.023276141099631786,
      "grad_norm": 0.09636351466178894,
      "learning_rate": 2.856652103526334e-05,
      "loss": 0.0248,
      "relation_loss": 0.0015023638145066798,
      "selected_pairs": 26648.7,
      "state_query_norm": 42.3125560760498,
      "state_relation_loss": 0.00010485873572179117,
      "step": 19770,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0015023638145066798,
      "weighted_relation_loss": 0.0015023638145066798
    },
    {
      "action_query_norm": 40.18775596618652,
      "action_relation_loss": 0.0013839578256011008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.50327408793265,
      "flow_matching_loss": 0.02197655849158764,
      "grad_norm": 0.10930989682674408,
      "learning_rate": 2.8516739198658755e-05,
      "loss": 0.0235,
      "relation_loss": 0.0014943378861062228,
      "selected_pairs": 27010.2,
      "state_query_norm": 42.30621185302734,
      "state_relation_loss": 0.00011038004740839824,
      "step": 19780,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014943378861062228,
      "weighted_relation_loss": 0.0014943378861062228
    },
    {
      "action_query_norm": 40.241829299926756,
      "action_relation_loss": 0.0014979080529883503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.512628624883067,
      "flow_matching_loss": 0.02175692468881607,
      "grad_norm": 0.09729553014039993,
      "learning_rate": 2.846698346624177e-05,
      "loss": 0.0234,
      "relation_loss": 0.0016043948009610177,
      "selected_pairs": 27197.8,
      "state_query_norm": 42.30910797119141,
      "state_relation_loss": 0.00010648672323441133,
      "step": 19790,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0016043948009610177,
      "weighted_relation_loss": 0.0016043948009610177
    },
    {
      "action_query_norm": 40.17873687744141,
      "action_relation_loss": 0.0013912546797655522,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.52198316183349,
      "flow_matching_loss": 0.02003807481378317,
      "grad_norm": 0.1308334618806839,
      "learning_rate": 2.841725389847032e-05,
      "loss": 0.0215,
      "relation_loss": 0.0014981841784901917,
      "selected_pairs": 27264.6,
      "state_query_norm": 42.29880828857422,
      "state_relation_loss": 0.00010692951254895889,
      "step": 19800,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0014981841784901917,
      "weighted_relation_loss": 0.0014981841784901917
    },
    {
      "action_query_norm": 40.23189468383789,
      "action_relation_loss": 0.001361657155212015,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.53133769878391,
      "flow_matching_loss": 0.02260552179068327,
      "grad_norm": 0.0928204283118248,
      "learning_rate": 2.8367550555770506e-05,
      "loss": 0.0241,
      "relation_loss": 0.001469384238589555,
      "selected_pairs": 26622.3,
      "state_query_norm": 42.300955581665036,
      "state_relation_loss": 0.00010772707755677403,
      "step": 19810,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.001469384238589555,
      "weighted_relation_loss": 0.001469384238589555
    },
    {
      "action_query_norm": 40.215049743652344,
      "action_relation_loss": 0.0013980006449855864,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.54069223573433,
      "flow_matching_loss": 0.021228431910276412,
      "grad_norm": 0.11500871181488037,
      "learning_rate": 2.831787349853655e-05,
      "loss": 0.0227,
      "relation_loss": 0.0015025999513454736,
      "selected_pairs": 27105.9,
      "state_query_norm": 42.29744415283203,
      "state_relation_loss": 0.0001045993107254617,
      "step": 19820,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0015025999513454736,
      "weighted_relation_loss": 0.0015025999513454736
    },
    {
      "action_query_norm": 40.139645385742185,
      "action_relation_loss": 0.0013996282243169845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.55004677268475,
      "flow_matching_loss": 0.02126796655356884,
      "grad_norm": 0.11528988927602768,
      "learning_rate": 2.8268222787130806e-05,
      "loss": 0.0228,
      "relation_loss": 0.0015047965222038328,
      "selected_pairs": 27079.4,
      "state_query_norm": 42.297998046875,
      "state_relation_loss": 0.00010516829643165692,
      "step": 19830,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0015047965222038328,
      "weighted_relation_loss": 0.0015047965222038328
    },
    {
      "action_query_norm": 40.23937911987305,
      "action_relation_loss": 0.001451208849903196,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.559401309635174,
      "flow_matching_loss": 0.022936986573040486,
      "grad_norm": 0.10377798974514008,
      "learning_rate": 2.821859848188355e-05,
      "loss": 0.0245,
      "relation_loss": 0.0015552338212728501,
      "selected_pairs": 26827.1,
      "state_query_norm": 42.308126068115236,
      "state_relation_loss": 0.00010402495463495143,
      "step": 19840,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0015552338212728501,
      "weighted_relation_loss": 0.0015552338212728501
    },
    {
      "action_query_norm": 40.17475547790527,
      "action_relation_loss": 0.0014712926698848606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.568755846585596,
      "flow_matching_loss": 0.022344865277409552,
      "grad_norm": 0.11079680919647217,
      "learning_rate": 2.816900064309299e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015774690895341338,
      "selected_pairs": 26459.3,
      "state_query_norm": 42.29661178588867,
      "state_relation_loss": 0.00010617642910801806,
      "step": 19850,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0015774690895341338,
      "weighted_relation_loss": 0.0015774690895341338
    },
    {
      "action_query_norm": 40.19297752380371,
      "action_relation_loss": 0.0014327313168905676,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.578110383536014,
      "flow_matching_loss": 0.024136600270867347,
      "grad_norm": 0.11143527179956436,
      "learning_rate": 2.811942933102517e-05,
      "loss": 0.0257,
      "relation_loss": 0.0015388657804578543,
      "selected_pairs": 26871.9,
      "state_query_norm": 42.3041805267334,
      "state_relation_loss": 0.00010613445701892487,
      "step": 19860,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015388657804578543,
      "weighted_relation_loss": 0.0015388657804578543
    },
    {
      "action_query_norm": 40.17347869873047,
      "action_relation_loss": 0.0013390114647336305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.587464920486436,
      "flow_matching_loss": 0.021635048277676105,
      "grad_norm": 0.18259915709495544,
      "learning_rate": 2.806988460591391e-05,
      "loss": 0.0231,
      "relation_loss": 0.001445460133254528,
      "selected_pairs": 26916.6,
      "state_query_norm": 42.305825805664064,
      "state_relation_loss": 0.00010644866852089762,
      "step": 19870,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001445460133254528,
      "weighted_relation_loss": 0.001445460133254528
    },
    {
      "action_query_norm": 40.213742065429685,
      "action_relation_loss": 0.0014344363007694483,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.596819457436858,
      "flow_matching_loss": 0.02288129646331072,
      "grad_norm": 0.10362064093351364,
      "learning_rate": 2.802036652796074e-05,
      "loss": 0.0244,
      "relation_loss": 0.00154242857825011,
      "selected_pairs": 27410.0,
      "state_query_norm": 42.300581741333005,
      "state_relation_loss": 0.00010799228693940677,
      "step": 19880,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.00154242857825011,
      "weighted_relation_loss": 0.00154242857825011
    },
    {
      "action_query_norm": 40.16099128723145,
      "action_relation_loss": 0.001442869147285819,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.606173994387277,
      "flow_matching_loss": 0.021729058213531972,
      "grad_norm": 0.0964299663901329,
      "learning_rate": 2.797087515733478e-05,
      "loss": 0.0233,
      "relation_loss": 0.001549803069792688,
      "selected_pairs": 26227.0,
      "state_query_norm": 42.305077362060544,
      "state_relation_loss": 0.00010693392614484765,
      "step": 19890,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.001549803069792688,
      "weighted_relation_loss": 0.001549803069792688
    },
    {
      "action_query_norm": 40.19576759338379,
      "action_relation_loss": 0.001426228869240731,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.6155285313377,
      "flow_matching_loss": 0.022340264916419984,
      "grad_norm": 0.11755610257387161,
      "learning_rate": 2.7921410554172723e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015324912499636411,
      "selected_pairs": 27180.8,
      "state_query_norm": 42.31032104492188,
      "state_relation_loss": 0.00010626238799886778,
      "step": 19900,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0015324912499636411,
      "weighted_relation_loss": 0.0015324912499636411
    },
    {
      "action_query_norm": 40.19167366027832,
      "action_relation_loss": 0.0013954953523352743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.62488306828812,
      "flow_matching_loss": 0.023710612021386624,
      "grad_norm": 0.10179456323385239,
      "learning_rate": 2.787197277857871e-05,
      "loss": 0.0252,
      "relation_loss": 0.001505012798588723,
      "selected_pairs": 26676.2,
      "state_query_norm": 42.295723342895506,
      "state_relation_loss": 0.00010951744043268264,
      "step": 19910,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001505012798588723,
      "weighted_relation_loss": 0.001505012798588723
    },
    {
      "action_query_norm": 40.16843757629395,
      "action_relation_loss": 0.0014095151913352312,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.63423760523854,
      "flow_matching_loss": 0.02224658094346523,
      "grad_norm": 0.12017007917165756,
      "learning_rate": 2.782256189062429e-05,
      "loss": 0.0238,
      "relation_loss": 0.0015108581283129752,
      "selected_pairs": 26990.8,
      "state_query_norm": 42.2983757019043,
      "state_relation_loss": 0.00010134295516763813,
      "step": 19920,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0015108581283129752,
      "weighted_relation_loss": 0.0015108581283129752
    },
    {
      "action_query_norm": 40.15366439819336,
      "action_relation_loss": 0.001394578954204917,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.64359214218896,
      "flow_matching_loss": 0.022167469561100005,
      "grad_norm": 0.08930999040603638,
      "learning_rate": 2.777317795034839e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015006502275355161,
      "selected_pairs": 27018.5,
      "state_query_norm": 42.291360473632814,
      "state_relation_loss": 0.00010607127769617364,
      "step": 19930,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0015006502275355161,
      "weighted_relation_loss": 0.0015006502275355161
    },
    {
      "action_query_norm": 40.16138801574707,
      "action_relation_loss": 0.0012539268587715923,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.652946679139383,
      "flow_matching_loss": 0.022335598804056646,
      "grad_norm": 0.09302274882793427,
      "learning_rate": 2.7723821017757112e-05,
      "loss": 0.0237,
      "relation_loss": 0.0013574561686255037,
      "selected_pairs": 27455.1,
      "state_query_norm": 42.30043182373047,
      "state_relation_loss": 0.00010352931058150716,
      "step": 19940,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0013574561686255037,
      "weighted_relation_loss": 0.0013574561686255037
    },
    {
      "action_query_norm": 40.19191360473633,
      "action_relation_loss": 0.0013588403817266226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.662301216089805,
      "flow_matching_loss": 0.023149785958230494,
      "grad_norm": 0.12698088586330414,
      "learning_rate": 2.7674491152823822e-05,
      "loss": 0.0246,
      "relation_loss": 0.0014639145229011774,
      "selected_pairs": 26758.1,
      "state_query_norm": 42.28712005615235,
      "state_relation_loss": 0.00010507412225706503,
      "step": 19950,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0014639145229011774,
      "weighted_relation_loss": 0.0014639145229011774
    },
    {
      "action_query_norm": 40.15682106018066,
      "action_relation_loss": 0.0014018983347341418,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.671655753040223,
      "flow_matching_loss": 0.02081078588962555,
      "grad_norm": 0.10113321989774704,
      "learning_rate": 2.7625188415488944e-05,
      "loss": 0.0223,
      "relation_loss": 0.001504420896526426,
      "selected_pairs": 27031.6,
      "state_query_norm": 42.281611633300784,
      "state_relation_loss": 0.00010252256106468849,
      "step": 19960,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001504420896526426,
      "weighted_relation_loss": 0.001504420896526426
    },
    {
      "action_query_norm": 40.17963829040527,
      "action_relation_loss": 0.0015116937458515168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.681010289990645,
      "flow_matching_loss": 0.022758184000849724,
      "grad_norm": 0.08920241892337799,
      "learning_rate": 2.7575912865659924e-05,
      "loss": 0.0244,
      "relation_loss": 0.0016148727852851152,
      "selected_pairs": 26490.1,
      "state_query_norm": 42.29311637878418,
      "state_relation_loss": 0.00010317905325791798,
      "step": 19970,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0016148727852851152,
      "weighted_relation_loss": 0.0016148727852851152
    },
    {
      "action_query_norm": 40.18570404052734,
      "action_relation_loss": 0.0014567383332177997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.690364826941067,
      "flow_matching_loss": 0.023466719314455986,
      "grad_norm": 0.12348461151123047,
      "learning_rate": 2.7526664563211245e-05,
      "loss": 0.025,
      "relation_loss": 0.0015641705133020878,
      "selected_pairs": 26576.1,
      "state_query_norm": 42.27484130859375,
      "state_relation_loss": 0.00010743217062554322,
      "step": 19980,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015641705133020878,
      "weighted_relation_loss": 0.0015641705133020878
    },
    {
      "action_query_norm": 40.137865447998045,
      "action_relation_loss": 0.0013721631374210119,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.699719363891486,
      "flow_matching_loss": 0.022713158279657364,
      "grad_norm": 0.09993593394756317,
      "learning_rate": 2.7477443567984224e-05,
      "loss": 0.0242,
      "relation_loss": 0.00147911497624591,
      "selected_pairs": 26978.5,
      "state_query_norm": 42.286970138549805,
      "state_relation_loss": 0.00010695184100768529,
      "step": 19990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.00147911497624591,
      "weighted_relation_loss": 0.00147911497624591
    },
    {
      "action_query_norm": 40.15261688232422,
      "action_relation_loss": 0.001450015907175839,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.709073900841908,
      "flow_matching_loss": 0.022411906346678733,
      "grad_norm": 0.10508792847394943,
      "learning_rate": 2.7428249939787e-05,
      "loss": 0.024,
      "relation_loss": 0.001555163215380162,
      "selected_pairs": 27381.1,
      "state_query_norm": 42.26598205566406,
      "state_relation_loss": 0.00010514730311115273,
      "step": 20000,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.001555163215380162,
      "weighted_relation_loss": 0.001555163215380162
    },
    {
      "action_query_norm": 40.15555419921875,
      "action_relation_loss": 0.001452616904862225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.71842843779233,
      "flow_matching_loss": 0.02149305585771799,
      "grad_norm": 0.1062445119023323,
      "learning_rate": 2.7379083738394484e-05,
      "loss": 0.0231,
      "relation_loss": 0.0015657490119338035,
      "selected_pairs": 27512.7,
      "state_query_norm": 42.28154335021973,
      "state_relation_loss": 0.00011313211070955731,
      "step": 20010,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0015657490119338035,
      "weighted_relation_loss": 0.0015657490119338035
    },
    {
      "action_query_norm": 40.10002250671387,
      "action_relation_loss": 0.0014102127053774894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.72778297474275,
      "flow_matching_loss": 0.023167971707880496,
      "grad_norm": 0.1995500773191452,
      "learning_rate": 2.732994502354823e-05,
      "loss": 0.0247,
      "relation_loss": 0.001518434623721987,
      "selected_pairs": 26662.0,
      "state_query_norm": 42.26899871826172,
      "state_relation_loss": 0.00010822191688930616,
      "step": 20020,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001518434623721987,
      "weighted_relation_loss": 0.001518434623721987
    },
    {
      "action_query_norm": 40.11453704833984,
      "action_relation_loss": 0.001409737765789032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.73713751169317,
      "flow_matching_loss": 0.022836418636143208,
      "grad_norm": 0.10742856562137604,
      "learning_rate": 2.72808338549564e-05,
      "loss": 0.0244,
      "relation_loss": 0.0015174963395111264,
      "selected_pairs": 26573.9,
      "state_query_norm": 42.27560691833496,
      "state_relation_loss": 0.0001077585606253706,
      "step": 20030,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0015174963395111264,
      "weighted_relation_loss": 0.0015174963395111264
    },
    {
      "action_query_norm": 40.13905906677246,
      "action_relation_loss": 0.001422806503251195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.746492048643592,
      "flow_matching_loss": 0.021055749990046023,
      "grad_norm": 0.1064237430691719,
      "learning_rate": 2.723175029229374e-05,
      "loss": 0.0226,
      "relation_loss": 0.00153336786897853,
      "selected_pairs": 26844.8,
      "state_query_norm": 42.27383460998535,
      "state_relation_loss": 0.00011056135117541999,
      "step": 20040,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.00153336786897853,
      "weighted_relation_loss": 0.00153336786897853
    },
    {
      "action_query_norm": 40.116868591308595,
      "action_relation_loss": 0.0014542387682013214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.755846585594014,
      "flow_matching_loss": 0.02148568965494633,
      "grad_norm": 0.12172218412160873,
      "learning_rate": 2.718269439520138e-05,
      "loss": 0.023,
      "relation_loss": 0.0015579711878672242,
      "selected_pairs": 26881.7,
      "state_query_norm": 42.2859432220459,
      "state_relation_loss": 0.00010373241602792404,
      "step": 20050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015579711878672242,
      "weighted_relation_loss": 0.0015579711878672242
    },
    {
      "action_query_norm": 40.12643203735352,
      "action_relation_loss": 0.0013994082808494567,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.765201122544433,
      "flow_matching_loss": 0.02181091867387295,
      "grad_norm": 0.16729846596717834,
      "learning_rate": 2.7133666223286858e-05,
      "loss": 0.0233,
      "relation_loss": 0.001504699920769781,
      "selected_pairs": 26004.6,
      "state_query_norm": 42.27265243530273,
      "state_relation_loss": 0.00010529162973398343,
      "step": 20060,
      "valid_action_samples": 228.5,
      "weighted_auxiliary_loss": 0.001504699920769781,
      "weighted_relation_loss": 0.001504699920769781
    },
    {
      "action_query_norm": 40.07857475280762,
      "action_relation_loss": 0.0014751679147593677,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.774555659494855,
      "flow_matching_loss": 0.02161341644823551,
      "grad_norm": 0.1748550832271576,
      "learning_rate": 2.7084665836124006e-05,
      "loss": 0.0232,
      "relation_loss": 0.001583347818814218,
      "selected_pairs": 27195.3,
      "state_query_norm": 42.27437629699707,
      "state_relation_loss": 0.00010817991569638253,
      "step": 20070,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001583347818814218,
      "weighted_relation_loss": 0.001583347818814218
    },
    {
      "action_query_norm": 40.104119873046876,
      "action_relation_loss": 0.0013819531071931125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.783910196445277,
      "flow_matching_loss": 0.020879419706761836,
      "grad_norm": 0.10430143028497696,
      "learning_rate": 2.703569329325296e-05,
      "loss": 0.0224,
      "relation_loss": 0.0014892773353494704,
      "selected_pairs": 27321.4,
      "state_query_norm": 42.26269721984863,
      "state_relation_loss": 0.00010732422597357072,
      "step": 20080,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0014892773353494704,
      "weighted_relation_loss": 0.0014892773353494704
    },
    {
      "action_query_norm": 40.13089141845703,
      "action_relation_loss": 0.0014144038199447096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.7932647333957,
      "flow_matching_loss": 0.02112663835287094,
      "grad_norm": 0.130309596657753,
      "learning_rate": 2.698674865417994e-05,
      "loss": 0.0226,
      "relation_loss": 0.0015131123713217677,
      "selected_pairs": 27007.5,
      "state_query_norm": 42.264983367919925,
      "state_relation_loss": 9.870857247733511e-05,
      "step": 20090,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0015131123713217677,
      "weighted_relation_loss": 0.0015131123713217677
    },
    {
      "action_query_norm": 40.16416931152344,
      "action_relation_loss": 0.001373257499653846,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.802619270346117,
      "flow_matching_loss": 0.02218882329761982,
      "grad_norm": 0.12145774811506271,
      "learning_rate": 2.6937831978377325e-05,
      "loss": 0.0237,
      "relation_loss": 0.0014870094251818954,
      "selected_pairs": 26550.1,
      "state_query_norm": 42.276970291137694,
      "state_relation_loss": 0.00011375191825209185,
      "step": 20100,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0014870094251818954,
      "weighted_relation_loss": 0.0014870094251818954
    },
    {
      "action_query_norm": 40.084841918945315,
      "action_relation_loss": 0.0014853429398499428,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.81197380729654,
      "flow_matching_loss": 0.02225979771465063,
      "grad_norm": 0.10968735814094543,
      "learning_rate": 2.6888943325283485e-05,
      "loss": 0.0239,
      "relation_loss": 0.0015916854608803988,
      "selected_pairs": 26858.8,
      "state_query_norm": 42.26235008239746,
      "state_relation_loss": 0.00010634252103045583,
      "step": 20110,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0015916854608803988,
      "weighted_relation_loss": 0.0015916854608803988
    },
    {
      "action_query_norm": 40.069588851928714,
      "action_relation_loss": 0.0014146480127237738,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.82132834424696,
      "flow_matching_loss": 0.023593760654330255,
      "grad_norm": 0.10151766985654831,
      "learning_rate": 2.6840082754302733e-05,
      "loss": 0.0251,
      "relation_loss": 0.0015210281126201153,
      "selected_pairs": 26878.1,
      "state_query_norm": 42.27841033935547,
      "state_relation_loss": 0.00010638010062393732,
      "step": 20120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0015210281126201153,
      "weighted_relation_loss": 0.0015210281126201153
    },
    {
      "action_query_norm": 40.0927619934082,
      "action_relation_loss": 0.0014717181795276702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.83068288119738,
      "flow_matching_loss": 0.02242367938160896,
      "grad_norm": 0.11481311172246933,
      "learning_rate": 2.679125032480525e-05,
      "loss": 0.024,
      "relation_loss": 0.0015692568849772215,
      "selected_pairs": 26710.1,
      "state_query_norm": 42.25675582885742,
      "state_relation_loss": 9.753870544955134e-05,
      "step": 20130,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0015692568849772215,
      "weighted_relation_loss": 0.0015692568849772215
    },
    {
      "action_query_norm": 40.152950286865234,
      "action_relation_loss": 0.0013421091600321234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.8400374181478,
      "flow_matching_loss": 0.021265413984656333,
      "grad_norm": 0.11071443557739258,
      "learning_rate": 2.674244609612708e-05,
      "loss": 0.0227,
      "relation_loss": 0.001448211434762925,
      "selected_pairs": 27152.8,
      "state_query_norm": 42.2558479309082,
      "state_relation_loss": 0.00010610226454446092,
      "step": 20140,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001448211434762925,
      "weighted_relation_loss": 0.001448211434762925
    },
    {
      "action_query_norm": 40.13160171508789,
      "action_relation_loss": 0.0013722726260311901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.849391955098223,
      "flow_matching_loss": 0.02019601073116064,
      "grad_norm": 0.10731911659240723,
      "learning_rate": 2.6693670127569958e-05,
      "loss": 0.0217,
      "relation_loss": 0.0014769668108783663,
      "selected_pairs": 27010.9,
      "state_query_norm": 42.254961013793945,
      "state_relation_loss": 0.00010469417611602694,
      "step": 20150,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014769668108783663,
      "weighted_relation_loss": 0.0014769668108783663
    },
    {
      "action_query_norm": 40.041290283203125,
      "action_relation_loss": 0.0014533112524077296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.858746492048645,
      "flow_matching_loss": 0.020627006143331527,
      "grad_norm": 0.09477534145116806,
      "learning_rate": 2.6644922478401268e-05,
      "loss": 0.0222,
      "relation_loss": 0.001554017118178308,
      "selected_pairs": 26413.8,
      "state_query_norm": 42.263502502441405,
      "state_relation_loss": 0.00010070587741211057,
      "step": 20160,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001554017118178308,
      "weighted_relation_loss": 0.001554017118178308
    },
    {
      "action_query_norm": 40.11244125366211,
      "action_relation_loss": 0.0013245392125099897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.868101028999064,
      "flow_matching_loss": 0.02257152795791626,
      "grad_norm": 0.10669118911027908,
      "learning_rate": 2.6596203207854004e-05,
      "loss": 0.024,
      "relation_loss": 0.0014296145527623594,
      "selected_pairs": 26964.3,
      "state_query_norm": 42.25798454284668,
      "state_relation_loss": 0.00010507532497285865,
      "step": 20170,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014296145527623594,
      "weighted_relation_loss": 0.0014296145527623594
    },
    {
      "action_query_norm": 40.10068054199219,
      "action_relation_loss": 0.0014149607508443297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.877455565949486,
      "flow_matching_loss": 0.02189452573657036,
      "grad_norm": 0.08171947300434113,
      "learning_rate": 2.654751237512666e-05,
      "loss": 0.0234,
      "relation_loss": 0.0015221539186313748,
      "selected_pairs": 26754.7,
      "state_query_norm": 42.25450820922852,
      "state_relation_loss": 0.0001071931648766622,
      "step": 20180,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0015221539186313748,
      "weighted_relation_loss": 0.0015221539186313748
    },
    {
      "action_query_norm": 40.120551681518556,
      "action_relation_loss": 0.0013959642150439323,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.886810102899908,
      "flow_matching_loss": 0.022762027196586133,
      "grad_norm": 0.144020676612854,
      "learning_rate": 2.649885003938323e-05,
      "loss": 0.0243,
      "relation_loss": 0.0014952717698179185,
      "selected_pairs": 27019.8,
      "state_query_norm": 42.2592056274414,
      "state_relation_loss": 9.930754531524144e-05,
      "step": 20190,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0014952717698179185,
      "weighted_relation_loss": 0.0014952717698179185
    },
    {
      "action_query_norm": 40.17403221130371,
      "action_relation_loss": 0.0013984489138238131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.896164639850326,
      "flow_matching_loss": 0.020900349877774716,
      "grad_norm": 0.1182369738817215,
      "learning_rate": 2.6450216259753e-05,
      "loss": 0.0224,
      "relation_loss": 0.0015027816989459097,
      "selected_pairs": 26439.1,
      "state_query_norm": 42.26953773498535,
      "state_relation_loss": 0.00010433278148411774,
      "step": 20200,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0015027816989459097,
      "weighted_relation_loss": 0.0015027816989459097
    },
    {
      "action_query_norm": 40.123344039916994,
      "action_relation_loss": 0.0013338398654013871,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.90551917680075,
      "flow_matching_loss": 0.020506368763744832,
      "grad_norm": 0.09305999428033829,
      "learning_rate": 2.640161109533063e-05,
      "loss": 0.0219,
      "relation_loss": 0.0014381949789822102,
      "selected_pairs": 27197.3,
      "state_query_norm": 42.265494918823244,
      "state_relation_loss": 0.00010435510193929076,
      "step": 20210,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0014381949789822102,
      "weighted_relation_loss": 0.0014381949789822102
    },
    {
      "action_query_norm": 40.111519241333006,
      "action_relation_loss": 0.0013514059013687075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.91487371375117,
      "flow_matching_loss": 0.02269350681453943,
      "grad_norm": 0.12915846705436707,
      "learning_rate": 2.6353034605175936e-05,
      "loss": 0.0241,
      "relation_loss": 0.0014545751851983368,
      "selected_pairs": 26914.6,
      "state_query_norm": 42.26505966186524,
      "state_relation_loss": 0.00010316927582607605,
      "step": 20220,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0014545751851983368,
      "weighted_relation_loss": 0.0014545751851983368
    },
    {
      "action_query_norm": 40.114126205444336,
      "action_relation_loss": 0.001385075063444674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.92422825070159,
      "flow_matching_loss": 0.025051955506205557,
      "grad_norm": 0.10815435647964478,
      "learning_rate": 2.630448684831398e-05,
      "loss": 0.0265,
      "relation_loss": 0.0014857450732961297,
      "selected_pairs": 26709.7,
      "state_query_norm": 42.25985221862793,
      "state_relation_loss": 0.0001006700040306896,
      "step": 20230,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0014857450732961297,
      "weighted_relation_loss": 0.0014857450732961297
    },
    {
      "action_query_norm": 40.05201377868652,
      "action_relation_loss": 0.0013985194847919047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.93358278765201,
      "flow_matching_loss": 0.023894476518034936,
      "grad_norm": 0.15166614949703217,
      "learning_rate": 2.625596788373482e-05,
      "loss": 0.0254,
      "relation_loss": 0.0014951360295526682,
      "selected_pairs": 27075.3,
      "state_query_norm": 42.260480117797854,
      "state_relation_loss": 9.66165382124018e-05,
      "step": 20240,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0014951360295526682,
      "weighted_relation_loss": 0.0014951360295526682
    },
    {
      "action_query_norm": 40.15405769348145,
      "action_relation_loss": 0.0013797161402180791,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.942937324602433,
      "flow_matching_loss": 0.022951208241283893,
      "grad_norm": 0.10329631716012955,
      "learning_rate": 2.620747777039363e-05,
      "loss": 0.0244,
      "relation_loss": 0.001478637068066746,
      "selected_pairs": 26990.8,
      "state_query_norm": 42.252191162109376,
      "state_relation_loss": 9.892092202790082e-05,
      "step": 20250,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001478637068066746,
      "weighted_relation_loss": 0.001478637068066746
    },
    {
      "action_query_norm": 40.04819107055664,
      "action_relation_loss": 0.0014381072483956813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.952291861552855,
      "flow_matching_loss": 0.02092039752751589,
      "grad_norm": 0.14930272102355957,
      "learning_rate": 2.6159016567210424e-05,
      "loss": 0.0225,
      "relation_loss": 0.0015421327669173479,
      "selected_pairs": 26409.9,
      "state_query_norm": 42.26364631652832,
      "state_relation_loss": 0.00010402552579762414,
      "step": 20260,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0015421327669173479,
      "weighted_relation_loss": 0.0015421327669173479
    },
    {
      "action_query_norm": 40.046102142333986,
      "action_relation_loss": 0.0014086702954955398,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.961646398503273,
      "flow_matching_loss": 0.020755892153829336,
      "grad_norm": 0.08702939748764038,
      "learning_rate": 2.6110584333070153e-05,
      "loss": 0.0223,
      "relation_loss": 0.001513431337662041,
      "selected_pairs": 26270.4,
      "state_query_norm": 42.244865036010744,
      "state_relation_loss": 0.00010476104071130977,
      "step": 20270,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.001513431337662041,
      "weighted_relation_loss": 0.001513431337662041
    },
    {
      "action_query_norm": 40.09291229248047,
      "action_relation_loss": 0.0013901988742873073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.971000935453695,
      "flow_matching_loss": 0.021925851702690125,
      "grad_norm": 0.11450410634279251,
      "learning_rate": 2.6062181126822537e-05,
      "loss": 0.0234,
      "relation_loss": 0.001494738517794758,
      "selected_pairs": 26590.9,
      "state_query_norm": 42.257863235473636,
      "state_relation_loss": 0.00010453964932821691,
      "step": 20280,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001494738517794758,
      "weighted_relation_loss": 0.001494738517794758
    },
    {
      "action_query_norm": 40.084075927734375,
      "action_relation_loss": 0.0014086184790357947,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.980355472404117,
      "flow_matching_loss": 0.02329396866261959,
      "grad_norm": 0.13038311898708344,
      "learning_rate": 2.601380700728203e-05,
      "loss": 0.0248,
      "relation_loss": 0.0015124870697036386,
      "selected_pairs": 26617.1,
      "state_query_norm": 42.25181427001953,
      "state_relation_loss": 0.00010386859357822687,
      "step": 20290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015124870697036386,
      "weighted_relation_loss": 0.0015124870697036386
    },
    {
      "action_query_norm": 40.079864501953125,
      "action_relation_loss": 0.0013923997757956386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.989710009354535,
      "flow_matching_loss": 0.022338545694947243,
      "grad_norm": 0.0989990308880806,
      "learning_rate": 2.596546203322777e-05,
      "loss": 0.0238,
      "relation_loss": 0.0014932200312614441,
      "selected_pairs": 27290.2,
      "state_query_norm": 42.259042739868164,
      "state_relation_loss": 0.00010082025328301824,
      "step": 20300,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0014932200312614441,
      "weighted_relation_loss": 0.0014932200312614441
    },
    {
      "action_query_norm": 40.13851432800293,
      "action_relation_loss": 0.0013748426572419703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 18.999064546304957,
      "flow_matching_loss": 0.020482726395130157,
      "grad_norm": 0.10884527117013931,
      "learning_rate": 2.5917146263403457e-05,
      "loss": 0.022,
      "relation_loss": 0.001480603008531034,
      "selected_pairs": 26641.6,
      "state_query_norm": 42.255283737182616,
      "state_relation_loss": 0.00010576036220300011,
      "step": 20310,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.001480603008531034,
      "weighted_relation_loss": 0.001480603008531034
    },
    {
      "action_query_norm": 40.09885787963867,
      "action_relation_loss": 0.001319006783887744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.00841908325538,
      "flow_matching_loss": 0.02047562450170517,
      "grad_norm": 0.09541600942611694,
      "learning_rate": 2.5868859756517296e-05,
      "loss": 0.0219,
      "relation_loss": 0.001419566769618541,
      "selected_pairs": 24783.1,
      "state_query_norm": 42.26679954528809,
      "state_relation_loss": 0.00010055999664473347,
      "step": 20320,
      "valid_action_samples": 216.0,
      "weighted_auxiliary_loss": 0.001419566769618541,
      "weighted_relation_loss": 0.001419566769618541
    },
    {
      "action_query_norm": 40.067073059082034,
      "action_relation_loss": 0.0013736481545493008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.0177736202058,
      "flow_matching_loss": 0.022423406317830085,
      "grad_norm": 0.0959477573633194,
      "learning_rate": 2.5820602571241947e-05,
      "loss": 0.0239,
      "relation_loss": 0.0014781635254621506,
      "selected_pairs": 26899.7,
      "state_query_norm": 42.23429069519043,
      "state_relation_loss": 0.00010451539128553122,
      "step": 20330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0014781635254621506,
      "weighted_relation_loss": 0.0014781635254621506
    },
    {
      "action_query_norm": 40.060477828979494,
      "action_relation_loss": 0.0014081583707593382,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.02712815715622,
      "flow_matching_loss": 0.021964338049292563,
      "grad_norm": 0.1065383329987526,
      "learning_rate": 2.577237476621442e-05,
      "loss": 0.0235,
      "relation_loss": 0.0015091194887645542,
      "selected_pairs": 27153.6,
      "state_query_norm": 42.237539291381836,
      "state_relation_loss": 0.00010096111727762037,
      "step": 20340,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0015091194887645542,
      "weighted_relation_loss": 0.0015091194887645542
    },
    {
      "action_query_norm": 40.1376449584961,
      "action_relation_loss": 0.0013125056633725763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.036482694106642,
      "flow_matching_loss": 0.021644390374422073,
      "grad_norm": 0.11656289547681808,
      "learning_rate": 2.572417640003609e-05,
      "loss": 0.0231,
      "relation_loss": 0.001413525000680238,
      "selected_pairs": 27034.5,
      "state_query_norm": 42.24070663452149,
      "state_relation_loss": 0.00010101932857651264,
      "step": 20350,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001413525000680238,
      "weighted_relation_loss": 0.001413525000680238
    },
    {
      "action_query_norm": 40.08486061096191,
      "action_relation_loss": 0.0013038792531006038,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.045837231057064,
      "flow_matching_loss": 0.019411059468984602,
      "grad_norm": 0.09436332434415817,
      "learning_rate": 2.5676007531272473e-05,
      "loss": 0.0208,
      "relation_loss": 0.0014059009728953243,
      "selected_pairs": 26386.6,
      "state_query_norm": 42.2351375579834,
      "state_relation_loss": 0.00010202172197750769,
      "step": 20360,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0014059009728953243,
      "weighted_relation_loss": 0.0014059009728953243
    },
    {
      "action_query_norm": 40.069936752319336,
      "action_relation_loss": 0.0013050223351456225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.055191768007482,
      "flow_matching_loss": 0.022113057784736156,
      "grad_norm": 0.10123591870069504,
      "learning_rate": 2.5627868218453332e-05,
      "loss": 0.0235,
      "relation_loss": 0.0014067568350583314,
      "selected_pairs": 26512.9,
      "state_query_norm": 42.23680191040039,
      "state_relation_loss": 0.00010173450136790052,
      "step": 20370,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0014067568350583314,
      "weighted_relation_loss": 0.0014067568350583314
    },
    {
      "action_query_norm": 40.086409378051755,
      "action_relation_loss": 0.0012562909047119319,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.064546304957904,
      "flow_matching_loss": 0.020613912679255008,
      "grad_norm": 0.09274624288082123,
      "learning_rate": 2.5579758520072443e-05,
      "loss": 0.022,
      "relation_loss": 0.0013539297855459155,
      "selected_pairs": 26735.2,
      "state_query_norm": 42.254035568237306,
      "state_relation_loss": 9.763887210283429e-05,
      "step": 20380,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013539297855459155,
      "weighted_relation_loss": 0.0013539297855459155
    },
    {
      "action_query_norm": 40.10427589416504,
      "action_relation_loss": 0.0012878429144620896,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.073900841908326,
      "flow_matching_loss": 0.021399139985442163,
      "grad_norm": 0.11534424871206284,
      "learning_rate": 2.5531678494587615e-05,
      "loss": 0.0228,
      "relation_loss": 0.0013885167427361012,
      "selected_pairs": 26735.0,
      "state_query_norm": 42.23031349182129,
      "state_relation_loss": 0.0001006738246360328,
      "step": 20390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013885167427361012,
      "weighted_relation_loss": 0.0013885167427361012
    },
    {
      "action_query_norm": 40.111904525756835,
      "action_relation_loss": 0.0013973431545309723,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.08325537885875,
      "flow_matching_loss": 0.020660190656781197,
      "grad_norm": 0.11288583278656006,
      "learning_rate": 2.5483628200420644e-05,
      "loss": 0.0222,
      "relation_loss": 0.001501277438364923,
      "selected_pairs": 26723.0,
      "state_query_norm": 42.23766021728515,
      "state_relation_loss": 0.00010393427583039738,
      "step": 20400,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001501277438364923,
      "weighted_relation_loss": 0.001501277438364923
    },
    {
      "action_query_norm": 40.1179931640625,
      "action_relation_loss": 0.001258835452608764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.092609915809167,
      "flow_matching_loss": 0.0211151460185647,
      "grad_norm": 0.11449307948350906,
      "learning_rate": 2.5435607695957154e-05,
      "loss": 0.0225,
      "relation_loss": 0.0013595160329714418,
      "selected_pairs": 26894.4,
      "state_query_norm": 42.2518928527832,
      "state_relation_loss": 0.00010068059928016737,
      "step": 20410,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013595160329714418,
      "weighted_relation_loss": 0.0013595160329714418
    },
    {
      "action_query_norm": 40.08245849609375,
      "action_relation_loss": 0.0013268788577988743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.10196445275959,
      "flow_matching_loss": 0.021592527627944946,
      "grad_norm": 0.09680818766355515,
      "learning_rate": 2.5387617039546586e-05,
      "loss": 0.023,
      "relation_loss": 0.001435838604811579,
      "selected_pairs": 27322.6,
      "state_query_norm": 42.234672927856444,
      "state_relation_loss": 0.00010895973173319362,
      "step": 20420,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.001435838604811579,
      "weighted_relation_loss": 0.001435838604811579
    },
    {
      "action_query_norm": 40.092309188842776,
      "action_relation_loss": 0.0012975360034033657,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.11131898971001,
      "flow_matching_loss": 0.020626729913055898,
      "grad_norm": 0.08422570675611496,
      "learning_rate": 2.5339656289502106e-05,
      "loss": 0.022,
      "relation_loss": 0.0014011301449500025,
      "selected_pairs": 26778.8,
      "state_query_norm": 42.21893310546875,
      "state_relation_loss": 0.00010359413790865802,
      "step": 20430,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0014011301449500025,
      "weighted_relation_loss": 0.0014011301449500025
    },
    {
      "action_query_norm": 40.08321685791016,
      "action_relation_loss": 0.0012750349240377545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.12067352666043,
      "flow_matching_loss": 0.020739953406155108,
      "grad_norm": 0.13280165195465088,
      "learning_rate": 2.529172550410056e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013825863366946577,
      "selected_pairs": 26789.0,
      "state_query_norm": 42.23453330993652,
      "state_relation_loss": 0.00010755139592220076,
      "step": 20440,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0013825863366946577,
      "weighted_relation_loss": 0.0013825863366946577
    },
    {
      "action_query_norm": 40.07276382446289,
      "action_relation_loss": 0.0013517701765522361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.13002806361085,
      "flow_matching_loss": 0.021639558486640453,
      "grad_norm": 0.1527574509382248,
      "learning_rate": 2.524382474158234e-05,
      "loss": 0.0231,
      "relation_loss": 0.001452904415782541,
      "selected_pairs": 27079.8,
      "state_query_norm": 42.228908920288085,
      "state_relation_loss": 0.00010113424723385833,
      "step": 20450,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001452904415782541,
      "weighted_relation_loss": 0.001452904415782541
    },
    {
      "action_query_norm": 40.09934272766113,
      "action_relation_loss": 0.001295656943693757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.139382600561273,
      "flow_matching_loss": 0.022776337154209614,
      "grad_norm": 0.10619693249464035,
      "learning_rate": 2.5195954060151433e-05,
      "loss": 0.0242,
      "relation_loss": 0.0013991406653076411,
      "selected_pairs": 26898.7,
      "state_query_norm": 42.23324279785156,
      "state_relation_loss": 0.00010348373543820344,
      "step": 20460,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013991406653076411,
      "weighted_relation_loss": 0.0013991406653076411
    },
    {
      "action_query_norm": 40.08894691467285,
      "action_relation_loss": 0.0014068585354834796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.14873713751169,
      "flow_matching_loss": 0.02217338439077139,
      "grad_norm": 0.1004716232419014,
      "learning_rate": 2.5148113517975212e-05,
      "loss": 0.0237,
      "relation_loss": 0.0015035839867778122,
      "selected_pairs": 26411.9,
      "state_query_norm": 42.25141487121582,
      "state_relation_loss": 9.672542801126837e-05,
      "step": 20470,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0015035839867778122,
      "weighted_relation_loss": 0.0015035839867778122
    },
    {
      "action_query_norm": 40.053208923339845,
      "action_relation_loss": 0.0013660352793522178,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.158091674462113,
      "flow_matching_loss": 0.022113690339028835,
      "grad_norm": 0.11743778735399246,
      "learning_rate": 2.5100303173184447e-05,
      "loss": 0.0236,
      "relation_loss": 0.00146638925652951,
      "selected_pairs": 27056.8,
      "state_query_norm": 42.2432559967041,
      "state_relation_loss": 0.00010035398372565396,
      "step": 20480,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.00146638925652951,
      "weighted_relation_loss": 0.00146638925652951
    },
    {
      "action_query_norm": 40.085353469848634,
      "action_relation_loss": 0.001301513414364308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.167446211412535,
      "flow_matching_loss": 0.020571309980005024,
      "grad_norm": 0.08795725554227829,
      "learning_rate": 2.5052523083873196e-05,
      "loss": 0.022,
      "relation_loss": 0.0014005177305079996,
      "selected_pairs": 26247.3,
      "state_query_norm": 42.22046585083008,
      "state_relation_loss": 9.900433069560677e-05,
      "step": 20490,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0014005177305079996,
      "weighted_relation_loss": 0.0014005177305079996
    },
    {
      "action_query_norm": 40.101366424560545,
      "action_relation_loss": 0.0012887170538306235,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.176800748362957,
      "flow_matching_loss": 0.02075356226414442,
      "grad_norm": 0.09152141958475113,
      "learning_rate": 2.500477330809881e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013864360633306205,
      "selected_pairs": 26597.2,
      "state_query_norm": 42.21895294189453,
      "state_relation_loss": 9.771901168278418e-05,
      "step": 20500,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013864360633306205,
      "weighted_relation_loss": 0.0013864360633306205
    },
    {
      "action_query_norm": 40.111993408203126,
      "action_relation_loss": 0.0012773411814123391,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.186155285313376,
      "flow_matching_loss": 0.021613445691764356,
      "grad_norm": 0.08965186774730682,
      "learning_rate": 2.4957053903881733e-05,
      "loss": 0.023,
      "relation_loss": 0.0013756476109847426,
      "selected_pairs": 26641.8,
      "state_query_norm": 42.23180618286133,
      "state_relation_loss": 9.830642593442463e-05,
      "step": 20510,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0013756476109847426,
      "weighted_relation_loss": 0.0013756476109847426
    },
    {
      "action_query_norm": 40.10631141662598,
      "action_relation_loss": 0.0013080176780931652,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.195509822263798,
      "flow_matching_loss": 0.02300888579338789,
      "grad_norm": 0.09154660999774933,
      "learning_rate": 2.4909364929205576e-05,
      "loss": 0.0244,
      "relation_loss": 0.0014108377159573137,
      "selected_pairs": 26461.8,
      "state_query_norm": 42.240849304199216,
      "state_relation_loss": 0.00010282003859174438,
      "step": 20520,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0014108377159573137,
      "weighted_relation_loss": 0.0014108377159573137
    },
    {
      "action_query_norm": 40.10603637695313,
      "action_relation_loss": 0.0013392670080065727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.20486435921422,
      "flow_matching_loss": 0.020798925869166852,
      "grad_norm": 0.1539260596036911,
      "learning_rate": 2.4861706442016923e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014374816790223121,
      "selected_pairs": 26049.9,
      "state_query_norm": 42.22804832458496,
      "state_relation_loss": 9.821467247093097e-05,
      "step": 20530,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0014374816790223121,
      "weighted_relation_loss": 0.0014374816790223121
    },
    {
      "action_query_norm": 40.041845703125,
      "action_relation_loss": 0.0012922648224048317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.21421889616464,
      "flow_matching_loss": 0.020854430086910725,
      "grad_norm": 0.09241099655628204,
      "learning_rate": 2.481407850022533e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013913633767515421,
      "selected_pairs": 26840.1,
      "state_query_norm": 42.220380783081055,
      "state_relation_loss": 9.909856598824262e-05,
      "step": 20540,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013913633767515421,
      "weighted_relation_loss": 0.0013913633767515421
    },
    {
      "action_query_norm": 40.040108108520506,
      "action_relation_loss": 0.0013565129251219333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.22357343311506,
      "flow_matching_loss": 0.021971153002232312,
      "grad_norm": 0.08545525372028351,
      "learning_rate": 2.4766481161703213e-05,
      "loss": 0.0234,
      "relation_loss": 0.0014572367654182017,
      "selected_pairs": 26528.2,
      "state_query_norm": 42.231829452514646,
      "state_relation_loss": 0.0001007238490274176,
      "step": 20550,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0014572367654182017,
      "weighted_relation_loss": 0.0014572367654182017
    },
    {
      "action_query_norm": 40.05028038024902,
      "action_relation_loss": 0.0013212902005761862,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.232927970065482,
      "flow_matching_loss": 0.020228638499975204,
      "grad_norm": 0.07751499116420746,
      "learning_rate": 2.4718914484285876e-05,
      "loss": 0.0217,
      "relation_loss": 0.0014294543769210577,
      "selected_pairs": 26800.8,
      "state_query_norm": 42.217828369140626,
      "state_relation_loss": 0.00010816419162438252,
      "step": 20560,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014294543769210577,
      "weighted_relation_loss": 0.0014294543769210577
    },
    {
      "action_query_norm": 40.042900466918944,
      "action_relation_loss": 0.0012652261182665824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.242282507015904,
      "flow_matching_loss": 0.019946575164794922,
      "grad_norm": 0.09632182866334915,
      "learning_rate": 2.467137852577129e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013592185219749807,
      "selected_pairs": 26780.2,
      "state_query_norm": 42.22005958557129,
      "state_relation_loss": 9.399240370839834e-05,
      "step": 20570,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0013592185219749807,
      "weighted_relation_loss": 0.0013592185219749807
    },
    {
      "action_query_norm": 40.04916534423828,
      "action_relation_loss": 0.0014065464725717901,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.251637043966323,
      "flow_matching_loss": 0.021700831316411495,
      "grad_norm": 0.1185702309012413,
      "learning_rate": 2.4623873343920123e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015083801234140992,
      "selected_pairs": 26623.3,
      "state_query_norm": 42.217499542236325,
      "state_relation_loss": 0.00010183364865952171,
      "step": 20580,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0015083801234140992,
      "weighted_relation_loss": 0.0015083801234140992
    },
    {
      "action_query_norm": 40.10880889892578,
      "action_relation_loss": 0.00125298936618492,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.260991580916745,
      "flow_matching_loss": 0.021382973808795214,
      "grad_norm": 0.11353112012147903,
      "learning_rate": 2.4576398996455656e-05,
      "loss": 0.0227,
      "relation_loss": 0.0013525947812013327,
      "selected_pairs": 27343.1,
      "state_query_norm": 42.20785789489746,
      "state_relation_loss": 9.960540410247631e-05,
      "step": 20590,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0013525947812013327,
      "weighted_relation_loss": 0.0013525947812013327
    },
    {
      "action_query_norm": 40.086250686645506,
      "action_relation_loss": 0.0013068035361357034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.270346117867167,
      "flow_matching_loss": 0.022336568031460047,
      "grad_norm": 0.10296259075403214,
      "learning_rate": 2.452895554106368e-05,
      "loss": 0.0237,
      "relation_loss": 0.0014093380654230715,
      "selected_pairs": 27548.9,
      "state_query_norm": 42.215410232543945,
      "state_relation_loss": 0.00010253452928736806,
      "step": 20600,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0014093380654230715,
      "weighted_relation_loss": 0.0014093380654230715
    },
    {
      "action_query_norm": 40.06990776062012,
      "action_relation_loss": 0.0013077075826004147,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.279700654817585,
      "flow_matching_loss": 0.02165697868913412,
      "grad_norm": 0.11522120982408524,
      "learning_rate": 2.4481543035392506e-05,
      "loss": 0.0231,
      "relation_loss": 0.0014079985907301308,
      "selected_pairs": 27197.5,
      "state_query_norm": 42.22456741333008,
      "state_relation_loss": 0.00010029100958490745,
      "step": 20610,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0014079985907301308,
      "weighted_relation_loss": 0.0014079985907301308
    },
    {
      "action_query_norm": 40.01699562072754,
      "action_relation_loss": 0.0013607087079435588,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.289055191768007,
      "flow_matching_loss": 0.021155243925750255,
      "grad_norm": 0.11339014023542404,
      "learning_rate": 2.4434161537052774e-05,
      "loss": 0.0226,
      "relation_loss": 0.001460757153108716,
      "selected_pairs": 26575.2,
      "state_query_norm": 42.21726989746094,
      "state_relation_loss": 0.0001000484568066895,
      "step": 20620,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.001460757153108716,
      "weighted_relation_loss": 0.001460757153108716
    },
    {
      "action_query_norm": 40.07863998413086,
      "action_relation_loss": 0.0013181254034861922,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.29840972871843,
      "flow_matching_loss": 0.02117799688130617,
      "grad_norm": 0.10198552161455154,
      "learning_rate": 2.4386811103617474e-05,
      "loss": 0.0226,
      "relation_loss": 0.001424579811282456,
      "selected_pairs": 26802.2,
      "state_query_norm": 42.19865760803223,
      "state_relation_loss": 0.00010645441652741283,
      "step": 20630,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001424579811282456,
      "weighted_relation_loss": 0.001424579811282456
    },
    {
      "action_query_norm": 40.04539260864258,
      "action_relation_loss": 0.0013447264907881617,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.30776426566885,
      "flow_matching_loss": 0.021518555656075477,
      "grad_norm": 0.10090965032577515,
      "learning_rate": 2.4339491792621833e-05,
      "loss": 0.023,
      "relation_loss": 0.001442728948313743,
      "selected_pairs": 27085.8,
      "state_query_norm": 42.221825408935544,
      "state_relation_loss": 9.800244588404894e-05,
      "step": 20640,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001442728948313743,
      "weighted_relation_loss": 0.001442728948313743
    },
    {
      "action_query_norm": 40.080308532714845,
      "action_relation_loss": 0.001418441196437925,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.31711880261927,
      "flow_matching_loss": 0.02044533584266901,
      "grad_norm": 0.11837119609117508,
      "learning_rate": 2.4292203661563312e-05,
      "loss": 0.022,
      "relation_loss": 0.0015146567136980593,
      "selected_pairs": 26920.3,
      "state_query_norm": 42.21439170837402,
      "state_relation_loss": 9.621551653253846e-05,
      "step": 20650,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0015146567136980593,
      "weighted_relation_loss": 0.0015146567136980593
    },
    {
      "action_query_norm": 40.07353477478027,
      "action_relation_loss": 0.0012928368058055639,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.32647333956969,
      "flow_matching_loss": 0.021158753708004952,
      "grad_norm": 0.11766690760850906,
      "learning_rate": 2.424494676790141e-05,
      "loss": 0.0226,
      "relation_loss": 0.001392388972453773,
      "selected_pairs": 27289.7,
      "state_query_norm": 42.20885543823242,
      "state_relation_loss": 9.955217246897519e-05,
      "step": 20660,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.001392388972453773,
      "weighted_relation_loss": 0.001392388972453773
    },
    {
      "action_query_norm": 40.072618865966795,
      "action_relation_loss": 0.0013176011969335377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.335827876520113,
      "flow_matching_loss": 0.021768789365887643,
      "grad_norm": 0.11678757518529892,
      "learning_rate": 2.4197721169057745e-05,
      "loss": 0.0232,
      "relation_loss": 0.0014158797916024924,
      "selected_pairs": 26524.0,
      "state_query_norm": 42.21716575622558,
      "state_relation_loss": 9.827860121731646e-05,
      "step": 20670,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0014158797916024924,
      "weighted_relation_loss": 0.0014158797916024924
    },
    {
      "action_query_norm": 40.06843719482422,
      "action_relation_loss": 0.0012895113439299166,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.345182413470532,
      "flow_matching_loss": 0.021375062689185143,
      "grad_norm": 0.08922084420919418,
      "learning_rate": 2.4150526922415857e-05,
      "loss": 0.0228,
      "relation_loss": 0.0013862719526514411,
      "selected_pairs": 27110.4,
      "state_query_norm": 42.22672233581543,
      "state_relation_loss": 9.676061235950329e-05,
      "step": 20680,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0013862719526514411,
      "weighted_relation_loss": 0.0013862719526514411
    },
    {
      "action_query_norm": 39.96810646057129,
      "action_relation_loss": 0.0014569075196050108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.354536950420954,
      "flow_matching_loss": 0.02164718434214592,
      "grad_norm": 0.10185568779706955,
      "learning_rate": 2.41033640853212e-05,
      "loss": 0.0232,
      "relation_loss": 0.0015524797374382616,
      "selected_pairs": 26951.1,
      "state_query_norm": 42.209817504882814,
      "state_relation_loss": 9.557222801959142e-05,
      "step": 20690,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0015524797374382616,
      "weighted_relation_loss": 0.0015524797374382616
    },
    {
      "action_query_norm": 40.020542526245116,
      "action_relation_loss": 0.0012651476310566068,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.363891487371376,
      "flow_matching_loss": 0.019690186157822608,
      "grad_norm": 0.09857837110757828,
      "learning_rate": 2.405623271508108e-05,
      "loss": 0.0211,
      "relation_loss": 0.0013664990314282478,
      "selected_pairs": 26876.6,
      "state_query_norm": 42.20096321105957,
      "state_relation_loss": 0.00010135141710634343,
      "step": 20700,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0013664990314282478,
      "weighted_relation_loss": 0.0013664990314282478
    },
    {
      "action_query_norm": 40.019290924072266,
      "action_relation_loss": 0.0013190158526413143,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.373246024321794,
      "flow_matching_loss": 0.019906075857579707,
      "grad_norm": 0.09994703531265259,
      "learning_rate": 2.4009132868964522e-05,
      "loss": 0.0213,
      "relation_loss": 0.001418433990329504,
      "selected_pairs": 26755.6,
      "state_query_norm": 42.20648765563965,
      "state_relation_loss": 9.941812022589147e-05,
      "step": 20710,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001418433990329504,
      "weighted_relation_loss": 0.001418433990329504
    },
    {
      "action_query_norm": 40.03602485656738,
      "action_relation_loss": 0.00124862837838009,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.382600561272216,
      "flow_matching_loss": 0.022161046043038368,
      "grad_norm": 0.1173321083188057,
      "learning_rate": 2.3962064604202327e-05,
      "loss": 0.0235,
      "relation_loss": 0.0013498562388122083,
      "selected_pairs": 26416.6,
      "state_query_norm": 42.16984825134277,
      "state_relation_loss": 0.00010122786261490546,
      "step": 20720,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0013498562388122083,
      "weighted_relation_loss": 0.0013498562388122083
    },
    {
      "action_query_norm": 40.06285514831543,
      "action_relation_loss": 0.0012235575006343425,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.39195509822264,
      "flow_matching_loss": 0.020010967366397382,
      "grad_norm": 0.09156343340873718,
      "learning_rate": 2.3915027977986858e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013229147880338132,
      "selected_pairs": 26696.2,
      "state_query_norm": 42.218495559692386,
      "state_relation_loss": 9.935728230630048e-05,
      "step": 20730,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0013229147880338132,
      "weighted_relation_loss": 0.0013229147880338132
    },
    {
      "action_query_norm": 40.01285858154297,
      "action_relation_loss": 0.0013410068000666796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.40130963517306,
      "flow_matching_loss": 0.022788980323821307,
      "grad_norm": 0.10455461591482162,
      "learning_rate": 2.386802304747205e-05,
      "loss": 0.0242,
      "relation_loss": 0.0014388888841494918,
      "selected_pairs": 27178.0,
      "state_query_norm": 42.2018367767334,
      "state_relation_loss": 9.788209281396121e-05,
      "step": 20740,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0014388888841494918,
      "weighted_relation_loss": 0.0014388888841494918
    },
    {
      "action_query_norm": 40.00283012390137,
      "action_relation_loss": 0.0012788184569217265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.41066417212348,
      "flow_matching_loss": 0.02121741697192192,
      "grad_norm": 0.09719541668891907,
      "learning_rate": 2.382104986977332e-05,
      "loss": 0.0226,
      "relation_loss": 0.001377422397490591,
      "selected_pairs": 26940.8,
      "state_query_norm": 42.187247085571286,
      "state_relation_loss": 9.860395875875838e-05,
      "step": 20750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001377422397490591,
      "weighted_relation_loss": 0.001377422397490591
    },
    {
      "action_query_norm": 40.07510871887207,
      "action_relation_loss": 0.001321626827120781,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.4200187090739,
      "flow_matching_loss": 0.02160322740674019,
      "grad_norm": 0.09681878238916397,
      "learning_rate": 2.377410850196749e-05,
      "loss": 0.023,
      "relation_loss": 0.0014200237230397761,
      "selected_pairs": 27525.3,
      "state_query_norm": 42.20620918273926,
      "state_relation_loss": 9.839689810178242e-05,
      "step": 20760,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0014200237230397761,
      "weighted_relation_loss": 0.0014200237230397761
    },
    {
      "action_query_norm": 40.06045379638672,
      "action_relation_loss": 0.0012879053596407176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.429373246024323,
      "flow_matching_loss": 0.02227500583976507,
      "grad_norm": 0.1395994871854782,
      "learning_rate": 2.3727199001092797e-05,
      "loss": 0.0237,
      "relation_loss": 0.0013882896746508777,
      "selected_pairs": 26920.2,
      "state_query_norm": 42.198554611206056,
      "state_relation_loss": 0.00010038429973064922,
      "step": 20770,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0013882896746508777,
      "weighted_relation_loss": 0.0013882896746508777
    },
    {
      "action_query_norm": 40.03269271850586,
      "action_relation_loss": 0.001329828437883407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.43872778297474,
      "flow_matching_loss": 0.01986945290118456,
      "grad_norm": 0.10444445163011551,
      "learning_rate": 2.3680321424148677e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014274317887611686,
      "selected_pairs": 27217.7,
      "state_query_norm": 42.21222457885742,
      "state_relation_loss": 9.760335669852793e-05,
      "step": 20780,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0014274317887611686,
      "weighted_relation_loss": 0.0014274317887611686
    },
    {
      "action_query_norm": 40.04070777893067,
      "action_relation_loss": 0.001392009574919939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.448082319925163,
      "flow_matching_loss": 0.021801043301820755,
      "grad_norm": 0.1029323861002922,
      "learning_rate": 2.3633475828095793e-05,
      "loss": 0.0233,
      "relation_loss": 0.0014907860313542187,
      "selected_pairs": 27456.6,
      "state_query_norm": 42.19256362915039,
      "state_relation_loss": 9.877646298264153e-05,
      "step": 20790,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0014907860313542187,
      "weighted_relation_loss": 0.0014907860313542187
    },
    {
      "action_query_norm": 39.99778289794922,
      "action_relation_loss": 0.001276712620165199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.457436856875585,
      "flow_matching_loss": 0.021482869796454907,
      "grad_norm": 0.09704841673374176,
      "learning_rate": 2.358666226985599e-05,
      "loss": 0.0229,
      "relation_loss": 0.0013781384681351484,
      "selected_pairs": 26966.4,
      "state_query_norm": 42.17320327758789,
      "state_relation_loss": 0.00010142585160792806,
      "step": 20800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0013781384681351484,
      "weighted_relation_loss": 0.0013781384681351484
    },
    {
      "action_query_norm": 40.010546112060545,
      "action_relation_loss": 0.0013174334657378496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.466791393826007,
      "flow_matching_loss": 0.02110801786184311,
      "grad_norm": 0.09572580456733704,
      "learning_rate": 2.353988080631213e-05,
      "loss": 0.0225,
      "relation_loss": 0.0014160066843032838,
      "selected_pairs": 26176.7,
      "state_query_norm": 42.21871757507324,
      "state_relation_loss": 9.857322802417912e-05,
      "step": 20810,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0014160066843032838,
      "weighted_relation_loss": 0.0014160066843032838
    },
    {
      "action_query_norm": 40.022661972045896,
      "action_relation_loss": 0.00122032358776778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.476145930776426,
      "flow_matching_loss": 0.019978230819106102,
      "grad_norm": 0.09497416764497757,
      "learning_rate": 2.3493131494308142e-05,
      "loss": 0.0213,
      "relation_loss": 0.001318402262404561,
      "selected_pairs": 27035.6,
      "state_query_norm": 42.20683364868164,
      "state_relation_loss": 9.80786862783134e-05,
      "step": 20820,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.001318402262404561,
      "weighted_relation_loss": 0.001318402262404561
    },
    {
      "action_query_norm": 40.04144134521484,
      "action_relation_loss": 0.0012575517874211072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.485500467726848,
      "flow_matching_loss": 0.020590661466121672,
      "grad_norm": 0.10689503699541092,
      "learning_rate": 2.344641439064881e-05,
      "loss": 0.0219,
      "relation_loss": 0.0013575884397141636,
      "selected_pairs": 26828.2,
      "state_query_norm": 42.2093147277832,
      "state_relation_loss": 0.00010003664792748168,
      "step": 20830,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013575884397141636,
      "weighted_relation_loss": 0.0013575884397141636
    },
    {
      "action_query_norm": 40.013540267944336,
      "action_relation_loss": 0.001303916657343507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.49485500467727,
      "flow_matching_loss": 0.020576723851263523,
      "grad_norm": 0.10933197289705276,
      "learning_rate": 2.339972955209984e-05,
      "loss": 0.022,
      "relation_loss": 0.0014103386434726416,
      "selected_pairs": 26813.1,
      "state_query_norm": 42.18937301635742,
      "state_relation_loss": 0.00010642199413268827,
      "step": 20840,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014103386434726416,
      "weighted_relation_loss": 0.0014103386434726416
    },
    {
      "action_query_norm": 39.98284797668457,
      "action_relation_loss": 0.001313617278356105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.504209541627688,
      "flow_matching_loss": 0.022970523126423358,
      "grad_norm": 0.08399084210395813,
      "learning_rate": 2.335307703538771e-05,
      "loss": 0.0244,
      "relation_loss": 0.0014170199166983366,
      "selected_pairs": 27317.5,
      "state_query_norm": 42.2082332611084,
      "state_relation_loss": 0.00010340262597310356,
      "step": 20850,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0014170199166983366,
      "weighted_relation_loss": 0.0014170199166983366
    },
    {
      "action_query_norm": 40.04384307861328,
      "action_relation_loss": 0.001310391863808036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.51356407857811,
      "flow_matching_loss": 0.02235785387456417,
      "grad_norm": 0.10099421441555023,
      "learning_rate": 2.330645689719962e-05,
      "loss": 0.0238,
      "relation_loss": 0.001405913825146854,
      "selected_pairs": 26987.8,
      "state_query_norm": 42.18905792236328,
      "state_relation_loss": 9.552195260766893e-05,
      "step": 20860,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001405913825146854,
      "weighted_relation_loss": 0.001405913825146854
    },
    {
      "action_query_norm": 40.03586196899414,
      "action_relation_loss": 0.0012772875721566378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.522918615528532,
      "flow_matching_loss": 0.02091411892324686,
      "grad_norm": 0.1260673999786377,
      "learning_rate": 2.3259869194183414e-05,
      "loss": 0.0223,
      "relation_loss": 0.0013742942828685045,
      "selected_pairs": 27150.1,
      "state_query_norm": 42.191019439697264,
      "state_relation_loss": 9.700672526378185e-05,
      "step": 20870,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013742942828685045,
      "weighted_relation_loss": 0.0013742942828685045
    },
    {
      "action_query_norm": 40.06666793823242,
      "action_relation_loss": 0.0012977569829672575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.532273152478954,
      "flow_matching_loss": 0.018691984843462704,
      "grad_norm": 0.08536294847726822,
      "learning_rate": 2.321331398294759e-05,
      "loss": 0.0201,
      "relation_loss": 0.0014021934010088443,
      "selected_pairs": 26482.7,
      "state_query_norm": 42.2010082244873,
      "state_relation_loss": 0.0001044364194967784,
      "step": 20880,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0014021934010088443,
      "weighted_relation_loss": 0.0014021934010088443
    },
    {
      "action_query_norm": 40.01126594543457,
      "action_relation_loss": 0.0013499928521923722,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.541627689429372,
      "flow_matching_loss": 0.020243713259696962,
      "grad_norm": 0.11201214790344238,
      "learning_rate": 2.3166791320061094e-05,
      "loss": 0.0217,
      "relation_loss": 0.0014456538832746447,
      "selected_pairs": 27058.6,
      "state_query_norm": 42.18721694946289,
      "state_relation_loss": 9.566103326505982e-05,
      "step": 20890,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0014456538832746447,
      "weighted_relation_loss": 0.0014456538832746447
    },
    {
      "action_query_norm": 40.052606201171876,
      "action_relation_loss": 0.0012334816972725093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.550982226379794,
      "flow_matching_loss": 0.020220549777150154,
      "grad_norm": 0.14758816361427307,
      "learning_rate": 2.312030126205335e-05,
      "loss": 0.0216,
      "relation_loss": 0.0013390091015025973,
      "selected_pairs": 27041.8,
      "state_query_norm": 42.18434638977051,
      "state_relation_loss": 0.00010552741514402442,
      "step": 20900,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0013390091015025973,
      "weighted_relation_loss": 0.0013390091015025973
    },
    {
      "action_query_norm": 39.98013496398926,
      "action_relation_loss": 0.001371682167518884,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.560336763330216,
      "flow_matching_loss": 0.019761410355567933,
      "grad_norm": 0.10196831822395325,
      "learning_rate": 2.3073843865414162e-05,
      "loss": 0.0212,
      "relation_loss": 0.0014703962486237287,
      "selected_pairs": 26666.1,
      "state_query_norm": 42.18526954650879,
      "state_relation_loss": 9.871407819446176e-05,
      "step": 20910,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0014703962486237287,
      "weighted_relation_loss": 0.0014703962486237287
    },
    {
      "action_query_norm": 40.047005844116214,
      "action_relation_loss": 0.0012575054541230201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.569691300280635,
      "flow_matching_loss": 0.020927250385284424,
      "grad_norm": 0.12422297149896622,
      "learning_rate": 2.302741918659363e-05,
      "loss": 0.0223,
      "relation_loss": 0.0013541771098971368,
      "selected_pairs": 26618.1,
      "state_query_norm": 42.19075469970703,
      "state_relation_loss": 9.667164631537162e-05,
      "step": 20920,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0013541771098971368,
      "weighted_relation_loss": 0.0013541771098971368
    },
    {
      "action_query_norm": 40.010514450073245,
      "action_relation_loss": 0.0013354883412830532,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.579045837231057,
      "flow_matching_loss": 0.019856381230056286,
      "grad_norm": 0.11611958593130112,
      "learning_rate": 2.2981027282002154e-05,
      "loss": 0.0213,
      "relation_loss": 0.0014302284456789494,
      "selected_pairs": 26803.9,
      "state_query_norm": 42.196118927001955,
      "state_relation_loss": 9.474012622376904e-05,
      "step": 20930,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0014302284456789494,
      "weighted_relation_loss": 0.0014302284456789494
    },
    {
      "action_query_norm": 40.017025756835935,
      "action_relation_loss": 0.0012366052716970444,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.58840037418148,
      "flow_matching_loss": 0.020237490069121122,
      "grad_norm": 0.08658161014318466,
      "learning_rate": 2.2934668208010236e-05,
      "loss": 0.0216,
      "relation_loss": 0.001335763616953045,
      "selected_pairs": 26559.1,
      "state_query_norm": 42.18731117248535,
      "state_relation_loss": 9.915834889397956e-05,
      "step": 20940,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001335763616953045,
      "weighted_relation_loss": 0.001335763616953045
    },
    {
      "action_query_norm": 39.996345138549806,
      "action_relation_loss": 0.0013557452475652099,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.5977549111319,
      "flow_matching_loss": 0.020397969894111155,
      "grad_norm": 0.09558843821287155,
      "learning_rate": 2.2888342020948555e-05,
      "loss": 0.0218,
      "relation_loss": 0.0014504337683320045,
      "selected_pairs": 26922.7,
      "state_query_norm": 42.17512664794922,
      "state_relation_loss": 9.468852658756077e-05,
      "step": 20950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0014504337683320045,
      "weighted_relation_loss": 0.0014504337683320045
    },
    {
      "action_query_norm": 39.97252655029297,
      "action_relation_loss": 0.0012929168180562556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.60710944808232,
      "flow_matching_loss": 0.020347415935248137,
      "grad_norm": 0.09157409518957138,
      "learning_rate": 2.2842048777107783e-05,
      "loss": 0.0217,
      "relation_loss": 0.0013907357351854444,
      "selected_pairs": 27180.3,
      "state_query_norm": 42.17630119323731,
      "state_relation_loss": 9.781891712918877e-05,
      "step": 20960,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0013907357351854444,
      "weighted_relation_loss": 0.0013907357351854444
    },
    {
      "action_query_norm": 40.001679229736325,
      "action_relation_loss": 0.0012377626029774547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.61646398503274,
      "flow_matching_loss": 0.019823375251144172,
      "grad_norm": 0.10525673627853394,
      "learning_rate": 2.2795788532738553e-05,
      "loss": 0.0212,
      "relation_loss": 0.0013308732653968037,
      "selected_pairs": 27013.6,
      "state_query_norm": 42.181967163085936,
      "state_relation_loss": 9.311065368819982e-05,
      "step": 20970,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013308732653968037,
      "weighted_relation_loss": 0.0013308732653968037
    },
    {
      "action_query_norm": 39.96468505859375,
      "action_relation_loss": 0.001335004682186991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.625818521983163,
      "flow_matching_loss": 0.021699564158916475,
      "grad_norm": 0.11488921195268631,
      "learning_rate": 2.2749561344051468e-05,
      "loss": 0.0231,
      "relation_loss": 0.0014273616834543645,
      "selected_pairs": 26831.8,
      "state_query_norm": 42.19570693969727,
      "state_relation_loss": 9.235700708813966e-05,
      "step": 20980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0014273616834543645,
      "weighted_relation_loss": 0.0014273616834543645
    },
    {
      "action_query_norm": 40.021589279174805,
      "action_relation_loss": 0.001251955213956535,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.63517305893358,
      "flow_matching_loss": 0.02084333561360836,
      "grad_norm": 0.1474754363298416,
      "learning_rate": 2.2703367267216896e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013527014409191907,
      "selected_pairs": 26964.2,
      "state_query_norm": 42.17602310180664,
      "state_relation_loss": 0.0001007462342386134,
      "step": 20990,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0013527014409191907,
      "weighted_relation_loss": 0.0013527014409191907
    },
    {
      "action_query_norm": 40.009961700439455,
      "action_relation_loss": 0.0013687901198863983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.644527595884004,
      "flow_matching_loss": 0.01949126534163952,
      "grad_norm": 0.10735218971967697,
      "learning_rate": 2.2657206358364997e-05,
      "loss": 0.021,
      "relation_loss": 0.0014662495232187212,
      "selected_pairs": 27180.4,
      "state_query_norm": 42.170694732666014,
      "state_relation_loss": 9.745941060828045e-05,
      "step": 21000,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0014662495232187212,
      "weighted_relation_loss": 0.0014662495232187212
    },
    {
      "action_query_norm": 40.00813179016113,
      "action_relation_loss": 0.0013475439394824208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.653882132834426,
      "flow_matching_loss": 0.02176994513720274,
      "grad_norm": 0.08872595429420471,
      "learning_rate": 2.261107867358563e-05,
      "loss": 0.0232,
      "relation_loss": 0.0014495996991172432,
      "selected_pairs": 26969.9,
      "state_query_norm": 42.1574047088623,
      "state_relation_loss": 0.00010205578219029121,
      "step": 21010,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0014495996991172432,
      "weighted_relation_loss": 0.0014495996991172432
    },
    {
      "action_query_norm": 40.003840637207034,
      "action_relation_loss": 0.0012292814324609935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.663236669784844,
      "flow_matching_loss": 0.02003413885831833,
      "grad_norm": 0.11200170964002609,
      "learning_rate": 2.2564984268928263e-05,
      "loss": 0.0214,
      "relation_loss": 0.001327153982128948,
      "selected_pairs": 26945.5,
      "state_query_norm": 42.2019775390625,
      "state_relation_loss": 9.787253220565617e-05,
      "step": 21020,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001327153982128948,
      "weighted_relation_loss": 0.001327153982128948
    },
    {
      "action_query_norm": 39.94755859375,
      "action_relation_loss": 0.0012536908965557815,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.672591206735266,
      "flow_matching_loss": 0.022445699758827688,
      "grad_norm": 0.11381013691425323,
      "learning_rate": 2.251892320040198e-05,
      "loss": 0.0238,
      "relation_loss": 0.0013520479667931795,
      "selected_pairs": 26432.8,
      "state_query_norm": 42.17272186279297,
      "state_relation_loss": 9.835708551690914e-05,
      "step": 21030,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0013520479667931795,
      "weighted_relation_loss": 0.0013520479667931795
    },
    {
      "action_query_norm": 39.9672248840332,
      "action_relation_loss": 0.0012448130175471306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.681945743685688,
      "flow_matching_loss": 0.01970468517392874,
      "grad_norm": 0.10039352625608444,
      "learning_rate": 2.2472895523975312e-05,
      "loss": 0.021,
      "relation_loss": 0.001342718058731407,
      "selected_pairs": 27149.0,
      "state_query_norm": 42.18057556152344,
      "state_relation_loss": 9.790503754629754e-05,
      "step": 21040,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001342718058731407,
      "weighted_relation_loss": 0.001342718058731407
    },
    {
      "action_query_norm": 40.01078720092774,
      "action_relation_loss": 0.0013489401317201555,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.69130028063611,
      "flow_matching_loss": 0.020796451810747386,
      "grad_norm": 0.07401132583618164,
      "learning_rate": 2.2426901295576212e-05,
      "loss": 0.0222,
      "relation_loss": 0.0014429567498154938,
      "selected_pairs": 26943.2,
      "state_query_norm": 42.17892913818359,
      "state_relation_loss": 9.401661591255107e-05,
      "step": 21050,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0014429567498154938,
      "weighted_relation_loss": 0.0014429567498154938
    },
    {
      "action_query_norm": 39.97065734863281,
      "action_relation_loss": 0.0013783090398646891,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.70065481758653,
      "flow_matching_loss": 0.020327564142644406,
      "grad_norm": 0.10580099374055862,
      "learning_rate": 2.2380940571092013e-05,
      "loss": 0.0218,
      "relation_loss": 0.0014782493351958692,
      "selected_pairs": 27365.3,
      "state_query_norm": 42.158032989501955,
      "state_relation_loss": 9.994029533118009e-05,
      "step": 21060,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0014782493351958692,
      "weighted_relation_loss": 0.0014782493351958692
    },
    {
      "action_query_norm": 39.98554649353027,
      "action_relation_loss": 0.001237171853426844,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.71000935453695,
      "flow_matching_loss": 0.022165879048407076,
      "grad_norm": 0.1072767823934555,
      "learning_rate": 2.2335013406369302e-05,
      "loss": 0.0235,
      "relation_loss": 0.0013382638455368579,
      "selected_pairs": 27344.0,
      "state_query_norm": 42.18340797424317,
      "state_relation_loss": 0.00010109200011356733,
      "step": 21070,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0013382638455368579,
      "weighted_relation_loss": 0.0013382638455368579
    },
    {
      "action_query_norm": 39.974357986450194,
      "action_relation_loss": 0.0013114058878272773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.719363891487372,
      "flow_matching_loss": 0.020203998312354088,
      "grad_norm": 0.09342700988054276,
      "learning_rate": 2.228911985721397e-05,
      "loss": 0.0216,
      "relation_loss": 0.001415973750408739,
      "selected_pairs": 26414.5,
      "state_query_norm": 42.17215843200684,
      "state_relation_loss": 0.000104567867674632,
      "step": 21080,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.001415973750408739,
      "weighted_relation_loss": 0.001415973750408739
    },
    {
      "action_query_norm": 39.96327285766601,
      "action_relation_loss": 0.0012788740219548344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.72871842843779,
      "flow_matching_loss": 0.020446811988949775,
      "grad_norm": 0.11356685310602188,
      "learning_rate": 2.224325997939095e-05,
      "loss": 0.0218,
      "relation_loss": 0.001377960224635899,
      "selected_pairs": 27083.4,
      "state_query_norm": 42.17789077758789,
      "state_relation_loss": 9.90862085018307e-05,
      "step": 21090,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001377960224635899,
      "weighted_relation_loss": 0.001377960224635899
    },
    {
      "action_query_norm": 39.91040954589844,
      "action_relation_loss": 0.0013073127600364387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.738072965388213,
      "flow_matching_loss": 0.021327379532158373,
      "grad_norm": 0.1297118067741394,
      "learning_rate": 2.219743382862437e-05,
      "loss": 0.0227,
      "relation_loss": 0.001403874100651592,
      "selected_pairs": 26762.6,
      "state_query_norm": 42.1734619140625,
      "state_relation_loss": 9.656134352553636e-05,
      "step": 21100,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001403874100651592,
      "weighted_relation_loss": 0.001403874100651592
    },
    {
      "action_query_norm": 39.973231887817384,
      "action_relation_loss": 0.0012414749711751939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.747427502338635,
      "flow_matching_loss": 0.02080450113862753,
      "grad_norm": 0.12201645225286484,
      "learning_rate": 2.2151641460597294e-05,
      "loss": 0.0221,
      "relation_loss": 0.001334384665824473,
      "selected_pairs": 26363.2,
      "state_query_norm": 42.169572067260745,
      "state_relation_loss": 9.290969173889608e-05,
      "step": 21110,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001334384665824473,
      "weighted_relation_loss": 0.001334384665824473
    },
    {
      "action_query_norm": 39.94908752441406,
      "action_relation_loss": 0.0012999315746128558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.756782039289057,
      "flow_matching_loss": 0.018183890916407108,
      "grad_norm": 0.1102883592247963,
      "learning_rate": 2.210588293095177e-05,
      "loss": 0.0196,
      "relation_loss": 0.0013930972665548325,
      "selected_pairs": 27502.1,
      "state_query_norm": 42.15956344604492,
      "state_relation_loss": 9.316567520727404e-05,
      "step": 21120,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0013930972665548325,
      "weighted_relation_loss": 0.0013930972665548325
    },
    {
      "action_query_norm": 39.9702392578125,
      "action_relation_loss": 0.001269520679488778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.766136576239475,
      "flow_matching_loss": 0.022033417038619517,
      "grad_norm": 0.0921308770775795,
      "learning_rate": 2.2060158295288714e-05,
      "loss": 0.0234,
      "relation_loss": 0.0013665650505572558,
      "selected_pairs": 27009.7,
      "state_query_norm": 42.17122383117676,
      "state_relation_loss": 9.704438780318014e-05,
      "step": 21130,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013665650505572558,
      "weighted_relation_loss": 0.0013665650505572558
    },
    {
      "action_query_norm": 40.07461738586426,
      "action_relation_loss": 0.0012478749151341618,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.775491113189897,
      "flow_matching_loss": 0.020376141276210548,
      "grad_norm": 0.12610109150409698,
      "learning_rate": 2.2014467609167905e-05,
      "loss": 0.0217,
      "relation_loss": 0.0013448200421407818,
      "selected_pairs": 27247.0,
      "state_query_norm": 42.157677459716794,
      "state_relation_loss": 9.694512409623713e-05,
      "step": 21140,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0013448200421407818,
      "weighted_relation_loss": 0.0013448200421407818
    },
    {
      "action_query_norm": 39.95561141967774,
      "action_relation_loss": 0.0012999893398955465,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.78484565014032,
      "flow_matching_loss": 0.020554067753255367,
      "grad_norm": 0.11602123826742172,
      "learning_rate": 2.1968810928107806e-05,
      "loss": 0.022,
      "relation_loss": 0.0013985559809952975,
      "selected_pairs": 26671.1,
      "state_query_norm": 42.153150177001955,
      "state_relation_loss": 9.856666147243232e-05,
      "step": 21150,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0013985559809952975,
      "weighted_relation_loss": 0.0013985559809952975
    },
    {
      "action_query_norm": 39.97653160095215,
      "action_relation_loss": 0.0012216299423016609,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.794200187090738,
      "flow_matching_loss": 0.020789687521755695,
      "grad_norm": 0.13471324741840363,
      "learning_rate": 2.1923188307585606e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013206006609834731,
      "selected_pairs": 26278.6,
      "state_query_norm": 42.16318244934082,
      "state_relation_loss": 9.897072668536566e-05,
      "step": 21160,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0013206006609834731,
      "weighted_relation_loss": 0.0013206006609834731
    },
    {
      "action_query_norm": 40.02685432434082,
      "action_relation_loss": 0.0012658439809456468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.80355472404116,
      "flow_matching_loss": 0.02036317586898804,
      "grad_norm": 0.12016962468624115,
      "learning_rate": 2.187759980303708e-05,
      "loss": 0.0217,
      "relation_loss": 0.001365643530152738,
      "selected_pairs": 27202.0,
      "state_query_norm": 42.17038345336914,
      "state_relation_loss": 9.979956230381504e-05,
      "step": 21170,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001365643530152738,
      "weighted_relation_loss": 0.001365643530152738
    },
    {
      "action_query_norm": 39.94098587036133,
      "action_relation_loss": 0.001316147076431662,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.81290926099158,
      "flow_matching_loss": 0.022670208103954793,
      "grad_norm": 0.11753913015127182,
      "learning_rate": 2.1832045469856545e-05,
      "loss": 0.0241,
      "relation_loss": 0.0014116168022155762,
      "selected_pairs": 26600.1,
      "state_query_norm": 42.150089263916016,
      "state_relation_loss": 9.546972796670161e-05,
      "step": 21180,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0014116168022155762,
      "weighted_relation_loss": 0.0014116168022155762
    },
    {
      "action_query_norm": 39.96160621643067,
      "action_relation_loss": 0.0012275519315153361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.822263797942004,
      "flow_matching_loss": 0.020219802297651766,
      "grad_norm": 0.09427033364772797,
      "learning_rate": 2.1786525363396844e-05,
      "loss": 0.0215,
      "relation_loss": 0.001327409001532942,
      "selected_pairs": 26992.0,
      "state_query_norm": 42.15147476196289,
      "state_relation_loss": 9.985708311432973e-05,
      "step": 21190,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.001327409001532942,
      "weighted_relation_loss": 0.001327409001532942
    },
    {
      "action_query_norm": 39.96085777282715,
      "action_relation_loss": 0.0012457962380722164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.831618334892422,
      "flow_matching_loss": 0.020794902369379996,
      "grad_norm": 0.10353577882051468,
      "learning_rate": 2.1741039538969183e-05,
      "loss": 0.0221,
      "relation_loss": 0.001346695888787508,
      "selected_pairs": 26465.1,
      "state_query_norm": 42.15063629150391,
      "state_relation_loss": 0.0001008996507152915,
      "step": 21200,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.001346695888787508,
      "weighted_relation_loss": 0.001346695888787508
    },
    {
      "action_query_norm": 39.932503128051756,
      "action_relation_loss": 0.0013089584768749773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.840972871842844,
      "flow_matching_loss": 0.02118366342037916,
      "grad_norm": 0.11659814417362213,
      "learning_rate": 2.169558805184313e-05,
      "loss": 0.0226,
      "relation_loss": 0.0014124080655165017,
      "selected_pairs": 26685.3,
      "state_query_norm": 42.139923858642575,
      "state_relation_loss": 0.00010344959300709889,
      "step": 21210,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0014124080655165017,
      "weighted_relation_loss": 0.0014124080655165017
    },
    {
      "action_query_norm": 39.977905654907225,
      "action_relation_loss": 0.001307241292670369,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.850327408793266,
      "flow_matching_loss": 0.021537570282816887,
      "grad_norm": 0.1063704639673233,
      "learning_rate": 2.165017095724651e-05,
      "loss": 0.0229,
      "relation_loss": 0.0014039500965736806,
      "selected_pairs": 26649.2,
      "state_query_norm": 42.15671653747559,
      "state_relation_loss": 9.670879881014116e-05,
      "step": 21220,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0014039500965736806,
      "weighted_relation_loss": 0.0014039500965736806
    },
    {
      "action_query_norm": 39.960968780517575,
      "action_relation_loss": 0.0012823599390685558,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.859681945743684,
      "flow_matching_loss": 0.020704510807991027,
      "grad_norm": 0.11247673630714417,
      "learning_rate": 2.1604788310365405e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013872992480173708,
      "selected_pairs": 26599.7,
      "state_query_norm": 42.157656860351565,
      "state_relation_loss": 0.00010493930531083607,
      "step": 21230,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013872992480173708,
      "weighted_relation_loss": 0.0013872992480173708
    },
    {
      "action_query_norm": 39.93736038208008,
      "action_relation_loss": 0.0012933094520121812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.869036482694106,
      "flow_matching_loss": 0.018733843602240084,
      "grad_norm": 0.09392204880714417,
      "learning_rate": 2.155944016634398e-05,
      "loss": 0.0201,
      "relation_loss": 0.0013944860897026957,
      "selected_pairs": 26420.0,
      "state_query_norm": 42.15266342163086,
      "state_relation_loss": 0.00010117664569406771,
      "step": 21240,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0013944860897026957,
      "weighted_relation_loss": 0.0013944860897026957
    },
    {
      "action_query_norm": 39.96587982177734,
      "action_relation_loss": 0.0012649771757423877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.87839101964453,
      "flow_matching_loss": 0.019339745119214057,
      "grad_norm": 0.09283824265003204,
      "learning_rate": 2.1514126580284538e-05,
      "loss": 0.0207,
      "relation_loss": 0.0013717402587644755,
      "selected_pairs": 26823.3,
      "state_query_norm": 42.14065933227539,
      "state_relation_loss": 0.00010676309975679032,
      "step": 21250,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013717402587644755,
      "weighted_relation_loss": 0.0013717402587644755
    },
    {
      "action_query_norm": 39.96311378479004,
      "action_relation_loss": 0.0012887789285741746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.88774555659495,
      "flow_matching_loss": 0.019449480809271336,
      "grad_norm": 0.13314153254032135,
      "learning_rate": 2.1468847607247344e-05,
      "loss": 0.0208,
      "relation_loss": 0.001385584648232907,
      "selected_pairs": 27406.0,
      "state_query_norm": 42.15354995727539,
      "state_relation_loss": 9.680571383796632e-05,
      "step": 21260,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001385584648232907,
      "weighted_relation_loss": 0.001385584648232907
    },
    {
      "action_query_norm": 39.99496955871582,
      "action_relation_loss": 0.0012077821418642997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.89710009354537,
      "flow_matching_loss": 0.020227291993796824,
      "grad_norm": 0.12056651711463928,
      "learning_rate": 2.1423603302250623e-05,
      "loss": 0.0215,
      "relation_loss": 0.0013037784723564983,
      "selected_pairs": 27173.8,
      "state_query_norm": 42.151367568969725,
      "state_relation_loss": 9.599632758181543e-05,
      "step": 21270,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0013037784723564983,
      "weighted_relation_loss": 0.0013037784723564983
    },
    {
      "action_query_norm": 40.00656967163086,
      "action_relation_loss": 0.0012557493173517287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.90645463049579,
      "flow_matching_loss": 0.021262295357882977,
      "grad_norm": 0.30829113721847534,
      "learning_rate": 2.1378393720270468e-05,
      "loss": 0.0226,
      "relation_loss": 0.0013517491519451142,
      "selected_pairs": 26893.0,
      "state_query_norm": 42.1614875793457,
      "state_relation_loss": 9.599983459338546e-05,
      "step": 21280,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0013517491519451142,
      "weighted_relation_loss": 0.0013517491519451142
    },
    {
      "action_query_norm": 39.97554969787598,
      "action_relation_loss": 0.0012358531006611885,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.915809167446213,
      "flow_matching_loss": 0.021126525849103926,
      "grad_norm": 0.11790797114372253,
      "learning_rate": 2.133321891624076e-05,
      "loss": 0.0225,
      "relation_loss": 0.0013320722617208957,
      "selected_pairs": 26942.4,
      "state_query_norm": 42.132463455200195,
      "state_relation_loss": 9.621915887692012e-05,
      "step": 21290,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013320722617208957,
      "weighted_relation_loss": 0.0013320722617208957
    },
    {
      "action_query_norm": 39.97203521728515,
      "action_relation_loss": 0.0012978353071957827,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.92516370439663,
      "flow_matching_loss": 0.020447548292577267,
      "grad_norm": 0.09101670980453491,
      "learning_rate": 2.1288078945053193e-05,
      "loss": 0.0218,
      "relation_loss": 0.001400456903502345,
      "selected_pairs": 26506.7,
      "state_query_norm": 42.132460021972655,
      "state_relation_loss": 0.00010262160285492427,
      "step": 21300,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.001400456903502345,
      "weighted_relation_loss": 0.001400456903502345
    },
    {
      "action_query_norm": 39.94186172485352,
      "action_relation_loss": 0.0012507562292739748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.934518241347053,
      "flow_matching_loss": 0.019197575375437738,
      "grad_norm": 0.09526655077934265,
      "learning_rate": 2.1242973861557064e-05,
      "loss": 0.0205,
      "relation_loss": 0.0013448367011733354,
      "selected_pairs": 26523.8,
      "state_query_norm": 42.14160919189453,
      "state_relation_loss": 9.408047553733922e-05,
      "step": 21310,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0013448367011733354,
      "weighted_relation_loss": 0.0013448367011733354
    },
    {
      "action_query_norm": 40.030877685546876,
      "action_relation_loss": 0.0012811258202418685,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.943872778297475,
      "flow_matching_loss": 0.019299982115626335,
      "grad_norm": 0.09781738370656967,
      "learning_rate": 2.1197903720559304e-05,
      "loss": 0.0207,
      "relation_loss": 0.0013809973024763167,
      "selected_pairs": 27152.9,
      "state_query_norm": 42.13494606018067,
      "state_relation_loss": 9.987149387598037e-05,
      "step": 21320,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0013809973024763167,
      "weighted_relation_loss": 0.0013809973024763167
    },
    {
      "action_query_norm": 39.96596450805664,
      "action_relation_loss": 0.0012795282760635017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.953227315247894,
      "flow_matching_loss": 0.019851512648165225,
      "grad_norm": 0.10717567801475525,
      "learning_rate": 2.115286857682438e-05,
      "loss": 0.0212,
      "relation_loss": 0.0013739783666096629,
      "selected_pairs": 26255.2,
      "state_query_norm": 42.138912963867185,
      "state_relation_loss": 9.445007453905419e-05,
      "step": 21330,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0013739783666096629,
      "weighted_relation_loss": 0.0013739783666096629
    },
    {
      "action_query_norm": 39.95992317199707,
      "action_relation_loss": 0.0011876642354764044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.962581852198316,
      "flow_matching_loss": 0.021654540300369264,
      "grad_norm": 0.10501206666231155,
      "learning_rate": 2.1107868485074228e-05,
      "loss": 0.0229,
      "relation_loss": 0.0012831983040086925,
      "selected_pairs": 26736.3,
      "state_query_norm": 42.136592864990234,
      "state_relation_loss": 9.553407217026688e-05,
      "step": 21340,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012831983040086925,
      "weighted_relation_loss": 0.0012831983040086925
    },
    {
      "action_query_norm": 39.94843940734863,
      "action_relation_loss": 0.0012792643276043237,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.971936389148738,
      "flow_matching_loss": 0.02317184740677476,
      "grad_norm": 0.12858878076076508,
      "learning_rate": 2.1062903499988234e-05,
      "loss": 0.0245,
      "relation_loss": 0.0013695070403628052,
      "selected_pairs": 27263.0,
      "state_query_norm": 42.1526985168457,
      "state_relation_loss": 9.024270038935356e-05,
      "step": 21350,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0013695070403628052,
      "weighted_relation_loss": 0.0013695070403628052
    },
    {
      "action_query_norm": 39.95327529907227,
      "action_relation_loss": 0.0011971765547059477,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.98129092609916,
      "flow_matching_loss": 0.019198240153491496,
      "grad_norm": 0.09121041744947433,
      "learning_rate": 2.101797367620308e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012975119869224727,
      "selected_pairs": 26132.5,
      "state_query_norm": 42.14197425842285,
      "state_relation_loss": 0.00010033541839220561,
      "step": 21360,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0012975119869224727,
      "weighted_relation_loss": 0.0012975119869224727
    },
    {
      "action_query_norm": 39.95530204772949,
      "action_relation_loss": 0.0012918217689730227,
      "auxiliary_loss_weight": 1.0,
      "epoch": 19.990645463049578,
      "flow_matching_loss": 0.020710357278585435,
      "grad_norm": 0.11281051486730576,
      "learning_rate": 2.0973079068312713e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013919610297307371,
      "selected_pairs": 26946.2,
      "state_query_norm": 42.15292320251465,
      "state_relation_loss": 0.00010013925784733147,
      "step": 21370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0013919610297307371,
      "weighted_relation_loss": 0.0013919610297307371
    },
    {
      "action_query_norm": 39.96372833251953,
      "action_relation_loss": 0.0012741588870994747,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.0,
      "flow_matching_loss": 0.0217569125816226,
      "grad_norm": 0.17454160749912262,
      "learning_rate": 2.0928219730868358e-05,
      "loss": 0.0231,
      "relation_loss": 0.001369844109285623,
      "selected_pairs": 24176.4,
      "state_query_norm": 42.14226112365723,
      "state_relation_loss": 9.568521127221175e-05,
      "step": 21380,
      "valid_action_samples": 213.9,
      "weighted_auxiliary_loss": 0.001369844109285623,
      "weighted_relation_loss": 0.001369844109285623
    },
    {
      "action_query_norm": 39.89007339477539,
      "action_relation_loss": 0.001263542112428695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.009354536950422,
      "flow_matching_loss": 0.021306365728378296,
      "grad_norm": 0.1944422721862793,
      "learning_rate": 2.0883395718378302e-05,
      "loss": 0.0227,
      "relation_loss": 0.0013601756654679775,
      "selected_pairs": 26738.5,
      "state_query_norm": 42.15898513793945,
      "state_relation_loss": 9.663355303928256e-05,
      "step": 21390,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013601756654679775,
      "weighted_relation_loss": 0.0013601756654679775
    },
    {
      "action_query_norm": 39.94220161437988,
      "action_relation_loss": 0.0011459582019597292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.01870907390084,
      "flow_matching_loss": 0.019134326372295618,
      "grad_norm": 0.08595533668994904,
      "learning_rate": 2.083860708530798e-05,
      "loss": 0.0204,
      "relation_loss": 0.001234751520678401,
      "selected_pairs": 27358.8,
      "state_query_norm": 42.160699462890626,
      "state_relation_loss": 8.87933092599269e-05,
      "step": 21400,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001234751520678401,
      "weighted_relation_loss": 0.001234751520678401
    },
    {
      "action_query_norm": 39.94772911071777,
      "action_relation_loss": 0.0011870717047713698,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.028063610851262,
      "flow_matching_loss": 0.02167115742340684,
      "grad_norm": 0.10390297323465347,
      "learning_rate": 2.0793853886079796e-05,
      "loss": 0.0229,
      "relation_loss": 0.0012786805862560868,
      "selected_pairs": 27408.7,
      "state_query_norm": 42.14315910339356,
      "state_relation_loss": 9.160886984318494e-05,
      "step": 21410,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0012786805862560868,
      "weighted_relation_loss": 0.0012786805862560868
    },
    {
      "action_query_norm": 39.96446647644043,
      "action_relation_loss": 0.001156642474234104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.037418147801684,
      "flow_matching_loss": 0.0186823520809412,
      "grad_norm": 0.1270509660243988,
      "learning_rate": 2.074913617507309e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012523608631454409,
      "selected_pairs": 26354.4,
      "state_query_norm": 42.144915390014646,
      "state_relation_loss": 9.571838163537904e-05,
      "step": 21420,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0012523608631454409,
      "weighted_relation_loss": 0.0012523608631454409
    },
    {
      "action_query_norm": 39.92111167907715,
      "action_relation_loss": 0.0012529383995570243,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.046772684752106,
      "flow_matching_loss": 0.02095804866403341,
      "grad_norm": 0.13922740519046783,
      "learning_rate": 2.070445400662411e-05,
      "loss": 0.0223,
      "relation_loss": 0.00134264713851735,
      "selected_pairs": 27312.9,
      "state_query_norm": 42.14449119567871,
      "state_relation_loss": 8.970872877398506e-05,
      "step": 21430,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.00134264713851735,
      "weighted_relation_loss": 0.00134264713851735
    },
    {
      "action_query_norm": 39.944580459594725,
      "action_relation_loss": 0.0011990739381872117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.056127221702525,
      "flow_matching_loss": 0.01895281858742237,
      "grad_norm": 0.10137995332479477,
      "learning_rate": 2.0659807435025907e-05,
      "loss": 0.0202,
      "relation_loss": 0.001294634025543928,
      "selected_pairs": 27452.7,
      "state_query_norm": 42.141627883911134,
      "state_relation_loss": 9.556008590152487e-05,
      "step": 21440,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.001294634025543928,
      "weighted_relation_loss": 0.001294634025543928
    },
    {
      "action_query_norm": 39.98094367980957,
      "action_relation_loss": 0.0011381675256416201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.065481758652947,
      "flow_matching_loss": 0.019041672069579362,
      "grad_norm": 0.11249582469463348,
      "learning_rate": 2.061519651452825e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012371636694297194,
      "selected_pairs": 26577.0,
      "state_query_norm": 42.122514724731445,
      "state_relation_loss": 9.899613287416287e-05,
      "step": 21450,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012371636694297194,
      "weighted_relation_loss": 0.0012371636694297194
    },
    {
      "action_query_norm": 39.968896484375,
      "action_relation_loss": 0.0011580251273699104,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.07483629560337,
      "flow_matching_loss": 0.01970114652067423,
      "grad_norm": 0.12561017274856567,
      "learning_rate": 2.0570621299337656e-05,
      "loss": 0.021,
      "relation_loss": 0.0012504118261858822,
      "selected_pairs": 27388.4,
      "state_query_norm": 42.15165901184082,
      "state_relation_loss": 9.238668935722672e-05,
      "step": 21460,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0012504118261858822,
      "weighted_relation_loss": 0.0012504118261858822
    },
    {
      "action_query_norm": 39.97566604614258,
      "action_relation_loss": 0.0012102088541723788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.084190832553787,
      "flow_matching_loss": 0.02087612133473158,
      "grad_norm": 0.10379065573215485,
      "learning_rate": 2.0526081843617183e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013076714123599232,
      "selected_pairs": 26558.4,
      "state_query_norm": 42.13507080078125,
      "state_relation_loss": 9.74625610979274e-05,
      "step": 21470,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013076714123599232,
      "weighted_relation_loss": 0.0013076714123599232
    },
    {
      "action_query_norm": 39.91232757568359,
      "action_relation_loss": 0.0012297102948650718,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.09354536950421,
      "flow_matching_loss": 0.021481387503445148,
      "grad_norm": 0.10269544273614883,
      "learning_rate": 2.0481578201486485e-05,
      "loss": 0.0228,
      "relation_loss": 0.0013237533159554005,
      "selected_pairs": 26984.8,
      "state_query_norm": 42.13330230712891,
      "state_relation_loss": 9.404300581081771e-05,
      "step": 21480,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0013237533159554005,
      "weighted_relation_loss": 0.0013237533159554005
    },
    {
      "action_query_norm": 39.90037841796875,
      "action_relation_loss": 0.0011893431190401315,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.10289990645463,
      "flow_matching_loss": 0.017836977355182172,
      "grad_norm": 0.08982047438621521,
      "learning_rate": 2.0437110427021677e-05,
      "loss": 0.0191,
      "relation_loss": 0.001285704062320292,
      "selected_pairs": 26360.7,
      "state_query_norm": 42.1354434967041,
      "state_relation_loss": 9.636093018343672e-05,
      "step": 21490,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001285704062320292,
      "weighted_relation_loss": 0.001285704062320292
    },
    {
      "action_query_norm": 39.93722686767578,
      "action_relation_loss": 0.0012518927920609712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.112254443405053,
      "flow_matching_loss": 0.01897554164752364,
      "grad_norm": 0.11887842416763306,
      "learning_rate": 2.0392678574255282e-05,
      "loss": 0.0203,
      "relation_loss": 0.0013521016458980738,
      "selected_pairs": 26643.8,
      "state_query_norm": 42.13096694946289,
      "state_relation_loss": 0.00010020884583354928,
      "step": 21500,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0013521016458980738,
      "weighted_relation_loss": 0.0013521016458980738
    },
    {
      "action_query_norm": 39.935355758666994,
      "action_relation_loss": 0.0012511160457506776,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.12160898035547,
      "flow_matching_loss": 0.0201396694406867,
      "grad_norm": 0.09702558815479279,
      "learning_rate": 2.0348282697176217e-05,
      "loss": 0.0215,
      "relation_loss": 0.0013464963063597678,
      "selected_pairs": 27246.9,
      "state_query_norm": 42.12908515930176,
      "state_relation_loss": 9.538026424706913e-05,
      "step": 21510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0013464963063597678,
      "weighted_relation_loss": 0.0013464963063597678
    },
    {
      "action_query_norm": 39.88538665771485,
      "action_relation_loss": 0.0012928323703818023,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.130963517305894,
      "flow_matching_loss": 0.02097083367407322,
      "grad_norm": 0.09526582807302475,
      "learning_rate": 2.030392284972964e-05,
      "loss": 0.0224,
      "relation_loss": 0.0013906221836805343,
      "selected_pairs": 27201.9,
      "state_query_norm": 42.12929420471191,
      "state_relation_loss": 9.778981620911509e-05,
      "step": 21520,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0013906221836805343,
      "weighted_relation_loss": 0.0013906221836805343
    },
    {
      "action_query_norm": 39.90708503723145,
      "action_relation_loss": 0.001178120751865208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.140318054256316,
      "flow_matching_loss": 0.019028378743678333,
      "grad_norm": 0.07397874444723129,
      "learning_rate": 2.0259599085816973e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012696652906015516,
      "selected_pairs": 26943.3,
      "state_query_norm": 42.133581924438474,
      "state_relation_loss": 9.15445365535561e-05,
      "step": 21530,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0012696652906015516,
      "weighted_relation_loss": 0.0012696652906015516
    },
    {
      "action_query_norm": 39.89813804626465,
      "action_relation_loss": 0.001214497280307114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.149672591206734,
      "flow_matching_loss": 0.020351259782910348,
      "grad_norm": 0.12170829623937607,
      "learning_rate": 2.0215311459295755e-05,
      "loss": 0.0217,
      "relation_loss": 0.0013032208080403507,
      "selected_pairs": 26604.4,
      "state_query_norm": 42.135083770751955,
      "state_relation_loss": 8.872353573678993e-05,
      "step": 21540,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013032208080403507,
      "weighted_relation_loss": 0.0013032208080403507
    },
    {
      "action_query_norm": 39.91637992858887,
      "action_relation_loss": 0.0011905113991815598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.159027128157156,
      "flow_matching_loss": 0.020706545375287534,
      "grad_norm": 0.10388089716434479,
      "learning_rate": 2.0171060023979606e-05,
      "loss": 0.022,
      "relation_loss": 0.0012845281627960504,
      "selected_pairs": 26413.9,
      "state_query_norm": 42.14214515686035,
      "state_relation_loss": 9.401677234563977e-05,
      "step": 21550,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0012845281627960504,
      "weighted_relation_loss": 0.0012845281627960504
    },
    {
      "action_query_norm": 39.93917732238769,
      "action_relation_loss": 0.0011624447710346431,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.168381665107578,
      "flow_matching_loss": 0.019836008176207544,
      "grad_norm": 0.10077174752950668,
      "learning_rate": 2.012684483363823e-05,
      "loss": 0.0211,
      "relation_loss": 0.00125814852071926,
      "selected_pairs": 26552.2,
      "state_query_norm": 42.11079559326172,
      "state_relation_loss": 9.570374750182964e-05,
      "step": 21560,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.00125814852071926,
      "weighted_relation_loss": 0.00125814852071926
    },
    {
      "action_query_norm": 39.936088943481444,
      "action_relation_loss": 0.0011748877703212201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.177736202057996,
      "flow_matching_loss": 0.01906461101025343,
      "grad_norm": 0.11775676161050797,
      "learning_rate": 2.0082665941997237e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012666489463299512,
      "selected_pairs": 26717.2,
      "state_query_norm": 42.130761337280276,
      "state_relation_loss": 9.176117964670993e-05,
      "step": 21570,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0012666489463299512,
      "weighted_relation_loss": 0.0012666489463299512
    },
    {
      "action_query_norm": 39.93612060546875,
      "action_relation_loss": 0.001189900829922408,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.18709073900842,
      "flow_matching_loss": 0.01998846959322691,
      "grad_norm": 0.12675562500953674,
      "learning_rate": 2.0038523402738146e-05,
      "loss": 0.0213,
      "relation_loss": 0.0012835322064347564,
      "selected_pairs": 27012.1,
      "state_query_norm": 42.12484436035156,
      "state_relation_loss": 9.363137723994441e-05,
      "step": 21580,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012835322064347564,
      "weighted_relation_loss": 0.0012835322064347564
    },
    {
      "action_query_norm": 39.95798034667969,
      "action_relation_loss": 0.0012036658823490142,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.19644527595884,
      "flow_matching_loss": 0.019309968687593937,
      "grad_norm": 0.1165536493062973,
      "learning_rate": 1.99944172694983e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012969928444363176,
      "selected_pairs": 26811.3,
      "state_query_norm": 42.12366409301758,
      "state_relation_loss": 9.332696718047372e-05,
      "step": 21590,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012969928444363176,
      "weighted_relation_loss": 0.0012969928444363176
    },
    {
      "action_query_norm": 39.95052375793457,
      "action_relation_loss": 0.0011814158642664552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.205799812909262,
      "flow_matching_loss": 0.020590866915881634,
      "grad_norm": 0.1118086650967598,
      "learning_rate": 1.99503475958708e-05,
      "loss": 0.0219,
      "relation_loss": 0.0012789497384801507,
      "selected_pairs": 27094.4,
      "state_query_norm": 42.11769828796387,
      "state_relation_loss": 9.753387639648281e-05,
      "step": 21600,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0012789497384801507,
      "weighted_relation_loss": 0.0012789497384801507
    },
    {
      "action_query_norm": 39.96512336730957,
      "action_relation_loss": 0.0011663827230222524,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.21515434985968,
      "flow_matching_loss": 0.019771352782845496,
      "grad_norm": 0.1340080350637436,
      "learning_rate": 1.9906314435404484e-05,
      "loss": 0.021,
      "relation_loss": 0.001258089707698673,
      "selected_pairs": 26645.8,
      "state_query_norm": 42.12712364196777,
      "state_relation_loss": 9.170698176603764e-05,
      "step": 21610,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.001258089707698673,
      "weighted_relation_loss": 0.001258089707698673
    },
    {
      "action_query_norm": 39.908224487304686,
      "action_relation_loss": 0.0011874091578647494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.224508886810103,
      "flow_matching_loss": 0.019721670635044576,
      "grad_norm": 0.07790432125329971,
      "learning_rate": 1.986231784160378e-05,
      "loss": 0.021,
      "relation_loss": 0.001282307947985828,
      "selected_pairs": 27220.9,
      "state_query_norm": 42.10391693115234,
      "state_relation_loss": 9.489881194895134e-05,
      "step": 21620,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.001282307947985828,
      "weighted_relation_loss": 0.001282307947985828
    },
    {
      "action_query_norm": 39.928092575073244,
      "action_relation_loss": 0.0012198311160318553,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.233863423760525,
      "flow_matching_loss": 0.020771611481904984,
      "grad_norm": 0.12599608302116394,
      "learning_rate": 1.9818357867928693e-05,
      "loss": 0.0221,
      "relation_loss": 0.0013121752417646348,
      "selected_pairs": 26689.7,
      "state_query_norm": 42.115119552612306,
      "state_relation_loss": 9.234413009835407e-05,
      "step": 21630,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0013121752417646348,
      "weighted_relation_loss": 0.0013121752417646348
    },
    {
      "action_query_norm": 39.94809684753418,
      "action_relation_loss": 0.001214770064689219,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.243217960710943,
      "flow_matching_loss": 0.020901925675570966,
      "grad_norm": 0.13528560101985931,
      "learning_rate": 1.977443456779474e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013118103030137717,
      "selected_pairs": 26761.5,
      "state_query_norm": 42.110638809204104,
      "state_relation_loss": 9.704022668302059e-05,
      "step": 21640,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0013118103030137717,
      "weighted_relation_loss": 0.0013118103030137717
    },
    {
      "action_query_norm": 39.94574127197266,
      "action_relation_loss": 0.0012124769389629364,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.252572497661365,
      "flow_matching_loss": 0.020967567805200815,
      "grad_norm": 0.10156827419996262,
      "learning_rate": 1.9730547994572858e-05,
      "loss": 0.0223,
      "relation_loss": 0.0013012596289627254,
      "selected_pairs": 26594.9,
      "state_query_norm": 42.137931442260744,
      "state_relation_loss": 8.878270236891694e-05,
      "step": 21650,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013012596289627254,
      "weighted_relation_loss": 0.0013012596289627254
    },
    {
      "action_query_norm": 39.94211196899414,
      "action_relation_loss": 0.0012278339941985904,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.261927034611787,
      "flow_matching_loss": 0.020669012516736984,
      "grad_norm": 0.09894713014364243,
      "learning_rate": 1.9686698201589393e-05,
      "loss": 0.022,
      "relation_loss": 0.001320530811790377,
      "selected_pairs": 26300.7,
      "state_query_norm": 42.12599334716797,
      "state_relation_loss": 9.269682632293553e-05,
      "step": 21660,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.001320530811790377,
      "weighted_relation_loss": 0.001320530811790377
    },
    {
      "action_query_norm": 39.94301261901855,
      "action_relation_loss": 0.0012602153467014432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.27128157156221,
      "flow_matching_loss": 0.02044845847412944,
      "grad_norm": 0.10365881025791168,
      "learning_rate": 1.9642885242125964e-05,
      "loss": 0.0218,
      "relation_loss": 0.0013515724567696452,
      "selected_pairs": 27103.9,
      "state_query_norm": 42.12279739379883,
      "state_relation_loss": 9.135710934060626e-05,
      "step": 21670,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0013515724567696452,
      "weighted_relation_loss": 0.0013515724567696452
    },
    {
      "action_query_norm": 39.89770431518555,
      "action_relation_loss": 0.0011932572699151934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.280636108512628,
      "flow_matching_loss": 0.02027155226096511,
      "grad_norm": 0.10154879093170166,
      "learning_rate": 1.9599109169419466e-05,
      "loss": 0.0216,
      "relation_loss": 0.001287307927850634,
      "selected_pairs": 26213.2,
      "state_query_norm": 42.10403327941894,
      "state_relation_loss": 9.405066739418544e-05,
      "step": 21680,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.001287307927850634,
      "weighted_relation_loss": 0.001287307927850634
    },
    {
      "action_query_norm": 39.94352493286133,
      "action_relation_loss": 0.0011852406547404825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.28999064546305,
      "flow_matching_loss": 0.019677024427801372,
      "grad_norm": 0.1612589955329895,
      "learning_rate": 1.9555370036661948e-05,
      "loss": 0.021,
      "relation_loss": 0.0012805170961655676,
      "selected_pairs": 26351.1,
      "state_query_norm": 42.1171573638916,
      "state_relation_loss": 9.527645233902149e-05,
      "step": 21690,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0012805170961655676,
      "weighted_relation_loss": 0.0012805170961655676
    },
    {
      "action_query_norm": 39.935424423217775,
      "action_relation_loss": 0.001119609666056931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.29934518241347,
      "flow_matching_loss": 0.019334164820611476,
      "grad_norm": 0.10255340486764908,
      "learning_rate": 1.9511667897000574e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012144611682742834,
      "selected_pairs": 27743.2,
      "state_query_norm": 42.107683181762695,
      "state_relation_loss": 9.485149203101174e-05,
      "step": 21700,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.0012144611682742834,
      "weighted_relation_loss": 0.0012144611682742834
    },
    {
      "action_query_norm": 39.900504684448244,
      "action_relation_loss": 0.0012105027912184597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.30869971936389,
      "flow_matching_loss": 0.0201939451508224,
      "grad_norm": 0.10651721060276031,
      "learning_rate": 1.946800280353755e-05,
      "loss": 0.0215,
      "relation_loss": 0.0013017675606533884,
      "selected_pairs": 26919.3,
      "state_query_norm": 42.099172592163086,
      "state_relation_loss": 9.126475415541791e-05,
      "step": 21710,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0013017675606533884,
      "weighted_relation_loss": 0.0013017675606533884
    },
    {
      "action_query_norm": 39.89994087219238,
      "action_relation_loss": 0.0011792245670221746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.318054256314312,
      "flow_matching_loss": 0.018935921415686608,
      "grad_norm": 0.12993702292442322,
      "learning_rate": 1.9424374809330114e-05,
      "loss": 0.0202,
      "relation_loss": 0.0012710008537396788,
      "selected_pairs": 26574.7,
      "state_query_norm": 42.11882400512695,
      "state_relation_loss": 9.177628453471698e-05,
      "step": 21720,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012710008537396788,
      "weighted_relation_loss": 0.0012710008537396788
    },
    {
      "action_query_norm": 39.9672248840332,
      "action_relation_loss": 0.001187653187662363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.327408793264734,
      "flow_matching_loss": 0.0202236901037395,
      "grad_norm": 0.10504137724637985,
      "learning_rate": 1.9380783967390382e-05,
      "loss": 0.0215,
      "relation_loss": 0.0012800081283785402,
      "selected_pairs": 26828.6,
      "state_query_norm": 42.10540618896484,
      "state_relation_loss": 9.235493635060266e-05,
      "step": 21730,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012800081283785402,
      "weighted_relation_loss": 0.0012800081283785402
    },
    {
      "action_query_norm": 39.9108283996582,
      "action_relation_loss": 0.0011877709184773266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.336763330215156,
      "flow_matching_loss": 0.01935883816331625,
      "grad_norm": 0.11892864853143692,
      "learning_rate": 1.933723033068533e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012830957886762917,
      "selected_pairs": 26989.8,
      "state_query_norm": 42.119862747192386,
      "state_relation_loss": 9.532485782983712e-05,
      "step": 21740,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0012830957886762917,
      "weighted_relation_loss": 0.0012830957886762917
    },
    {
      "action_query_norm": 39.90371742248535,
      "action_relation_loss": 0.0012348280870355666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.346117867165574,
      "flow_matching_loss": 0.01942331586033106,
      "grad_norm": 0.14085181057453156,
      "learning_rate": 1.929371395213674e-05,
      "loss": 0.0207,
      "relation_loss": 0.001326076511759311,
      "selected_pairs": 26476.8,
      "state_query_norm": 42.12416725158691,
      "state_relation_loss": 9.124841380980797e-05,
      "step": 21750,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001326076511759311,
      "weighted_relation_loss": 0.001326076511759311
    },
    {
      "action_query_norm": 39.88364524841309,
      "action_relation_loss": 0.0011820114217698575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.355472404115996,
      "flow_matching_loss": 0.018751507718116046,
      "grad_norm": 0.09868934005498886,
      "learning_rate": 1.9250234884621092e-05,
      "loss": 0.02,
      "relation_loss": 0.001275154936593026,
      "selected_pairs": 26893.6,
      "state_query_norm": 42.10732307434082,
      "state_relation_loss": 9.314351627836004e-05,
      "step": 21760,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001275154936593026,
      "weighted_relation_loss": 0.001275154936593026
    },
    {
      "action_query_norm": 39.890719985961915,
      "action_relation_loss": 0.0011630691180471331,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.36482694106642,
      "flow_matching_loss": 0.019536394253373147,
      "grad_norm": 0.08295107632875443,
      "learning_rate": 1.920679318096959e-05,
      "loss": 0.0208,
      "relation_loss": 0.0012586457189172507,
      "selected_pairs": 26781.5,
      "state_query_norm": 42.10471649169922,
      "state_relation_loss": 9.557659723213874e-05,
      "step": 21770,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0012586457189172507,
      "weighted_relation_loss": 0.0012586457189172507
    },
    {
      "action_query_norm": 39.8419303894043,
      "action_relation_loss": 0.0012209031614474953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.374181478016837,
      "flow_matching_loss": 0.020849515125155448,
      "grad_norm": 0.160101056098938,
      "learning_rate": 1.9163388893967982e-05,
      "loss": 0.0222,
      "relation_loss": 0.0013149885809980333,
      "selected_pairs": 26751.3,
      "state_query_norm": 42.11827545166015,
      "state_relation_loss": 9.408542391611263e-05,
      "step": 21780,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013149885809980333,
      "weighted_relation_loss": 0.0013149885809980333
    },
    {
      "action_query_norm": 39.95222053527832,
      "action_relation_loss": 0.001132342719938606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.38353601496726,
      "flow_matching_loss": 0.02015575272962451,
      "grad_norm": 0.13769392669200897,
      "learning_rate": 1.9120022076356575e-05,
      "loss": 0.0214,
      "relation_loss": 0.0012297824840061366,
      "selected_pairs": 26578.3,
      "state_query_norm": 42.09641456604004,
      "state_relation_loss": 9.743975460878574e-05,
      "step": 21790,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012297824840061366,
      "weighted_relation_loss": 0.0012297824840061366
    },
    {
      "action_query_norm": 39.89014854431152,
      "action_relation_loss": 0.0011487432639114559,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.39289055191768,
      "flow_matching_loss": 0.019702461548149586,
      "grad_norm": 0.09196335077285767,
      "learning_rate": 1.9076692780830114e-05,
      "loss": 0.0209,
      "relation_loss": 0.001239239308051765,
      "selected_pairs": 26872.9,
      "state_query_norm": 42.11336898803711,
      "state_relation_loss": 9.049604632309638e-05,
      "step": 21800,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001239239308051765,
      "weighted_relation_loss": 0.001239239308051765
    },
    {
      "action_query_norm": 39.916662979125974,
      "action_relation_loss": 0.001184418157208711,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.4022450888681,
      "flow_matching_loss": 0.01845280881971121,
      "grad_norm": 0.11238210648298264,
      "learning_rate": 1.903340106003782e-05,
      "loss": 0.0197,
      "relation_loss": 0.0012753291055560112,
      "selected_pairs": 27103.6,
      "state_query_norm": 42.10560493469238,
      "state_relation_loss": 9.091094616451301e-05,
      "step": 21810,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012753291055560112,
      "weighted_relation_loss": 0.0012753291055560112
    },
    {
      "action_query_norm": 39.88643188476563,
      "action_relation_loss": 0.0011290973285213112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.41159962581852,
      "flow_matching_loss": 0.01993736168369651,
      "grad_norm": 0.14602312445640564,
      "learning_rate": 1.8990146966583183e-05,
      "loss": 0.0212,
      "relation_loss": 0.0012245686026290058,
      "selected_pairs": 26508.3,
      "state_query_norm": 42.107849502563475,
      "state_relation_loss": 9.547126683173701e-05,
      "step": 21820,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0012245686026290058,
      "weighted_relation_loss": 0.0012245686026290058
    },
    {
      "action_query_norm": 39.88421058654785,
      "action_relation_loss": 0.0011603342834860086,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.420954162768943,
      "flow_matching_loss": 0.02093568667769432,
      "grad_norm": 0.09783883392810822,
      "learning_rate": 1.8946930553024035e-05,
      "loss": 0.0222,
      "relation_loss": 0.0012561383191496134,
      "selected_pairs": 26831.3,
      "state_query_norm": 42.10419044494629,
      "state_relation_loss": 9.580402766005137e-05,
      "step": 21830,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012561383191496134,
      "weighted_relation_loss": 0.0012561383191496134
    },
    {
      "action_query_norm": 39.910116577148436,
      "action_relation_loss": 0.0011983965523540974,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.430308699719365,
      "flow_matching_loss": 0.020905156247317792,
      "grad_norm": 0.19348381459712982,
      "learning_rate": 1.8903751871872378e-05,
      "loss": 0.0222,
      "relation_loss": 0.00129384488100186,
      "selected_pairs": 27083.6,
      "state_query_norm": 42.11476860046387,
      "state_relation_loss": 9.544832719257101e-05,
      "step": 21840,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00129384488100186,
      "weighted_relation_loss": 0.00129384488100186
    },
    {
      "action_query_norm": 39.90668563842773,
      "action_relation_loss": 0.001174066832754761,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.439663236669784,
      "flow_matching_loss": 0.020071793347597122,
      "grad_norm": 0.12124663591384888,
      "learning_rate": 1.8860610975594382e-05,
      "loss": 0.0213,
      "relation_loss": 0.0012700202642008662,
      "selected_pairs": 27196.7,
      "state_query_norm": 42.076184844970705,
      "state_relation_loss": 9.595342125976459e-05,
      "step": 21850,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0012700202642008662,
      "weighted_relation_loss": 0.0012700202642008662
    },
    {
      "action_query_norm": 39.918152236938475,
      "action_relation_loss": 0.0011797794490121306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.449017773620206,
      "flow_matching_loss": 0.019856681022793053,
      "grad_norm": 0.09853626787662506,
      "learning_rate": 1.8817507916610307e-05,
      "loss": 0.0211,
      "relation_loss": 0.0012719082995317876,
      "selected_pairs": 26897.6,
      "state_query_norm": 42.11157722473145,
      "state_relation_loss": 9.21288286917843e-05,
      "step": 21860,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0012719082995317876,
      "weighted_relation_loss": 0.0012719082995317876
    },
    {
      "action_query_norm": 39.856899642944335,
      "action_relation_loss": 0.0012415032600983978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.458372310570628,
      "flow_matching_loss": 0.02156205791980028,
      "grad_norm": 0.10913456231355667,
      "learning_rate": 1.8774442747294407e-05,
      "loss": 0.0229,
      "relation_loss": 0.001330332038924098,
      "selected_pairs": 26890.6,
      "state_query_norm": 42.10062599182129,
      "state_relation_loss": 8.882880429155194e-05,
      "step": 21870,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001330332038924098,
      "weighted_relation_loss": 0.001330332038924098
    },
    {
      "action_query_norm": 39.91998825073242,
      "action_relation_loss": 0.0012020054156892002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.467726847521046,
      "flow_matching_loss": 0.021754127554595472,
      "grad_norm": 0.17195384204387665,
      "learning_rate": 1.8731415519974965e-05,
      "loss": 0.023,
      "relation_loss": 0.0012951232376508415,
      "selected_pairs": 27057.9,
      "state_query_norm": 42.10163841247559,
      "state_relation_loss": 9.311781977885402e-05,
      "step": 21880,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0012951232376508415,
      "weighted_relation_loss": 0.0012951232376508415
    },
    {
      "action_query_norm": 39.91995506286621,
      "action_relation_loss": 0.0011781318113207816,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.477081384471468,
      "flow_matching_loss": 0.019747721124440433,
      "grad_norm": 0.1676366925239563,
      "learning_rate": 1.86884262869341e-05,
      "loss": 0.021,
      "relation_loss": 0.0012721555307507516,
      "selected_pairs": 26800.1,
      "state_query_norm": 42.10296287536621,
      "state_relation_loss": 9.402372888871469e-05,
      "step": 21890,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012721555307507516,
      "weighted_relation_loss": 0.0012721555307507516
    },
    {
      "action_query_norm": 39.924250793457034,
      "action_relation_loss": 0.0011980204959399997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.48643592142189,
      "flow_matching_loss": 0.019853494595736265,
      "grad_norm": 0.09548521041870117,
      "learning_rate": 1.8645475100407788e-05,
      "loss": 0.0211,
      "relation_loss": 0.0012926724855788052,
      "selected_pairs": 27361.3,
      "state_query_norm": 42.09858512878418,
      "state_relation_loss": 9.46519838180393e-05,
      "step": 21900,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0012926724855788052,
      "weighted_relation_loss": 0.0012926724855788052
    },
    {
      "action_query_norm": 39.92050666809082,
      "action_relation_loss": 0.0011763039859943093,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.495790458372312,
      "flow_matching_loss": 0.019353221450001002,
      "grad_norm": 0.1080058217048645,
      "learning_rate": 1.8602562012585768e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012641661101952194,
      "selected_pairs": 26368.4,
      "state_query_norm": 42.117157745361325,
      "state_relation_loss": 8.786211255937815e-05,
      "step": 21910,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0012641661101952194,
      "weighted_relation_loss": 0.0012641661101952194
    },
    {
      "action_query_norm": 39.85926742553711,
      "action_relation_loss": 0.0012736983597278595,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.50514499532273,
      "flow_matching_loss": 0.02106285635381937,
      "grad_norm": 0.15944203734397888,
      "learning_rate": 1.8559687075611466e-05,
      "loss": 0.0224,
      "relation_loss": 0.0013673272216692567,
      "selected_pairs": 27059.1,
      "state_query_norm": 42.1022274017334,
      "state_relation_loss": 9.362887212773784e-05,
      "step": 21920,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0013673272216692567,
      "weighted_relation_loss": 0.0013673272216692567
    },
    {
      "action_query_norm": 39.889198303222656,
      "action_relation_loss": 0.0011748028453439474,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.514499532273152,
      "flow_matching_loss": 0.02113397605717182,
      "grad_norm": 0.09215513616800308,
      "learning_rate": 1.8516850341582017e-05,
      "loss": 0.0224,
      "relation_loss": 0.001268610474653542,
      "selected_pairs": 26295.7,
      "state_query_norm": 42.08602294921875,
      "state_relation_loss": 9.380763003719039e-05,
      "step": 21930,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.001268610474653542,
      "weighted_relation_loss": 0.001268610474653542
    },
    {
      "action_query_norm": 39.941212844848636,
      "action_relation_loss": 0.0011680416646413505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.523854069223574,
      "flow_matching_loss": 0.0191294901072979,
      "grad_norm": 0.13075804710388184,
      "learning_rate": 1.847405186254807e-05,
      "loss": 0.0204,
      "relation_loss": 0.0012626662966795266,
      "selected_pairs": 27030.2,
      "state_query_norm": 42.096872329711914,
      "state_relation_loss": 9.462462330702692e-05,
      "step": 21940,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0012626662966795266,
      "weighted_relation_loss": 0.0012626662966795266
    },
    {
      "action_query_norm": 39.91260185241699,
      "action_relation_loss": 0.0011767750373110175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.533208606173993,
      "flow_matching_loss": 0.019242784194648265,
      "grad_norm": 0.0952269583940506,
      "learning_rate": 1.843129169051379e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012646841467358172,
      "selected_pairs": 27551.2,
      "state_query_norm": 42.09426498413086,
      "state_relation_loss": 8.790910651441663e-05,
      "step": 21950,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0012646841467358172,
      "weighted_relation_loss": 0.0012646841467358172
    },
    {
      "action_query_norm": 39.871238708496094,
      "action_relation_loss": 0.0011708749341778457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.542563143124415,
      "flow_matching_loss": 0.018594741355627775,
      "grad_norm": 0.1184527799487114,
      "learning_rate": 1.8388569877436863e-05,
      "loss": 0.0199,
      "relation_loss": 0.001257384940981865,
      "selected_pairs": 27107.6,
      "state_query_norm": 42.084221267700194,
      "state_relation_loss": 8.650998934172094e-05,
      "step": 21960,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001257384940981865,
      "weighted_relation_loss": 0.001257384940981865
    },
    {
      "action_query_norm": 39.88885917663574,
      "action_relation_loss": 0.0011846101027913392,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.551917680074837,
      "flow_matching_loss": 0.019397650193423033,
      "grad_norm": 0.09216894954442978,
      "learning_rate": 1.834588647522828e-05,
      "loss": 0.0207,
      "relation_loss": 0.0012708997121080757,
      "selected_pairs": 26734.5,
      "state_query_norm": 42.10640983581543,
      "state_relation_loss": 8.628962968941778e-05,
      "step": 21970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012708997121080757,
      "weighted_relation_loss": 0.0012708997121080757
    },
    {
      "action_query_norm": 39.89340400695801,
      "action_relation_loss": 0.001235084910877049,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.56127221702526,
      "flow_matching_loss": 0.018743593618273734,
      "grad_norm": 0.09586241096258163,
      "learning_rate": 1.8303241535752435e-05,
      "loss": 0.0201,
      "relation_loss": 0.0013252189033664762,
      "selected_pairs": 26545.4,
      "state_query_norm": 42.0960807800293,
      "state_relation_loss": 9.013398666866124e-05,
      "step": 21980,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0013252189033664762,
      "weighted_relation_loss": 0.0013252189033664762
    },
    {
      "action_query_norm": 39.90721092224121,
      "action_relation_loss": 0.0011398114671465009,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.570626753975677,
      "flow_matching_loss": 0.01889881109818816,
      "grad_norm": 0.10110089927911758,
      "learning_rate": 1.8260635110826934e-05,
      "loss": 0.0201,
      "relation_loss": 0.0012339382432401181,
      "selected_pairs": 26805.9,
      "state_query_norm": 42.09711112976074,
      "state_relation_loss": 9.412678555236198e-05,
      "step": 21990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0012339382432401181,
      "weighted_relation_loss": 0.0012339382432401181
    },
    {
      "action_query_norm": 39.8880558013916,
      "action_relation_loss": 0.0011477481632027775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.5799812909261,
      "flow_matching_loss": 0.019436429534107447,
      "grad_norm": 0.11389895528554916,
      "learning_rate": 1.8218067252222597e-05,
      "loss": 0.0207,
      "relation_loss": 0.0012403644737787544,
      "selected_pairs": 27059.3,
      "state_query_norm": 42.089718627929685,
      "state_relation_loss": 9.261629747925326e-05,
      "step": 22000,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0012403644737787544,
      "weighted_relation_loss": 0.0012403644737787544
    },
    {
      "action_query_norm": 39.848022079467775,
      "action_relation_loss": 0.0011168682016432286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.58933582787652,
      "flow_matching_loss": 0.01882464624941349,
      "grad_norm": 0.1407187283039093,
      "learning_rate": 1.817553801166339e-05,
      "loss": 0.02,
      "relation_loss": 0.001209439500235021,
      "selected_pairs": 26858.9,
      "state_query_norm": 42.070296096801755,
      "state_relation_loss": 9.257129495381378e-05,
      "step": 22010,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001209439500235021,
      "weighted_relation_loss": 0.001209439500235021
    },
    {
      "action_query_norm": 39.8734691619873,
      "action_relation_loss": 0.0011741985450498759,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.59869036482694,
      "flow_matching_loss": 0.01958540640771389,
      "grad_norm": 0.11455220729112625,
      "learning_rate": 1.8133047440826333e-05,
      "loss": 0.0208,
      "relation_loss": 0.0012613717932254075,
      "selected_pairs": 26898.8,
      "state_query_norm": 42.104689025878905,
      "state_relation_loss": 8.717324890312738e-05,
      "step": 22020,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0012613717932254075,
      "weighted_relation_loss": 0.0012613717932254075
    },
    {
      "action_query_norm": 39.89753684997559,
      "action_relation_loss": 0.001166919048409909,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.60804490177736,
      "flow_matching_loss": 0.01928237620741129,
      "grad_norm": 0.10529626160860062,
      "learning_rate": 1.809059559134151e-05,
      "loss": 0.0205,
      "relation_loss": 0.001256764424033463,
      "selected_pairs": 27251.2,
      "state_query_norm": 42.0820873260498,
      "state_relation_loss": 8.984537707874551e-05,
      "step": 22030,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001256764424033463,
      "weighted_relation_loss": 0.001256764424033463
    },
    {
      "action_query_norm": 39.86851119995117,
      "action_relation_loss": 0.0011640542245004327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.617399438727784,
      "flow_matching_loss": 0.020899300277233124,
      "grad_norm": 0.08140456676483154,
      "learning_rate": 1.80481825147919e-05,
      "loss": 0.0222,
      "relation_loss": 0.0012566572171635927,
      "selected_pairs": 26663.5,
      "state_query_norm": 42.084699630737305,
      "state_relation_loss": 9.260299120796844e-05,
      "step": 22040,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012566572171635927,
      "weighted_relation_loss": 0.0012566572171635927
    },
    {
      "action_query_norm": 39.93634300231933,
      "action_relation_loss": 0.0011767807765863836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.626753975678206,
      "flow_matching_loss": 0.02049115430563688,
      "grad_norm": 0.08639233559370041,
      "learning_rate": 1.8005808262713396e-05,
      "loss": 0.0218,
      "relation_loss": 0.001262453116942197,
      "selected_pairs": 26439.1,
      "state_query_norm": 42.09005355834961,
      "state_relation_loss": 8.567233453504741e-05,
      "step": 22050,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.001262453116942197,
      "weighted_relation_loss": 0.001262453116942197
    },
    {
      "action_query_norm": 39.8566291809082,
      "action_relation_loss": 0.0012428126879967748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.636108512628624,
      "flow_matching_loss": 0.019513110630214214,
      "grad_norm": 0.10517283529043198,
      "learning_rate": 1.7963472886594714e-05,
      "loss": 0.0208,
      "relation_loss": 0.0013334576971828938,
      "selected_pairs": 26734.3,
      "state_query_norm": 42.07399444580078,
      "state_relation_loss": 9.064500482054427e-05,
      "step": 22060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013334576971828938,
      "weighted_relation_loss": 0.0013334576971828938
    },
    {
      "action_query_norm": 39.862128829956056,
      "action_relation_loss": 0.0011500808876007795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.645463049579046,
      "flow_matching_loss": 0.019721695221960545,
      "grad_norm": 0.08548979461193085,
      "learning_rate": 1.7921176437877302e-05,
      "loss": 0.021,
      "relation_loss": 0.0012407129514031112,
      "selected_pairs": 27153.4,
      "state_query_norm": 42.08785629272461,
      "state_relation_loss": 9.063205725396983e-05,
      "step": 22070,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0012407129514031112,
      "weighted_relation_loss": 0.0012407129514031112
    },
    {
      "action_query_norm": 39.871022415161136,
      "action_relation_loss": 0.0012097172788344325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.654817586529468,
      "flow_matching_loss": 0.019507694896310566,
      "grad_norm": 0.07987385988235474,
      "learning_rate": 1.7878918967955366e-05,
      "loss": 0.0208,
      "relation_loss": 0.0013034364441409708,
      "selected_pairs": 27012.4,
      "state_query_norm": 42.08498077392578,
      "state_relation_loss": 9.37191623961553e-05,
      "step": 22080,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0013034364441409708,
      "weighted_relation_loss": 0.0013034364441409708
    },
    {
      "action_query_norm": 39.855163955688475,
      "action_relation_loss": 0.001164591929409653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.664172123479887,
      "flow_matching_loss": 0.019038283452391624,
      "grad_norm": 0.11137112975120544,
      "learning_rate": 1.7836700528175693e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012548875529319048,
      "selected_pairs": 27285.6,
      "state_query_norm": 42.0818790435791,
      "state_relation_loss": 9.029561770148575e-05,
      "step": 22090,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0012548875529319048,
      "weighted_relation_loss": 0.0012548875529319048
    },
    {
      "action_query_norm": 39.878048706054685,
      "action_relation_loss": 0.0011475008097477258,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.67352666043031,
      "flow_matching_loss": 0.019528300780802966,
      "grad_norm": 0.101679228246212,
      "learning_rate": 1.779452116983769e-05,
      "loss": 0.0208,
      "relation_loss": 0.0012407043250277638,
      "selected_pairs": 26159.4,
      "state_query_norm": 42.070181655883786,
      "state_relation_loss": 9.320351964561268e-05,
      "step": 22100,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0012407043250277638,
      "weighted_relation_loss": 0.0012407043250277638
    },
    {
      "action_query_norm": 39.864125442504886,
      "action_relation_loss": 0.0012573803309351207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.68288119738073,
      "flow_matching_loss": 0.01972589921206236,
      "grad_norm": 0.14457416534423828,
      "learning_rate": 1.775238094419325e-05,
      "loss": 0.0211,
      "relation_loss": 0.001351860084105283,
      "selected_pairs": 27571.0,
      "state_query_norm": 42.09025535583496,
      "state_relation_loss": 9.447976553929038e-05,
      "step": 22110,
      "valid_action_samples": 235.3,
      "weighted_auxiliary_loss": 0.001351860084105283,
      "weighted_relation_loss": 0.001351860084105283
    },
    {
      "action_query_norm": 39.906032943725585,
      "action_relation_loss": 0.001212619082070887,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.69223573433115,
      "flow_matching_loss": 0.019294170755892993,
      "grad_norm": 0.08314509689807892,
      "learning_rate": 1.7710279902446708e-05,
      "loss": 0.0206,
      "relation_loss": 0.001303805026691407,
      "selected_pairs": 26951.5,
      "state_query_norm": 42.08912696838379,
      "state_relation_loss": 9.118593588937074e-05,
      "step": 22120,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001303805026691407,
      "weighted_relation_loss": 0.001303805026691407
    },
    {
      "action_query_norm": 39.84491653442383,
      "action_relation_loss": 0.0011844675405882298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.70159027128157,
      "flow_matching_loss": 0.020137766283005475,
      "grad_norm": 0.12252543866634369,
      "learning_rate": 1.7668218095754795e-05,
      "loss": 0.0214,
      "relation_loss": 0.0012728273635730148,
      "selected_pairs": 26456.8,
      "state_query_norm": 42.070291137695314,
      "state_relation_loss": 8.835982443997637e-05,
      "step": 22130,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012728273635730148,
      "weighted_relation_loss": 0.0012728273635730148
    },
    {
      "action_query_norm": 39.88418731689453,
      "action_relation_loss": 0.0011178718996234238,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.710944808231993,
      "flow_matching_loss": 0.019128915574401616,
      "grad_norm": 0.11263710260391235,
      "learning_rate": 1.7626195575226594e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012165694148279726,
      "selected_pairs": 26898.1,
      "state_query_norm": 42.08078842163086,
      "state_relation_loss": 9.869751811493188e-05,
      "step": 22140,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0012165694148279726,
      "weighted_relation_loss": 0.0012165694148279726
    },
    {
      "action_query_norm": 39.853858184814456,
      "action_relation_loss": 0.0012247386504895986,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.720299345182415,
      "flow_matching_loss": 0.02001360710710287,
      "grad_norm": 0.10620418190956116,
      "learning_rate": 1.758421239192343e-05,
      "loss": 0.0213,
      "relation_loss": 0.0013172139413654804,
      "selected_pairs": 26824.9,
      "state_query_norm": 42.067026138305664,
      "state_relation_loss": 9.247529378626495e-05,
      "step": 22150,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0013172139413654804,
      "weighted_relation_loss": 0.0013172139413654804
    },
    {
      "action_query_norm": 39.83842430114746,
      "action_relation_loss": 0.0011533154174685478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.729653882132833,
      "flow_matching_loss": 0.020644481480121612,
      "grad_norm": 0.13841794431209564,
      "learning_rate": 1.7542268596858814e-05,
      "loss": 0.0219,
      "relation_loss": 0.001242156547959894,
      "selected_pairs": 27184.2,
      "state_query_norm": 42.08290481567383,
      "state_relation_loss": 8.884112830855883e-05,
      "step": 22160,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001242156547959894,
      "weighted_relation_loss": 0.001242156547959894
    },
    {
      "action_query_norm": 39.908953475952146,
      "action_relation_loss": 0.001195555634330958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.739008419083255,
      "flow_matching_loss": 0.019991403724998236,
      "grad_norm": 0.1129116639494896,
      "learning_rate": 1.750036424099841e-05,
      "loss": 0.0213,
      "relation_loss": 0.0012903671013191342,
      "selected_pairs": 26924.0,
      "state_query_norm": 42.05722579956055,
      "state_relation_loss": 9.481147135375068e-05,
      "step": 22170,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0012903671013191342,
      "weighted_relation_loss": 0.0012903671013191342
    },
    {
      "action_query_norm": 39.8659912109375,
      "action_relation_loss": 0.0011825693189166486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.748362956033677,
      "flow_matching_loss": 0.017723724339157344,
      "grad_norm": 0.11959957331418991,
      "learning_rate": 1.7458499375259956e-05,
      "loss": 0.019,
      "relation_loss": 0.0012737319921143353,
      "selected_pairs": 26410.8,
      "state_query_norm": 42.076577758789064,
      "state_relation_loss": 9.11626651941333e-05,
      "step": 22180,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0012737319921143353,
      "weighted_relation_loss": 0.0012737319921143353
    },
    {
      "action_query_norm": 39.924024963378905,
      "action_relation_loss": 0.001176617026794702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.757717492984096,
      "flow_matching_loss": 0.019671791791915895,
      "grad_norm": 0.07785038650035858,
      "learning_rate": 1.7416674050513245e-05,
      "loss": 0.0209,
      "relation_loss": 0.0012654707301408053,
      "selected_pairs": 26833.5,
      "state_query_norm": 42.10017471313476,
      "state_relation_loss": 8.885369170457125e-05,
      "step": 22190,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0012654707301408053,
      "weighted_relation_loss": 0.0012654707301408053
    },
    {
      "action_query_norm": 39.83973083496094,
      "action_relation_loss": 0.0011767877847887575,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.767072029934518,
      "flow_matching_loss": 0.020980792306363582,
      "grad_norm": 0.11357539147138596,
      "learning_rate": 1.737488831757997e-05,
      "loss": 0.0222,
      "relation_loss": 0.0012688142713159322,
      "selected_pairs": 27476.9,
      "state_query_norm": 42.07484817504883,
      "state_relation_loss": 9.202649889630265e-05,
      "step": 22200,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0012688142713159322,
      "weighted_relation_loss": 0.0012688142713159322
    },
    {
      "action_query_norm": 39.866828536987306,
      "action_relation_loss": 0.00110461397562176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.77642656688494,
      "flow_matching_loss": 0.019740344770252705,
      "grad_norm": 0.08611289411783218,
      "learning_rate": 1.733314222723373e-05,
      "loss": 0.0209,
      "relation_loss": 0.0011947431950829924,
      "selected_pairs": 27080.9,
      "state_query_norm": 42.0736587524414,
      "state_relation_loss": 9.012923328555189e-05,
      "step": 22210,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011947431950829924,
      "weighted_relation_loss": 0.0011947431950829924
    },
    {
      "action_query_norm": 39.852613830566405,
      "action_relation_loss": 0.001164929149672389,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.78578110383536,
      "flow_matching_loss": 0.019004374649375676,
      "grad_norm": 0.10080893337726593,
      "learning_rate": 1.7291435830199954e-05,
      "loss": 0.0203,
      "relation_loss": 0.0012512189685367047,
      "selected_pairs": 26988.9,
      "state_query_norm": 42.08311462402344,
      "state_relation_loss": 8.62898217746988e-05,
      "step": 22220,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0012512189685367047,
      "weighted_relation_loss": 0.0012512189685367047
    },
    {
      "action_query_norm": 39.89927101135254,
      "action_relation_loss": 0.0011797804152593016,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.79513564078578,
      "flow_matching_loss": 0.021240981481969356,
      "grad_norm": 0.11706636101007462,
      "learning_rate": 1.7249769177155878e-05,
      "loss": 0.0225,
      "relation_loss": 0.001265003706794232,
      "selected_pairs": 27081.1,
      "state_query_norm": 42.082648849487306,
      "state_relation_loss": 8.52233039040584e-05,
      "step": 22230,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001265003706794232,
      "weighted_relation_loss": 0.001265003706794232
    },
    {
      "action_query_norm": 39.88802947998047,
      "action_relation_loss": 0.001105463970452547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.804490177736202,
      "flow_matching_loss": 0.020682015083730222,
      "grad_norm": 0.10272037237882614,
      "learning_rate": 1.720814231873038e-05,
      "loss": 0.0219,
      "relation_loss": 0.0011932426365092396,
      "selected_pairs": 26829.7,
      "state_query_norm": 42.06550636291504,
      "state_relation_loss": 8.777866241871379e-05,
      "step": 22240,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011932426365092396,
      "weighted_relation_loss": 0.0011932426365092396
    },
    {
      "action_query_norm": 39.90739974975586,
      "action_relation_loss": 0.0011312640155665577,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.813844714686624,
      "flow_matching_loss": 0.02028668951243162,
      "grad_norm": 0.12141944468021393,
      "learning_rate": 1.716655530550405e-05,
      "loss": 0.0215,
      "relation_loss": 0.0012170428177341818,
      "selected_pairs": 27316.2,
      "state_query_norm": 42.07130928039551,
      "state_relation_loss": 8.577880871598608e-05,
      "step": 22250,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0012170428177341818,
      "weighted_relation_loss": 0.0012170428177341818
    },
    {
      "action_query_norm": 39.86751174926758,
      "action_relation_loss": 0.001182664278894663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.823199251637043,
      "flow_matching_loss": 0.01993656763806939,
      "grad_norm": 0.10768131911754608,
      "learning_rate": 1.7125008188009017e-05,
      "loss": 0.0212,
      "relation_loss": 0.0012671898817643522,
      "selected_pairs": 27339.7,
      "state_query_norm": 42.07936248779297,
      "state_relation_loss": 8.452561669400894e-05,
      "step": 22260,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0012671898817643522,
      "weighted_relation_loss": 0.0012671898817643522
    },
    {
      "action_query_norm": 39.86580047607422,
      "action_relation_loss": 0.001089563814457506,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.832553788587465,
      "flow_matching_loss": 0.019347755052149297,
      "grad_norm": 0.10162319242954254,
      "learning_rate": 1.708350101672894e-05,
      "loss": 0.0205,
      "relation_loss": 0.0011782101704739035,
      "selected_pairs": 26755.1,
      "state_query_norm": 42.063772201538086,
      "state_relation_loss": 8.864635456120595e-05,
      "step": 22270,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011782101704739035,
      "weighted_relation_loss": 0.0011782101704739035
    },
    {
      "action_query_norm": 39.83780059814453,
      "action_relation_loss": 0.0010977469675708563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.841908325537887,
      "flow_matching_loss": 0.020253073796629904,
      "grad_norm": 0.10126600414514542,
      "learning_rate": 1.7042033842098957e-05,
      "loss": 0.0214,
      "relation_loss": 0.0011834025965072215,
      "selected_pairs": 26828.5,
      "state_query_norm": 42.05396194458008,
      "state_relation_loss": 8.565562529838644e-05,
      "step": 22280,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011834025965072215,
      "weighted_relation_loss": 0.0011834025965072215
    },
    {
      "action_query_norm": 39.823768615722656,
      "action_relation_loss": 0.0011346932966262101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.85126286248831,
      "flow_matching_loss": 0.021658735163509846,
      "grad_norm": 0.11056192964315414,
      "learning_rate": 1.7000606714505567e-05,
      "loss": 0.0229,
      "relation_loss": 0.0012259696377441287,
      "selected_pairs": 26630.9,
      "state_query_norm": 42.06168174743652,
      "state_relation_loss": 9.127634693868458e-05,
      "step": 22290,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0012259696377441287,
      "weighted_relation_loss": 0.0012259696377441287
    },
    {
      "action_query_norm": 39.85982666015625,
      "action_relation_loss": 0.0011697747861035169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.860617399438727,
      "flow_matching_loss": 0.02108119335025549,
      "grad_norm": 0.09369876235723495,
      "learning_rate": 1.695921968428668e-05,
      "loss": 0.0223,
      "relation_loss": 0.0012600080692209304,
      "selected_pairs": 27014.8,
      "state_query_norm": 42.07468376159668,
      "state_relation_loss": 9.023328821058385e-05,
      "step": 22300,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012600080692209304,
      "weighted_relation_loss": 0.0012600080692209304
    },
    {
      "action_query_norm": 39.84619407653808,
      "action_relation_loss": 0.0011401767667848616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.86997193638915,
      "flow_matching_loss": 0.01806099181994796,
      "grad_norm": 0.09461948275566101,
      "learning_rate": 1.6917872801731416e-05,
      "loss": 0.0193,
      "relation_loss": 0.0012355553568340837,
      "selected_pairs": 26419.6,
      "state_query_norm": 42.05478401184082,
      "state_relation_loss": 9.537859150441364e-05,
      "step": 22310,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0012355553568340837,
      "weighted_relation_loss": 0.0012355553568340837
    },
    {
      "action_query_norm": 39.8018383026123,
      "action_relation_loss": 0.0011797839193604886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.87932647333957,
      "flow_matching_loss": 0.019634875748306514,
      "grad_norm": 0.09340649843215942,
      "learning_rate": 1.687656611708014e-05,
      "loss": 0.0209,
      "relation_loss": 0.0012736355187371374,
      "selected_pairs": 26506.9,
      "state_query_norm": 42.04716339111328,
      "state_relation_loss": 9.385161174577661e-05,
      "step": 22320,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0012736355187371374,
      "weighted_relation_loss": 0.0012736355187371374
    },
    {
      "action_query_norm": 39.83327522277832,
      "action_relation_loss": 0.0011050860804971308,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.88868101028999,
      "flow_matching_loss": 0.020371394325047733,
      "grad_norm": 0.09622068703174591,
      "learning_rate": 1.6835299680524368e-05,
      "loss": 0.0216,
      "relation_loss": 0.0011958271148614585,
      "selected_pairs": 26667.1,
      "state_query_norm": 42.04539833068848,
      "state_relation_loss": 9.074102199519984e-05,
      "step": 22330,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011958271148614585,
      "weighted_relation_loss": 0.0011958271148614585
    },
    {
      "action_query_norm": 39.81298942565918,
      "action_relation_loss": 0.001119909790577367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.89803554724041,
      "flow_matching_loss": 0.01864277832210064,
      "grad_norm": 0.09588658809661865,
      "learning_rate": 1.679407354220669e-05,
      "loss": 0.0199,
      "relation_loss": 0.0012088482442777603,
      "selected_pairs": 26601.8,
      "state_query_norm": 42.06144828796387,
      "state_relation_loss": 8.893845079001039e-05,
      "step": 22340,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0012088482442777603,
      "weighted_relation_loss": 0.0012088482442777603
    },
    {
      "action_query_norm": 39.880502700805664,
      "action_relation_loss": 0.0011507600778713823,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.907390084190833,
      "flow_matching_loss": 0.017880260199308395,
      "grad_norm": 0.10862695425748825,
      "learning_rate": 1.6752887752220793e-05,
      "loss": 0.0191,
      "relation_loss": 0.001240003330167383,
      "selected_pairs": 26535.4,
      "state_query_norm": 42.04285087585449,
      "state_relation_loss": 8.924325011321344e-05,
      "step": 22350,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001240003330167383,
      "weighted_relation_loss": 0.001240003330167383
    },
    {
      "action_query_norm": 39.83062591552734,
      "action_relation_loss": 0.0011853717966005205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.916744621141255,
      "flow_matching_loss": 0.017513329070061444,
      "grad_norm": 0.11586498469114304,
      "learning_rate": 1.6711742360611277e-05,
      "loss": 0.0188,
      "relation_loss": 0.001275510562118143,
      "selected_pairs": 26734.7,
      "state_query_norm": 42.0635009765625,
      "state_relation_loss": 9.013876333483495e-05,
      "step": 22360,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001275510562118143,
      "weighted_relation_loss": 0.001275510562118143
    },
    {
      "action_query_norm": 39.86397361755371,
      "action_relation_loss": 0.0011845050496049226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.926099158091674,
      "flow_matching_loss": 0.019242705218493937,
      "grad_norm": 0.10676119476556778,
      "learning_rate": 1.6670637417373652e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012727499124594033,
      "selected_pairs": 26741.2,
      "state_query_norm": 42.058614349365236,
      "state_relation_loss": 8.824488322716207e-05,
      "step": 22370,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0012727499124594033,
      "weighted_relation_loss": 0.0012727499124594033
    },
    {
      "action_query_norm": 39.82983589172363,
      "action_relation_loss": 0.0011498127947561442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.935453695042096,
      "flow_matching_loss": 0.02042264919728041,
      "grad_norm": 0.11413487792015076,
      "learning_rate": 1.6629572972454333e-05,
      "loss": 0.0217,
      "relation_loss": 0.0012368438066914678,
      "selected_pairs": 27665.1,
      "state_query_norm": 42.06040573120117,
      "state_relation_loss": 8.703101557330228e-05,
      "step": 22380,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0012368438066914678,
      "weighted_relation_loss": 0.0012368438066914678
    },
    {
      "action_query_norm": 39.79772033691406,
      "action_relation_loss": 0.0011427446035668254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.944808231992518,
      "flow_matching_loss": 0.01798674650490284,
      "grad_norm": 0.13249695301055908,
      "learning_rate": 1.6588549075750464e-05,
      "loss": 0.0192,
      "relation_loss": 0.0012331986450590192,
      "selected_pairs": 26462.8,
      "state_query_norm": 42.056541442871094,
      "state_relation_loss": 9.04540589544922e-05,
      "step": 22390,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012331986450590192,
      "weighted_relation_loss": 0.0012331986450590192
    },
    {
      "action_query_norm": 39.81215324401855,
      "action_relation_loss": 0.001165394892450422,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.954162768942936,
      "flow_matching_loss": 0.01838631257414818,
      "grad_norm": 0.10038178414106369,
      "learning_rate": 1.6547565777109976e-05,
      "loss": 0.0196,
      "relation_loss": 0.0012600207701325417,
      "selected_pairs": 26876.1,
      "state_query_norm": 42.04518623352051,
      "state_relation_loss": 9.462588495807722e-05,
      "step": 22400,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012600207701325417,
      "weighted_relation_loss": 0.0012600207701325417
    },
    {
      "action_query_norm": 39.85442276000977,
      "action_relation_loss": 0.0012174143746960908,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.963517305893358,
      "flow_matching_loss": 0.01858246177434921,
      "grad_norm": 0.0906248614192009,
      "learning_rate": 1.650662312633143e-05,
      "loss": 0.0199,
      "relation_loss": 0.0013051858986727893,
      "selected_pairs": 26590.7,
      "state_query_norm": 42.06554527282715,
      "state_relation_loss": 8.777150433161295e-05,
      "step": 22410,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0013051858986727893,
      "weighted_relation_loss": 0.0013051858986727893
    },
    {
      "action_query_norm": 39.844857406616214,
      "action_relation_loss": 0.0011676484253257513,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.97287184284378,
      "flow_matching_loss": 0.018918200489133598,
      "grad_norm": 0.08418577164411545,
      "learning_rate": 1.6465721173164002e-05,
      "loss": 0.0202,
      "relation_loss": 0.0012569870566949248,
      "selected_pairs": 26179.4,
      "state_query_norm": 42.05394058227539,
      "state_relation_loss": 8.933862991398201e-05,
      "step": 22420,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0012569870566949248,
      "weighted_relation_loss": 0.0012569870566949248
    },
    {
      "action_query_norm": 39.770534133911134,
      "action_relation_loss": 0.0012209750479087234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.9822263797942,
      "flow_matching_loss": 0.020056501030921936,
      "grad_norm": 0.11818265914916992,
      "learning_rate": 1.6424859967307428e-05,
      "loss": 0.0214,
      "relation_loss": 0.0013093913206830622,
      "selected_pairs": 26741.2,
      "state_query_norm": 42.037483596801756,
      "state_relation_loss": 8.841626622597687e-05,
      "step": 22430,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0013093913206830622,
      "weighted_relation_loss": 0.0013093913206830622
    },
    {
      "action_query_norm": 39.86119918823242,
      "action_relation_loss": 0.0011368837207555772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 20.99158091674462,
      "flow_matching_loss": 0.017874223086982964,
      "grad_norm": 0.10638588666915894,
      "learning_rate": 1.6384039558411903e-05,
      "loss": 0.0191,
      "relation_loss": 0.0012266952311620117,
      "selected_pairs": 27182.4,
      "state_query_norm": 42.04277572631836,
      "state_relation_loss": 8.981148421298713e-05,
      "step": 22440,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0012266952311620117,
      "weighted_relation_loss": 0.0012266952311620117
    },
    {
      "action_query_norm": 39.788415145874026,
      "action_relation_loss": 0.0011543847504071891,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.000935453695043,
      "flow_matching_loss": 0.019221434369683266,
      "grad_norm": 0.1675845831632614,
      "learning_rate": 1.634325999607811e-05,
      "loss": 0.0205,
      "relation_loss": 0.0012411636766046286,
      "selected_pairs": 24271.0,
      "state_query_norm": 42.0532642364502,
      "state_relation_loss": 8.677891455590724e-05,
      "step": 22450,
      "valid_action_samples": 214.0,
      "weighted_auxiliary_loss": 0.0012411636766046286,
      "weighted_relation_loss": 0.0012411636766046286
    },
    {
      "action_query_norm": 39.83963508605957,
      "action_relation_loss": 0.0011065241647884249,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.010289990645465,
      "flow_matching_loss": 0.020204853639006613,
      "grad_norm": 0.10173111408948898,
      "learning_rate": 1.6302521329857045e-05,
      "loss": 0.0214,
      "relation_loss": 0.001195068343076855,
      "selected_pairs": 27089.3,
      "state_query_norm": 42.06206512451172,
      "state_relation_loss": 8.854416591930203e-05,
      "step": 22460,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001195068343076855,
      "weighted_relation_loss": 0.001195068343076855
    },
    {
      "action_query_norm": 39.87146835327148,
      "action_relation_loss": 0.0011001378705259412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.019644527595883,
      "flow_matching_loss": 0.01984315812587738,
      "grad_norm": 0.09839575737714767,
      "learning_rate": 1.6261823609250027e-05,
      "loss": 0.021,
      "relation_loss": 0.0011915775015950204,
      "selected_pairs": 27010.6,
      "state_query_norm": 42.040896606445315,
      "state_relation_loss": 9.143961215158924e-05,
      "step": 22470,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011915775015950204,
      "weighted_relation_loss": 0.0011915775015950204
    },
    {
      "action_query_norm": 39.85570793151855,
      "action_relation_loss": 0.001153268152847886,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.028999064546305,
      "flow_matching_loss": 0.0172686193138361,
      "grad_norm": 0.0889023020863533,
      "learning_rate": 1.6221166883708627e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012392717995680868,
      "selected_pairs": 27222.1,
      "state_query_norm": 42.07388877868652,
      "state_relation_loss": 8.600366854807361e-05,
      "step": 22480,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0012392717995680868,
      "weighted_relation_loss": 0.0012392717995680868
    },
    {
      "action_query_norm": 39.85129051208496,
      "action_relation_loss": 0.0010661054169759155,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.038353601496727,
      "flow_matching_loss": 0.019366515520960092,
      "grad_norm": 0.11957789212465286,
      "learning_rate": 1.6180551202634603e-05,
      "loss": 0.0205,
      "relation_loss": 0.0011540481238625944,
      "selected_pairs": 26906.2,
      "state_query_norm": 42.05274925231934,
      "state_relation_loss": 8.79427170730196e-05,
      "step": 22490,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011540481238625944,
      "weighted_relation_loss": 0.0011540481238625944
    },
    {
      "action_query_norm": 39.83329086303711,
      "action_relation_loss": 0.0011316496296785773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.047708138447145,
      "flow_matching_loss": 0.018063928186893462,
      "grad_norm": 0.11393474787473679,
      "learning_rate": 1.613997661537981e-05,
      "loss": 0.0193,
      "relation_loss": 0.0012163210252765567,
      "selected_pairs": 26996.3,
      "state_query_norm": 42.05557861328125,
      "state_relation_loss": 8.467139778076672e-05,
      "step": 22500,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0012163210252765567,
      "weighted_relation_loss": 0.0012163210252765567
    },
    {
      "action_query_norm": 39.829250717163085,
      "action_relation_loss": 0.001146384235471487,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.057062675397567,
      "flow_matching_loss": 0.018849935196340085,
      "grad_norm": 0.10116621851921082,
      "learning_rate": 1.609944317124624e-05,
      "loss": 0.0201,
      "relation_loss": 0.0012317090178839863,
      "selected_pairs": 27057.3,
      "state_query_norm": 42.0414176940918,
      "state_relation_loss": 8.532477222615853e-05,
      "step": 22510,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0012317090178839863,
      "weighted_relation_loss": 0.0012317090178839863
    },
    {
      "action_query_norm": 39.8850658416748,
      "action_relation_loss": 0.0010841009207069874,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.06641721234799,
      "flow_matching_loss": 0.018625330924987794,
      "grad_norm": 0.12263134121894836,
      "learning_rate": 1.605895091948582e-05,
      "loss": 0.0198,
      "relation_loss": 0.0011753249797038733,
      "selected_pairs": 26638.9,
      "state_query_norm": 42.048495864868165,
      "state_relation_loss": 9.122405244852417e-05,
      "step": 22520,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0011753249797038733,
      "weighted_relation_loss": 0.0011753249797038733
    },
    {
      "action_query_norm": 39.82520370483398,
      "action_relation_loss": 0.0011246551061049105,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.07577174929841,
      "flow_matching_loss": 0.018568696361035107,
      "grad_norm": 0.09926764667034149,
      "learning_rate": 1.6018499909300477e-05,
      "loss": 0.0198,
      "relation_loss": 0.0012150467140600086,
      "selected_pairs": 27110.1,
      "state_query_norm": 42.055078125,
      "state_relation_loss": 9.039162177941762e-05,
      "step": 22530,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012150467140600086,
      "weighted_relation_loss": 0.0012150467140600086
    },
    {
      "action_query_norm": 39.83082313537598,
      "action_relation_loss": 0.0011156933207530529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.08512628624883,
      "flow_matching_loss": 0.017949278093874456,
      "grad_norm": 0.10063138604164124,
      "learning_rate": 1.597809018984199e-05,
      "loss": 0.0192,
      "relation_loss": 0.0012030964833684265,
      "selected_pairs": 27153.8,
      "state_query_norm": 42.04786224365235,
      "state_relation_loss": 8.740316552575677e-05,
      "step": 22540,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0012030964833684265,
      "weighted_relation_loss": 0.0012030964833684265
    },
    {
      "action_query_norm": 39.777988052368165,
      "action_relation_loss": 0.0011128879676107317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.09448082319925,
      "flow_matching_loss": 0.019350812863558532,
      "grad_norm": 0.10134536027908325,
      "learning_rate": 1.593772181021196e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012023106217384338,
      "selected_pairs": 27152.7,
      "state_query_norm": 42.03762855529785,
      "state_relation_loss": 8.94226657692343e-05,
      "step": 22550,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0012023106217384338,
      "weighted_relation_loss": 0.0012023106217384338
    },
    {
      "action_query_norm": 39.856529998779294,
      "action_relation_loss": 0.0010627339710481466,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.103835360149674,
      "flow_matching_loss": 0.019226191844791175,
      "grad_norm": 0.103861965239048,
      "learning_rate": 1.5897394819461813e-05,
      "loss": 0.0204,
      "relation_loss": 0.001154114177916199,
      "selected_pairs": 27009.7,
      "state_query_norm": 42.03685684204102,
      "state_relation_loss": 9.138019740930758e-05,
      "step": 22560,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.001154114177916199,
      "weighted_relation_loss": 0.001154114177916199
    },
    {
      "action_query_norm": 39.87486038208008,
      "action_relation_loss": 0.0010717094293795526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.113189897100092,
      "flow_matching_loss": 0.01795833110809326,
      "grad_norm": 0.09339717775583267,
      "learning_rate": 1.5857109266592625e-05,
      "loss": 0.0191,
      "relation_loss": 0.001158351404592395,
      "selected_pairs": 27103.4,
      "state_query_norm": 42.05350456237793,
      "state_relation_loss": 8.664198103360831e-05,
      "step": 22570,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001158351404592395,
      "weighted_relation_loss": 0.001158351404592395
    },
    {
      "action_query_norm": 39.810307312011716,
      "action_relation_loss": 0.001178140880074352,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.122544434050514,
      "flow_matching_loss": 0.019641842041164637,
      "grad_norm": 0.08980627357959747,
      "learning_rate": 1.581686520055514e-05,
      "loss": 0.0209,
      "relation_loss": 0.0012688523274846375,
      "selected_pairs": 26226.8,
      "state_query_norm": 42.04888572692871,
      "state_relation_loss": 9.07114525034558e-05,
      "step": 22580,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0012688523274846375,
      "weighted_relation_loss": 0.0012688523274846375
    },
    {
      "action_query_norm": 39.827877807617185,
      "action_relation_loss": 0.0011422977317124606,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.131898971000936,
      "flow_matching_loss": 0.020187816210091115,
      "grad_norm": 0.11165879666805267,
      "learning_rate": 1.5776662670249704e-05,
      "loss": 0.0214,
      "relation_loss": 0.0012315265368670226,
      "selected_pairs": 26575.8,
      "state_query_norm": 42.05212554931641,
      "state_relation_loss": 8.922882261686027e-05,
      "step": 22590,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012315265368670226,
      "weighted_relation_loss": 0.0012315265368670226
    },
    {
      "action_query_norm": 39.8558765411377,
      "action_relation_loss": 0.0010973117838148027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.141253507951358,
      "flow_matching_loss": 0.019219586066901682,
      "grad_norm": 0.10603855550289154,
      "learning_rate": 1.573650172452615e-05,
      "loss": 0.0204,
      "relation_loss": 0.0011842351406812668,
      "selected_pairs": 26731.9,
      "state_query_norm": 42.045073699951175,
      "state_relation_loss": 8.692335177329369e-05,
      "step": 22600,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011842351406812668,
      "weighted_relation_loss": 0.0011842351406812668
    },
    {
      "action_query_norm": 39.8322696685791,
      "action_relation_loss": 0.0011158846318721771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.150608044901777,
      "flow_matching_loss": 0.017222186643630268,
      "grad_norm": 0.10669178515672684,
      "learning_rate": 1.5696382412183852e-05,
      "loss": 0.0184,
      "relation_loss": 0.0012024404131807387,
      "selected_pairs": 27125.9,
      "state_query_norm": 42.03313217163086,
      "state_relation_loss": 8.655578785692342e-05,
      "step": 22610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0012024404131807387,
      "weighted_relation_loss": 0.0012024404131807387
    },
    {
      "action_query_norm": 39.809730529785156,
      "action_relation_loss": 0.0010516118025407195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.1599625818522,
      "flow_matching_loss": 0.019394486490637065,
      "grad_norm": 0.10606975853443146,
      "learning_rate": 1.565630478197155e-05,
      "loss": 0.0205,
      "relation_loss": 0.0011416631867177784,
      "selected_pairs": 26965.0,
      "state_query_norm": 42.02724342346191,
      "state_relation_loss": 9.005138199427165e-05,
      "step": 22620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0011416631867177784,
      "weighted_relation_loss": 0.0011416631867177784
    },
    {
      "action_query_norm": 39.761285400390626,
      "action_relation_loss": 0.0011194636696018278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.16931711880262,
      "flow_matching_loss": 0.01987971495836973,
      "grad_norm": 0.1107151210308075,
      "learning_rate": 1.5616268882587332e-05,
      "loss": 0.0211,
      "relation_loss": 0.0012058188673108815,
      "selected_pairs": 27688.0,
      "state_query_norm": 42.028371047973636,
      "state_relation_loss": 8.635519334347919e-05,
      "step": 22630,
      "valid_action_samples": 235.8,
      "weighted_auxiliary_loss": 0.0012058188673108815,
      "weighted_relation_loss": 0.0012058188673108815
    },
    {
      "action_query_norm": 39.78442306518555,
      "action_relation_loss": 0.001098405523225665,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.17867165575304,
      "flow_matching_loss": 0.019186143949627876,
      "grad_norm": 0.08257641643285751,
      "learning_rate": 1.5576274762678594e-05,
      "loss": 0.0204,
      "relation_loss": 0.001188240316696465,
      "selected_pairs": 26874.7,
      "state_query_norm": 42.04295196533203,
      "state_relation_loss": 8.983478401205502e-05,
      "step": 22640,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001188240316696465,
      "weighted_relation_loss": 0.001188240316696465
    },
    {
      "action_query_norm": 39.803504943847656,
      "action_relation_loss": 0.0011049791995901615,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.18802619270346,
      "flow_matching_loss": 0.018515618238598108,
      "grad_norm": 0.08340942114591599,
      "learning_rate": 1.5536322470841952e-05,
      "loss": 0.0197,
      "relation_loss": 0.0011909896857105196,
      "selected_pairs": 26579.3,
      "state_query_norm": 42.041330337524414,
      "state_relation_loss": 8.601047811680473e-05,
      "step": 22650,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011909896857105196,
      "weighted_relation_loss": 0.0011909896857105196
    },
    {
      "action_query_norm": 39.797637176513675,
      "action_relation_loss": 0.0011498151696287095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.197380729653883,
      "flow_matching_loss": 0.020098244305700064,
      "grad_norm": 0.09616244584321976,
      "learning_rate": 1.549641205562324e-05,
      "loss": 0.0213,
      "relation_loss": 0.0012338334345258772,
      "selected_pairs": 26673.1,
      "state_query_norm": 42.039954376220706,
      "state_relation_loss": 8.401826416957193e-05,
      "step": 22660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0012338334345258772,
      "weighted_relation_loss": 0.0012338334345258772
    },
    {
      "action_query_norm": 39.82438926696777,
      "action_relation_loss": 0.001082605222472921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.2067352666043,
      "flow_matching_loss": 0.01927179265767336,
      "grad_norm": 0.09444893151521683,
      "learning_rate": 1.5456543565517334e-05,
      "loss": 0.0204,
      "relation_loss": 0.0011713662650436163,
      "selected_pairs": 26879.4,
      "state_query_norm": 42.0370906829834,
      "state_relation_loss": 8.876102947397158e-05,
      "step": 22670,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011713662650436163,
      "weighted_relation_loss": 0.0011713662650436163
    },
    {
      "action_query_norm": 39.827632904052734,
      "action_relation_loss": 0.0011170255835168063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.216089803554723,
      "flow_matching_loss": 0.018448689952492714,
      "grad_norm": 0.09608365595340729,
      "learning_rate": 1.541671704896825e-05,
      "loss": 0.0197,
      "relation_loss": 0.0012034189887344838,
      "selected_pairs": 26164.2,
      "state_query_norm": 42.05548439025879,
      "state_relation_loss": 8.639340521767735e-05,
      "step": 22680,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0012034189887344838,
      "weighted_relation_loss": 0.0012034189887344838
    },
    {
      "action_query_norm": 39.77980537414551,
      "action_relation_loss": 0.0011241735657677055,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.225444340505145,
      "flow_matching_loss": 0.01986585259437561,
      "grad_norm": 0.16475751996040344,
      "learning_rate": 1.5376932554368938e-05,
      "loss": 0.0211,
      "relation_loss": 0.0012153556919656694,
      "selected_pairs": 27015.0,
      "state_query_norm": 42.03275337219238,
      "state_relation_loss": 9.118212474277244e-05,
      "step": 22690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0012153556919656694,
      "weighted_relation_loss": 0.0012153556919656694
    },
    {
      "action_query_norm": 39.80298614501953,
      "action_relation_loss": 0.001129765948280692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.234798877455567,
      "flow_matching_loss": 0.018191643804311753,
      "grad_norm": 0.31376537680625916,
      "learning_rate": 1.5337190130061318e-05,
      "loss": 0.0194,
      "relation_loss": 0.001218643004540354,
      "selected_pairs": 27407.3,
      "state_query_norm": 42.03549919128418,
      "state_relation_loss": 8.887706208042801e-05,
      "step": 22700,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.001218643004540354,
      "weighted_relation_loss": 0.001218643004540354
    },
    {
      "action_query_norm": 39.842637634277345,
      "action_relation_loss": 0.0010971511073876172,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.244153414405986,
      "flow_matching_loss": 0.017939768824726345,
      "grad_norm": 0.0850176215171814,
      "learning_rate": 1.5297489824336165e-05,
      "loss": 0.0191,
      "relation_loss": 0.0011876755394041537,
      "selected_pairs": 26504.0,
      "state_query_norm": 42.029248046875,
      "state_relation_loss": 9.052443056134507e-05,
      "step": 22710,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011876755394041537,
      "weighted_relation_loss": 0.0011876755394041537
    },
    {
      "action_query_norm": 39.78263282775879,
      "action_relation_loss": 0.0010793578112497926,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.253507951356408,
      "flow_matching_loss": 0.017657398618757723,
      "grad_norm": 0.10217177867889404,
      "learning_rate": 1.5257831685433121e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011668538441881537,
      "selected_pairs": 27012.1,
      "state_query_norm": 42.04249267578125,
      "state_relation_loss": 8.749602493480779e-05,
      "step": 22720,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011668538441881537,
      "weighted_relation_loss": 0.0011668538441881537
    },
    {
      "action_query_norm": 39.75512275695801,
      "action_relation_loss": 0.0010816991853062063,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.26286248830683,
      "flow_matching_loss": 0.018181321583688258,
      "grad_norm": 0.1596570611000061,
      "learning_rate": 1.521821576154055e-05,
      "loss": 0.0194,
      "relation_loss": 0.0011704751988872886,
      "selected_pairs": 26580.1,
      "state_query_norm": 42.01858253479004,
      "state_relation_loss": 8.877603468135931e-05,
      "step": 22730,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011704751988872886,
      "weighted_relation_loss": 0.0011704751988872886
    },
    {
      "action_query_norm": 39.80730209350586,
      "action_relation_loss": 0.001136660308111459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.27221702525725,
      "flow_matching_loss": 0.017585919611155985,
      "grad_norm": 0.11093737930059433,
      "learning_rate": 1.5178642100795542e-05,
      "loss": 0.0188,
      "relation_loss": 0.001221535832155496,
      "selected_pairs": 26806.9,
      "state_query_norm": 42.0428653717041,
      "state_relation_loss": 8.487553423037752e-05,
      "step": 22740,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001221535832155496,
      "weighted_relation_loss": 0.001221535832155496
    },
    {
      "action_query_norm": 39.765364456176755,
      "action_relation_loss": 0.0010973730648402124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.28157156220767,
      "flow_matching_loss": 0.018338640127331017,
      "grad_norm": 0.09453481435775757,
      "learning_rate": 1.5139110751283819e-05,
      "loss": 0.0195,
      "relation_loss": 0.001187687914352864,
      "selected_pairs": 26852.8,
      "state_query_norm": 42.03750801086426,
      "state_relation_loss": 9.031485897139646e-05,
      "step": 22750,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001187687914352864,
      "weighted_relation_loss": 0.001187687914352864
    },
    {
      "action_query_norm": 39.77870063781738,
      "action_relation_loss": 0.0011237642145715653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.290926099158092,
      "flow_matching_loss": 0.02118939533829689,
      "grad_norm": 0.10111355036497116,
      "learning_rate": 1.5099621761039683e-05,
      "loss": 0.0224,
      "relation_loss": 0.0012068800977431237,
      "selected_pairs": 27105.1,
      "state_query_norm": 42.03443450927735,
      "state_relation_loss": 8.311587953357957e-05,
      "step": 22760,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012068800977431237,
      "weighted_relation_loss": 0.0012068800977431237
    },
    {
      "action_query_norm": 39.82431411743164,
      "action_relation_loss": 0.0011178952408954501,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.300280636108514,
      "flow_matching_loss": 0.01896734591573477,
      "grad_norm": 0.0880555659532547,
      "learning_rate": 1.5060175178046016e-05,
      "loss": 0.0202,
      "relation_loss": 0.001198203209787607,
      "selected_pairs": 27452.8,
      "state_query_norm": 42.03726692199707,
      "state_relation_loss": 8.030796816456131e-05,
      "step": 22770,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.001198203209787607,
      "weighted_relation_loss": 0.001198203209787607
    },
    {
      "action_query_norm": 39.85994529724121,
      "action_relation_loss": 0.0010379644925706088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.309635173058933,
      "flow_matching_loss": 0.020398352947086095,
      "grad_norm": 0.11148712784051895,
      "learning_rate": 1.5020771050234117e-05,
      "loss": 0.0215,
      "relation_loss": 0.0011299293430056423,
      "selected_pairs": 26911.3,
      "state_query_norm": 42.02523918151856,
      "state_relation_loss": 9.196484534186311e-05,
      "step": 22780,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011299293430056423,
      "weighted_relation_loss": 0.0011299293430056423
    },
    {
      "action_query_norm": 39.816705322265626,
      "action_relation_loss": 0.0010569401958491654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.318989710009355,
      "flow_matching_loss": 0.018762765824794768,
      "grad_norm": 0.142292320728302,
      "learning_rate": 1.4981409425483716e-05,
      "loss": 0.0199,
      "relation_loss": 0.001143888710066676,
      "selected_pairs": 26138.1,
      "state_query_norm": 42.03047676086426,
      "state_relation_loss": 8.694850985193625e-05,
      "step": 22790,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.001143888710066676,
      "weighted_relation_loss": 0.001143888710066676
    },
    {
      "action_query_norm": 39.81938438415527,
      "action_relation_loss": 0.0010946301277726888,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.328344246959777,
      "flow_matching_loss": 0.017108514811843632,
      "grad_norm": 0.08863844722509384,
      "learning_rate": 1.4942090351622883e-05,
      "loss": 0.0183,
      "relation_loss": 0.0011822950677014888,
      "selected_pairs": 26670.9,
      "state_query_norm": 42.03073577880859,
      "state_relation_loss": 8.766493701841682e-05,
      "step": 22800,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011822950677014888,
      "weighted_relation_loss": 0.0011822950677014888
    },
    {
      "action_query_norm": 39.792562103271486,
      "action_relation_loss": 0.0011197500978596508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.337698783910195,
      "flow_matching_loss": 0.01773949209600687,
      "grad_norm": 0.12515433132648468,
      "learning_rate": 1.490281387642804e-05,
      "loss": 0.0189,
      "relation_loss": 0.0012084916816093028,
      "selected_pairs": 26455.4,
      "state_query_norm": 42.022928619384764,
      "state_relation_loss": 8.874157865648157e-05,
      "step": 22810,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0012084916816093028,
      "weighted_relation_loss": 0.0012084916816093028
    },
    {
      "action_query_norm": 39.79748306274414,
      "action_relation_loss": 0.001132976682856679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.347053320860617,
      "flow_matching_loss": 0.017540425434708594,
      "grad_norm": 0.11357526481151581,
      "learning_rate": 1.4863580047623765e-05,
      "loss": 0.0188,
      "relation_loss": 0.0012197078904137015,
      "selected_pairs": 26869.5,
      "state_query_norm": 42.01991767883301,
      "state_relation_loss": 8.67311995534692e-05,
      "step": 22820,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0012197078904137015,
      "weighted_relation_loss": 0.0012197078904137015
    },
    {
      "action_query_norm": 39.79890480041504,
      "action_relation_loss": 0.0010657232138328253,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.35640785781104,
      "flow_matching_loss": 0.017106151953339577,
      "grad_norm": 0.09407047927379608,
      "learning_rate": 1.4824388912882897e-05,
      "loss": 0.0183,
      "relation_loss": 0.0011500390828587115,
      "selected_pairs": 26595.4,
      "state_query_norm": 42.027448654174805,
      "state_relation_loss": 8.431587339146063e-05,
      "step": 22830,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0011500390828587115,
      "weighted_relation_loss": 0.0011500390828587115
    },
    {
      "action_query_norm": 39.78021087646484,
      "action_relation_loss": 0.0011067089741118251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.36576239476146,
      "flow_matching_loss": 0.018363599106669425,
      "grad_norm": 0.11384513974189758,
      "learning_rate": 1.4785240519826344e-05,
      "loss": 0.0196,
      "relation_loss": 0.0011925438651815058,
      "selected_pairs": 27242.7,
      "state_query_norm": 42.024264907836915,
      "state_relation_loss": 8.583488670410588e-05,
      "step": 22840,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011925438651815058,
      "weighted_relation_loss": 0.0011925438651815058
    },
    {
      "action_query_norm": 39.815355682373045,
      "action_relation_loss": 0.001111368794227019,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.37511693171188,
      "flow_matching_loss": 0.016848244424909353,
      "grad_norm": 0.12573030591011047,
      "learning_rate": 1.4746134916023097e-05,
      "loss": 0.0181,
      "relation_loss": 0.0012024989118799567,
      "selected_pairs": 26755.4,
      "state_query_norm": 42.03193664550781,
      "state_relation_loss": 9.11301096493844e-05,
      "step": 22850,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012024989118799567,
      "weighted_relation_loss": 0.0012024989118799567
    },
    {
      "action_query_norm": 39.74312744140625,
      "action_relation_loss": 0.0010889901604969054,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.3844714686623,
      "flow_matching_loss": 0.018319286964833738,
      "grad_norm": 0.11153418570756912,
      "learning_rate": 1.4707072148990141e-05,
      "loss": 0.0195,
      "relation_loss": 0.0011742018163204194,
      "selected_pairs": 26438.2,
      "state_query_norm": 42.03190803527832,
      "state_relation_loss": 8.52116740134079e-05,
      "step": 22860,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011742018163204194,
      "weighted_relation_loss": 0.0011742018163204194
    },
    {
      "action_query_norm": 39.7892147064209,
      "action_relation_loss": 0.0011050002241972834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.393826005612723,
      "flow_matching_loss": 0.018893126770853995,
      "grad_norm": 0.10141215473413467,
      "learning_rate": 1.4668052266192423e-05,
      "loss": 0.0201,
      "relation_loss": 0.00119034739327617,
      "selected_pairs": 26616.0,
      "state_query_norm": 42.01991691589355,
      "state_relation_loss": 8.534716762369499e-05,
      "step": 22870,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.00119034739327617,
      "weighted_relation_loss": 0.00119034739327617
    },
    {
      "action_query_norm": 39.779443359375,
      "action_relation_loss": 0.001086511858738959,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.403180542563142,
      "flow_matching_loss": 0.01961708487942815,
      "grad_norm": 0.11077021807432175,
      "learning_rate": 1.4629075315042795e-05,
      "loss": 0.0208,
      "relation_loss": 0.0011745474766939878,
      "selected_pairs": 27481.5,
      "state_query_norm": 42.011549377441405,
      "state_relation_loss": 8.803562595858239e-05,
      "step": 22880,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0011745474766939878,
      "weighted_relation_loss": 0.0011745474766939878
    },
    {
      "action_query_norm": 39.77124214172363,
      "action_relation_loss": 0.001139863021671772,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.412535079513564,
      "flow_matching_loss": 0.01940820086747408,
      "grad_norm": 0.10426491498947144,
      "learning_rate": 1.4590141342901925e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012220034492202102,
      "selected_pairs": 27038.0,
      "state_query_norm": 42.02195243835449,
      "state_relation_loss": 8.21404333692044e-05,
      "step": 22890,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0012220034492202102,
      "weighted_relation_loss": 0.0012220034492202102
    },
    {
      "action_query_norm": 39.85368690490723,
      "action_relation_loss": 0.0009788594208657742,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.421889616463986,
      "flow_matching_loss": 0.019546015560626982,
      "grad_norm": 0.09191517531871796,
      "learning_rate": 1.4551250397078253e-05,
      "loss": 0.0206,
      "relation_loss": 0.001066029816865921,
      "selected_pairs": 26945.3,
      "state_query_norm": 42.02243270874023,
      "state_relation_loss": 8.717038363101893e-05,
      "step": 22900,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001066029816865921,
      "weighted_relation_loss": 0.001066029816865921
    },
    {
      "action_query_norm": 39.854455184936526,
      "action_relation_loss": 0.0010890328790992498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.431244153414404,
      "flow_matching_loss": 0.017598225735127926,
      "grad_norm": 0.1323424130678177,
      "learning_rate": 1.4512402524827945e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011747926473617554,
      "selected_pairs": 26899.5,
      "state_query_norm": 42.03607025146484,
      "state_relation_loss": 8.575977408327162e-05,
      "step": 22910,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011747926473617554,
      "weighted_relation_loss": 0.0011747926473617554
    },
    {
      "action_query_norm": 39.79614906311035,
      "action_relation_loss": 0.0010983455867972225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.440598690364826,
      "flow_matching_loss": 0.0186497800052166,
      "grad_norm": 0.11452062427997589,
      "learning_rate": 1.447359777335482e-05,
      "loss": 0.0198,
      "relation_loss": 0.0011875563417561353,
      "selected_pairs": 27034.0,
      "state_query_norm": 42.029849243164065,
      "state_relation_loss": 8.921075277612545e-05,
      "step": 22920,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011875563417561353,
      "weighted_relation_loss": 0.0011875563417561353
    },
    {
      "action_query_norm": 39.79177398681641,
      "action_relation_loss": 0.0010410590912215412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.44995322731525,
      "flow_matching_loss": 0.017997810430824757,
      "grad_norm": 0.11146176606416702,
      "learning_rate": 1.4434836189810335e-05,
      "loss": 0.0191,
      "relation_loss": 0.0011258199927397072,
      "selected_pairs": 26735.0,
      "state_query_norm": 42.02063674926758,
      "state_relation_loss": 8.476090224576182e-05,
      "step": 22930,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011258199927397072,
      "weighted_relation_loss": 0.0011258199927397072
    },
    {
      "action_query_norm": 39.8107349395752,
      "action_relation_loss": 0.001131850329693407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.45930776426567,
      "flow_matching_loss": 0.01893150694668293,
      "grad_norm": 0.09550285339355469,
      "learning_rate": 1.4396117821293453e-05,
      "loss": 0.0202,
      "relation_loss": 0.0012201089528389275,
      "selected_pairs": 26759.8,
      "state_query_norm": 42.01531448364258,
      "state_relation_loss": 8.825863624224439e-05,
      "step": 22940,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012201089528389275,
      "weighted_relation_loss": 0.0012201089528389275
    },
    {
      "action_query_norm": 39.811990356445314,
      "action_relation_loss": 0.0010756007803138345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.46866230121609,
      "flow_matching_loss": 0.01909661516547203,
      "grad_norm": 0.12464912980794907,
      "learning_rate": 1.4357442714850633e-05,
      "loss": 0.0203,
      "relation_loss": 0.00116340066306293,
      "selected_pairs": 26393.2,
      "state_query_norm": 42.013358688354494,
      "state_relation_loss": 8.779989148024469e-05,
      "step": 22950,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.00116340066306293,
      "weighted_relation_loss": 0.00116340066306293
    },
    {
      "action_query_norm": 39.826036834716795,
      "action_relation_loss": 0.0010519903153181076,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.47801683816651,
      "flow_matching_loss": 0.018243505991995333,
      "grad_norm": 0.09142779558897018,
      "learning_rate": 1.43188109174758e-05,
      "loss": 0.0194,
      "relation_loss": 0.0011365087353624403,
      "selected_pairs": 27010.7,
      "state_query_norm": 42.030526351928714,
      "state_relation_loss": 8.451842004433274e-05,
      "step": 22960,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011365087353624403,
      "weighted_relation_loss": 0.0011365087353624403
    },
    {
      "action_query_norm": 39.79905281066895,
      "action_relation_loss": 0.0011277595709543674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.487371375116933,
      "flow_matching_loss": 0.02042835596948862,
      "grad_norm": 0.09990331530570984,
      "learning_rate": 1.4280222476110206e-05,
      "loss": 0.0216,
      "relation_loss": 0.0012167419190518558,
      "selected_pairs": 26364.4,
      "state_query_norm": 42.0212287902832,
      "state_relation_loss": 8.898232845240272e-05,
      "step": 22970,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0012167419190518558,
      "weighted_relation_loss": 0.0012167419190518558
    },
    {
      "action_query_norm": 39.81229705810547,
      "action_relation_loss": 0.0010663764202035963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.49672591206735,
      "flow_matching_loss": 0.018124186806380747,
      "grad_norm": 0.1178542971611023,
      "learning_rate": 1.4241677437642469e-05,
      "loss": 0.0193,
      "relation_loss": 0.001148803485557437,
      "selected_pairs": 26737.7,
      "state_query_norm": 42.02126045227051,
      "state_relation_loss": 8.242706971941515e-05,
      "step": 22980,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.001148803485557437,
      "weighted_relation_loss": 0.001148803485557437
    },
    {
      "action_query_norm": 39.810503768920896,
      "action_relation_loss": 0.0010616069310344755,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.506080449017773,
      "flow_matching_loss": 0.017647144943475725,
      "grad_norm": 0.11774029582738876,
      "learning_rate": 1.420317584890844e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011509634088724852,
      "selected_pairs": 26805.0,
      "state_query_norm": 42.00688781738281,
      "state_relation_loss": 8.935646765166894e-05,
      "step": 22990,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011509634088724852,
      "weighted_relation_loss": 0.0011509634088724852
    },
    {
      "action_query_norm": 39.77757034301758,
      "action_relation_loss": 0.0010775973903946578,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.515434985968195,
      "flow_matching_loss": 0.018579324707388878,
      "grad_norm": 0.08722905814647675,
      "learning_rate": 1.4164717756691176e-05,
      "loss": 0.0197,
      "relation_loss": 0.0011652874061837792,
      "selected_pairs": 27086.5,
      "state_query_norm": 42.015241622924805,
      "state_relation_loss": 8.768999396124855e-05,
      "step": 23000,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011652874061837792,
      "weighted_relation_loss": 0.0011652874061837792
    },
    {
      "action_query_norm": 39.808824920654295,
      "action_relation_loss": 0.001098367921076715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.524789522918617,
      "flow_matching_loss": 0.020188340079039337,
      "grad_norm": 0.11155957728624344,
      "learning_rate": 1.4126303207720882e-05,
      "loss": 0.0214,
      "relation_loss": 0.0011844727443531156,
      "selected_pairs": 27081.2,
      "state_query_norm": 42.02012977600098,
      "state_relation_loss": 8.610483273514546e-05,
      "step": 23010,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011844727443531156,
      "weighted_relation_loss": 0.0011844727443531156
    },
    {
      "action_query_norm": 39.81669692993164,
      "action_relation_loss": 0.0010835538618266583,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.534144059869035,
      "flow_matching_loss": 0.019574782997369766,
      "grad_norm": 0.10878009349107742,
      "learning_rate": 1.408793224867484e-05,
      "loss": 0.0207,
      "relation_loss": 0.0011696833884343506,
      "selected_pairs": 26716.1,
      "state_query_norm": 42.02715492248535,
      "state_relation_loss": 8.612952588009648e-05,
      "step": 23020,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011696833884343506,
      "weighted_relation_loss": 0.0011696833884343506
    },
    {
      "action_query_norm": 39.799221420288085,
      "action_relation_loss": 0.0011359559663105755,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.543498596819457,
      "flow_matching_loss": 0.019352123141288757,
      "grad_norm": 0.09571518003940582,
      "learning_rate": 1.4049604926177423e-05,
      "loss": 0.0206,
      "relation_loss": 0.0012253097724169494,
      "selected_pairs": 26599.0,
      "state_query_norm": 42.002716827392575,
      "state_relation_loss": 8.935379373724573e-05,
      "step": 23030,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0012253097724169494,
      "weighted_relation_loss": 0.0012253097724169494
    },
    {
      "action_query_norm": 39.79106674194336,
      "action_relation_loss": 0.0011168801051098854,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.55285313376988,
      "flow_matching_loss": 0.019628067314624787,
      "grad_norm": 0.10216598212718964,
      "learning_rate": 1.4011321286799916e-05,
      "loss": 0.0208,
      "relation_loss": 0.001203261490445584,
      "selected_pairs": 27154.9,
      "state_query_norm": 42.010399627685544,
      "state_relation_loss": 8.638140352559275e-05,
      "step": 23040,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001203261490445584,
      "weighted_relation_loss": 0.001203261490445584
    },
    {
      "action_query_norm": 39.792632293701175,
      "action_relation_loss": 0.0010750766610726713,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.562207670720298,
      "flow_matching_loss": 0.017296303622424604,
      "grad_norm": 0.09716382622718811,
      "learning_rate": 1.3973081377060565e-05,
      "loss": 0.0185,
      "relation_loss": 0.001156823441851884,
      "selected_pairs": 26348.0,
      "state_query_norm": 42.035103225708006,
      "state_relation_loss": 8.174678223440423e-05,
      "step": 23050,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.001156823441851884,
      "weighted_relation_loss": 0.001156823441851884
    },
    {
      "action_query_norm": 39.79881629943848,
      "action_relation_loss": 0.0011166484910063446,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.57156220767072,
      "flow_matching_loss": 0.02067976789548993,
      "grad_norm": 0.09923826158046722,
      "learning_rate": 1.393488524342445e-05,
      "loss": 0.0219,
      "relation_loss": 0.0012037179316394032,
      "selected_pairs": 27112.2,
      "state_query_norm": 42.019676208496094,
      "state_relation_loss": 8.706946318852715e-05,
      "step": 23060,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0012037179316394032,
      "weighted_relation_loss": 0.0012037179316394032
    },
    {
      "action_query_norm": 39.773841094970706,
      "action_relation_loss": 0.0010228758095763625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.580916744621142,
      "flow_matching_loss": 0.020638060569763184,
      "grad_norm": 0.12523327767848969,
      "learning_rate": 1.3896732932303486e-05,
      "loss": 0.0217,
      "relation_loss": 0.0011081387929152697,
      "selected_pairs": 26231.7,
      "state_query_norm": 42.01980400085449,
      "state_relation_loss": 8.526297460775822e-05,
      "step": 23070,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0011081387929152697,
      "weighted_relation_loss": 0.0011081387929152697
    },
    {
      "action_query_norm": 39.767629623413086,
      "action_relation_loss": 0.0010985493659973144,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.590271281571564,
      "flow_matching_loss": 0.018800773471593858,
      "grad_norm": 0.1160275936126709,
      "learning_rate": 1.3858624490056304e-05,
      "loss": 0.02,
      "relation_loss": 0.0011833289405331015,
      "selected_pairs": 27064.2,
      "state_query_norm": 42.01988182067871,
      "state_relation_loss": 8.477956944261678e-05,
      "step": 23080,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011833289405331015,
      "weighted_relation_loss": 0.0011833289405331015
    },
    {
      "action_query_norm": 39.79287223815918,
      "action_relation_loss": 0.00106029327143915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.599625818521982,
      "flow_matching_loss": 0.017489262763410805,
      "grad_norm": 0.09122922271490097,
      "learning_rate": 1.3820559962988299e-05,
      "loss": 0.0186,
      "relation_loss": 0.0011415737331844865,
      "selected_pairs": 26916.9,
      "state_query_norm": 42.01946487426758,
      "state_relation_loss": 8.128045374178328e-05,
      "step": 23090,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011415737331844865,
      "weighted_relation_loss": 0.0011415737331844865
    },
    {
      "action_query_norm": 39.767146301269534,
      "action_relation_loss": 0.0011052776011638344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.608980355472404,
      "flow_matching_loss": 0.01893613561987877,
      "grad_norm": 0.09341946244239807,
      "learning_rate": 1.3782539397351418e-05,
      "loss": 0.0201,
      "relation_loss": 0.0011884533101692795,
      "selected_pairs": 26618.8,
      "state_query_norm": 42.0198413848877,
      "state_relation_loss": 8.317570463987068e-05,
      "step": 23100,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011884533101692795,
      "weighted_relation_loss": 0.0011884533101692795
    },
    {
      "action_query_norm": 39.80807342529297,
      "action_relation_loss": 0.001089226349722594,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.618334892422826,
      "flow_matching_loss": 0.01973752398043871,
      "grad_norm": 0.10282707959413528,
      "learning_rate": 1.3744562839344266e-05,
      "loss": 0.0209,
      "relation_loss": 0.0011767479358240963,
      "selected_pairs": 27317.3,
      "state_query_norm": 42.02276420593262,
      "state_relation_loss": 8.752157445997e-05,
      "step": 23110,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0011767479358240963,
      "weighted_relation_loss": 0.0011767479358240963
    },
    {
      "action_query_norm": 39.79063186645508,
      "action_relation_loss": 0.0010920892586000265,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.627689429373245,
      "flow_matching_loss": 0.018853116407990457,
      "grad_norm": 0.2138594537973404,
      "learning_rate": 1.370663033511193e-05,
      "loss": 0.02,
      "relation_loss": 0.0011795752448961138,
      "selected_pairs": 26872.3,
      "state_query_norm": 42.019646072387694,
      "state_relation_loss": 8.748599357204512e-05,
      "step": 23120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011795752448961138,
      "weighted_relation_loss": 0.0011795752448961138
    },
    {
      "action_query_norm": 39.79189453125,
      "action_relation_loss": 0.001023543713381514,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.637043966323667,
      "flow_matching_loss": 0.01901165032759309,
      "grad_norm": 0.13765567541122437,
      "learning_rate": 1.3668741930745964e-05,
      "loss": 0.0201,
      "relation_loss": 0.0011130478116683662,
      "selected_pairs": 26998.6,
      "state_query_norm": 42.00665435791016,
      "state_relation_loss": 8.950410992838442e-05,
      "step": 23130,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011130478116683662,
      "weighted_relation_loss": 0.0011130478116683662
    },
    {
      "action_query_norm": 39.797847747802734,
      "action_relation_loss": 0.0010647089744452386,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.64639850327409,
      "flow_matching_loss": 0.0176380948163569,
      "grad_norm": 0.08863706141710281,
      "learning_rate": 1.3630897672284382e-05,
      "loss": 0.0188,
      "relation_loss": 0.001149588858243078,
      "selected_pairs": 26917.9,
      "state_query_norm": 42.01531295776367,
      "state_relation_loss": 8.487989689456298e-05,
      "step": 23140,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001149588858243078,
      "weighted_relation_loss": 0.001149588858243078
    },
    {
      "action_query_norm": 39.77934303283691,
      "action_relation_loss": 0.0010504600242711604,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.65575304022451,
      "flow_matching_loss": 0.018262625299394132,
      "grad_norm": 0.09131366014480591,
      "learning_rate": 1.3593097605711508e-05,
      "loss": 0.0194,
      "relation_loss": 0.0011383887263946234,
      "selected_pairs": 26144.3,
      "state_query_norm": 42.00807876586914,
      "state_relation_loss": 8.792870139586739e-05,
      "step": 23150,
      "valid_action_samples": 229.1,
      "weighted_auxiliary_loss": 0.0011383887263946234,
      "weighted_relation_loss": 0.0011383887263946234
    },
    {
      "action_query_norm": 39.736043548583986,
      "action_relation_loss": 0.0011637703573796898,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.66510757717493,
      "flow_matching_loss": 0.02066269312053919,
      "grad_norm": 0.12992966175079346,
      "learning_rate": 1.3555341776957992e-05,
      "loss": 0.0219,
      "relation_loss": 0.001244967826642096,
      "selected_pairs": 26964.3,
      "state_query_norm": 42.00300178527832,
      "state_relation_loss": 8.119745107251219e-05,
      "step": 23160,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001244967826642096,
      "weighted_relation_loss": 0.001244967826642096
    },
    {
      "action_query_norm": 39.76946678161621,
      "action_relation_loss": 0.0011088119528722018,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.67446211412535,
      "flow_matching_loss": 0.01745036132633686,
      "grad_norm": 0.09666892886161804,
      "learning_rate": 1.3517630231900724e-05,
      "loss": 0.0186,
      "relation_loss": 0.0011945619946345687,
      "selected_pairs": 26755.4,
      "state_query_norm": 42.01742172241211,
      "state_relation_loss": 8.57500490383245e-05,
      "step": 23170,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011945619946345687,
      "weighted_relation_loss": 0.0011945619946345687
    },
    {
      "action_query_norm": 39.79783592224121,
      "action_relation_loss": 0.0010172285605221988,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.683816651075773,
      "flow_matching_loss": 0.017532916460186242,
      "grad_norm": 0.08844225108623505,
      "learning_rate": 1.3479963016362767e-05,
      "loss": 0.0186,
      "relation_loss": 0.0011045425198972225,
      "selected_pairs": 26801.5,
      "state_query_norm": 42.01809997558594,
      "state_relation_loss": 8.731396301300264e-05,
      "step": 23180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011045425198972225,
      "weighted_relation_loss": 0.0011045425198972225
    },
    {
      "action_query_norm": 39.78375053405762,
      "action_relation_loss": 0.0010966880712658168,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.69317118802619,
      "flow_matching_loss": 0.01948898984119296,
      "grad_norm": 0.08285185694694519,
      "learning_rate": 1.3442340176113377e-05,
      "loss": 0.0207,
      "relation_loss": 0.0011808566865511238,
      "selected_pairs": 27126.0,
      "state_query_norm": 42.01703300476074,
      "state_relation_loss": 8.41686240164563e-05,
      "step": 23190,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011808566865511238,
      "weighted_relation_loss": 0.0011808566865511238
    },
    {
      "action_query_norm": 39.774845886230466,
      "action_relation_loss": 0.0010981455678120255,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.702525724976613,
      "flow_matching_loss": 0.0197519158013165,
      "grad_norm": 0.09179160743951797,
      "learning_rate": 1.340476175686784e-05,
      "loss": 0.0209,
      "relation_loss": 0.0011826430330984295,
      "selected_pairs": 27014.2,
      "state_query_norm": 42.01085357666015,
      "state_relation_loss": 8.449746965197847e-05,
      "step": 23200,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011826430330984295,
      "weighted_relation_loss": 0.0011826430330984295
    },
    {
      "action_query_norm": 39.768233108520505,
      "action_relation_loss": 0.001085682987468317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.711880261927035,
      "flow_matching_loss": 0.01929917251691222,
      "grad_norm": 0.13054296374320984,
      "learning_rate": 1.3367227804287469e-05,
      "loss": 0.0205,
      "relation_loss": 0.0011728875455446542,
      "selected_pairs": 26758.7,
      "state_query_norm": 41.999564743041994,
      "state_relation_loss": 8.720453406567686e-05,
      "step": 23210,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0011728875455446542,
      "weighted_relation_loss": 0.0011728875455446542
    },
    {
      "action_query_norm": 39.8570556640625,
      "action_relation_loss": 0.0010233615117613225,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.721234798877454,
      "flow_matching_loss": 0.018723865691572427,
      "grad_norm": 0.07202190160751343,
      "learning_rate": 1.332973836397956e-05,
      "loss": 0.0198,
      "relation_loss": 0.0011091858497820795,
      "selected_pairs": 27199.2,
      "state_query_norm": 42.00620803833008,
      "state_relation_loss": 8.582434456911869e-05,
      "step": 23220,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0011091858497820795,
      "weighted_relation_loss": 0.0011091858497820795
    },
    {
      "action_query_norm": 39.790719985961914,
      "action_relation_loss": 0.001110196695663035,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.730589335827876,
      "flow_matching_loss": 0.019162285141646862,
      "grad_norm": 0.08567442744970322,
      "learning_rate": 1.329229348149731e-05,
      "loss": 0.0204,
      "relation_loss": 0.0011928796186111868,
      "selected_pairs": 27246.6,
      "state_query_norm": 42.002302169799805,
      "state_relation_loss": 8.268293822766282e-05,
      "step": 23230,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0011928796186111868,
      "weighted_relation_loss": 0.0011928796186111868
    },
    {
      "action_query_norm": 39.757551193237305,
      "action_relation_loss": 0.001085044740466401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.739943872778298,
      "flow_matching_loss": 0.01866537444293499,
      "grad_norm": 0.09706369787454605,
      "learning_rate": 1.3254893202339796e-05,
      "loss": 0.0198,
      "relation_loss": 0.0011687855701893568,
      "selected_pairs": 27058.4,
      "state_query_norm": 41.99518356323242,
      "state_relation_loss": 8.37408093502745e-05,
      "step": 23240,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0011687855701893568,
      "weighted_relation_loss": 0.0011687855701893568
    },
    {
      "action_query_norm": 39.768694686889646,
      "action_relation_loss": 0.001082133420277387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.74929840972872,
      "flow_matching_loss": 0.019938627630472182,
      "grad_norm": 0.1436966508626938,
      "learning_rate": 1.3217537571951872e-05,
      "loss": 0.0211,
      "relation_loss": 0.0011707891710102559,
      "selected_pairs": 26618.5,
      "state_query_norm": 41.99990730285644,
      "state_relation_loss": 8.865576310199686e-05,
      "step": 23250,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011707891710102559,
      "weighted_relation_loss": 0.0011707891710102559
    },
    {
      "action_query_norm": 39.793609619140625,
      "action_relation_loss": 0.001073954429011792,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.75865294667914,
      "flow_matching_loss": 0.01787415109574795,
      "grad_norm": 0.09713089466094971,
      "learning_rate": 1.3180226635724169e-05,
      "loss": 0.019,
      "relation_loss": 0.0011555712902918459,
      "selected_pairs": 26738.3,
      "state_query_norm": 41.98037261962891,
      "state_relation_loss": 8.1616854731692e-05,
      "step": 23260,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011555712902918459,
      "weighted_relation_loss": 0.0011555712902918459
    },
    {
      "action_query_norm": 39.81262550354004,
      "action_relation_loss": 0.0010064585076179355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.76800748362956,
      "flow_matching_loss": 0.018527879007160665,
      "grad_norm": 0.07739806920289993,
      "learning_rate": 1.314296043899298e-05,
      "loss": 0.0196,
      "relation_loss": 0.0010923909372650086,
      "selected_pairs": 27341.7,
      "state_query_norm": 41.99545478820801,
      "state_relation_loss": 8.593244056100957e-05,
      "step": 23270,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0010923909372650086,
      "weighted_relation_loss": 0.0010923909372650086
    },
    {
      "action_query_norm": 39.8042610168457,
      "action_relation_loss": 0.0010429761605337262,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.777362020579982,
      "flow_matching_loss": 0.018453894183039665,
      "grad_norm": 0.12468446046113968,
      "learning_rate": 1.3105739027040248e-05,
      "loss": 0.0196,
      "relation_loss": 0.0011295008356682957,
      "selected_pairs": 26989.5,
      "state_query_norm": 41.99381294250488,
      "state_relation_loss": 8.652467367937788e-05,
      "step": 23280,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0011295008356682957,
      "weighted_relation_loss": 0.0011295008356682957
    },
    {
      "action_query_norm": 39.74731788635254,
      "action_relation_loss": 0.0010486851504538208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.7867165575304,
      "flow_matching_loss": 0.018583497125655414,
      "grad_norm": 0.11130590736865997,
      "learning_rate": 1.3068562445093491e-05,
      "loss": 0.0197,
      "relation_loss": 0.001129957428202033,
      "selected_pairs": 26484.2,
      "state_query_norm": 42.01268692016602,
      "state_relation_loss": 8.127229157253168e-05,
      "step": 23290,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001129957428202033,
      "weighted_relation_loss": 0.001129957428202033
    },
    {
      "action_query_norm": 39.80763931274414,
      "action_relation_loss": 0.0010544934018980711,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.796071094480823,
      "flow_matching_loss": 0.019198376778513195,
      "grad_norm": 0.11800230294466019,
      "learning_rate": 1.3031430738325778e-05,
      "loss": 0.0203,
      "relation_loss": 0.0011369076208211481,
      "selected_pairs": 26737.0,
      "state_query_norm": 42.00257759094238,
      "state_relation_loss": 8.241421019192786e-05,
      "step": 23300,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011369076208211481,
      "weighted_relation_loss": 0.0011369076208211481
    },
    {
      "action_query_norm": 39.77725257873535,
      "action_relation_loss": 0.0010714410338550806,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.805425631431245,
      "flow_matching_loss": 0.01861215867102146,
      "grad_norm": 0.09569915384054184,
      "learning_rate": 1.299434395185563e-05,
      "loss": 0.0198,
      "relation_loss": 0.001152989116962999,
      "selected_pairs": 26374.4,
      "state_query_norm": 41.98934669494629,
      "state_relation_loss": 8.154808529070579e-05,
      "step": 23310,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.001152989116962999,
      "weighted_relation_loss": 0.001152989116962999
    },
    {
      "action_query_norm": 39.75683174133301,
      "action_relation_loss": 0.0011152428400237114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.814780168381667,
      "flow_matching_loss": 0.017270498257130385,
      "grad_norm": 0.13768672943115234,
      "learning_rate": 1.295730213074699e-05,
      "loss": 0.0185,
      "relation_loss": 0.0012005646713078022,
      "selected_pairs": 26757.3,
      "state_query_norm": 41.99159240722656,
      "state_relation_loss": 8.532181236660108e-05,
      "step": 23320,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0012005646713078022,
      "weighted_relation_loss": 0.0012005646713078022
    },
    {
      "action_query_norm": 39.79883041381836,
      "action_relation_loss": 0.0011245188186876475,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.824134705332085,
      "flow_matching_loss": 0.018149302527308463,
      "grad_norm": 0.1047331765294075,
      "learning_rate": 1.2920305320009152e-05,
      "loss": 0.0194,
      "relation_loss": 0.0012116646743379534,
      "selected_pairs": 26567.8,
      "state_query_norm": 42.008266830444335,
      "state_relation_loss": 8.71458621986676e-05,
      "step": 23330,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0012116646743379534,
      "weighted_relation_loss": 0.0012116646743379534
    },
    {
      "action_query_norm": 39.773847579956055,
      "action_relation_loss": 0.0010724900115747006,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.833489242282507,
      "flow_matching_loss": 0.018114336766302586,
      "grad_norm": 0.09331481158733368,
      "learning_rate": 1.2883353564596729e-05,
      "loss": 0.0193,
      "relation_loss": 0.0011570418486371637,
      "selected_pairs": 26917.6,
      "state_query_norm": 41.99961547851562,
      "state_relation_loss": 8.455180941382423e-05,
      "step": 23340,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0011570418486371637,
      "weighted_relation_loss": 0.0011570418486371637
    },
    {
      "action_query_norm": 39.75648078918457,
      "action_relation_loss": 0.0010446140018757432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.84284377923293,
      "flow_matching_loss": 0.019971388392150402,
      "grad_norm": 0.11235278099775314,
      "learning_rate": 1.2846446909409604e-05,
      "loss": 0.0211,
      "relation_loss": 0.0011334773909766227,
      "selected_pairs": 26803.6,
      "state_query_norm": 41.984191513061525,
      "state_relation_loss": 8.886338473530486e-05,
      "step": 23350,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011334773909766227,
      "weighted_relation_loss": 0.0011334773909766227
    },
    {
      "action_query_norm": 39.763298416137694,
      "action_relation_loss": 0.0010607678443193436,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.852198316183348,
      "flow_matching_loss": 0.018829462863504887,
      "grad_norm": 0.12505307793617249,
      "learning_rate": 1.280958539929284e-05,
      "loss": 0.02,
      "relation_loss": 0.0011424205382354557,
      "selected_pairs": 26615.1,
      "state_query_norm": 42.008527755737305,
      "state_relation_loss": 8.165267718140967e-05,
      "step": 23360,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011424205382354557,
      "weighted_relation_loss": 0.0011424205382354557
    },
    {
      "action_query_norm": 39.79360084533691,
      "action_relation_loss": 0.0010570929269306361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.86155285313377,
      "flow_matching_loss": 0.01725142803043127,
      "grad_norm": 0.08841611444950104,
      "learning_rate": 1.277276907903664e-05,
      "loss": 0.0184,
      "relation_loss": 0.001142298080958426,
      "selected_pairs": 26872.7,
      "state_query_norm": 42.00956001281738,
      "state_relation_loss": 8.520516348653473e-05,
      "step": 23370,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001142298080958426,
      "weighted_relation_loss": 0.001142298080958426
    },
    {
      "action_query_norm": 39.736680603027345,
      "action_relation_loss": 0.001073633646592498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.87090739008419,
      "flow_matching_loss": 0.019511476811021566,
      "grad_norm": 0.07390958070755005,
      "learning_rate": 1.2735997993376298e-05,
      "loss": 0.0207,
      "relation_loss": 0.001160163147142157,
      "selected_pairs": 26530.6,
      "state_query_norm": 41.989571762084964,
      "state_relation_loss": 8.652949545648881e-05,
      "step": 23380,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001160163147142157,
      "weighted_relation_loss": 0.001160163147142157
    },
    {
      "action_query_norm": 39.78179244995117,
      "action_relation_loss": 0.0010955601639579982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.880261927034613,
      "flow_matching_loss": 0.017632776405662298,
      "grad_norm": 0.08364836871623993,
      "learning_rate": 1.2699272186992166e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011792682635132223,
      "selected_pairs": 26803.8,
      "state_query_norm": 41.99346389770508,
      "state_relation_loss": 8.370808354811743e-05,
      "step": 23390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0011792682635132223,
      "weighted_relation_loss": 0.0011792682635132223
    },
    {
      "action_query_norm": 39.78335456848144,
      "action_relation_loss": 0.0011081518488936127,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.889616463985032,
      "flow_matching_loss": 0.01678032260388136,
      "grad_norm": 0.08623135834932327,
      "learning_rate": 1.2662591704509547e-05,
      "loss": 0.018,
      "relation_loss": 0.001191822881810367,
      "selected_pairs": 26523.3,
      "state_query_norm": 42.00090217590332,
      "state_relation_loss": 8.36710307339672e-05,
      "step": 23400,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001191822881810367,
      "weighted_relation_loss": 0.001191822881810367
    },
    {
      "action_query_norm": 39.79430656433105,
      "action_relation_loss": 0.0010130906128324568,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.898971000935454,
      "flow_matching_loss": 0.019720050692558288,
      "grad_norm": 0.07633781433105469,
      "learning_rate": 1.2625956590498711e-05,
      "loss": 0.0208,
      "relation_loss": 0.0010949291288852691,
      "selected_pairs": 26500.1,
      "state_query_norm": 42.003210830688474,
      "state_relation_loss": 8.18385167804081e-05,
      "step": 23410,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0010949291288852691,
      "weighted_relation_loss": 0.0010949291288852691
    },
    {
      "action_query_norm": 39.800017547607425,
      "action_relation_loss": 0.0010192785994149745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.908325537885876,
      "flow_matching_loss": 0.016853472217917442,
      "grad_norm": 0.1369825005531311,
      "learning_rate": 1.2589366889474757e-05,
      "loss": 0.018,
      "relation_loss": 0.0011029452900402248,
      "selected_pairs": 26940.0,
      "state_query_norm": 41.997093963623044,
      "state_relation_loss": 8.366667971131392e-05,
      "step": 23420,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0011029452900402248,
      "weighted_relation_loss": 0.0011029452900402248
    },
    {
      "action_query_norm": 39.815865707397464,
      "action_relation_loss": 0.001044727029511705,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.917680074836294,
      "flow_matching_loss": 0.019972456246614458,
      "grad_norm": 0.2637016177177429,
      "learning_rate": 1.2552822645897621e-05,
      "loss": 0.0211,
      "relation_loss": 0.001129544829018414,
      "selected_pairs": 27174.7,
      "state_query_norm": 41.97679443359375,
      "state_relation_loss": 8.481780387228354e-05,
      "step": 23430,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001129544829018414,
      "weighted_relation_loss": 0.001129544829018414
    },
    {
      "action_query_norm": 39.77906341552735,
      "action_relation_loss": 0.0010286140721291303,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.927034611786716,
      "flow_matching_loss": 0.01826417837291956,
      "grad_norm": 0.09957195818424225,
      "learning_rate": 1.2516323904172e-05,
      "loss": 0.0194,
      "relation_loss": 0.0011133129824884235,
      "selected_pairs": 27317.0,
      "state_query_norm": 41.98057060241699,
      "state_relation_loss": 8.469891618005932e-05,
      "step": 23440,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0011133129824884235,
      "weighted_relation_loss": 0.0011133129824884235
    },
    {
      "action_query_norm": 39.755851364135744,
      "action_relation_loss": 0.0010664348141290248,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.93638914873714,
      "flow_matching_loss": 0.01875059762969613,
      "grad_norm": 0.1044197604060173,
      "learning_rate": 1.2479870708647324e-05,
      "loss": 0.0199,
      "relation_loss": 0.001153024530503899,
      "selected_pairs": 27244.7,
      "state_query_norm": 41.99115409851074,
      "state_relation_loss": 8.658972874400205e-05,
      "step": 23450,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001153024530503899,
      "weighted_relation_loss": 0.001153024530503899
    },
    {
      "action_query_norm": 39.70305824279785,
      "action_relation_loss": 0.0011119652772322297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.94574368568756,
      "flow_matching_loss": 0.018807446397840977,
      "grad_norm": 0.11617080867290497,
      "learning_rate": 1.2443463103617658e-05,
      "loss": 0.02,
      "relation_loss": 0.0011914719245396555,
      "selected_pairs": 27040.2,
      "state_query_norm": 41.98812408447266,
      "state_relation_loss": 7.95066291175317e-05,
      "step": 23460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0011914719245396555,
      "weighted_relation_loss": 0.0011914719245396555
    },
    {
      "action_query_norm": 39.75185089111328,
      "action_relation_loss": 0.0010483833961188794,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.95509822263798,
      "flow_matching_loss": 0.017010899633169173,
      "grad_norm": 0.10588321089744568,
      "learning_rate": 1.240710113332167e-05,
      "loss": 0.0181,
      "relation_loss": 0.0011334894457831978,
      "selected_pairs": 26668.8,
      "state_query_norm": 41.99703712463379,
      "state_relation_loss": 8.510607658536173e-05,
      "step": 23470,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011334894457831978,
      "weighted_relation_loss": 0.0011334894457831978
    },
    {
      "action_query_norm": 39.770342636108396,
      "action_relation_loss": 0.0010891654994338751,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.9644527595884,
      "flow_matching_loss": 0.019187624100595713,
      "grad_norm": 0.09536147862672806,
      "learning_rate": 1.2370784841942584e-05,
      "loss": 0.0204,
      "relation_loss": 0.0011721660324838013,
      "selected_pairs": 27175.5,
      "state_query_norm": 41.97898406982422,
      "state_relation_loss": 8.300053741550073e-05,
      "step": 23480,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0011721660324838013,
      "weighted_relation_loss": 0.0011721660324838013
    },
    {
      "action_query_norm": 39.75643463134766,
      "action_relation_loss": 0.001075145014328882,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.973807296538823,
      "flow_matching_loss": 0.016622941941022873,
      "grad_norm": 0.09771284461021423,
      "learning_rate": 1.2334514273608117e-05,
      "loss": 0.0178,
      "relation_loss": 0.0011592957773245872,
      "selected_pairs": 26445.2,
      "state_query_norm": 41.9849437713623,
      "state_relation_loss": 8.415077536483295e-05,
      "step": 23490,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011592957773245872,
      "weighted_relation_loss": 0.0011592957773245872
    },
    {
      "action_query_norm": 39.75989875793457,
      "action_relation_loss": 0.001118795556249097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.98316183348924,
      "flow_matching_loss": 0.01832532212138176,
      "grad_norm": 0.09088929742574692,
      "learning_rate": 1.2298289472390417e-05,
      "loss": 0.0195,
      "relation_loss": 0.001198259787634015,
      "selected_pairs": 26881.2,
      "state_query_norm": 41.985233306884766,
      "state_relation_loss": 7.946422483655624e-05,
      "step": 23500,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.001198259787634015,
      "weighted_relation_loss": 0.001198259787634015
    },
    {
      "action_query_norm": 39.78120498657226,
      "action_relation_loss": 0.0010486432758625597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 21.992516370439663,
      "flow_matching_loss": 0.018164874613285066,
      "grad_norm": 0.09428282827138901,
      "learning_rate": 1.226211048230606e-05,
      "loss": 0.0193,
      "relation_loss": 0.0011370553052984177,
      "selected_pairs": 27470.3,
      "state_query_norm": 41.98567008972168,
      "state_relation_loss": 8.841204544296488e-05,
      "step": 23510,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0011370553052984177,
      "weighted_relation_loss": 0.0011370553052984177
    },
    {
      "action_query_norm": 39.77042808532715,
      "action_relation_loss": 0.0011086360726039856,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.001870907390085,
      "flow_matching_loss": 0.017617117241024972,
      "grad_norm": 0.09532542526721954,
      "learning_rate": 1.2225977347315921e-05,
      "loss": 0.0188,
      "relation_loss": 0.0011922995618078858,
      "selected_pairs": 24058.7,
      "state_query_norm": 41.977958297729494,
      "state_relation_loss": 8.366351103177294e-05,
      "step": 23520,
      "valid_action_samples": 213.3,
      "weighted_auxiliary_loss": 0.0011922995618078858,
      "weighted_relation_loss": 0.0011922995618078858
    },
    {
      "action_query_norm": 39.73729705810547,
      "action_relation_loss": 0.0010534278699196875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.011225444340504,
      "flow_matching_loss": 0.018850527238100767,
      "grad_norm": 0.09916691482067108,
      "learning_rate": 1.2189890111325148e-05,
      "loss": 0.02,
      "relation_loss": 0.0011360435048118234,
      "selected_pairs": 25891.5,
      "state_query_norm": 41.97619209289551,
      "state_relation_loss": 8.26156472612638e-05,
      "step": 23530,
      "valid_action_samples": 228.0,
      "weighted_auxiliary_loss": 0.0011360435048118234,
      "weighted_relation_loss": 0.0011360435048118234
    },
    {
      "action_query_norm": 39.74915733337402,
      "action_relation_loss": 0.0010185670340433716,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.020579981290926,
      "flow_matching_loss": 0.01943188477307558,
      "grad_norm": 0.11829774081707001,
      "learning_rate": 1.2153848818183161e-05,
      "loss": 0.0205,
      "relation_loss": 0.0011010689311660826,
      "selected_pairs": 26090.6,
      "state_query_norm": 41.96765365600586,
      "state_relation_loss": 8.250190730905161e-05,
      "step": 23540,
      "valid_action_samples": 228.9,
      "weighted_auxiliary_loss": 0.0011010689311660826,
      "weighted_relation_loss": 0.0011010689311660826
    },
    {
      "action_query_norm": 39.79847030639648,
      "action_relation_loss": 0.0009741863759700209,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.029934518241348,
      "flow_matching_loss": 0.01772466255351901,
      "grad_norm": 0.15455491840839386,
      "learning_rate": 1.2117853511683507e-05,
      "loss": 0.0188,
      "relation_loss": 0.0010578245273791254,
      "selected_pairs": 26874.1,
      "state_query_norm": 41.982188034057614,
      "state_relation_loss": 8.363815068150871e-05,
      "step": 23550,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010578245273791254,
      "weighted_relation_loss": 0.0010578245273791254
    },
    {
      "action_query_norm": 39.767917251586915,
      "action_relation_loss": 0.0010356347833294422,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.03928905519177,
      "flow_matching_loss": 0.01685748025774956,
      "grad_norm": 0.08834570646286011,
      "learning_rate": 1.2081904235563906e-05,
      "loss": 0.018,
      "relation_loss": 0.0011187378200702368,
      "selected_pairs": 27342.8,
      "state_query_norm": 41.967755508422854,
      "state_relation_loss": 8.310304692713544e-05,
      "step": 23560,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0011187378200702368,
      "weighted_relation_loss": 0.0011187378200702368
    },
    {
      "action_query_norm": 39.76545867919922,
      "action_relation_loss": 0.0010274423460941763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.048643592142188,
      "flow_matching_loss": 0.018077305145561696,
      "grad_norm": 0.08580750972032547,
      "learning_rate": 1.20460010335061e-05,
      "loss": 0.0192,
      "relation_loss": 0.0011121335090138018,
      "selected_pairs": 26782.3,
      "state_query_norm": 41.991600799560544,
      "state_relation_loss": 8.469115491607226e-05,
      "step": 23570,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0011121335090138018,
      "weighted_relation_loss": 0.0011121335090138018
    },
    {
      "action_query_norm": 39.81836929321289,
      "action_relation_loss": 0.0010195444221608341,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.05799812909261,
      "flow_matching_loss": 0.01762772025540471,
      "grad_norm": 0.08745988458395004,
      "learning_rate": 1.2010143949135865e-05,
      "loss": 0.0187,
      "relation_loss": 0.001104375848080963,
      "selected_pairs": 26564.5,
      "state_query_norm": 41.98999214172363,
      "state_relation_loss": 8.483142082695849e-05,
      "step": 23580,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001104375848080963,
      "weighted_relation_loss": 0.001104375848080963
    },
    {
      "action_query_norm": 39.82676277160645,
      "action_relation_loss": 0.00098505481146276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.067352666043032,
      "flow_matching_loss": 0.018416799884289504,
      "grad_norm": 0.09883110225200653,
      "learning_rate": 1.1974333026022938e-05,
      "loss": 0.0195,
      "relation_loss": 0.0010701109655201434,
      "selected_pairs": 27461.7,
      "state_query_norm": 42.00027770996094,
      "state_relation_loss": 8.505616497131996e-05,
      "step": 23590,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0010701109655201434,
      "weighted_relation_loss": 0.0010701109655201434
    },
    {
      "action_query_norm": 39.7992244720459,
      "action_relation_loss": 0.0009801248204894363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.07670720299345,
      "flow_matching_loss": 0.019696478359401227,
      "grad_norm": 0.09780317544937134,
      "learning_rate": 1.1938568307680964e-05,
      "loss": 0.0208,
      "relation_loss": 0.001061503152595833,
      "selected_pairs": 27181.0,
      "state_query_norm": 41.96623344421387,
      "state_relation_loss": 8.137832119246014e-05,
      "step": 23600,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.001061503152595833,
      "weighted_relation_loss": 0.001061503152595833
    },
    {
      "action_query_norm": 39.757849502563474,
      "action_relation_loss": 0.0010285070049576462,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.086061739943872,
      "flow_matching_loss": 0.016731109097599985,
      "grad_norm": 0.11833162605762482,
      "learning_rate": 1.1902849837567464e-05,
      "loss": 0.0178,
      "relation_loss": 0.0011072834546212106,
      "selected_pairs": 26660.5,
      "state_query_norm": 41.98709945678711,
      "state_relation_loss": 7.87764583947137e-05,
      "step": 23610,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0011072834546212106,
      "weighted_relation_loss": 0.0011072834546212106
    },
    {
      "action_query_norm": 39.74167938232422,
      "action_relation_loss": 0.0010524486715439707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.095416276894294,
      "flow_matching_loss": 0.018434777203947306,
      "grad_norm": 0.11398444324731827,
      "learning_rate": 1.1867177659083739e-05,
      "loss": 0.0196,
      "relation_loss": 0.0011316385818645359,
      "selected_pairs": 26571.1,
      "state_query_norm": 41.9827766418457,
      "state_relation_loss": 7.918991541373543e-05,
      "step": 23620,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011316385818645359,
      "weighted_relation_loss": 0.0011316385818645359
    },
    {
      "action_query_norm": 39.751565551757814,
      "action_relation_loss": 0.0010796859918627889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.104770813844716,
      "flow_matching_loss": 0.01896302802488208,
      "grad_norm": 0.08565645664930344,
      "learning_rate": 1.1831551815574848e-05,
      "loss": 0.0201,
      "relation_loss": 0.0011690508341416716,
      "selected_pairs": 27736.0,
      "state_query_norm": 41.97537040710449,
      "state_relation_loss": 8.936484446167014e-05,
      "step": 23630,
      "valid_action_samples": 236.0,
      "weighted_auxiliary_loss": 0.0011690508341416716,
      "weighted_relation_loss": 0.0011690508341416716
    },
    {
      "action_query_norm": 39.787469482421876,
      "action_relation_loss": 0.001042894780403003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.114125350795135,
      "flow_matching_loss": 0.018834071699529886,
      "grad_norm": 0.10716520994901657,
      "learning_rate": 1.1795972350329554e-05,
      "loss": 0.02,
      "relation_loss": 0.0011296648590359837,
      "selected_pairs": 26579.5,
      "state_query_norm": 41.97365188598633,
      "state_relation_loss": 8.677006771904417e-05,
      "step": 23640,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011296648590359837,
      "weighted_relation_loss": 0.0011296648590359837
    },
    {
      "action_query_norm": 39.695652389526366,
      "action_relation_loss": 0.0010022171307355165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.123479887745557,
      "flow_matching_loss": 0.018406437430530786,
      "grad_norm": 0.10306337475776672,
      "learning_rate": 1.176043930658024e-05,
      "loss": 0.0195,
      "relation_loss": 0.0010860686830710619,
      "selected_pairs": 26965.6,
      "state_query_norm": 41.96667404174805,
      "state_relation_loss": 8.385152905248106e-05,
      "step": 23650,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010860686830710619,
      "weighted_relation_loss": 0.0010860686830710619
    },
    {
      "action_query_norm": 39.749383544921876,
      "action_relation_loss": 0.0009962726966477931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.13283442469598,
      "flow_matching_loss": 0.018131538573652507,
      "grad_norm": 0.093605637550354,
      "learning_rate": 1.1724952727502942e-05,
      "loss": 0.0192,
      "relation_loss": 0.0010773412650451064,
      "selected_pairs": 26669.5,
      "state_query_norm": 41.96820411682129,
      "state_relation_loss": 8.106857276288793e-05,
      "step": 23660,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010773412650451064,
      "weighted_relation_loss": 0.0010773412650451064
    },
    {
      "action_query_norm": 39.785447692871095,
      "action_relation_loss": 0.001014193327864632,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.142188961646397,
      "flow_matching_loss": 0.01808310430496931,
      "grad_norm": 0.09808924794197083,
      "learning_rate": 1.1689512656217178e-05,
      "loss": 0.0192,
      "relation_loss": 0.0010975776531267911,
      "selected_pairs": 26948.6,
      "state_query_norm": 41.97146797180176,
      "state_relation_loss": 8.338433326571248e-05,
      "step": 23670,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010975776531267911,
      "weighted_relation_loss": 0.0010975776531267911
    },
    {
      "action_query_norm": 39.75105361938476,
      "action_relation_loss": 0.0009934039146173746,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.15154349859682,
      "flow_matching_loss": 0.01829218929633498,
      "grad_norm": 0.12345608323812485,
      "learning_rate": 1.1654119135785962e-05,
      "loss": 0.0194,
      "relation_loss": 0.0010742625687271356,
      "selected_pairs": 26555.9,
      "state_query_norm": 41.98348884582519,
      "state_relation_loss": 8.085864683380351e-05,
      "step": 23680,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010742625687271356,
      "weighted_relation_loss": 0.0010742625687271356
    },
    {
      "action_query_norm": 39.74268341064453,
      "action_relation_loss": 0.0010137474222574384,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.16089803554724,
      "flow_matching_loss": 0.018141576368361712,
      "grad_norm": 0.10807108879089355,
      "learning_rate": 1.1618772209215794e-05,
      "loss": 0.0192,
      "relation_loss": 0.0010947536095045507,
      "selected_pairs": 27019.2,
      "state_query_norm": 41.96256866455078,
      "state_relation_loss": 8.100619743345305e-05,
      "step": 23690,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010947536095045507,
      "weighted_relation_loss": 0.0010947536095045507
    },
    {
      "action_query_norm": 39.768647003173825,
      "action_relation_loss": 0.0009846480505075305,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.170252572497663,
      "flow_matching_loss": 0.016911105066537858,
      "grad_norm": 0.08287233859300613,
      "learning_rate": 1.1583471919456506e-05,
      "loss": 0.018,
      "relation_loss": 0.0010677919432055204,
      "selected_pairs": 27126.0,
      "state_query_norm": 41.9680477142334,
      "state_relation_loss": 8.314389488077723e-05,
      "step": 23700,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010677919432055204,
      "weighted_relation_loss": 0.0010677919432055204
    },
    {
      "action_query_norm": 39.77962532043457,
      "action_relation_loss": 0.0010080299922265112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.17960710944808,
      "flow_matching_loss": 0.017262809816747902,
      "grad_norm": 0.14439551532268524,
      "learning_rate": 1.1548218309401266e-05,
      "loss": 0.0184,
      "relation_loss": 0.001088367379270494,
      "selected_pairs": 26476.7,
      "state_query_norm": 41.982546997070315,
      "state_relation_loss": 8.033737176447176e-05,
      "step": 23710,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001088367379270494,
      "weighted_relation_loss": 0.001088367379270494
    },
    {
      "action_query_norm": 39.768018341064455,
      "action_relation_loss": 0.0010318516229745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.188961646398504,
      "flow_matching_loss": 0.017162440903484823,
      "grad_norm": 0.09456747025251389,
      "learning_rate": 1.1513011421886555e-05,
      "loss": 0.0183,
      "relation_loss": 0.0011138718109577894,
      "selected_pairs": 26556.4,
      "state_query_norm": 41.9614616394043,
      "state_relation_loss": 8.202018871088513e-05,
      "step": 23720,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0011138718109577894,
      "weighted_relation_loss": 0.0011138718109577894
    },
    {
      "action_query_norm": 39.7603702545166,
      "action_relation_loss": 0.0009828691021539272,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.198316183348926,
      "flow_matching_loss": 0.01595891546458006,
      "grad_norm": 0.1875593215227127,
      "learning_rate": 1.1477851299692056e-05,
      "loss": 0.017,
      "relation_loss": 0.0010679605649784208,
      "selected_pairs": 27320.2,
      "state_query_norm": 41.992150115966794,
      "state_relation_loss": 8.509147082804702e-05,
      "step": 23730,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0010679605649784208,
      "weighted_relation_loss": 0.0010679605649784208
    },
    {
      "action_query_norm": 39.76219940185547,
      "action_relation_loss": 0.00103672263212502,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.207670720299344,
      "flow_matching_loss": 0.017272148001939058,
      "grad_norm": 0.1336297243833542,
      "learning_rate": 1.144273798554063e-05,
      "loss": 0.0184,
      "relation_loss": 0.0011187727563083173,
      "selected_pairs": 27289.5,
      "state_query_norm": 41.98259735107422,
      "state_relation_loss": 8.205011836253107e-05,
      "step": 23740,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011187727563083173,
      "weighted_relation_loss": 0.0011187727563083173
    },
    {
      "action_query_norm": 39.77691764831543,
      "action_relation_loss": 0.001013126171892509,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.217025257249766,
      "flow_matching_loss": 0.017760601453483106,
      "grad_norm": 0.09719469398260117,
      "learning_rate": 1.1407671522098262e-05,
      "loss": 0.0189,
      "relation_loss": 0.0010929336887784303,
      "selected_pairs": 26846.0,
      "state_query_norm": 41.992203903198245,
      "state_relation_loss": 7.980751252034679e-05,
      "step": 23750,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010929336887784303,
      "weighted_relation_loss": 0.0010929336887784303
    },
    {
      "action_query_norm": 39.707832717895506,
      "action_relation_loss": 0.0010055781924165786,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.226379794200188,
      "flow_matching_loss": 0.017507341131567954,
      "grad_norm": 0.10219544917345047,
      "learning_rate": 1.1372651951974e-05,
      "loss": 0.0186,
      "relation_loss": 0.0010844828560948373,
      "selected_pairs": 27152.3,
      "state_query_norm": 41.98970108032226,
      "state_relation_loss": 7.890467459219508e-05,
      "step": 23760,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0010844828560948373,
      "weighted_relation_loss": 0.0010844828560948373
    },
    {
      "action_query_norm": 39.73435745239258,
      "action_relation_loss": 0.0009878444543574005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.235734331150606,
      "flow_matching_loss": 0.017351291328668594,
      "grad_norm": 0.08503735065460205,
      "learning_rate": 1.1337679317719952e-05,
      "loss": 0.0184,
      "relation_loss": 0.001069192070281133,
      "selected_pairs": 26909.5,
      "state_query_norm": 41.974596405029295,
      "state_relation_loss": 8.134762101690285e-05,
      "step": 23770,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.001069192070281133,
      "weighted_relation_loss": 0.001069192070281133
    },
    {
      "action_query_norm": 39.753289794921876,
      "action_relation_loss": 0.001046180841512978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.24508886810103,
      "flow_matching_loss": 0.01646540258079767,
      "grad_norm": 0.08207350969314575,
      "learning_rate": 1.1302753661831151e-05,
      "loss": 0.0176,
      "relation_loss": 0.001129319635219872,
      "selected_pairs": 26299.2,
      "state_query_norm": 41.980498504638675,
      "state_relation_loss": 8.313879152410664e-05,
      "step": 23780,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.001129319635219872,
      "weighted_relation_loss": 0.001129319635219872
    },
    {
      "action_query_norm": 39.73287467956543,
      "action_relation_loss": 0.0009948630409780889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.25444340505145,
      "flow_matching_loss": 0.01742381285876036,
      "grad_norm": 0.12237343192100525,
      "learning_rate": 1.1267875026745561e-05,
      "loss": 0.0185,
      "relation_loss": 0.0010742714046500623,
      "selected_pairs": 26805.6,
      "state_query_norm": 41.97034034729004,
      "state_relation_loss": 7.940834329929203e-05,
      "step": 23790,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010742714046500623,
      "weighted_relation_loss": 0.0010742714046500623
    },
    {
      "action_query_norm": 39.828985595703124,
      "action_relation_loss": 0.0009691128507256508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.263797942001872,
      "flow_matching_loss": 0.017066853959113358,
      "grad_norm": 0.10851491242647171,
      "learning_rate": 1.1233043454844016e-05,
      "loss": 0.0181,
      "relation_loss": 0.0010496554197743535,
      "selected_pairs": 26963.9,
      "state_query_norm": 41.98011169433594,
      "state_relation_loss": 8.05425581347663e-05,
      "step": 23800,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010496554197743535,
      "weighted_relation_loss": 0.0010496554197743535
    },
    {
      "action_query_norm": 39.76765098571777,
      "action_relation_loss": 0.0010458696051500738,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.27315247895229,
      "flow_matching_loss": 0.01869614003226161,
      "grad_norm": 0.09790556877851486,
      "learning_rate": 1.1198258988450144e-05,
      "loss": 0.0198,
      "relation_loss": 0.0011292080860584974,
      "selected_pairs": 27015.4,
      "state_query_norm": 41.97003059387207,
      "state_relation_loss": 8.33384765428491e-05,
      "step": 23810,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011292080860584974,
      "weighted_relation_loss": 0.0011292080860584974
    },
    {
      "action_query_norm": 39.75080757141113,
      "action_relation_loss": 0.0010178770462516695,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.282507015902713,
      "flow_matching_loss": 0.017579589132219554,
      "grad_norm": 0.10069730132818222,
      "learning_rate": 1.1163521669830368e-05,
      "loss": 0.0187,
      "relation_loss": 0.0011012051312718541,
      "selected_pairs": 27405.1,
      "state_query_norm": 41.974396514892575,
      "state_relation_loss": 8.332809520652517e-05,
      "step": 23820,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0011012051312718541,
      "weighted_relation_loss": 0.0011012051312718541
    },
    {
      "action_query_norm": 39.71369895935059,
      "action_relation_loss": 0.0010386905167251825,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.291861552853135,
      "flow_matching_loss": 0.018589503969997166,
      "grad_norm": 0.11745359748601913,
      "learning_rate": 1.1128831541193818e-05,
      "loss": 0.0197,
      "relation_loss": 0.0011235850979574025,
      "selected_pairs": 26051.9,
      "state_query_norm": 41.962222290039065,
      "state_relation_loss": 8.489458705298602e-05,
      "step": 23830,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0011235850979574025,
      "weighted_relation_loss": 0.0011235850979574025
    },
    {
      "action_query_norm": 39.71987533569336,
      "action_relation_loss": 0.0010663057677447797,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.301216089803553,
      "flow_matching_loss": 0.018392862193286418,
      "grad_norm": 0.11307962983846664,
      "learning_rate": 1.1094188644692255e-05,
      "loss": 0.0195,
      "relation_loss": 0.0011475192033685744,
      "selected_pairs": 27009.0,
      "state_query_norm": 41.97544212341309,
      "state_relation_loss": 8.121344144456089e-05,
      "step": 23840,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011475192033685744,
      "weighted_relation_loss": 0.0011475192033685744
    },
    {
      "action_query_norm": 39.706868743896486,
      "action_relation_loss": 0.0010303790797479451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.310570626753975,
      "flow_matching_loss": 0.01827324079349637,
      "grad_norm": 0.14802570641040802,
      "learning_rate": 1.1059593022420067e-05,
      "loss": 0.0194,
      "relation_loss": 0.0011121325951535255,
      "selected_pairs": 26184.5,
      "state_query_norm": 41.950422286987305,
      "state_relation_loss": 8.175353359547443e-05,
      "step": 23850,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0011121325951535255,
      "weighted_relation_loss": 0.0011121325951535255
    },
    {
      "action_query_norm": 39.75473480224609,
      "action_relation_loss": 0.000997972657205537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.319925163704397,
      "flow_matching_loss": 0.017833287362009287,
      "grad_norm": 0.1216316968202591,
      "learning_rate": 1.1025044716414184e-05,
      "loss": 0.0189,
      "relation_loss": 0.001079283596482128,
      "selected_pairs": 26757.7,
      "state_query_norm": 41.97067756652832,
      "state_relation_loss": 8.131093272822909e-05,
      "step": 23860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001079283596482128,
      "weighted_relation_loss": 0.001079283596482128
    },
    {
      "action_query_norm": 39.75717964172363,
      "action_relation_loss": 0.0009520598046947271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.32927970065482,
      "flow_matching_loss": 0.017206365149468182,
      "grad_norm": 0.11169828474521637,
      "learning_rate": 1.0990543768654082e-05,
      "loss": 0.0182,
      "relation_loss": 0.0010379384504631162,
      "selected_pairs": 26804.4,
      "state_query_norm": 41.96765480041504,
      "state_relation_loss": 8.587865304434671e-05,
      "step": 23870,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010379384504631162,
      "weighted_relation_loss": 0.0010379384504631162
    },
    {
      "action_query_norm": 39.68753852844238,
      "action_relation_loss": 0.001051700668176636,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.338634237605238,
      "flow_matching_loss": 0.016918948385864496,
      "grad_norm": 0.11662351340055466,
      "learning_rate": 1.0956090221061655e-05,
      "loss": 0.0181,
      "relation_loss": 0.0011327223852276801,
      "selected_pairs": 26689.7,
      "state_query_norm": 41.96588554382324,
      "state_relation_loss": 8.102170977508649e-05,
      "step": 23880,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011327223852276801,
      "weighted_relation_loss": 0.0011327223852276801
    },
    {
      "action_query_norm": 39.72278022766113,
      "action_relation_loss": 0.000989371136529371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.34798877455566,
      "flow_matching_loss": 0.018121918383985757,
      "grad_norm": 0.11579793691635132,
      "learning_rate": 1.0921684115501208e-05,
      "loss": 0.0192,
      "relation_loss": 0.0010713852709159255,
      "selected_pairs": 26859.9,
      "state_query_norm": 42.000796508789065,
      "state_relation_loss": 8.201412201742641e-05,
      "step": 23890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010713852709159255,
      "weighted_relation_loss": 0.0010713852709159255
    },
    {
      "action_query_norm": 39.71739273071289,
      "action_relation_loss": 0.0010248150501865894,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.35734331150608,
      "flow_matching_loss": 0.01815984370186925,
      "grad_norm": 0.11227857321500778,
      "learning_rate": 1.0887325493779405e-05,
      "loss": 0.0193,
      "relation_loss": 0.0011056952585931866,
      "selected_pairs": 26838.5,
      "state_query_norm": 41.95223274230957,
      "state_relation_loss": 8.088020113063976e-05,
      "step": 23900,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0011056952585931866,
      "weighted_relation_loss": 0.0011056952585931866
    },
    {
      "action_query_norm": 39.70790100097656,
      "action_relation_loss": 0.0009985114098526537,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.3666978484565,
      "flow_matching_loss": 0.016884837578982115,
      "grad_norm": 0.12490645051002502,
      "learning_rate": 1.0853014397645207e-05,
      "loss": 0.018,
      "relation_loss": 0.0010804622899740935,
      "selected_pairs": 26695.7,
      "state_query_norm": 41.977626419067384,
      "state_relation_loss": 8.195087139029056e-05,
      "step": 23910,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010804622899740935,
      "weighted_relation_loss": 0.0010804622899740935
    },
    {
      "action_query_norm": 39.71197319030762,
      "action_relation_loss": 0.0010554467735346407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.376052385406922,
      "flow_matching_loss": 0.01778225991874933,
      "grad_norm": 0.11957956105470657,
      "learning_rate": 1.0818750868789829e-05,
      "loss": 0.0189,
      "relation_loss": 0.0011365687241777778,
      "selected_pairs": 26433.6,
      "state_query_norm": 41.98069038391113,
      "state_relation_loss": 8.112194045679643e-05,
      "step": 23920,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011365687241777778,
      "weighted_relation_loss": 0.0011365687241777778
    },
    {
      "action_query_norm": 39.73735008239746,
      "action_relation_loss": 0.0009767539217136801,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.385406922357344,
      "flow_matching_loss": 0.01771612660959363,
      "grad_norm": 0.1443977952003479,
      "learning_rate": 1.0784534948846703e-05,
      "loss": 0.0188,
      "relation_loss": 0.001062962197465822,
      "selected_pairs": 27245.9,
      "state_query_norm": 41.95077362060547,
      "state_relation_loss": 8.620829685241915e-05,
      "step": 23930,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.001062962197465822,
      "weighted_relation_loss": 0.001062962197465822
    },
    {
      "action_query_norm": 39.703546142578126,
      "action_relation_loss": 0.001045527815585956,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.394761459307766,
      "flow_matching_loss": 0.01886375555768609,
      "grad_norm": 0.13636992871761322,
      "learning_rate": 1.0750366679391393e-05,
      "loss": 0.02,
      "relation_loss": 0.0011244932771660387,
      "selected_pairs": 26440.8,
      "state_query_norm": 41.979143905639646,
      "state_relation_loss": 7.896546812844463e-05,
      "step": 23940,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0011244932771660387,
      "weighted_relation_loss": 0.0011244932771660387
    },
    {
      "action_query_norm": 39.764452362060545,
      "action_relation_loss": 0.0009717935463413596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.404115996258184,
      "flow_matching_loss": 0.017677629087120294,
      "grad_norm": 0.08196175843477249,
      "learning_rate": 1.0716246101941557e-05,
      "loss": 0.0187,
      "relation_loss": 0.0010528234648518263,
      "selected_pairs": 27157.4,
      "state_query_norm": 41.95665321350098,
      "state_relation_loss": 8.102993670036084e-05,
      "step": 23950,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0010528234648518263,
      "weighted_relation_loss": 0.0010528234648518263
    },
    {
      "action_query_norm": 39.70772438049316,
      "action_relation_loss": 0.00102819909225218,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.413470533208606,
      "flow_matching_loss": 0.016790640633553268,
      "grad_norm": 0.10485374182462692,
      "learning_rate": 1.0682173257956934e-05,
      "loss": 0.0179,
      "relation_loss": 0.0011107741273008287,
      "selected_pairs": 26709.9,
      "state_query_norm": 41.982876205444335,
      "state_relation_loss": 8.257502486230806e-05,
      "step": 23960,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0011107741273008287,
      "weighted_relation_loss": 0.0011107741273008287
    },
    {
      "action_query_norm": 39.76261100769043,
      "action_relation_loss": 0.0010439708712510765,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.42282507015903,
      "flow_matching_loss": 0.016176443547010422,
      "grad_norm": 0.08877741545438766,
      "learning_rate": 1.064814818883922e-05,
      "loss": 0.0173,
      "relation_loss": 0.0011242099921219052,
      "selected_pairs": 26502.5,
      "state_query_norm": 41.9572509765625,
      "state_relation_loss": 8.023913032957352e-05,
      "step": 23970,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0011242099921219052,
      "weighted_relation_loss": 0.0011242099921219052
    },
    {
      "action_query_norm": 39.7468376159668,
      "action_relation_loss": 0.0010325360752176493,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.432179607109447,
      "flow_matching_loss": 0.018359017092734575,
      "grad_norm": 0.09067875891923904,
      "learning_rate": 1.06141709359321e-05,
      "loss": 0.0195,
      "relation_loss": 0.001110864186193794,
      "selected_pairs": 26835.2,
      "state_query_norm": 41.9635066986084,
      "state_relation_loss": 7.832810879335739e-05,
      "step": 23980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001110864186193794,
      "weighted_relation_loss": 0.001110864186193794
    },
    {
      "action_query_norm": 39.68438453674317,
      "action_relation_loss": 0.0009733809856697917,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.44153414405987,
      "flow_matching_loss": 0.01685713492333889,
      "grad_norm": 0.09456644207239151,
      "learning_rate": 1.058024154052114e-05,
      "loss": 0.0179,
      "relation_loss": 0.001056525786407292,
      "selected_pairs": 26475.9,
      "state_query_norm": 41.951454162597656,
      "state_relation_loss": 8.314478036481887e-05,
      "step": 23990,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001056525786407292,
      "weighted_relation_loss": 0.001056525786407292
    },
    {
      "action_query_norm": 39.74827919006348,
      "action_relation_loss": 0.001039366703480482,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.45088868101029,
      "flow_matching_loss": 0.017032478004693985,
      "grad_norm": 0.1606200486421585,
      "learning_rate": 1.0546360043833747e-05,
      "loss": 0.0182,
      "relation_loss": 0.0011191284633241595,
      "selected_pairs": 27126.5,
      "state_query_norm": 41.95385856628418,
      "state_relation_loss": 7.976174310897477e-05,
      "step": 24000,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011191284633241595,
      "weighted_relation_loss": 0.0011191284633241595
    },
    {
      "action_query_norm": 39.70759544372559,
      "action_relation_loss": 0.0010420045990031212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.46024321796071,
      "flow_matching_loss": 0.018746483698487283,
      "grad_norm": 0.09896688908338547,
      "learning_rate": 1.0512526487039138e-05,
      "loss": 0.0199,
      "relation_loss": 0.0011258664075285197,
      "selected_pairs": 27335.3,
      "state_query_norm": 41.953116226196286,
      "state_relation_loss": 8.386180343222804e-05,
      "step": 24010,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0011258664075285197,
      "weighted_relation_loss": 0.0011258664075285197
    },
    {
      "action_query_norm": 39.74924583435059,
      "action_relation_loss": 0.0009723590745124965,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.46959775491113,
      "flow_matching_loss": 0.016059981565922497,
      "grad_norm": 0.0997300073504448,
      "learning_rate": 1.0478740911248258e-05,
      "loss": 0.0171,
      "relation_loss": 0.0010497503622900694,
      "selected_pairs": 27018.1,
      "state_query_norm": 41.96144676208496,
      "state_relation_loss": 7.739128777757287e-05,
      "step": 24020,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010497503622900694,
      "weighted_relation_loss": 0.0010497503622900694
    },
    {
      "action_query_norm": 39.71985549926758,
      "action_relation_loss": 0.0009849713649600744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.478952291861553,
      "flow_matching_loss": 0.016903980169445276,
      "grad_norm": 0.11164191365242004,
      "learning_rate": 1.0445003357513793e-05,
      "loss": 0.018,
      "relation_loss": 0.001060460542794317,
      "selected_pairs": 27104.7,
      "state_query_norm": 41.95976676940918,
      "state_relation_loss": 7.548918147222139e-05,
      "step": 24030,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001060460542794317,
      "weighted_relation_loss": 0.001060460542794317
    },
    {
      "action_query_norm": 39.69972877502441,
      "action_relation_loss": 0.0010105133813340216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.488306828811975,
      "flow_matching_loss": 0.01712256483733654,
      "grad_norm": 0.08912608027458191,
      "learning_rate": 1.0411313866830042e-05,
      "loss": 0.0182,
      "relation_loss": 0.0010908750351518392,
      "selected_pairs": 27182.1,
      "state_query_norm": 41.97027015686035,
      "state_relation_loss": 8.036164581426419e-05,
      "step": 24040,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0010908750351518392,
      "weighted_relation_loss": 0.0010908750351518392
    },
    {
      "action_query_norm": 39.73134536743164,
      "action_relation_loss": 0.0010508094157557935,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.497661365762394,
      "flow_matching_loss": 0.015727414563298226,
      "grad_norm": 0.08758427202701569,
      "learning_rate": 1.0377672480132915e-05,
      "loss": 0.0169,
      "relation_loss": 0.0011318922857753932,
      "selected_pairs": 27081.9,
      "state_query_norm": 41.96602554321289,
      "state_relation_loss": 8.108287875074893e-05,
      "step": 24050,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011318922857753932,
      "weighted_relation_loss": 0.0011318922857753932
    },
    {
      "action_query_norm": 39.73380699157715,
      "action_relation_loss": 0.001063871307997033,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.507015902712816,
      "flow_matching_loss": 0.016301808692514897,
      "grad_norm": 0.08257541805505753,
      "learning_rate": 1.0344079238299864e-05,
      "loss": 0.0174,
      "relation_loss": 0.0011405201628804206,
      "selected_pairs": 27527.0,
      "state_query_norm": 41.973697662353516,
      "state_relation_loss": 7.664884615223854e-05,
      "step": 24060,
      "valid_action_samples": 235.1,
      "weighted_auxiliary_loss": 0.0011405201628804206,
      "weighted_relation_loss": 0.0011405201628804206
    },
    {
      "action_query_norm": 39.754625701904295,
      "action_relation_loss": 0.0009458362881559878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.516370439663238,
      "flow_matching_loss": 0.01700954120606184,
      "grad_norm": 0.09635550528764725,
      "learning_rate": 1.0310534182149833e-05,
      "loss": 0.018,
      "relation_loss": 0.0010259992675855755,
      "selected_pairs": 26483.9,
      "state_query_norm": 41.96261138916016,
      "state_relation_loss": 8.016298597794958e-05,
      "step": 24070,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0010259992675855755,
      "weighted_relation_loss": 0.0010259992675855755
    },
    {
      "action_query_norm": 39.70009651184082,
      "action_relation_loss": 0.0010139819118194283,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.525724976613656,
      "flow_matching_loss": 0.016865310538560153,
      "grad_norm": 0.09860321134328842,
      "learning_rate": 1.0277037352443258e-05,
      "loss": 0.018,
      "relation_loss": 0.0010977452038787305,
      "selected_pairs": 26625.2,
      "state_query_norm": 41.95175476074219,
      "state_relation_loss": 8.376330151804723e-05,
      "step": 24080,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010977452038787305,
      "weighted_relation_loss": 0.0010977452038787305
    },
    {
      "action_query_norm": 39.71679191589355,
      "action_relation_loss": 0.0010381516767665744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.535079513564078,
      "flow_matching_loss": 0.017864498775452375,
      "grad_norm": 0.08599435538053513,
      "learning_rate": 1.024358878988193e-05,
      "loss": 0.019,
      "relation_loss": 0.0011207720555830747,
      "selected_pairs": 27366.9,
      "state_query_norm": 41.96562385559082,
      "state_relation_loss": 8.262037445092574e-05,
      "step": 24090,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0011207720555830747,
      "weighted_relation_loss": 0.0011207720555830747
    },
    {
      "action_query_norm": 39.70295600891113,
      "action_relation_loss": 0.0010451409965753556,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.5444340505145,
      "flow_matching_loss": 0.01784273227676749,
      "grad_norm": 0.1153043732047081,
      "learning_rate": 1.0210188535108995e-05,
      "loss": 0.019,
      "relation_loss": 0.0011237355356570333,
      "selected_pairs": 26328.0,
      "state_query_norm": 41.95855979919433,
      "state_relation_loss": 7.85945296229329e-05,
      "step": 24100,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0011237355356570333,
      "weighted_relation_loss": 0.0011237355356570333
    },
    {
      "action_query_norm": 39.76699028015137,
      "action_relation_loss": 0.001037954818457365,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.553788587464922,
      "flow_matching_loss": 0.0158383309841156,
      "grad_norm": 0.10836442559957504,
      "learning_rate": 1.0176836628708935e-05,
      "loss": 0.017,
      "relation_loss": 0.0011192686099093408,
      "selected_pairs": 27130.6,
      "state_query_norm": 41.947838592529294,
      "state_relation_loss": 8.131379872793332e-05,
      "step": 24110,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0011192686099093408,
      "weighted_relation_loss": 0.0011192686099093408
    },
    {
      "action_query_norm": 39.69805488586426,
      "action_relation_loss": 0.000998124707257375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.56314312441534,
      "flow_matching_loss": 0.018125064577907322,
      "grad_norm": 0.10357891023159027,
      "learning_rate": 1.0143533111207454e-05,
      "loss": 0.0192,
      "relation_loss": 0.0010799386771395802,
      "selected_pairs": 26878.9,
      "state_query_norm": 41.95799293518066,
      "state_relation_loss": 8.181396115105599e-05,
      "step": 24120,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010799386771395802,
      "weighted_relation_loss": 0.0010799386771395802
    },
    {
      "action_query_norm": 39.7051025390625,
      "action_relation_loss": 0.0009996304928790779,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.572497661365762,
      "flow_matching_loss": 0.018568797782063485,
      "grad_norm": 0.11974745243787766,
      "learning_rate": 1.0110278023071446e-05,
      "loss": 0.0197,
      "relation_loss": 0.0010817572590894997,
      "selected_pairs": 26956.0,
      "state_query_norm": 41.964352416992185,
      "state_relation_loss": 8.212676693801768e-05,
      "step": 24130,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010817572590894997,
      "weighted_relation_loss": 0.0010817572590894997
    },
    {
      "action_query_norm": 39.68078269958496,
      "action_relation_loss": 0.0010374750883784146,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.581852198316184,
      "flow_matching_loss": 0.018093716260045765,
      "grad_norm": 0.12051954865455627,
      "learning_rate": 1.007707140470901e-05,
      "loss": 0.0192,
      "relation_loss": 0.0011185625102370978,
      "selected_pairs": 27023.3,
      "state_query_norm": 41.95468292236328,
      "state_relation_loss": 8.108743495540693e-05,
      "step": 24140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0011185625102370978,
      "weighted_relation_loss": 0.0011185625102370978
    },
    {
      "action_query_norm": 39.69868202209473,
      "action_relation_loss": 0.0009830949478782714,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.591206735266603,
      "flow_matching_loss": 0.016355026233941316,
      "grad_norm": 0.08283624053001404,
      "learning_rate": 1.0043913296469298e-05,
      "loss": 0.0174,
      "relation_loss": 0.0010668462549801917,
      "selected_pairs": 26734.7,
      "state_query_norm": 41.93808364868164,
      "state_relation_loss": 8.375131437787786e-05,
      "step": 24150,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010668462549801917,
      "weighted_relation_loss": 0.0010668462549801917
    },
    {
      "action_query_norm": 39.72596817016601,
      "action_relation_loss": 0.001003393594874069,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.600561272217025,
      "flow_matching_loss": 0.01699995920062065,
      "grad_norm": 0.1142897978425026,
      "learning_rate": 1.001080373864255e-05,
      "loss": 0.0181,
      "relation_loss": 0.001085711584892124,
      "selected_pairs": 26760.6,
      "state_query_norm": 41.94758987426758,
      "state_relation_loss": 8.231800384237432e-05,
      "step": 24160,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001085711584892124,
      "weighted_relation_loss": 0.001085711584892124
    },
    {
      "action_query_norm": 39.72498474121094,
      "action_relation_loss": 0.001086423051310703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.609915809167447,
      "flow_matching_loss": 0.01736116614192724,
      "grad_norm": 0.10755244642496109,
      "learning_rate": 9.977742771459991e-06,
      "loss": 0.0185,
      "relation_loss": 0.0011663471232168376,
      "selected_pairs": 26699.1,
      "state_query_norm": 41.947700119018556,
      "state_relation_loss": 7.992408136487938e-05,
      "step": 24170,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011663471232168376,
      "weighted_relation_loss": 0.0011663471232168376
    },
    {
      "action_query_norm": 39.70413246154785,
      "action_relation_loss": 0.0009686912526376545,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.61927034611787,
      "flow_matching_loss": 0.0181826108135283,
      "grad_norm": 0.08421898633241653,
      "learning_rate": 9.944730435093802e-06,
      "loss": 0.0192,
      "relation_loss": 0.0010518645867705345,
      "selected_pairs": 27487.1,
      "state_query_norm": 41.94311676025391,
      "state_relation_loss": 8.317333631566726e-05,
      "step": 24180,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0010518645867705345,
      "weighted_relation_loss": 0.0010518645867705345
    },
    {
      "action_query_norm": 39.71761207580566,
      "action_relation_loss": 0.0010109358583576978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.628624883068287,
      "flow_matching_loss": 0.017479291278868915,
      "grad_norm": 0.1155361607670784,
      "learning_rate": 9.911766769657115e-06,
      "loss": 0.0186,
      "relation_loss": 0.0010955089470371603,
      "selected_pairs": 26993.2,
      "state_query_norm": 41.960687255859376,
      "state_relation_loss": 8.457310686935671e-05,
      "step": 24190,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010955089470371603,
      "weighted_relation_loss": 0.0010955089470371603
    },
    {
      "action_query_norm": 39.72328109741211,
      "action_relation_loss": 0.0010154340241570025,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.63797942001871,
      "flow_matching_loss": 0.01730703990906477,
      "grad_norm": 0.07494785636663437,
      "learning_rate": 9.878851815203882e-06,
      "loss": 0.0184,
      "relation_loss": 0.0011012297007255256,
      "selected_pairs": 27096.7,
      "state_query_norm": 41.955580139160155,
      "state_relation_loss": 8.579567511333153e-05,
      "step": 24200,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0011012297007255256,
      "weighted_relation_loss": 0.0011012297007255256
    },
    {
      "action_query_norm": 39.72511367797851,
      "action_relation_loss": 0.0010190976900048554,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.64733395696913,
      "flow_matching_loss": 0.017217968590557575,
      "grad_norm": 0.09083437919616699,
      "learning_rate": 9.845985611728864e-06,
      "loss": 0.0183,
      "relation_loss": 0.0011012824717909098,
      "selected_pairs": 26617.4,
      "state_query_norm": 41.95591239929199,
      "state_relation_loss": 8.218478251365013e-05,
      "step": 24210,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0011012824717909098,
      "weighted_relation_loss": 0.0011012824717909098
    },
    {
      "action_query_norm": 39.71522064208985,
      "action_relation_loss": 0.0010378000210039318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.65668849391955,
      "flow_matching_loss": 0.019383166451007128,
      "grad_norm": 0.15087361633777618,
      "learning_rate": 9.813168199167606e-06,
      "loss": 0.0205,
      "relation_loss": 0.001116683630971238,
      "selected_pairs": 26825.5,
      "state_query_norm": 41.95067825317383,
      "state_relation_loss": 7.88836223364342e-05,
      "step": 24220,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.001116683630971238,
      "weighted_relation_loss": 0.001116683630971238
    },
    {
      "action_query_norm": 39.687683486938475,
      "action_relation_loss": 0.0009996667329687625,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.66604303086997,
      "flow_matching_loss": 0.01924122404307127,
      "grad_norm": 0.09551931917667389,
      "learning_rate": 9.780399617396341e-06,
      "loss": 0.0203,
      "relation_loss": 0.0010850409104023128,
      "selected_pairs": 26780.5,
      "state_query_norm": 41.935275650024415,
      "state_relation_loss": 8.537417597835884e-05,
      "step": 24230,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0010850409104023128,
      "weighted_relation_loss": 0.0010850409104023128
    },
    {
      "action_query_norm": 39.697195053100586,
      "action_relation_loss": 0.0010307453747373074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.675397567820394,
      "flow_matching_loss": 0.017987215891480446,
      "grad_norm": 0.08793091028928757,
      "learning_rate": 9.747679906232016e-06,
      "loss": 0.0191,
      "relation_loss": 0.0011073128785938024,
      "selected_pairs": 26877.4,
      "state_query_norm": 41.94592933654785,
      "state_relation_loss": 7.656749585294165e-05,
      "step": 24240,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0011073128785938024,
      "weighted_relation_loss": 0.0011073128785938024
    },
    {
      "action_query_norm": 39.680148696899415,
      "action_relation_loss": 0.0010191265144385397,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.684752104770816,
      "flow_matching_loss": 0.01745537146925926,
      "grad_norm": 0.09697732329368591,
      "learning_rate": 9.715009105432138e-06,
      "loss": 0.0186,
      "relation_loss": 0.0010988349793478846,
      "selected_pairs": 27467.2,
      "state_query_norm": 41.94358749389649,
      "state_relation_loss": 7.970846054377035e-05,
      "step": 24250,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0010988349793478846,
      "weighted_relation_loss": 0.0010988349793478846
    },
    {
      "action_query_norm": 39.75750999450683,
      "action_relation_loss": 0.0009853840339928866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.694106641721234,
      "flow_matching_loss": 0.01806683437898755,
      "grad_norm": 0.10245823860168457,
      "learning_rate": 9.682387254694835e-06,
      "loss": 0.0191,
      "relation_loss": 0.0010652478551492095,
      "selected_pairs": 26482.2,
      "state_query_norm": 41.95750350952149,
      "state_relation_loss": 7.986381315276958e-05,
      "step": 24260,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0010652478551492095,
      "weighted_relation_loss": 0.0010652478551492095
    },
    {
      "action_query_norm": 39.69523315429687,
      "action_relation_loss": 0.000995835947105661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.703461178671656,
      "flow_matching_loss": 0.018019119463860988,
      "grad_norm": 0.1140490397810936,
      "learning_rate": 9.649814393658723e-06,
      "loss": 0.0191,
      "relation_loss": 0.0010764872131403536,
      "selected_pairs": 26526.5,
      "state_query_norm": 41.9552978515625,
      "state_relation_loss": 8.065125002758578e-05,
      "step": 24270,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010764872131403536,
      "weighted_relation_loss": 0.0010764872131403536
    },
    {
      "action_query_norm": 39.72907485961914,
      "action_relation_loss": 0.0011005215987097473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.712815715622078,
      "flow_matching_loss": 0.01886727437376976,
      "grad_norm": 0.16059955954551697,
      "learning_rate": 9.617290561902881e-06,
      "loss": 0.02,
      "relation_loss": 0.0011792404809966683,
      "selected_pairs": 27218.8,
      "state_query_norm": 41.95141868591308,
      "state_relation_loss": 7.87188699177932e-05,
      "step": 24280,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0011792404809966683,
      "weighted_relation_loss": 0.0011792404809966683
    },
    {
      "action_query_norm": 39.70104484558105,
      "action_relation_loss": 0.0010036458203103392,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.722170252572496,
      "flow_matching_loss": 0.01876626517623663,
      "grad_norm": 0.1247098371386528,
      "learning_rate": 9.584815798946861e-06,
      "loss": 0.0199,
      "relation_loss": 0.0010860584559850394,
      "selected_pairs": 26989.9,
      "state_query_norm": 41.93894538879395,
      "state_relation_loss": 8.241261748480611e-05,
      "step": 24290,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010860584559850394,
      "weighted_relation_loss": 0.0010860584559850394
    },
    {
      "action_query_norm": 39.7095890045166,
      "action_relation_loss": 0.0009917783550918101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.73152478952292,
      "flow_matching_loss": 0.015956768952310085,
      "grad_norm": 0.08947592228651047,
      "learning_rate": 9.55239014425055e-06,
      "loss": 0.017,
      "relation_loss": 0.0010731942544225602,
      "selected_pairs": 26968.6,
      "state_query_norm": 41.94835243225098,
      "state_relation_loss": 8.141590078594163e-05,
      "step": 24300,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010731942544225602,
      "weighted_relation_loss": 0.0010731942544225602
    },
    {
      "action_query_norm": 39.6979377746582,
      "action_relation_loss": 0.0010602429159916938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.74087932647334,
      "flow_matching_loss": 0.018997376970946788,
      "grad_norm": 0.11578025668859482,
      "learning_rate": 9.520013637214176e-06,
      "loss": 0.0201,
      "relation_loss": 0.0011392102809622885,
      "selected_pairs": 26895.7,
      "state_query_norm": 41.94543113708496,
      "state_relation_loss": 7.89673656981904e-05,
      "step": 24310,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0011392102809622885,
      "weighted_relation_loss": 0.0011392102809622885
    },
    {
      "action_query_norm": 39.73965225219727,
      "action_relation_loss": 0.0010077504557557404,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.75023386342376,
      "flow_matching_loss": 0.017243224009871484,
      "grad_norm": 0.08983619511127472,
      "learning_rate": 9.48768631717824e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010842978837899864,
      "selected_pairs": 27479.5,
      "state_query_norm": 41.95321044921875,
      "state_relation_loss": 7.654743021703325e-05,
      "step": 24320,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0010842978837899864,
      "weighted_relation_loss": 0.0010842978837899864
    },
    {
      "action_query_norm": 39.71240653991699,
      "action_relation_loss": 0.0009855197626166047,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.75958840037418,
      "flow_matching_loss": 0.0171981462277472,
      "grad_norm": 0.08698026090860367,
      "learning_rate": 9.455408223423496e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010637217375915498,
      "selected_pairs": 26052.3,
      "state_query_norm": 41.970937728881836,
      "state_relation_loss": 7.82019822509028e-05,
      "step": 24330,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0010637217375915498,
      "weighted_relation_loss": 0.0010637217375915498
    },
    {
      "action_query_norm": 39.71134033203125,
      "action_relation_loss": 0.0009766888513695448,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.768942937324603,
      "flow_matching_loss": 0.019049595482647418,
      "grad_norm": 0.12417446821928024,
      "learning_rate": 9.423179395170845e-06,
      "loss": 0.0201,
      "relation_loss": 0.0010592377977445722,
      "selected_pairs": 26851.1,
      "state_query_norm": 41.93567962646485,
      "state_relation_loss": 8.254895510617644e-05,
      "step": 24340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010592377977445722,
      "weighted_relation_loss": 0.0010592377977445722
    },
    {
      "action_query_norm": 39.73500137329101,
      "action_relation_loss": 0.0010572110302746295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.778297474275025,
      "flow_matching_loss": 0.01805750923231244,
      "grad_norm": 0.10815940052270889,
      "learning_rate": 9.390999871581391e-06,
      "loss": 0.0192,
      "relation_loss": 0.0011361206881701947,
      "selected_pairs": 27927.1,
      "state_query_norm": 41.94517936706543,
      "state_relation_loss": 7.89096586231608e-05,
      "step": 24350,
      "valid_action_samples": 236.8,
      "weighted_auxiliary_loss": 0.0011361206881701947,
      "weighted_relation_loss": 0.0011361206881701947
    },
    {
      "action_query_norm": 39.774850082397464,
      "action_relation_loss": 0.0010126894805580378,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.787652011225443,
      "flow_matching_loss": 0.018808032292872666,
      "grad_norm": 0.10376865416765213,
      "learning_rate": 9.358869691756273e-06,
      "loss": 0.0199,
      "relation_loss": 0.0010951204865705221,
      "selected_pairs": 26719.0,
      "state_query_norm": 41.93794975280762,
      "state_relation_loss": 8.24309914605692e-05,
      "step": 24360,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0010951204865705221,
      "weighted_relation_loss": 0.0010951204865705221
    },
    {
      "action_query_norm": 39.706781005859376,
      "action_relation_loss": 0.001011536963051185,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.797006548175865,
      "flow_matching_loss": 0.01761011630296707,
      "grad_norm": 0.09286697208881378,
      "learning_rate": 9.326788894736688e-06,
      "loss": 0.0187,
      "relation_loss": 0.001094256341457367,
      "selected_pairs": 26320.1,
      "state_query_norm": 41.93559341430664,
      "state_relation_loss": 8.271938422694802e-05,
      "step": 24370,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.001094256341457367,
      "weighted_relation_loss": 0.001094256341457367
    },
    {
      "action_query_norm": 39.69621391296387,
      "action_relation_loss": 0.0009989128564484417,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.806361085126287,
      "flow_matching_loss": 0.017279120348393916,
      "grad_norm": 0.1350928097963333,
      "learning_rate": 9.294757519503811e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010823489166796208,
      "selected_pairs": 26846.6,
      "state_query_norm": 41.91949462890625,
      "state_relation_loss": 8.343603403773158e-05,
      "step": 24380,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0010823489166796208,
      "weighted_relation_loss": 0.0010823489166796208
    },
    {
      "action_query_norm": 39.659163284301755,
      "action_relation_loss": 0.0009912179899401962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.815715622076706,
      "flow_matching_loss": 0.017164711840450764,
      "grad_norm": 0.09244880080223083,
      "learning_rate": 9.262775604978819e-06,
      "loss": 0.0182,
      "relation_loss": 0.0010735765157733113,
      "selected_pairs": 26919.4,
      "state_query_norm": 41.947994232177734,
      "state_relation_loss": 8.235852073994465e-05,
      "step": 24390,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010735765157733113,
      "weighted_relation_loss": 0.0010735765157733113
    },
    {
      "action_query_norm": 39.72959518432617,
      "action_relation_loss": 0.001032184192445129,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.825070159027128,
      "flow_matching_loss": 0.0185561116784811,
      "grad_norm": 0.09482643753290176,
      "learning_rate": 9.230843190022725e-06,
      "loss": 0.0197,
      "relation_loss": 0.0011153562925755977,
      "selected_pairs": 26739.8,
      "state_query_norm": 41.92979774475098,
      "state_relation_loss": 8.317210085806437e-05,
      "step": 24400,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011153562925755977,
      "weighted_relation_loss": 0.0011153562925755977
    },
    {
      "action_query_norm": 39.70566520690918,
      "action_relation_loss": 0.0010286496544722468,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.83442469597755,
      "flow_matching_loss": 0.01713187713176012,
      "grad_norm": 0.11341793090105057,
      "learning_rate": 9.198960313436445e-06,
      "loss": 0.0182,
      "relation_loss": 0.001108308241236955,
      "selected_pairs": 26754.1,
      "state_query_norm": 41.94126968383789,
      "state_relation_loss": 7.965860713738948e-05,
      "step": 24410,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.001108308241236955,
      "weighted_relation_loss": 0.001108308241236955
    },
    {
      "action_query_norm": 39.762742614746095,
      "action_relation_loss": 0.0009984837903175502,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.84377923292797,
      "flow_matching_loss": 0.01722568366676569,
      "grad_norm": 0.09226430207490921,
      "learning_rate": 9.16712701396067e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010785324964672327,
      "selected_pairs": 27244.9,
      "state_query_norm": 41.95594749450684,
      "state_relation_loss": 8.004869814612903e-05,
      "step": 24420,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0010785324964672327,
      "weighted_relation_loss": 0.0010785324964672327
    },
    {
      "action_query_norm": 39.70823631286621,
      "action_relation_loss": 0.0010521020856685936,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.85313376987839,
      "flow_matching_loss": 0.01601997036486864,
      "grad_norm": 0.09967223554849625,
      "learning_rate": 9.135343330275864e-06,
      "loss": 0.0172,
      "relation_loss": 0.0011315144365653396,
      "selected_pairs": 26730.6,
      "state_query_norm": 41.965650939941405,
      "state_relation_loss": 7.941235817270354e-05,
      "step": 24430,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0011315144365653396,
      "weighted_relation_loss": 0.0011315144365653396
    },
    {
      "action_query_norm": 39.704684829711915,
      "action_relation_loss": 0.000990947661921382,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.862488306828812,
      "flow_matching_loss": 0.016124062426388263,
      "grad_norm": 0.11547400802373886,
      "learning_rate": 9.10360930100218e-06,
      "loss": 0.0172,
      "relation_loss": 0.001071242440957576,
      "selected_pairs": 27199.3,
      "state_query_norm": 41.94038009643555,
      "state_relation_loss": 8.02947637566831e-05,
      "step": 24440,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001071242440957576,
      "weighted_relation_loss": 0.001071242440957576
    },
    {
      "action_query_norm": 39.77089309692383,
      "action_relation_loss": 0.0009430536709260195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.871842843779234,
      "flow_matching_loss": 0.01733042746782303,
      "grad_norm": 0.0872577577829361,
      "learning_rate": 9.07192496469949e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010282851406373083,
      "selected_pairs": 26649.6,
      "state_query_norm": 41.95259017944336,
      "state_relation_loss": 8.523147989762947e-05,
      "step": 24450,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0010282851406373083,
      "weighted_relation_loss": 0.0010282851406373083
    },
    {
      "action_query_norm": 39.73502159118652,
      "action_relation_loss": 0.0009927735023666172,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.881197380729652,
      "flow_matching_loss": 0.016181620955467223,
      "grad_norm": 0.07962623983621597,
      "learning_rate": 9.040290359867231e-06,
      "loss": 0.0173,
      "relation_loss": 0.001075489621143788,
      "selected_pairs": 26925.1,
      "state_query_norm": 41.939033889770506,
      "state_relation_loss": 8.271610568044707e-05,
      "step": 24460,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.001075489621143788,
      "weighted_relation_loss": 0.001075489621143788
    },
    {
      "action_query_norm": 39.70727767944336,
      "action_relation_loss": 0.00095019027357921,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.890551917680074,
      "flow_matching_loss": 0.018077158834785223,
      "grad_norm": 0.09521390497684479,
      "learning_rate": 9.008705524944438e-06,
      "loss": 0.0191,
      "relation_loss": 0.0010286882519721986,
      "selected_pairs": 27174.9,
      "state_query_norm": 41.931447982788086,
      "state_relation_loss": 7.849797621020115e-05,
      "step": 24470,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0010286882519721986,
      "weighted_relation_loss": 0.0010286882519721986
    },
    {
      "action_query_norm": 39.66926574707031,
      "action_relation_loss": 0.0010060587723273785,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.899906454630496,
      "flow_matching_loss": 0.017333020363003014,
      "grad_norm": 0.1163005381822586,
      "learning_rate": 8.977170498309651e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010905141709372402,
      "selected_pairs": 26758.1,
      "state_query_norm": 41.93853569030762,
      "state_relation_loss": 8.44553804199677e-05,
      "step": 24480,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010905141709372402,
      "weighted_relation_loss": 0.0010905141709372402
    },
    {
      "action_query_norm": 39.67102699279785,
      "action_relation_loss": 0.0009997644112445414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.90926099158092,
      "flow_matching_loss": 0.017406282108277084,
      "grad_norm": 0.1380196064710617,
      "learning_rate": 8.945685318280916e-06,
      "loss": 0.0185,
      "relation_loss": 0.0010749061941169203,
      "selected_pairs": 26942.2,
      "state_query_norm": 41.9366870880127,
      "state_relation_loss": 7.514176395488903e-05,
      "step": 24490,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0010749061941169203,
      "weighted_relation_loss": 0.0010749061941169203
    },
    {
      "action_query_norm": 39.718641662597655,
      "action_relation_loss": 0.001048121094936505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.918615528531337,
      "flow_matching_loss": 0.017844355199486017,
      "grad_norm": 0.10390353202819824,
      "learning_rate": 8.914250023115672e-06,
      "loss": 0.019,
      "relation_loss": 0.0011307912587653845,
      "selected_pairs": 27293.2,
      "state_query_norm": 41.93806457519531,
      "state_relation_loss": 8.267013909062371e-05,
      "step": 24500,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0011307912587653845,
      "weighted_relation_loss": 0.0011307912587653845
    },
    {
      "action_query_norm": 39.73142852783203,
      "action_relation_loss": 0.0009809000184759498,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.92797006548176,
      "flow_matching_loss": 0.015972321294248103,
      "grad_norm": 0.09553246945142746,
      "learning_rate": 8.882864651010798e-06,
      "loss": 0.017,
      "relation_loss": 0.0010610407043714076,
      "selected_pairs": 26600.5,
      "state_query_norm": 41.95263748168945,
      "state_relation_loss": 8.01407040853519e-05,
      "step": 24510,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0010610407043714076,
      "weighted_relation_loss": 0.0010610407043714076
    },
    {
      "action_query_norm": 39.71081466674805,
      "action_relation_loss": 0.0009802562126424164,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.93732460243218,
      "flow_matching_loss": 0.01731667583808303,
      "grad_norm": 0.10892724245786667,
      "learning_rate": 8.851529240102462e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010611039760988206,
      "selected_pairs": 26551.2,
      "state_query_norm": 41.92352523803711,
      "state_relation_loss": 8.08477561804466e-05,
      "step": 24520,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010611039760988206,
      "weighted_relation_loss": 0.0010611039760988206
    },
    {
      "action_query_norm": 39.70070915222168,
      "action_relation_loss": 0.0009966986661311239,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.9466791393826,
      "flow_matching_loss": 0.018142865598201753,
      "grad_norm": 0.13237303495407104,
      "learning_rate": 8.820243828466134e-06,
      "loss": 0.0192,
      "relation_loss": 0.0010823507909663022,
      "selected_pairs": 26626.4,
      "state_query_norm": 41.937832260131835,
      "state_relation_loss": 8.565212701796553e-05,
      "step": 24530,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010823507909663022,
      "weighted_relation_loss": 0.0010823507909663022
    },
    {
      "action_query_norm": 39.751119995117186,
      "action_relation_loss": 0.0010568487516138703,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.95603367633302,
      "flow_matching_loss": 0.017095286399126053,
      "grad_norm": 0.11023540794849396,
      "learning_rate": 8.789008454116566e-06,
      "loss": 0.0182,
      "relation_loss": 0.001136605767533183,
      "selected_pairs": 27127.0,
      "state_query_norm": 41.951272583007814,
      "state_relation_loss": 7.975702610565349e-05,
      "step": 24540,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.001136605767533183,
      "weighted_relation_loss": 0.001136605767533183
    },
    {
      "action_query_norm": 39.70788803100586,
      "action_relation_loss": 0.0010254273889586329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.965388213283443,
      "flow_matching_loss": 0.01741157453507185,
      "grad_norm": 0.11519578844308853,
      "learning_rate": 8.757823155007655e-06,
      "loss": 0.0185,
      "relation_loss": 0.0011035782052204012,
      "selected_pairs": 26574.6,
      "state_query_norm": 41.93530960083008,
      "state_relation_loss": 7.815081989974715e-05,
      "step": 24550,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0011035782052204012,
      "weighted_relation_loss": 0.0011035782052204012
    },
    {
      "action_query_norm": 39.71805076599121,
      "action_relation_loss": 0.0009926935250405221,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.974742750233865,
      "flow_matching_loss": 0.015400297194719314,
      "grad_norm": 0.10165591537952423,
      "learning_rate": 8.72668796903251e-06,
      "loss": 0.0165,
      "relation_loss": 0.0010791856679134072,
      "selected_pairs": 26961.0,
      "state_query_norm": 41.908447647094725,
      "state_relation_loss": 8.649212832096964e-05,
      "step": 24560,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010791856679134072,
      "weighted_relation_loss": 0.0010791856679134072
    },
    {
      "action_query_norm": 39.69211921691895,
      "action_relation_loss": 0.0010068000643514096,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.984097287184284,
      "flow_matching_loss": 0.017519069276750086,
      "grad_norm": 0.09148778021335602,
      "learning_rate": 8.695602934023301e-06,
      "loss": 0.0186,
      "relation_loss": 0.0010905884439125658,
      "selected_pairs": 26544.5,
      "state_query_norm": 41.93736877441406,
      "state_relation_loss": 8.378838319913484e-05,
      "step": 24570,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010905884439125658,
      "weighted_relation_loss": 0.0010905884439125658
    },
    {
      "action_query_norm": 39.71873321533203,
      "action_relation_loss": 0.000986471399664879,
      "auxiliary_loss_weight": 1.0,
      "epoch": 22.993451824134706,
      "flow_matching_loss": 0.015748247783631087,
      "grad_norm": 0.1319027841091156,
      "learning_rate": 8.664568087751274e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010658165614586324,
      "selected_pairs": 26387.5,
      "state_query_norm": 41.96380958557129,
      "state_relation_loss": 7.934515379020013e-05,
      "step": 24580,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0010658165614586324,
      "weighted_relation_loss": 0.0010658165614586324
    },
    {
      "action_query_norm": 39.710527420043945,
      "action_relation_loss": 0.000966254179365933,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.002806361085128,
      "flow_matching_loss": 0.01725577563047409,
      "grad_norm": 0.09237148612737656,
      "learning_rate": 8.633583467926697e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010438647062983365,
      "selected_pairs": 24489.3,
      "state_query_norm": 41.93260154724121,
      "state_relation_loss": 7.761052838759497e-05,
      "step": 24590,
      "valid_action_samples": 214.7,
      "weighted_auxiliary_loss": 0.0010438647062983365,
      "weighted_relation_loss": 0.0010438647062983365
    },
    {
      "action_query_norm": 39.690036392211915,
      "action_relation_loss": 0.0009893342212308198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.012160898035546,
      "flow_matching_loss": 0.01765823271125555,
      "grad_norm": 0.09176318347454071,
      "learning_rate": 8.602649112198796e-06,
      "loss": 0.0187,
      "relation_loss": 0.0010657735227141529,
      "selected_pairs": 26872.9,
      "state_query_norm": 41.93732604980469,
      "state_relation_loss": 7.643929347977974e-05,
      "step": 24600,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010657735227141529,
      "weighted_relation_loss": 0.0010657735227141529
    },
    {
      "action_query_norm": 39.702265930175784,
      "action_relation_loss": 0.0009911983041092753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.021515434985968,
      "flow_matching_loss": 0.018166171759366988,
      "grad_norm": 0.16077731549739838,
      "learning_rate": 8.571765058155745e-06,
      "loss": 0.0192,
      "relation_loss": 0.0010725174215622246,
      "selected_pairs": 26931.1,
      "state_query_norm": 41.94622917175293,
      "state_relation_loss": 8.131911381497048e-05,
      "step": 24610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010725174215622246,
      "weighted_relation_loss": 0.0010725174215622246
    },
    {
      "action_query_norm": 39.63570137023926,
      "action_relation_loss": 0.0010037833766546101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.03086997193639,
      "flow_matching_loss": 0.017129746451973914,
      "grad_norm": 0.1366763859987259,
      "learning_rate": 8.540931343324583e-06,
      "loss": 0.0182,
      "relation_loss": 0.0010802701057400555,
      "selected_pairs": 26990.8,
      "state_query_norm": 41.92492828369141,
      "state_relation_loss": 7.648674509255215e-05,
      "step": 24620,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010802701057400555,
      "weighted_relation_loss": 0.0010802701057400555
    },
    {
      "action_query_norm": 39.72854423522949,
      "action_relation_loss": 0.0009599959012120962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.04022450888681,
      "flow_matching_loss": 0.016927235294133425,
      "grad_norm": 0.08730122447013855,
      "learning_rate": 8.51014800517117e-06,
      "loss": 0.018,
      "relation_loss": 0.0010395200457423925,
      "selected_pairs": 26803.9,
      "state_query_norm": 41.927659606933595,
      "state_relation_loss": 7.952416635816916e-05,
      "step": 24630,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010395200457423925,
      "weighted_relation_loss": 0.0010395200457423925
    },
    {
      "action_query_norm": 39.7442684173584,
      "action_relation_loss": 0.0009671844891272485,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.04957904583723,
      "flow_matching_loss": 0.016076173260807992,
      "grad_norm": 0.12123698741197586,
      "learning_rate": 8.479415081100167e-06,
      "loss": 0.0171,
      "relation_loss": 0.0010458292497787624,
      "selected_pairs": 27299.5,
      "state_query_norm": 41.93623657226563,
      "state_relation_loss": 7.864477520342916e-05,
      "step": 24640,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0010458292497787624,
      "weighted_relation_loss": 0.0010458292497787624
    },
    {
      "action_query_norm": 39.73409576416016,
      "action_relation_loss": 0.001003173051867634,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.058933582787652,
      "flow_matching_loss": 0.01763293147087097,
      "grad_norm": 0.11094264686107635,
      "learning_rate": 8.448732608454967e-06,
      "loss": 0.0187,
      "relation_loss": 0.0010829730425029993,
      "selected_pairs": 27127.9,
      "state_query_norm": 41.93856201171875,
      "state_relation_loss": 7.980000227689744e-05,
      "step": 24650,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010829730425029993,
      "weighted_relation_loss": 0.0010829730425029993
    },
    {
      "action_query_norm": 39.6459457397461,
      "action_relation_loss": 0.0010327669675461947,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.068288119738074,
      "flow_matching_loss": 0.017101946473121642,
      "grad_norm": 0.15711982548236847,
      "learning_rate": 8.418100624517688e-06,
      "loss": 0.0182,
      "relation_loss": 0.0011136131943203509,
      "selected_pairs": 26635.6,
      "state_query_norm": 41.92236747741699,
      "state_relation_loss": 8.084620931185782e-05,
      "step": 24660,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0011136131943203509,
      "weighted_relation_loss": 0.0011136131943203509
    },
    {
      "action_query_norm": 39.701637268066406,
      "action_relation_loss": 0.0010051131772343068,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.077642656688493,
      "flow_matching_loss": 0.017337315250188112,
      "grad_norm": 0.10038381814956665,
      "learning_rate": 8.387519166509062e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010826561891008168,
      "selected_pairs": 26813.1,
      "state_query_norm": 41.92668342590332,
      "state_relation_loss": 7.754301259410568e-05,
      "step": 24670,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010826561891008168,
      "weighted_relation_loss": 0.0010826561891008168
    },
    {
      "action_query_norm": 39.73464508056641,
      "action_relation_loss": 0.0010014572471845895,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.086997193638915,
      "flow_matching_loss": 0.016381190344691276,
      "grad_norm": 0.14208678901195526,
      "learning_rate": 8.356988271588445e-06,
      "loss": 0.0175,
      "relation_loss": 0.0010809243307448924,
      "selected_pairs": 27224.4,
      "state_query_norm": 41.9147762298584,
      "state_relation_loss": 7.946708938106894e-05,
      "step": 24680,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0010809243307448924,
      "weighted_relation_loss": 0.0010809243307448924
    },
    {
      "action_query_norm": 39.73642997741699,
      "action_relation_loss": 0.0008849082922097295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.096351730589337,
      "flow_matching_loss": 0.017616307828575373,
      "grad_norm": 0.10603190958499908,
      "learning_rate": 8.326507976853765e-06,
      "loss": 0.0186,
      "relation_loss": 0.0009661600284744054,
      "selected_pairs": 26491.7,
      "state_query_norm": 41.929212951660155,
      "state_relation_loss": 8.125173117150553e-05,
      "step": 24690,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0009661600284744054,
      "weighted_relation_loss": 0.0009661600284744054
    },
    {
      "action_query_norm": 39.70630035400391,
      "action_relation_loss": 0.0010025993222370743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.105706267539755,
      "flow_matching_loss": 0.016923460736870767,
      "grad_norm": 0.1244068369269371,
      "learning_rate": 8.296078319341443e-06,
      "loss": 0.018,
      "relation_loss": 0.001081872044596821,
      "selected_pairs": 27105.7,
      "state_query_norm": 41.921540069580075,
      "state_relation_loss": 7.927272163215094e-05,
      "step": 24700,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001081872044596821,
      "weighted_relation_loss": 0.001081872044596821
    },
    {
      "action_query_norm": 39.73672676086426,
      "action_relation_loss": 0.0009426024975255132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.115060804490177,
      "flow_matching_loss": 0.016261069942265748,
      "grad_norm": 0.07589580863714218,
      "learning_rate": 8.265699336026383e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010237858688924462,
      "selected_pairs": 26964.4,
      "state_query_norm": 41.94712371826172,
      "state_relation_loss": 8.118336845654994e-05,
      "step": 24710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010237858688924462,
      "weighted_relation_loss": 0.0010237858688924462
    },
    {
      "action_query_norm": 39.69388427734375,
      "action_relation_loss": 0.0010111771465744824,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.1244153414406,
      "flow_matching_loss": 0.017066133860498666,
      "grad_norm": 0.11133739352226257,
      "learning_rate": 8.235371063821923e-06,
      "loss": 0.0182,
      "relation_loss": 0.0010888847580645233,
      "selected_pairs": 26925.7,
      "state_query_norm": 41.92860298156738,
      "state_relation_loss": 7.770760348648764e-05,
      "step": 24720,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010888847580645233,
      "weighted_relation_loss": 0.0010888847580645233
    },
    {
      "action_query_norm": 39.697608184814456,
      "action_relation_loss": 0.0009572078299243004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.13376987839102,
      "flow_matching_loss": 0.016697518154978754,
      "grad_norm": 0.10079853981733322,
      "learning_rate": 8.205093539579767e-06,
      "loss": 0.0177,
      "relation_loss": 0.001032652601134032,
      "selected_pairs": 27083.4,
      "state_query_norm": 41.93456497192383,
      "state_relation_loss": 7.544476247858256e-05,
      "step": 24730,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.001032652601134032,
      "weighted_relation_loss": 0.001032652601134032
    },
    {
      "action_query_norm": 39.71998672485351,
      "action_relation_loss": 0.0009687514975667,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.14312441534144,
      "flow_matching_loss": 0.017997209168970584,
      "grad_norm": 0.0948769673705101,
      "learning_rate": 8.174866800089963e-06,
      "loss": 0.019,
      "relation_loss": 0.0010470235312823207,
      "selected_pairs": 27128.5,
      "state_query_norm": 41.922299194335935,
      "state_relation_loss": 7.827203953638672e-05,
      "step": 24740,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010470235312823207,
      "weighted_relation_loss": 0.0010470235312823207
    },
    {
      "action_query_norm": 39.724077987670896,
      "action_relation_loss": 0.0009476900158915669,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.15247895229186,
      "flow_matching_loss": 0.016853957250714303,
      "grad_norm": 0.08649444580078125,
      "learning_rate": 8.144690882080853e-06,
      "loss": 0.0179,
      "relation_loss": 0.001023108692606911,
      "selected_pairs": 26945.2,
      "state_query_norm": 41.938083267211915,
      "state_relation_loss": 7.541869199485518e-05,
      "step": 24750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001023108692606911,
      "weighted_relation_loss": 0.001023108692606911
    },
    {
      "action_query_norm": 39.67191123962402,
      "action_relation_loss": 0.0009368385595735163,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.161833489242284,
      "flow_matching_loss": 0.017351495567709208,
      "grad_norm": 0.0812646746635437,
      "learning_rate": 8.114565822219006e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010181825200561434,
      "selected_pairs": 26463.0,
      "state_query_norm": 41.934548950195314,
      "state_relation_loss": 8.134396921377629e-05,
      "step": 24760,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010181825200561434,
      "weighted_relation_loss": 0.0010181825200561434
    },
    {
      "action_query_norm": 39.697713470458986,
      "action_relation_loss": 0.000949105602921918,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.171188026192702,
      "flow_matching_loss": 0.016448028292506932,
      "grad_norm": 0.11522048711776733,
      "learning_rate": 8.084491657109233e-06,
      "loss": 0.0175,
      "relation_loss": 0.0010295894520822913,
      "selected_pairs": 26872.1,
      "state_query_norm": 41.92080917358398,
      "state_relation_loss": 8.048384697758592e-05,
      "step": 24770,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010295894520822913,
      "weighted_relation_loss": 0.0010295894520822913
    },
    {
      "action_query_norm": 39.67138595581055,
      "action_relation_loss": 0.0009619735297746957,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.180542563143124,
      "flow_matching_loss": 0.016403211373835803,
      "grad_norm": 0.12468627095222473,
      "learning_rate": 8.054468423294476e-06,
      "loss": 0.0174,
      "relation_loss": 0.0010396109602879733,
      "selected_pairs": 26274.3,
      "state_query_norm": 41.93196754455566,
      "state_relation_loss": 7.763744724798016e-05,
      "step": 24780,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0010396109602879733,
      "weighted_relation_loss": 0.0010396109602879733
    },
    {
      "action_query_norm": 39.68479499816895,
      "action_relation_loss": 0.0009408125188201666,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.189897100093546,
      "flow_matching_loss": 0.015059210453182459,
      "grad_norm": 0.09111512452363968,
      "learning_rate": 8.024496157255785e-06,
      "loss": 0.0161,
      "relation_loss": 0.001016970816999674,
      "selected_pairs": 26669.2,
      "state_query_norm": 41.94255256652832,
      "state_relation_loss": 7.615829017595387e-05,
      "step": 24790,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.001016970816999674,
      "weighted_relation_loss": 0.001016970816999674
    },
    {
      "action_query_norm": 39.68480529785156,
      "action_relation_loss": 0.0009489646006841212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.199251637043965,
      "flow_matching_loss": 0.01769034480676055,
      "grad_norm": 0.10627492517232895,
      "learning_rate": 7.994574895412293e-06,
      "loss": 0.0187,
      "relation_loss": 0.0010284520220011472,
      "selected_pairs": 26687.5,
      "state_query_norm": 41.933118438720705,
      "state_relation_loss": 7.94874329585582e-05,
      "step": 24800,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010284520220011472,
      "weighted_relation_loss": 0.0010284520220011472
    },
    {
      "action_query_norm": 39.69638023376465,
      "action_relation_loss": 0.0009465354203712195,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.208606173994387,
      "flow_matching_loss": 0.01765797883272171,
      "grad_norm": 0.10219278186559677,
      "learning_rate": 7.964704674121149e-06,
      "loss": 0.0187,
      "relation_loss": 0.0010219139629043638,
      "selected_pairs": 26806.3,
      "state_query_norm": 41.93371658325195,
      "state_relation_loss": 7.537855126429349e-05,
      "step": 24810,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010219139629043638,
      "weighted_relation_loss": 0.0010219139629043638
    },
    {
      "action_query_norm": 39.71791381835938,
      "action_relation_loss": 0.0009512536460533738,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.21796071094481,
      "flow_matching_loss": 0.015932276379317045,
      "grad_norm": 0.08831585943698883,
      "learning_rate": 7.934885529677505e-06,
      "loss": 0.017,
      "relation_loss": 0.0010254282969981432,
      "selected_pairs": 26762.7,
      "state_query_norm": 41.909518814086915,
      "state_relation_loss": 7.417464439640753e-05,
      "step": 24820,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010254282969981432,
      "weighted_relation_loss": 0.0010254282969981432
    },
    {
      "action_query_norm": 39.72546653747558,
      "action_relation_loss": 0.0009616446739528328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.22731524789523,
      "flow_matching_loss": 0.018781695514917374,
      "grad_norm": 0.08834998309612274,
      "learning_rate": 7.905117498314412e-06,
      "loss": 0.0198,
      "relation_loss": 0.0010416788281872869,
      "selected_pairs": 26683.5,
      "state_query_norm": 41.90184898376465,
      "state_relation_loss": 8.00341455033049e-05,
      "step": 24830,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010416788281872869,
      "weighted_relation_loss": 0.0010416788281872869
    },
    {
      "action_query_norm": 39.70991477966309,
      "action_relation_loss": 0.0009362291835714132,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.23666978484565,
      "flow_matching_loss": 0.016752371564507484,
      "grad_norm": 0.10614525526762009,
      "learning_rate": 7.87540061620286e-06,
      "loss": 0.0178,
      "relation_loss": 0.0010143477702513337,
      "selected_pairs": 27226.3,
      "state_query_norm": 41.93370094299316,
      "state_relation_loss": 7.811858449713327e-05,
      "step": 24840,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0010143477702513337,
      "weighted_relation_loss": 0.0010143477702513337
    },
    {
      "action_query_norm": 39.729590606689456,
      "action_relation_loss": 0.0009767731768079101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.24602432179607,
      "flow_matching_loss": 0.017320015374571084,
      "grad_norm": 0.08342788368463516,
      "learning_rate": 7.845734919451647e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010591203870717436,
      "selected_pairs": 26690.9,
      "state_query_norm": 41.92531318664551,
      "state_relation_loss": 8.234721462940797e-05,
      "step": 24850,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010591203870717436,
      "weighted_relation_loss": 0.0010591203870717436
    },
    {
      "action_query_norm": 39.693611526489256,
      "action_relation_loss": 0.001021070993738249,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.255378858746493,
      "flow_matching_loss": 0.015895969606935977,
      "grad_norm": 0.1209993064403534,
      "learning_rate": 7.816120444107383e-06,
      "loss": 0.017,
      "relation_loss": 0.0010988467605784536,
      "selected_pairs": 26548.6,
      "state_query_norm": 41.94213218688965,
      "state_relation_loss": 7.777575447107666e-05,
      "step": 24860,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010988467605784536,
      "weighted_relation_loss": 0.0010988467605784536
    },
    {
      "action_query_norm": 39.703210067749026,
      "action_relation_loss": 0.0009660669893492013,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.26473339569691,
      "flow_matching_loss": 0.017311422061175107,
      "grad_norm": 0.11233805865049362,
      "learning_rate": 7.786557226154473e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010422360035590827,
      "selected_pairs": 26778.5,
      "state_query_norm": 41.92249031066895,
      "state_relation_loss": 7.616901566507295e-05,
      "step": 24870,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0010422360035590827,
      "weighted_relation_loss": 0.0010422360035590827
    },
    {
      "action_query_norm": 39.68880157470703,
      "action_relation_loss": 0.000984493235591799,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.274087932647333,
      "flow_matching_loss": 0.018588898051530123,
      "grad_norm": 0.1084299087524414,
      "learning_rate": 7.757045301514998e-06,
      "loss": 0.0196,
      "relation_loss": 0.0010585436131805182,
      "selected_pairs": 26249.9,
      "state_query_norm": 41.92413215637207,
      "state_relation_loss": 7.405037977150641e-05,
      "step": 24880,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0010585436131805182,
      "weighted_relation_loss": 0.0010585436131805182
    },
    {
      "action_query_norm": 39.67740440368652,
      "action_relation_loss": 0.0009945188357960433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.283442469597755,
      "flow_matching_loss": 0.017262246552854778,
      "grad_norm": 0.19024431705474854,
      "learning_rate": 7.727584706048735e-06,
      "loss": 0.0183,
      "relation_loss": 0.001076411409303546,
      "selected_pairs": 26849.3,
      "state_query_norm": 41.922706604003906,
      "state_relation_loss": 8.189258296624758e-05,
      "step": 24890,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001076411409303546,
      "weighted_relation_loss": 0.001076411409303546
    },
    {
      "action_query_norm": 39.67488441467285,
      "action_relation_loss": 0.0009432903083506971,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.292797006548177,
      "flow_matching_loss": 0.01766716446727514,
      "grad_norm": 0.09560330957174301,
      "learning_rate": 7.698175475553077e-06,
      "loss": 0.0187,
      "relation_loss": 0.0010232649801764637,
      "selected_pairs": 26786.7,
      "state_query_norm": 41.90690155029297,
      "state_relation_loss": 7.997467910172417e-05,
      "step": 24900,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0010232649801764637,
      "weighted_relation_loss": 0.0010232649801764637
    },
    {
      "action_query_norm": 39.72893142700195,
      "action_relation_loss": 0.0009520641295239329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.302151543498596,
      "flow_matching_loss": 0.01827007457613945,
      "grad_norm": 0.11317537724971771,
      "learning_rate": 7.668817645763021e-06,
      "loss": 0.0193,
      "relation_loss": 0.0010308368247933686,
      "selected_pairs": 27193.4,
      "state_query_norm": 41.904563903808594,
      "state_relation_loss": 7.877270036260597e-05,
      "step": 24910,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0010308368247933686,
      "weighted_relation_loss": 0.0010308368247933686
    },
    {
      "action_query_norm": 39.692185974121095,
      "action_relation_loss": 0.0009469856566283851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.311506080449018,
      "flow_matching_loss": 0.01817869786173105,
      "grad_norm": 0.10203620791435242,
      "learning_rate": 7.639511252351088e-06,
      "loss": 0.0192,
      "relation_loss": 0.0010293672326952218,
      "selected_pairs": 26798.3,
      "state_query_norm": 41.91692008972168,
      "state_relation_loss": 8.238157388404944e-05,
      "step": 24920,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010293672326952218,
      "weighted_relation_loss": 0.0010293672326952218
    },
    {
      "action_query_norm": 39.689534759521486,
      "action_relation_loss": 0.0010222411772701889,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.32086061739944,
      "flow_matching_loss": 0.01783743128180504,
      "grad_norm": 0.10524125397205353,
      "learning_rate": 7.610256330927324e-06,
      "loss": 0.0189,
      "relation_loss": 0.0011049045308027417,
      "selected_pairs": 26691.1,
      "state_query_norm": 41.91400566101074,
      "state_relation_loss": 8.266334334621206e-05,
      "step": 24930,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0011049045308027417,
      "weighted_relation_loss": 0.0011049045308027417
    },
    {
      "action_query_norm": 39.629490280151366,
      "action_relation_loss": 0.0010259652219247073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.330215154349858,
      "flow_matching_loss": 0.015851793345063925,
      "grad_norm": 0.0875633955001831,
      "learning_rate": 7.581052917039211e-06,
      "loss": 0.017,
      "relation_loss": 0.0011060639109928161,
      "selected_pairs": 26368.4,
      "state_query_norm": 41.91307144165039,
      "state_relation_loss": 8.009870289242827e-05,
      "step": 24940,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0011060639109928161,
      "weighted_relation_loss": 0.0011060639109928161
    },
    {
      "action_query_norm": 39.7085521697998,
      "action_relation_loss": 0.00094819514779374,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.33956969130028,
      "flow_matching_loss": 0.017279003094881774,
      "grad_norm": 0.10766055434942245,
      "learning_rate": 7.551901046171645e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010278499743435533,
      "selected_pairs": 26924.3,
      "state_query_norm": 41.91467056274414,
      "state_relation_loss": 7.965483600855806e-05,
      "step": 24950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010278499743435533,
      "weighted_relation_loss": 0.0010278499743435533
    },
    {
      "action_query_norm": 39.670824432373045,
      "action_relation_loss": 0.000951485411496833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.348924228250702,
      "flow_matching_loss": 0.01598517261445522,
      "grad_norm": 0.09711792320013046,
      "learning_rate": 7.522800753746895e-06,
      "loss": 0.017,
      "relation_loss": 0.0010311754071153701,
      "selected_pairs": 26442.2,
      "state_query_norm": 41.89550132751465,
      "state_relation_loss": 7.96899912529625e-05,
      "step": 24960,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0010311754071153701,
      "weighted_relation_loss": 0.0010311754071153701
    },
    {
      "action_query_norm": 39.69314956665039,
      "action_relation_loss": 0.0009747891570441424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.358278765201124,
      "flow_matching_loss": 0.01722134593874216,
      "grad_norm": 0.10009127110242844,
      "learning_rate": 7.493752075124577e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010552335414104163,
      "selected_pairs": 26739.0,
      "state_query_norm": 41.92089195251465,
      "state_relation_loss": 8.044438727665693e-05,
      "step": 24970,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0010552335414104163,
      "weighted_relation_loss": 0.0010552335414104163
    },
    {
      "action_query_norm": 39.667429733276364,
      "action_relation_loss": 0.0009787322662305087,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.367633302151543,
      "flow_matching_loss": 0.016831851843744517,
      "grad_norm": 0.11570078879594803,
      "learning_rate": 7.464755045601557e-06,
      "loss": 0.0179,
      "relation_loss": 0.001053417578805238,
      "selected_pairs": 26484.2,
      "state_query_norm": 41.915409088134766,
      "state_relation_loss": 7.468530675396323e-05,
      "step": 24980,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.001053417578805238,
      "weighted_relation_loss": 0.001053417578805238
    },
    {
      "action_query_norm": 39.72080192565918,
      "action_relation_loss": 0.0009093107306398452,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.376987839101965,
      "flow_matching_loss": 0.015788505785167217,
      "grad_norm": 0.08481389284133911,
      "learning_rate": 7.435809700411972e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009886578191071749,
      "selected_pairs": 26875.9,
      "state_query_norm": 41.90184364318848,
      "state_relation_loss": 7.934708482935094e-05,
      "step": 24990,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009886578191071749,
      "weighted_relation_loss": 0.0009886578191071749
    },
    {
      "action_query_norm": 39.705529403686526,
      "action_relation_loss": 0.0009458643035031855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.386342376052387,
      "flow_matching_loss": 0.016796227544546127,
      "grad_norm": 0.1225740984082222,
      "learning_rate": 7.40691607472715e-06,
      "loss": 0.0178,
      "relation_loss": 0.0010240846953820438,
      "selected_pairs": 27314.7,
      "state_query_norm": 41.92297248840332,
      "state_relation_loss": 7.822039260645397e-05,
      "step": 25000,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0010240846953820438,
      "weighted_relation_loss": 0.0010240846953820438
    },
    {
      "action_query_norm": 39.641763305664064,
      "action_relation_loss": 0.0010107495065312833,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.395696913002805,
      "flow_matching_loss": 0.01590848099440336,
      "grad_norm": 0.08963741362094879,
      "learning_rate": 7.3780742036555606e-06,
      "loss": 0.017,
      "relation_loss": 0.0010873828490730376,
      "selected_pairs": 27018.6,
      "state_query_norm": 41.909746170043945,
      "state_relation_loss": 7.663335200049914e-05,
      "step": 25010,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010873828490730376,
      "weighted_relation_loss": 0.0010873828490730376
    },
    {
      "action_query_norm": 39.702402877807614,
      "action_relation_loss": 0.0009778115549124778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.405051449953227,
      "flow_matching_loss": 0.017374583147466184,
      "grad_norm": 0.14821450412273407,
      "learning_rate": 7.349284122242783e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010546421515755355,
      "selected_pairs": 27034.2,
      "state_query_norm": 41.905339050292966,
      "state_relation_loss": 7.683060321141965e-05,
      "step": 25020,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010546421515755355,
      "weighted_relation_loss": 0.0010546421515755355
    },
    {
      "action_query_norm": 39.68633766174317,
      "action_relation_loss": 0.0009522502252366394,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.41440598690365,
      "flow_matching_loss": 0.01659406991675496,
      "grad_norm": 0.21200643479824066,
      "learning_rate": 7.320545865471512e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010313571547158062,
      "selected_pairs": 26298.4,
      "state_query_norm": 41.90762710571289,
      "state_relation_loss": 7.910693529993296e-05,
      "step": 25030,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0010313571547158062,
      "weighted_relation_loss": 0.0010313571547158062
    },
    {
      "action_query_norm": 39.684807205200194,
      "action_relation_loss": 0.0009264036954846234,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.42376052385407,
      "flow_matching_loss": 0.01664597475901246,
      "grad_norm": 0.09008621424436569,
      "learning_rate": 7.291859468261425e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010009389836341143,
      "selected_pairs": 27138.2,
      "state_query_norm": 41.909813690185544,
      "state_relation_loss": 7.453528814949095e-05,
      "step": 25040,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0010009389836341143,
      "weighted_relation_loss": 0.0010009389836341143
    },
    {
      "action_query_norm": 39.721140670776364,
      "action_relation_loss": 0.0009498676168732345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.43311506080449,
      "flow_matching_loss": 0.01574443355202675,
      "grad_norm": 0.1535022258758545,
      "learning_rate": 7.263224965469195e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010273268562741578,
      "selected_pairs": 26711.3,
      "state_query_norm": 41.914140701293945,
      "state_relation_loss": 7.745921757305041e-05,
      "step": 25050,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0010273268562741578,
      "weighted_relation_loss": 0.0010273268562741578
    },
    {
      "action_query_norm": 39.70622177124024,
      "action_relation_loss": 0.0009448627708479763,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.44246959775491,
      "flow_matching_loss": 0.017238289583474398,
      "grad_norm": 0.19993922114372253,
      "learning_rate": 7.234642391888446e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010131613584235311,
      "selected_pairs": 26623.0,
      "state_query_norm": 41.915901565551756,
      "state_relation_loss": 6.829859194112942e-05,
      "step": 25060,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010131613584235311,
      "weighted_relation_loss": 0.0010131613584235311
    },
    {
      "action_query_norm": 39.69436416625977,
      "action_relation_loss": 0.0009971973253414036,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.451824134705333,
      "flow_matching_loss": 0.01731413323432207,
      "grad_norm": 0.10589569807052612,
      "learning_rate": 7.206111782249697e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010732949944213034,
      "selected_pairs": 26804.4,
      "state_query_norm": 41.917945861816406,
      "state_relation_loss": 7.609766762470827e-05,
      "step": 25070,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010732949944213034,
      "weighted_relation_loss": 0.0010732949944213034
    },
    {
      "action_query_norm": 39.700993347167966,
      "action_relation_loss": 0.000997214362723753,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.46117867165575,
      "flow_matching_loss": 0.01712658293545246,
      "grad_norm": 0.16889390349388123,
      "learning_rate": 7.1776331712203385e-06,
      "loss": 0.0182,
      "relation_loss": 0.0010728011024184526,
      "selected_pairs": 26987.8,
      "state_query_norm": 41.929435348510744,
      "state_relation_loss": 7.558675424661488e-05,
      "step": 25080,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010728011024184526,
      "weighted_relation_loss": 0.0010728011024184526
    },
    {
      "action_query_norm": 39.69880828857422,
      "action_relation_loss": 0.000920715375104919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.470533208606174,
      "flow_matching_loss": 0.01669216053560376,
      "grad_norm": 0.09204605221748352,
      "learning_rate": 7.149206593404561e-06,
      "loss": 0.0177,
      "relation_loss": 0.0009963169868569821,
      "selected_pairs": 26871.2,
      "state_query_norm": 41.93029708862305,
      "state_relation_loss": 7.560161466244609e-05,
      "step": 25090,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009963169868569821,
      "weighted_relation_loss": 0.0009963169868569821
    },
    {
      "action_query_norm": 39.655609893798825,
      "action_relation_loss": 0.0009429801080841571,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.479887745556596,
      "flow_matching_loss": 0.016392614133656024,
      "grad_norm": 0.16142524778842926,
      "learning_rate": 7.120832083343337e-06,
      "loss": 0.0174,
      "relation_loss": 0.0010202229663264007,
      "selected_pairs": 26639.2,
      "state_query_norm": 41.91943016052246,
      "state_relation_loss": 7.72428662457969e-05,
      "step": 25100,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0010202229663264007,
      "weighted_relation_loss": 0.0010202229663264007
    },
    {
      "action_query_norm": 39.707664489746094,
      "action_relation_loss": 0.0009796273370739073,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.489242282507014,
      "flow_matching_loss": 0.014822978992015123,
      "grad_norm": 0.0843890979886055,
      "learning_rate": 7.092509675514369e-06,
      "loss": 0.0159,
      "relation_loss": 0.001057749311439693,
      "selected_pairs": 27055.6,
      "state_query_norm": 41.92751617431641,
      "state_relation_loss": 7.81219823693391e-05,
      "step": 25110,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001057749311439693,
      "weighted_relation_loss": 0.001057749311439693
    },
    {
      "action_query_norm": 39.71786384582519,
      "action_relation_loss": 0.0009836978744715453,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.498596819457436,
      "flow_matching_loss": 0.016203485149890184,
      "grad_norm": 0.07878898829221725,
      "learning_rate": 7.064239404332063e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010639135260134935,
      "selected_pairs": 27012.7,
      "state_query_norm": 41.89441757202148,
      "state_relation_loss": 8.021566027309746e-05,
      "step": 25120,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010639135260134935,
      "weighted_relation_loss": 0.0010639135260134935
    },
    {
      "action_query_norm": 39.71389884948731,
      "action_relation_loss": 0.000982116104569286,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.507951356407858,
      "flow_matching_loss": 0.017282334715127946,
      "grad_norm": 0.09543311595916748,
      "learning_rate": 7.03602130414745e-06,
      "loss": 0.0183,
      "relation_loss": 0.00106186083285138,
      "selected_pairs": 27225.8,
      "state_query_norm": 41.908694458007815,
      "state_relation_loss": 7.974472173373215e-05,
      "step": 25130,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.00106186083285138,
      "weighted_relation_loss": 0.00106186083285138
    },
    {
      "action_query_norm": 39.6880126953125,
      "action_relation_loss": 0.0009561746614053845,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.51730589335828,
      "flow_matching_loss": 0.017201822344213726,
      "grad_norm": 0.10723523795604706,
      "learning_rate": 7.007855409248198e-06,
      "loss": 0.0182,
      "relation_loss": 0.0010334872873499989,
      "selected_pairs": 27010.0,
      "state_query_norm": 41.927802658081056,
      "state_relation_loss": 7.731263613095508e-05,
      "step": 25140,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010334872873499989,
      "weighted_relation_loss": 0.0010334872873499989
    },
    {
      "action_query_norm": 39.75736999511719,
      "action_relation_loss": 0.0009487829229328781,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.5266604303087,
      "flow_matching_loss": 0.016192885395139456,
      "grad_norm": 0.10427264124155045,
      "learning_rate": 6.979741753858521e-06,
      "loss": 0.0172,
      "relation_loss": 0.001024389051599428,
      "selected_pairs": 27117.5,
      "state_query_norm": 41.92077713012695,
      "state_relation_loss": 7.560613521491177e-05,
      "step": 25150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001024389051599428,
      "weighted_relation_loss": 0.001024389051599428
    },
    {
      "action_query_norm": 39.69319648742676,
      "action_relation_loss": 0.0009522702835965902,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.53601496725912,
      "flow_matching_loss": 0.017264725267887117,
      "grad_norm": 0.10182023048400879,
      "learning_rate": 6.951680372139158e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010281629511155188,
      "selected_pairs": 26711.1,
      "state_query_norm": 41.91506156921387,
      "state_relation_loss": 7.589266242575831e-05,
      "step": 25160,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0010281629511155188,
      "weighted_relation_loss": 0.0010281629511155188
    },
    {
      "action_query_norm": 39.66038017272949,
      "action_relation_loss": 0.0009160828078165651,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.545369504209543,
      "flow_matching_loss": 0.017582740169018507,
      "grad_norm": 0.10776125639677048,
      "learning_rate": 6.923671298187334e-06,
      "loss": 0.0186,
      "relation_loss": 0.000988477875944227,
      "selected_pairs": 26988.1,
      "state_query_norm": 41.91904411315918,
      "state_relation_loss": 7.239505430334248e-05,
      "step": 25170,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.000988477875944227,
      "weighted_relation_loss": 0.000988477875944227
    },
    {
      "action_query_norm": 39.70379791259766,
      "action_relation_loss": 0.0009666245430707931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.55472404115996,
      "flow_matching_loss": 0.01603042995557189,
      "grad_norm": 0.08934856951236725,
      "learning_rate": 6.895714566036704e-06,
      "loss": 0.0171,
      "relation_loss": 0.0010431626462377608,
      "selected_pairs": 26915.5,
      "state_query_norm": 41.90569381713867,
      "state_relation_loss": 7.653809152543544e-05,
      "step": 25180,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010431626462377608,
      "weighted_relation_loss": 0.0010431626462377608
    },
    {
      "action_query_norm": 39.7245677947998,
      "action_relation_loss": 0.0009465923765674233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.564078578110383,
      "flow_matching_loss": 0.016133626271039248,
      "grad_norm": 0.1385444700717926,
      "learning_rate": 6.86781020965736e-06,
      "loss": 0.0172,
      "relation_loss": 0.001025882741669193,
      "selected_pairs": 27198.2,
      "state_query_norm": 41.90718154907226,
      "state_relation_loss": 7.929036946734414e-05,
      "step": 25190,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.001025882741669193,
      "weighted_relation_loss": 0.001025882741669193
    },
    {
      "action_query_norm": 39.696349716186525,
      "action_relation_loss": 0.0009597425814718008,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.573433115060805,
      "flow_matching_loss": 0.01626170901581645,
      "grad_norm": 0.09341365098953247,
      "learning_rate": 6.839958262955709e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010379868559539317,
      "selected_pairs": 26762.5,
      "state_query_norm": 41.90476417541504,
      "state_relation_loss": 7.824427302693948e-05,
      "step": 25200,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010379868559539317,
      "weighted_relation_loss": 0.0010379868559539317
    },
    {
      "action_query_norm": 39.713922119140626,
      "action_relation_loss": 0.0009158577886410058,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.582787652011227,
      "flow_matching_loss": 0.01650810269638896,
      "grad_norm": 0.16768907010555267,
      "learning_rate": 6.8121587597744894e-06,
      "loss": 0.0175,
      "relation_loss": 0.0009957747475709765,
      "selected_pairs": 27359.0,
      "state_query_norm": 41.90202598571777,
      "state_relation_loss": 7.991695674718358e-05,
      "step": 25210,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009957747475709765,
      "weighted_relation_loss": 0.0009957747475709765
    },
    {
      "action_query_norm": 39.690617370605466,
      "action_relation_loss": 0.0009518401231616736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.592142188961645,
      "flow_matching_loss": 0.016421213187277316,
      "grad_norm": 0.11080756783485413,
      "learning_rate": 6.784411733892732e-06,
      "loss": 0.0175,
      "relation_loss": 0.001031685760244727,
      "selected_pairs": 26344.1,
      "state_query_norm": 41.90157470703125,
      "state_relation_loss": 7.984564799699002e-05,
      "step": 25220,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.001031685760244727,
      "weighted_relation_loss": 0.001031685760244727
    },
    {
      "action_query_norm": 39.685677719116214,
      "action_relation_loss": 0.0010010956088081003,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.601496725912067,
      "flow_matching_loss": 0.015571255329996347,
      "grad_norm": 0.09801051020622253,
      "learning_rate": 6.756717219025666e-06,
      "loss": 0.0166,
      "relation_loss": 0.0010773445304948837,
      "selected_pairs": 27274.5,
      "state_query_norm": 41.91691246032715,
      "state_relation_loss": 7.624892459716648e-05,
      "step": 25230,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0010773445304948837,
      "weighted_relation_loss": 0.0010773445304948837
    },
    {
      "action_query_norm": 39.67551536560059,
      "action_relation_loss": 0.0009483530651777982,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.61085126286249,
      "flow_matching_loss": 0.016581910941749812,
      "grad_norm": 0.15036001801490784,
      "learning_rate": 6.729075248824762e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010235769266728311,
      "selected_pairs": 27363.7,
      "state_query_norm": 41.90083885192871,
      "state_relation_loss": 7.522385494667106e-05,
      "step": 25240,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0010235769266728311,
      "weighted_relation_loss": 0.0010235769266728311
    },
    {
      "action_query_norm": 39.68457717895508,
      "action_relation_loss": 0.0009441889065783471,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.620205799812908,
      "flow_matching_loss": 0.017365800496190787,
      "grad_norm": 0.10537104308605194,
      "learning_rate": 6.701485856877615e-06,
      "loss": 0.0184,
      "relation_loss": 0.001016424474073574,
      "selected_pairs": 27063.2,
      "state_query_norm": 41.908211135864256,
      "state_relation_loss": 7.223556822282262e-05,
      "step": 25250,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.001016424474073574,
      "weighted_relation_loss": 0.001016424474073574
    },
    {
      "action_query_norm": 39.701655578613284,
      "action_relation_loss": 0.0008845925389323383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.62956033676333,
      "flow_matching_loss": 0.016774624213576318,
      "grad_norm": 0.11326432973146439,
      "learning_rate": 6.673949076707925e-06,
      "loss": 0.0177,
      "relation_loss": 0.0009621843579225242,
      "selected_pairs": 26298.0,
      "state_query_norm": 41.908146667480466,
      "state_relation_loss": 7.759182699373923e-05,
      "step": 25260,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0009621843579225242,
      "weighted_relation_loss": 0.0009621843579225242
    },
    {
      "action_query_norm": 39.713549041748045,
      "action_relation_loss": 0.0009292459348216653,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.63891487371375,
      "flow_matching_loss": 0.016971915401518343,
      "grad_norm": 0.11623521894216537,
      "learning_rate": 6.646464941775499e-06,
      "loss": 0.018,
      "relation_loss": 0.0010092246229760349,
      "selected_pairs": 26523.9,
      "state_query_norm": 41.90506019592285,
      "state_relation_loss": 7.997868015081621e-05,
      "step": 25270,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0010092246229760349,
      "weighted_relation_loss": 0.0010092246229760349
    },
    {
      "action_query_norm": 39.665704345703126,
      "action_relation_loss": 0.0009108206606470049,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.648269410664174,
      "flow_matching_loss": 0.014766175858676433,
      "grad_norm": 0.1001083105802536,
      "learning_rate": 6.619033485476128e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009872775757685305,
      "selected_pairs": 26828.6,
      "state_query_norm": 41.93347511291504,
      "state_relation_loss": 7.645691439392977e-05,
      "step": 25280,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009872775757685305,
      "weighted_relation_loss": 0.0009872775757685305
    },
    {
      "action_query_norm": 39.67116508483887,
      "action_relation_loss": 0.0009542370273265988,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.657623947614592,
      "flow_matching_loss": 0.017428241576999425,
      "grad_norm": 0.09287511557340622,
      "learning_rate": 6.59165474114164e-06,
      "loss": 0.0185,
      "relation_loss": 0.0010302441951353103,
      "selected_pairs": 27009.3,
      "state_query_norm": 41.908878326416016,
      "state_relation_loss": 7.600716635352001e-05,
      "step": 25290,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010302441951353103,
      "weighted_relation_loss": 0.0010302441951353103
    },
    {
      "action_query_norm": 39.68903121948242,
      "action_relation_loss": 0.0009181556000839919,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.666978484565014,
      "flow_matching_loss": 0.016151896584779023,
      "grad_norm": 0.0907447561621666,
      "learning_rate": 6.564328742039783e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009966614074073733,
      "selected_pairs": 26507.6,
      "state_query_norm": 41.903221130371094,
      "state_relation_loss": 7.850580514059402e-05,
      "step": 25300,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009966614074073733,
      "weighted_relation_loss": 0.0009966614074073733
    },
    {
      "action_query_norm": 39.67229385375977,
      "action_relation_loss": 0.0009497312421444803,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.676333021515436,
      "flow_matching_loss": 0.01657682042568922,
      "grad_norm": 0.09879659116268158,
      "learning_rate": 6.537055521374219e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010243175260256977,
      "selected_pairs": 26662.2,
      "state_query_norm": 41.91727409362793,
      "state_relation_loss": 7.458626787411049e-05,
      "step": 25310,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010243175260256977,
      "weighted_relation_loss": 0.0010243175260256977
    },
    {
      "action_query_norm": 39.6475643157959,
      "action_relation_loss": 0.0009412359911948442,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.685687558465855,
      "flow_matching_loss": 0.017870545107871293,
      "grad_norm": 0.12298698723316193,
      "learning_rate": 6.509835112284485e-06,
      "loss": 0.0189,
      "relation_loss": 0.0010178073891438544,
      "selected_pairs": 26413.3,
      "state_query_norm": 41.92442932128906,
      "state_relation_loss": 7.657140522496775e-05,
      "step": 25320,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0010178073891438544,
      "weighted_relation_loss": 0.0010178073891438544
    },
    {
      "action_query_norm": 39.65859413146973,
      "action_relation_loss": 0.001023811288177967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.695042095416277,
      "flow_matching_loss": 0.016379269119352103,
      "grad_norm": 0.09561502933502197,
      "learning_rate": 6.482667547845944e-06,
      "loss": 0.0175,
      "relation_loss": 0.0010988411260768772,
      "selected_pairs": 26928.2,
      "state_query_norm": 41.91933631896973,
      "state_relation_loss": 7.50298408092931e-05,
      "step": 25330,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010988411260768772,
      "weighted_relation_loss": 0.0010988411260768772
    },
    {
      "action_query_norm": 39.64445686340332,
      "action_relation_loss": 0.0009195686259772628,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.7043966323667,
      "flow_matching_loss": 0.016376872826367617,
      "grad_norm": 0.09401977062225342,
      "learning_rate": 6.455552861069736e-06,
      "loss": 0.0174,
      "relation_loss": 0.000995151058305055,
      "selected_pairs": 26608.2,
      "state_query_norm": 41.90612258911133,
      "state_relation_loss": 7.558243305538781e-05,
      "step": 25340,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.000995151058305055,
      "weighted_relation_loss": 0.000995151058305055
    },
    {
      "action_query_norm": 39.70607376098633,
      "action_relation_loss": 0.0009321701421868056,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.71375116931712,
      "flow_matching_loss": 0.01741784093901515,
      "grad_norm": 0.09149685502052307,
      "learning_rate": 6.428491084902788e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010074745456222444,
      "selected_pairs": 26758.3,
      "state_query_norm": 41.90549736022949,
      "state_relation_loss": 7.53043896111194e-05,
      "step": 25350,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0010074745456222444,
      "weighted_relation_loss": 0.0010074745456222444
    },
    {
      "action_query_norm": 39.711922073364256,
      "action_relation_loss": 0.0009883378574158997,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.72310570626754,
      "flow_matching_loss": 0.01568760871887207,
      "grad_norm": 0.07215581089258194,
      "learning_rate": 6.401482252227698e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010631348879542202,
      "selected_pairs": 27056.5,
      "state_query_norm": 41.908992767333984,
      "state_relation_loss": 7.47970389056718e-05,
      "step": 25360,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010631348879542202,
      "weighted_relation_loss": 0.0010631348879542202
    },
    {
      "action_query_norm": 39.72408790588379,
      "action_relation_loss": 0.000941814825637266,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.73246024321796,
      "flow_matching_loss": 0.01484192255884409,
      "grad_norm": 0.08792341500520706,
      "learning_rate": 6.37452639586274e-06,
      "loss": 0.0159,
      "relation_loss": 0.0010164168896153569,
      "selected_pairs": 26917.3,
      "state_query_norm": 41.91289749145508,
      "state_relation_loss": 7.460206325049512e-05,
      "step": 25370,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010164168896153569,
      "weighted_relation_loss": 0.0010164168896153569
    },
    {
      "action_query_norm": 39.73591003417969,
      "action_relation_loss": 0.000950689660385251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.741814780168383,
      "flow_matching_loss": 0.015518696792423726,
      "grad_norm": 0.10522913932800293,
      "learning_rate": 6.347623548561826e-06,
      "loss": 0.0165,
      "relation_loss": 0.0010258890106342733,
      "selected_pairs": 26919.8,
      "state_query_norm": 41.90010795593262,
      "state_relation_loss": 7.519935243180954e-05,
      "step": 25380,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010258890106342733,
      "weighted_relation_loss": 0.0010258890106342733
    },
    {
      "action_query_norm": 39.70856399536133,
      "action_relation_loss": 0.0008892254962120205,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.7511693171188,
      "flow_matching_loss": 0.017020615004003047,
      "grad_norm": 0.07554858922958374,
      "learning_rate": 6.320773743014441e-06,
      "loss": 0.018,
      "relation_loss": 0.0009668018377851695,
      "selected_pairs": 26946.2,
      "state_query_norm": 41.907888412475586,
      "state_relation_loss": 7.757634157314896e-05,
      "step": 25390,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009668018377851695,
      "weighted_relation_loss": 0.0009668018377851695
    },
    {
      "action_query_norm": 39.689654159545896,
      "action_relation_loss": 0.0009369789913762361,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.760523854069223,
      "flow_matching_loss": 0.016263643652200697,
      "grad_norm": 0.08853886276483536,
      "learning_rate": 6.293977011845648e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010182585334405302,
      "selected_pairs": 26923.0,
      "state_query_norm": 41.885562133789065,
      "state_relation_loss": 8.127952896757052e-05,
      "step": 25400,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0010182585334405302,
      "weighted_relation_loss": 0.0010182585334405302
    },
    {
      "action_query_norm": 39.650259017944336,
      "action_relation_loss": 0.0009461577807087451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.769878391019645,
      "flow_matching_loss": 0.01586575973778963,
      "grad_norm": 0.0971282422542572,
      "learning_rate": 6.2672333876159835e-06,
      "loss": 0.0169,
      "relation_loss": 0.001021707832114771,
      "selected_pairs": 26533.1,
      "state_query_norm": 41.90036811828613,
      "state_relation_loss": 7.55500550440047e-05,
      "step": 25410,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.001021707832114771,
      "weighted_relation_loss": 0.001021707832114771
    },
    {
      "action_query_norm": 39.67029724121094,
      "action_relation_loss": 0.0009514707082416863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.779232927970064,
      "flow_matching_loss": 0.017058067210018634,
      "grad_norm": 0.11157429963350296,
      "learning_rate": 6.2405429028215e-06,
      "loss": 0.0181,
      "relation_loss": 0.0010282234172336758,
      "selected_pairs": 27150.1,
      "state_query_norm": 41.90723838806152,
      "state_relation_loss": 7.675270608160644e-05,
      "step": 25420,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0010282234172336758,
      "weighted_relation_loss": 0.0010282234172336758
    },
    {
      "action_query_norm": 39.63569679260254,
      "action_relation_loss": 0.0009057186427526176,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.788587464920486,
      "flow_matching_loss": 0.0165349586866796,
      "grad_norm": 0.12686015665531158,
      "learning_rate": 6.213905589893631e-06,
      "loss": 0.0175,
      "relation_loss": 0.00098210105788894,
      "selected_pairs": 26458.2,
      "state_query_norm": 41.89723167419434,
      "state_relation_loss": 7.638241368113086e-05,
      "step": 25430,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.00098210105788894,
      "weighted_relation_loss": 0.00098210105788894
    },
    {
      "action_query_norm": 39.71802673339844,
      "action_relation_loss": 0.0009320235170889646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.797942001870908,
      "flow_matching_loss": 0.015951099805533886,
      "grad_norm": 0.12463165819644928,
      "learning_rate": 6.187321481199221e-06,
      "loss": 0.017,
      "relation_loss": 0.001006182242417708,
      "selected_pairs": 26967.1,
      "state_query_norm": 41.908148193359374,
      "state_relation_loss": 7.415872896672226e-05,
      "step": 25440,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.001006182242417708,
      "weighted_relation_loss": 0.001006182242417708
    },
    {
      "action_query_norm": 39.70748596191406,
      "action_relation_loss": 0.0009289635985624045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.80729653882133,
      "flow_matching_loss": 0.018561911396682264,
      "grad_norm": 0.09455933421850204,
      "learning_rate": 6.1607906090404895e-06,
      "loss": 0.0196,
      "relation_loss": 0.0010059970139991493,
      "selected_pairs": 26824.3,
      "state_query_norm": 41.89935111999512,
      "state_relation_loss": 7.703340161242522e-05,
      "step": 25450,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0010059970139991493,
      "weighted_relation_loss": 0.0010059970139991493
    },
    {
      "action_query_norm": 39.67877197265625,
      "action_relation_loss": 0.0009121405077166855,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.81665107577175,
      "flow_matching_loss": 0.01729087522253394,
      "grad_norm": 0.10045889019966125,
      "learning_rate": 6.134313005654929e-06,
      "loss": 0.0183,
      "relation_loss": 0.0009882203477900476,
      "selected_pairs": 27002.4,
      "state_query_norm": 41.92327995300293,
      "state_relation_loss": 7.607983643538318e-05,
      "step": 25460,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009882203477900476,
      "weighted_relation_loss": 0.0009882203477900476
    },
    {
      "action_query_norm": 39.64623603820801,
      "action_relation_loss": 0.0010143071936909108,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.82600561272217,
      "flow_matching_loss": 0.016447951924055813,
      "grad_norm": 0.08205709606409073,
      "learning_rate": 6.107888703215337e-06,
      "loss": 0.0175,
      "relation_loss": 0.001095985749270767,
      "selected_pairs": 26602.2,
      "state_query_norm": 41.887963104248044,
      "state_relation_loss": 8.167856067302637e-05,
      "step": 25470,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.001095985749270767,
      "weighted_relation_loss": 0.001095985749270767
    },
    {
      "action_query_norm": 39.64433212280274,
      "action_relation_loss": 0.0009853273862972856,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.835360149672592,
      "flow_matching_loss": 0.016710163187235595,
      "grad_norm": 0.09194526076316833,
      "learning_rate": 6.081517733829722e-06,
      "loss": 0.0178,
      "relation_loss": 0.001060218084603548,
      "selected_pairs": 26549.8,
      "state_query_norm": 41.919380569458006,
      "state_relation_loss": 7.489069466828368e-05,
      "step": 25480,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.001060218084603548,
      "weighted_relation_loss": 0.001060218084603548
    },
    {
      "action_query_norm": 39.68564643859863,
      "action_relation_loss": 0.0009536879020743072,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.84471468662301,
      "flow_matching_loss": 0.016330520063638686,
      "grad_norm": 0.08346209675073624,
      "learning_rate": 6.055200129541294e-06,
      "loss": 0.0174,
      "relation_loss": 0.0010329710959922523,
      "selected_pairs": 27551.4,
      "state_query_norm": 41.91967887878418,
      "state_relation_loss": 7.928317281766795e-05,
      "step": 25490,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.0010329710959922523,
      "weighted_relation_loss": 0.0010329710959922523
    },
    {
      "action_query_norm": 39.61943054199219,
      "action_relation_loss": 0.0009461756388191133,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.854069223573433,
      "flow_matching_loss": 0.01645562732592225,
      "grad_norm": 0.10272466391324997,
      "learning_rate": 6.0289359223284434e-06,
      "loss": 0.0175,
      "relation_loss": 0.0010204964375589043,
      "selected_pairs": 26616.7,
      "state_query_norm": 41.88733329772949,
      "state_relation_loss": 7.432080601574854e-05,
      "step": 25500,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0010204964375589043,
      "weighted_relation_loss": 0.0010204964375589043
    },
    {
      "action_query_norm": 39.68127250671387,
      "action_relation_loss": 0.0008996304241009057,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.863423760523855,
      "flow_matching_loss": 0.016325759049504996,
      "grad_norm": 0.14084956049919128,
      "learning_rate": 6.002725144104648e-06,
      "loss": 0.0173,
      "relation_loss": 0.0009735155501402915,
      "selected_pairs": 27244.8,
      "state_query_norm": 41.924865341186525,
      "state_relation_loss": 7.388512822217307e-05,
      "step": 25510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009735155501402915,
      "weighted_relation_loss": 0.0009735155501402915
    },
    {
      "action_query_norm": 39.65528259277344,
      "action_relation_loss": 0.0009350656007882207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.872778297474277,
      "flow_matching_loss": 0.01737461881712079,
      "grad_norm": 0.14749716222286224,
      "learning_rate": 5.976567826718477e-06,
      "loss": 0.0184,
      "relation_loss": 0.0010077136510517448,
      "selected_pairs": 26802.6,
      "state_query_norm": 41.89930267333985,
      "state_relation_loss": 7.264806918101385e-05,
      "step": 25520,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0010077136510517448,
      "weighted_relation_loss": 0.0010077136510517448
    },
    {
      "action_query_norm": 39.69267539978027,
      "action_relation_loss": 0.0009187637188006192,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.882132834424695,
      "flow_matching_loss": 0.017659652698785066,
      "grad_norm": 0.1065027117729187,
      "learning_rate": 5.9504640019535315e-06,
      "loss": 0.0187,
      "relation_loss": 0.0009983351337723434,
      "selected_pairs": 27007.7,
      "state_query_norm": 41.886986923217776,
      "state_relation_loss": 7.957142370287328e-05,
      "step": 25530,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009983351337723434,
      "weighted_relation_loss": 0.0009983351337723434
    },
    {
      "action_query_norm": 39.69814834594727,
      "action_relation_loss": 0.0009867372456938029,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.891487371375117,
      "flow_matching_loss": 0.016268324013799428,
      "grad_norm": 0.11007126420736313,
      "learning_rate": 5.924413701528414e-06,
      "loss": 0.0173,
      "relation_loss": 0.001060742838308215,
      "selected_pairs": 26937.4,
      "state_query_norm": 41.91331024169922,
      "state_relation_loss": 7.400557733490132e-05,
      "step": 25540,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.001060742838308215,
      "weighted_relation_loss": 0.001060742838308215
    },
    {
      "action_query_norm": 39.69603958129883,
      "action_relation_loss": 0.0009511274925898761,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.90084190832554,
      "flow_matching_loss": 0.01674850294366479,
      "grad_norm": 0.09801102429628372,
      "learning_rate": 5.898416957096703e-06,
      "loss": 0.0178,
      "relation_loss": 0.001026119280140847,
      "selected_pairs": 26711.8,
      "state_query_norm": 41.905708694458006,
      "state_relation_loss": 7.499178100260906e-05,
      "step": 25550,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.001026119280140847,
      "weighted_relation_loss": 0.001026119280140847
    },
    {
      "action_query_norm": 39.65695343017578,
      "action_relation_loss": 0.0009375341760460287,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.910196445275957,
      "flow_matching_loss": 0.01512013077735901,
      "grad_norm": 0.0926288366317749,
      "learning_rate": 5.872473800246913e-06,
      "loss": 0.0161,
      "relation_loss": 0.001015355036361143,
      "selected_pairs": 27112.2,
      "state_query_norm": 41.88598289489746,
      "state_relation_loss": 7.78208595875185e-05,
      "step": 25560,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.001015355036361143,
      "weighted_relation_loss": 0.001015355036361143
    },
    {
      "action_query_norm": 39.62289733886719,
      "action_relation_loss": 0.0009693557920400053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.91955098222638,
      "flow_matching_loss": 0.01565450392663479,
      "grad_norm": 0.15952418744564056,
      "learning_rate": 5.846584262502402e-06,
      "loss": 0.0167,
      "relation_loss": 0.0010472920432221145,
      "selected_pairs": 26278.7,
      "state_query_norm": 41.917654037475586,
      "state_relation_loss": 7.793624899932183e-05,
      "step": 25570,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0010472920432221145,
      "weighted_relation_loss": 0.0010472920432221145
    },
    {
      "action_query_norm": 39.68215065002441,
      "action_relation_loss": 0.0009621929377317429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.9289055191768,
      "flow_matching_loss": 0.015840541943907736,
      "grad_norm": 0.11159028112888336,
      "learning_rate": 5.820748375321411e-06,
      "loss": 0.0169,
      "relation_loss": 0.00103925684816204,
      "selected_pairs": 27037.7,
      "state_query_norm": 41.88649444580078,
      "state_relation_loss": 7.706391043029725e-05,
      "step": 25580,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.00103925684816204,
      "weighted_relation_loss": 0.00103925684816204
    },
    {
      "action_query_norm": 39.63639602661133,
      "action_relation_loss": 0.0009581641759723424,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.938260056127223,
      "flow_matching_loss": 0.016294459532946347,
      "grad_norm": 0.11165256053209305,
      "learning_rate": 5.794966170096977e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010356838931329549,
      "selected_pairs": 27087.0,
      "state_query_norm": 41.899071884155276,
      "state_relation_loss": 7.751970333629288e-05,
      "step": 25590,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010356838931329549,
      "weighted_relation_loss": 0.0010356838931329549
    },
    {
      "action_query_norm": 39.68008575439453,
      "action_relation_loss": 0.0009288993664085865,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.947614593077642,
      "flow_matching_loss": 0.01618984742090106,
      "grad_norm": 0.08780328929424286,
      "learning_rate": 5.769237678156897e-06,
      "loss": 0.0172,
      "relation_loss": 0.0010068592615425586,
      "selected_pairs": 27060.3,
      "state_query_norm": 41.890248489379886,
      "state_relation_loss": 7.795989949954674e-05,
      "step": 25600,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0010068592615425586,
      "weighted_relation_loss": 0.0010068592615425586
    },
    {
      "action_query_norm": 39.68936882019043,
      "action_relation_loss": 0.0009209813084453345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.956969130028064,
      "flow_matching_loss": 0.015502476412802935,
      "grad_norm": 0.15751338005065918,
      "learning_rate": 5.743562930763735e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009942923730704933,
      "selected_pairs": 27312.3,
      "state_query_norm": 41.88192863464356,
      "state_relation_loss": 7.331105589400977e-05,
      "step": 25610,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0009942923730704933,
      "weighted_relation_loss": 0.0009942923730704933
    },
    {
      "action_query_norm": 39.652397537231444,
      "action_relation_loss": 0.0009091031097341329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.966323666978486,
      "flow_matching_loss": 0.016474297363311054,
      "grad_norm": 0.09962229430675507,
      "learning_rate": 5.717941959114726e-06,
      "loss": 0.0175,
      "relation_loss": 0.000982413423480466,
      "selected_pairs": 26976.4,
      "state_query_norm": 41.906724548339845,
      "state_relation_loss": 7.331032538786531e-05,
      "step": 25620,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.000982413423480466,
      "weighted_relation_loss": 0.000982413423480466
    },
    {
      "action_query_norm": 39.6842544555664,
      "action_relation_loss": 0.0009106343379244209,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.975678203928904,
      "flow_matching_loss": 0.01607030052691698,
      "grad_norm": 0.10946979373693466,
      "learning_rate": 5.69237479434176e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009892763453535735,
      "selected_pairs": 26990.2,
      "state_query_norm": 41.903703689575195,
      "state_relation_loss": 7.864200961194001e-05,
      "step": 25630,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009892763453535735,
      "weighted_relation_loss": 0.0009892763453535735
    },
    {
      "action_query_norm": 39.68876647949219,
      "action_relation_loss": 0.0009414628788363188,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.985032740879326,
      "flow_matching_loss": 0.01967687690630555,
      "grad_norm": 0.09698235988616943,
      "learning_rate": 5.666861467511353e-06,
      "loss": 0.0207,
      "relation_loss": 0.0010160345234908163,
      "selected_pairs": 27268.8,
      "state_query_norm": 41.92625045776367,
      "state_relation_loss": 7.457165957021061e-05,
      "step": 25640,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.0010160345234908163,
      "weighted_relation_loss": 0.0010160345234908163
    },
    {
      "action_query_norm": 39.671790313720706,
      "action_relation_loss": 0.000910889747319743,
      "auxiliary_loss_weight": 1.0,
      "epoch": 23.99438727782975,
      "flow_matching_loss": 0.0178937797434628,
      "grad_norm": 0.20359933376312256,
      "learning_rate": 5.641402009624591e-06,
      "loss": 0.0189,
      "relation_loss": 0.0009855521901044994,
      "selected_pairs": 26721.6,
      "state_query_norm": 41.88068313598633,
      "state_relation_loss": 7.466243114322424e-05,
      "step": 25650,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0009855521901044994,
      "weighted_relation_loss": 0.0009855521901044994
    },
    {
      "action_query_norm": 39.662025451660156,
      "action_relation_loss": 0.0009161478548776359,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.003741814780167,
      "flow_matching_loss": 0.01775009064003825,
      "grad_norm": 0.09397012740373611,
      "learning_rate": 5.615996451617145e-06,
      "loss": 0.0187,
      "relation_loss": 0.0009899817989207805,
      "selected_pairs": 24154.5,
      "state_query_norm": 41.887044525146486,
      "state_relation_loss": 7.383393822237849e-05,
      "step": 25660,
      "valid_action_samples": 213.5,
      "weighted_auxiliary_loss": 0.0009899817989207805,
      "weighted_relation_loss": 0.0009899817989207805
    },
    {
      "action_query_norm": 39.69148216247559,
      "action_relation_loss": 0.000899951282190159,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.01309635173059,
      "flow_matching_loss": 0.017610474955290556,
      "grad_norm": 0.10923763364553452,
      "learning_rate": 5.590644824359148e-06,
      "loss": 0.0186,
      "relation_loss": 0.0009763136389665305,
      "selected_pairs": 26990.7,
      "state_query_norm": 41.90104713439941,
      "state_relation_loss": 7.636234950041399e-05,
      "step": 25670,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009763136389665305,
      "weighted_relation_loss": 0.0009763136389665305
    },
    {
      "action_query_norm": 39.625698471069335,
      "action_relation_loss": 0.0009253360331058502,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.02245088868101,
      "flow_matching_loss": 0.016549708507955074,
      "grad_norm": 0.11201848089694977,
      "learning_rate": 5.56534715865521e-06,
      "loss": 0.0175,
      "relation_loss": 0.0009999599657021462,
      "selected_pairs": 26950.9,
      "state_query_norm": 41.87650604248047,
      "state_relation_loss": 7.462393114110455e-05,
      "step": 25680,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009999599657021462,
      "weighted_relation_loss": 0.0009999599657021462
    },
    {
      "action_query_norm": 39.69631080627441,
      "action_relation_loss": 0.0009383523312862963,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.031805425631433,
      "flow_matching_loss": 0.016348697245121002,
      "grad_norm": 0.07672383636236191,
      "learning_rate": 5.5401034852443965e-06,
      "loss": 0.0174,
      "relation_loss": 0.0010143275430891663,
      "selected_pairs": 26873.9,
      "state_query_norm": 41.87900619506836,
      "state_relation_loss": 7.597520379931666e-05,
      "step": 25690,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0010143275430891663,
      "weighted_relation_loss": 0.0010143275430891663
    },
    {
      "action_query_norm": 39.701520919799805,
      "action_relation_loss": 0.0009061744844075292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.04115996258185,
      "flow_matching_loss": 0.01565540498122573,
      "grad_norm": 0.13065403699874878,
      "learning_rate": 5.514913834800134e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009831241331994534,
      "selected_pairs": 26285.0,
      "state_query_norm": 41.898817443847655,
      "state_relation_loss": 7.694965388509445e-05,
      "step": 25700,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0009831241331994534,
      "weighted_relation_loss": 0.0009831241331994534
    },
    {
      "action_query_norm": 39.703346252441406,
      "action_relation_loss": 0.0009033907437697053,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.050514499532273,
      "flow_matching_loss": 0.017785881645977497,
      "grad_norm": 0.08757340162992477,
      "learning_rate": 5.489778237930238e-06,
      "loss": 0.0188,
      "relation_loss": 0.0009764532267581671,
      "selected_pairs": 27012.2,
      "state_query_norm": 41.89662322998047,
      "state_relation_loss": 7.306248662644065e-05,
      "step": 25710,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009764532267581671,
      "weighted_relation_loss": 0.0009764532267581671
    },
    {
      "action_query_norm": 39.66253929138183,
      "action_relation_loss": 0.0008817366324365139,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.059869036482695,
      "flow_matching_loss": 0.01612231004983187,
      "grad_norm": 0.10458578914403915,
      "learning_rate": 5.464696725176827e-06,
      "loss": 0.0171,
      "relation_loss": 0.000953846424818039,
      "selected_pairs": 26459.1,
      "state_query_norm": 41.88710327148438,
      "state_relation_loss": 7.210979529190809e-05,
      "step": 25720,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.000953846424818039,
      "weighted_relation_loss": 0.000953846424818039
    },
    {
      "action_query_norm": 39.72230529785156,
      "action_relation_loss": 0.0009288752102293074,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.069223573433113,
      "flow_matching_loss": 0.01470136847347021,
      "grad_norm": 0.13789910078048706,
      "learning_rate": 5.439669327016294e-06,
      "loss": 0.0157,
      "relation_loss": 0.0010039970162324607,
      "selected_pairs": 27454.3,
      "state_query_norm": 41.89512825012207,
      "state_relation_loss": 7.512179799959995e-05,
      "step": 25730,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0010039970162324607,
      "weighted_relation_loss": 0.0010039970162324607
    },
    {
      "action_query_norm": 39.66071319580078,
      "action_relation_loss": 0.0009785599948372691,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.078578110383535,
      "flow_matching_loss": 0.01634378554299474,
      "grad_norm": 0.10122183710336685,
      "learning_rate": 5.4146960738592985e-06,
      "loss": 0.0174,
      "relation_loss": 0.0010535518522374332,
      "selected_pairs": 27090.9,
      "state_query_norm": 41.88715858459473,
      "state_relation_loss": 7.499184139305725e-05,
      "step": 25740,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0010535518522374332,
      "weighted_relation_loss": 0.0010535518522374332
    },
    {
      "action_query_norm": 39.719848251342775,
      "action_relation_loss": 0.0008797624788712711,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.087932647333957,
      "flow_matching_loss": 0.01613318482413888,
      "grad_norm": 0.09584373980760574,
      "learning_rate": 5.389776996050694e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009534749027807265,
      "selected_pairs": 26375.8,
      "state_query_norm": 41.90085563659668,
      "state_relation_loss": 7.371242900262587e-05,
      "step": 25750,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0009534749027807265,
      "weighted_relation_loss": 0.0009534749027807265
    },
    {
      "action_query_norm": 39.66449737548828,
      "action_relation_loss": 0.0009753519494552165,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.09728718428438,
      "flow_matching_loss": 0.016192120406776665,
      "grad_norm": 0.09686724096536636,
      "learning_rate": 5.364912123869492e-06,
      "loss": 0.0172,
      "relation_loss": 0.0010491682798601687,
      "selected_pairs": 26707.1,
      "state_query_norm": 41.907836532592775,
      "state_relation_loss": 7.381632058240939e-05,
      "step": 25760,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0010491682798601687,
      "weighted_relation_loss": 0.0010491682798601687
    },
    {
      "action_query_norm": 39.6283576965332,
      "action_relation_loss": 0.0009705446369480342,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.106641721234798,
      "flow_matching_loss": 0.014975713472813368,
      "grad_norm": 0.07460983097553253,
      "learning_rate": 5.3401014875288864e-06,
      "loss": 0.016,
      "relation_loss": 0.0010481978300958873,
      "selected_pairs": 26504.7,
      "state_query_norm": 41.89406509399414,
      "state_relation_loss": 7.765320042381063e-05,
      "step": 25770,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0010481978300958873,
      "weighted_relation_loss": 0.0010481978300958873
    },
    {
      "action_query_norm": 39.67293014526367,
      "action_relation_loss": 0.0008996002899948508,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.11599625818522,
      "flow_matching_loss": 0.018277844227850438,
      "grad_norm": 0.193153515458107,
      "learning_rate": 5.315345117176129e-06,
      "loss": 0.0193,
      "relation_loss": 0.000972792919492349,
      "selected_pairs": 27503.5,
      "state_query_norm": 41.898646926879884,
      "state_relation_loss": 7.319263604586013e-05,
      "step": 25780,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.000972792919492349,
      "weighted_relation_loss": 0.000972792919492349
    },
    {
      "action_query_norm": 39.686951065063475,
      "action_relation_loss": 0.0008911266922950745,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.125350795135642,
      "flow_matching_loss": 0.015739069785922766,
      "grad_norm": 0.08178716152906418,
      "learning_rate": 5.2906430428925405e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009693939355202019,
      "selected_pairs": 26430.9,
      "state_query_norm": 41.87835693359375,
      "state_relation_loss": 7.826725050108507e-05,
      "step": 25790,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0009693939355202019,
      "weighted_relation_loss": 0.0009693939355202019
    },
    {
      "action_query_norm": 39.671269226074216,
      "action_relation_loss": 0.0009674716100562363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.13470533208606,
      "flow_matching_loss": 0.016229684464633465,
      "grad_norm": 0.09762227535247803,
      "learning_rate": 5.265995294693471e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010442057508043944,
      "selected_pairs": 27016.5,
      "state_query_norm": 41.88839111328125,
      "state_relation_loss": 7.673415748286062e-05,
      "step": 25800,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0010442057508043944,
      "weighted_relation_loss": 0.0010442057508043944
    },
    {
      "action_query_norm": 39.640943145751955,
      "action_relation_loss": 0.0009062355034984648,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.144059869036482,
      "flow_matching_loss": 0.014738761819899082,
      "grad_norm": 0.1293977051973343,
      "learning_rate": 5.241401902528254e-06,
      "loss": 0.0157,
      "relation_loss": 0.000980569195235148,
      "selected_pairs": 26362.7,
      "state_query_norm": 41.89525032043457,
      "state_relation_loss": 7.433369246427901e-05,
      "step": 25810,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.000980569195235148,
      "weighted_relation_loss": 0.000980569195235148
    },
    {
      "action_query_norm": 39.69685173034668,
      "action_relation_loss": 0.0009525317407678812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.153414405986904,
      "flow_matching_loss": 0.01706666946411133,
      "grad_norm": 0.1288733035326004,
      "learning_rate": 5.216862896280189e-06,
      "loss": 0.0181,
      "relation_loss": 0.001030203903792426,
      "selected_pairs": 26848.0,
      "state_query_norm": 41.893856048583984,
      "state_relation_loss": 7.767217030050234e-05,
      "step": 25820,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.001030203903792426,
      "weighted_relation_loss": 0.001030203903792426
    },
    {
      "action_query_norm": 39.653632736206056,
      "action_relation_loss": 0.0009017324657179416,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.162768942937326,
      "flow_matching_loss": 0.01657930379733443,
      "grad_norm": 0.11701013147830963,
      "learning_rate": 5.192378305766471e-06,
      "loss": 0.0176,
      "relation_loss": 0.0009761805995367467,
      "selected_pairs": 26735.3,
      "state_query_norm": 41.882089614868164,
      "state_relation_loss": 7.444813163601794e-05,
      "step": 25830,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009761805995367467,
      "weighted_relation_loss": 0.0009761805995367467
    },
    {
      "action_query_norm": 39.649207305908206,
      "action_relation_loss": 0.0009221905726008117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.172123479887745,
      "flow_matching_loss": 0.017578457947820425,
      "grad_norm": 0.14400097727775574,
      "learning_rate": 5.1679481607382065e-06,
      "loss": 0.0186,
      "relation_loss": 0.0009988084610085934,
      "selected_pairs": 26625.9,
      "state_query_norm": 41.88404579162598,
      "state_relation_loss": 7.661787822144106e-05,
      "step": 25840,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009988084610085934,
      "weighted_relation_loss": 0.0009988084610085934
    },
    {
      "action_query_norm": 39.65030479431152,
      "action_relation_loss": 0.0009023445309139788,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.181478016838167,
      "flow_matching_loss": 0.014475986082106828,
      "grad_norm": 0.11479668319225311,
      "learning_rate": 5.1435724908803105e-06,
      "loss": 0.0155,
      "relation_loss": 0.0009771254670340567,
      "selected_pairs": 26507.2,
      "state_query_norm": 41.89052543640137,
      "state_relation_loss": 7.478092920791824e-05,
      "step": 25850,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009771254670340567,
      "weighted_relation_loss": 0.0009771254670340567
    },
    {
      "action_query_norm": 39.63629417419433,
      "action_relation_loss": 0.0009916863404214383,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.19083255378859,
      "flow_matching_loss": 0.017199798859655856,
      "grad_norm": 0.13077400624752045,
      "learning_rate": 5.119251325811514e-06,
      "loss": 0.0183,
      "relation_loss": 0.0010652244265656919,
      "selected_pairs": 26467.7,
      "state_query_norm": 41.89350204467773,
      "state_relation_loss": 7.353807231993414e-05,
      "step": 25860,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010652244265656919,
      "weighted_relation_loss": 0.0010652244265656919
    },
    {
      "action_query_norm": 39.69156150817871,
      "action_relation_loss": 0.0009186307375784964,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.200187090739007,
      "flow_matching_loss": 0.015756568126380443,
      "grad_norm": 0.12731315195560455,
      "learning_rate": 5.094984695084348e-06,
      "loss": 0.0167,
      "relation_loss": 0.000989830691833049,
      "selected_pairs": 27246.6,
      "state_query_norm": 41.908010482788086,
      "state_relation_loss": 7.119995971152093e-05,
      "step": 25870,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.000989830691833049,
      "weighted_relation_loss": 0.000989830691833049
    },
    {
      "action_query_norm": 39.69096488952637,
      "action_relation_loss": 0.0009188158845063299,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.20954162768943,
      "flow_matching_loss": 0.017157287895679475,
      "grad_norm": 0.12408697605133057,
      "learning_rate": 5.070772628185039e-06,
      "loss": 0.0182,
      "relation_loss": 0.0009934905334375798,
      "selected_pairs": 26524.4,
      "state_query_norm": 41.89634017944336,
      "state_relation_loss": 7.467465329682454e-05,
      "step": 25880,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009934905334375798,
      "weighted_relation_loss": 0.0009934905334375798
    },
    {
      "action_query_norm": 39.68547248840332,
      "action_relation_loss": 0.0008927944640163332,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.21889616463985,
      "flow_matching_loss": 0.016212848108261825,
      "grad_norm": 0.08426681905984879,
      "learning_rate": 5.046615154533535e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009672540763858706,
      "selected_pairs": 26999.5,
      "state_query_norm": 41.90148277282715,
      "state_relation_loss": 7.445961673511193e-05,
      "step": 25890,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009672540763858706,
      "weighted_relation_loss": 0.0009672540763858706
    },
    {
      "action_query_norm": 39.69652137756348,
      "action_relation_loss": 0.0009396684356033802,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.22825070159027,
      "flow_matching_loss": 0.01617100592702627,
      "grad_norm": 0.08802233636379242,
      "learning_rate": 5.022512303483451e-06,
      "loss": 0.0172,
      "relation_loss": 0.0010141198406927287,
      "selected_pairs": 26480.8,
      "state_query_norm": 41.877825546264646,
      "state_relation_loss": 7.445141891366802e-05,
      "step": 25900,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0010141198406927287,
      "weighted_relation_loss": 0.0010141198406927287
    },
    {
      "action_query_norm": 39.64250717163086,
      "action_relation_loss": 0.0009144052222836762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.23760523854069,
      "flow_matching_loss": 0.017256596963852643,
      "grad_norm": 0.09453233331441879,
      "learning_rate": 4.998464104322015e-06,
      "loss": 0.0182,
      "relation_loss": 0.0009888926986604929,
      "selected_pairs": 26760.8,
      "state_query_norm": 41.88874397277832,
      "state_relation_loss": 7.448746109730564e-05,
      "step": 25910,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009888926986604929,
      "weighted_relation_loss": 0.0009888926986604929
    },
    {
      "action_query_norm": 39.69557647705078,
      "action_relation_loss": 0.0009287389402743429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.246959775491113,
      "flow_matching_loss": 0.015642730332911015,
      "grad_norm": 0.12671302258968353,
      "learning_rate": 4.974470586270047e-06,
      "loss": 0.0166,
      "relation_loss": 0.0010051753837615252,
      "selected_pairs": 26667.5,
      "state_query_norm": 41.884716796875,
      "state_relation_loss": 7.643645003554411e-05,
      "step": 25920,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010051753837615252,
      "weighted_relation_loss": 0.0010051753837615252
    },
    {
      "action_query_norm": 39.677880096435544,
      "action_relation_loss": 0.0008803647651802748,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.256314312441535,
      "flow_matching_loss": 0.015782323013991118,
      "grad_norm": 0.09182846546173096,
      "learning_rate": 4.950531778481965e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009571495058480651,
      "selected_pairs": 27333.7,
      "state_query_norm": 41.892998123168944,
      "state_relation_loss": 7.678474648855626e-05,
      "step": 25930,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009571495058480651,
      "weighted_relation_loss": 0.0009571495058480651
    },
    {
      "action_query_norm": 39.6439567565918,
      "action_relation_loss": 0.0009480163105763495,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.265668849391954,
      "flow_matching_loss": 0.016260078642517328,
      "grad_norm": 0.08934293687343597,
      "learning_rate": 4.926647710045651e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010253907588776202,
      "selected_pairs": 27007.0,
      "state_query_norm": 41.883365631103516,
      "state_relation_loss": 7.73744315665681e-05,
      "step": 25940,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010253907588776202,
      "weighted_relation_loss": 0.0010253907588776202
    },
    {
      "action_query_norm": 39.63840560913086,
      "action_relation_loss": 0.0009313100774306804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.275023386342376,
      "flow_matching_loss": 0.015805058274418117,
      "grad_norm": 0.10926070064306259,
      "learning_rate": 4.9028184099825125e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010062926507089288,
      "selected_pairs": 26547.5,
      "state_query_norm": 41.90928573608399,
      "state_relation_loss": 7.498256491089705e-05,
      "step": 25950,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010062926507089288,
      "weighted_relation_loss": 0.0010062926507089288
    },
    {
      "action_query_norm": 39.641697311401366,
      "action_relation_loss": 0.0009124468895606696,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.284377923292798,
      "flow_matching_loss": 0.015766737051308155,
      "grad_norm": 0.0961008369922638,
      "learning_rate": 4.879043907247382e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009869784873444587,
      "selected_pairs": 26857.4,
      "state_query_norm": 41.891300201416016,
      "state_relation_loss": 7.453159996657632e-05,
      "step": 25960,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009869784873444587,
      "weighted_relation_loss": 0.0009869784873444587
    },
    {
      "action_query_norm": 39.65362739562988,
      "action_relation_loss": 0.0009274953743442893,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.293732460243216,
      "flow_matching_loss": 0.016858568880707026,
      "grad_norm": 0.10042767971754074,
      "learning_rate": 4.8553242307285414e-06,
      "loss": 0.0179,
      "relation_loss": 0.0010069734242279082,
      "selected_pairs": 26987.7,
      "state_query_norm": 41.89443435668945,
      "state_relation_loss": 7.947805133881047e-05,
      "step": 25970,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010069734242279082,
      "weighted_relation_loss": 0.0010069734242279082
    },
    {
      "action_query_norm": 39.690883255004884,
      "action_relation_loss": 0.0009641813521739095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.30308699719364,
      "flow_matching_loss": 0.01536238333210349,
      "grad_norm": 0.08461390435695648,
      "learning_rate": 4.8316594092476185e-06,
      "loss": 0.0164,
      "relation_loss": 0.0010432835260871798,
      "selected_pairs": 26414.4,
      "state_query_norm": 41.874280548095705,
      "state_relation_loss": 7.910218264441937e-05,
      "step": 25980,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0010432835260871798,
      "weighted_relation_loss": 0.0010432835260871798
    },
    {
      "action_query_norm": 39.706007385253905,
      "action_relation_loss": 0.0008761813805904239,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.31244153414406,
      "flow_matching_loss": 0.01679346114397049,
      "grad_norm": 0.0940195694565773,
      "learning_rate": 4.808049471559617e-06,
      "loss": 0.0177,
      "relation_loss": 0.0009500819549430161,
      "selected_pairs": 27194.8,
      "state_query_norm": 41.89002876281738,
      "state_relation_loss": 7.390057362499647e-05,
      "step": 25990,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009500819549430161,
      "weighted_relation_loss": 0.0009500819549430161
    },
    {
      "action_query_norm": 39.621782302856445,
      "action_relation_loss": 0.0009789377509150654,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.321796071094482,
      "flow_matching_loss": 0.01613986324518919,
      "grad_norm": 0.10706847161054611,
      "learning_rate": 4.784494446352833e-06,
      "loss": 0.0172,
      "relation_loss": 0.0010561686765868216,
      "selected_pairs": 27298.0,
      "state_query_norm": 41.87450103759765,
      "state_relation_loss": 7.723092421656475e-05,
      "step": 26000,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0010561686765868216,
      "weighted_relation_loss": 0.0010561686765868216
    },
    {
      "action_query_norm": 39.65608901977539,
      "action_relation_loss": 0.0009209206560626626,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.3311506080449,
      "flow_matching_loss": 0.016352496389299632,
      "grad_norm": 0.11572922021150589,
      "learning_rate": 4.760994362248833e-06,
      "loss": 0.0173,
      "relation_loss": 0.000996361835859716,
      "selected_pairs": 26464.2,
      "state_query_norm": 41.88842544555664,
      "state_relation_loss": 7.544117397628725e-05,
      "step": 26010,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.000996361835859716,
      "weighted_relation_loss": 0.000996361835859716
    },
    {
      "action_query_norm": 39.64642868041992,
      "action_relation_loss": 0.000896476034540683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.340505144995323,
      "flow_matching_loss": 0.015177736803889275,
      "grad_norm": 0.09752769768238068,
      "learning_rate": 4.737549247802442e-06,
      "loss": 0.0161,
      "relation_loss": 0.000970828632125631,
      "selected_pairs": 27081.7,
      "state_query_norm": 41.86679878234863,
      "state_relation_loss": 7.435258521582e-05,
      "step": 26020,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.000970828632125631,
      "weighted_relation_loss": 0.000970828632125631
    },
    {
      "action_query_norm": 39.69514427185059,
      "action_relation_loss": 0.0009225445101037621,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.349859681945745,
      "flow_matching_loss": 0.015432624332606792,
      "grad_norm": 0.08823415637016296,
      "learning_rate": 4.714159131501689e-06,
      "loss": 0.0164,
      "relation_loss": 0.0009965303994249552,
      "selected_pairs": 26393.5,
      "state_query_norm": 41.892493057250974,
      "state_relation_loss": 7.39858929591719e-05,
      "step": 26030,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0009965303994249552,
      "weighted_relation_loss": 0.0009965303994249552
    },
    {
      "action_query_norm": 39.687605667114255,
      "action_relation_loss": 0.0009265264554414899,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.359214218896163,
      "flow_matching_loss": 0.014997070003300906,
      "grad_norm": 0.10684335976839066,
      "learning_rate": 4.690824041767766e-06,
      "loss": 0.016,
      "relation_loss": 0.0009992619161494076,
      "selected_pairs": 26807.1,
      "state_query_norm": 41.88260002136231,
      "state_relation_loss": 7.273544870258774e-05,
      "step": 26040,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009992619161494076,
      "weighted_relation_loss": 0.0009992619161494076
    },
    {
      "action_query_norm": 39.68712463378906,
      "action_relation_loss": 0.0008636825426947325,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.368568755846585,
      "flow_matching_loss": 0.015138589125126601,
      "grad_norm": 0.10285694152116776,
      "learning_rate": 4.6675440069550105e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009404637501575052,
      "selected_pairs": 27058.9,
      "state_query_norm": 41.88386459350586,
      "state_relation_loss": 7.678121619392186e-05,
      "step": 26050,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009404637501575052,
      "weighted_relation_loss": 0.0009404637501575052
    },
    {
      "action_query_norm": 39.67099227905273,
      "action_relation_loss": 0.0009026744693983347,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.377923292797007,
      "flow_matching_loss": 0.016223990079015493,
      "grad_norm": 0.15986138582229614,
      "learning_rate": 4.64431905535086e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009749658755026757,
      "selected_pairs": 27034.8,
      "state_query_norm": 41.89328384399414,
      "state_relation_loss": 7.229140755953267e-05,
      "step": 26060,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009749658755026757,
      "weighted_relation_loss": 0.0009749658755026757
    },
    {
      "action_query_norm": 39.661896514892575,
      "action_relation_loss": 0.0009214681922458112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.38727782974743,
      "flow_matching_loss": 0.016910497285425662,
      "grad_norm": 0.1264793425798416,
      "learning_rate": 4.621149215175808e-06,
      "loss": 0.0179,
      "relation_loss": 0.0009977628476917743,
      "selected_pairs": 27197.5,
      "state_query_norm": 41.8909049987793,
      "state_relation_loss": 7.629465253558009e-05,
      "step": 26070,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009977628476917743,
      "weighted_relation_loss": 0.0009977628476917743
    },
    {
      "action_query_norm": 39.68835220336914,
      "action_relation_loss": 0.0008618860563728958,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.396632366697848,
      "flow_matching_loss": 0.015538631193339824,
      "grad_norm": 0.13722160458564758,
      "learning_rate": 4.5980345145834154e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009353980130981654,
      "selected_pairs": 27361.6,
      "state_query_norm": 41.87461395263672,
      "state_relation_loss": 7.351195235969499e-05,
      "step": 26080,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009353980130981654,
      "weighted_relation_loss": 0.0009353980130981654
    },
    {
      "action_query_norm": 39.65159187316895,
      "action_relation_loss": 0.0008899226842913777,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.40598690364827,
      "flow_matching_loss": 0.016614740807563068,
      "grad_norm": 0.08918162435293198,
      "learning_rate": 4.574974981660213e-06,
      "loss": 0.0176,
      "relation_loss": 0.0009638057614210993,
      "selected_pairs": 26965.4,
      "state_query_norm": 41.88748359680176,
      "state_relation_loss": 7.388306730717886e-05,
      "step": 26090,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009638057614210993,
      "weighted_relation_loss": 0.0009638057614210993
    },
    {
      "action_query_norm": 39.67583808898926,
      "action_relation_loss": 0.0008971829665824771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.41534144059869,
      "flow_matching_loss": 0.016061503998935224,
      "grad_norm": 0.11258161813020706,
      "learning_rate": 4.551970644425707e-06,
      "loss": 0.017,
      "relation_loss": 0.0009726573945954442,
      "selected_pairs": 27200.8,
      "state_query_norm": 41.89193420410156,
      "state_relation_loss": 7.54744149162434e-05,
      "step": 26100,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009726573945954442,
      "weighted_relation_loss": 0.0009726573945954442
    },
    {
      "action_query_norm": 39.65623893737793,
      "action_relation_loss": 0.000940439518308267,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.42469597754911,
      "flow_matching_loss": 0.014998508710414171,
      "grad_norm": 0.09431520849466324,
      "learning_rate": 4.529021530832328e-06,
      "loss": 0.016,
      "relation_loss": 0.001013840432278812,
      "selected_pairs": 27288.5,
      "state_query_norm": 41.89607925415039,
      "state_relation_loss": 7.340092051890678e-05,
      "step": 26110,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.001013840432278812,
      "weighted_relation_loss": 0.001013840432278812
    },
    {
      "action_query_norm": 39.71219749450684,
      "action_relation_loss": 0.0009017631586175412,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.434050514499532,
      "flow_matching_loss": 0.017460406199097632,
      "grad_norm": 0.1105441078543663,
      "learning_rate": 4.5061276687654306e-06,
      "loss": 0.0184,
      "relation_loss": 0.000979541492415592,
      "selected_pairs": 26460.5,
      "state_query_norm": 41.885834503173825,
      "state_relation_loss": 7.777833016007208e-05,
      "step": 26120,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.000979541492415592,
      "weighted_relation_loss": 0.000979541492415592
    },
    {
      "action_query_norm": 39.6360481262207,
      "action_relation_loss": 0.0009071754233445972,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.443405051449954,
      "flow_matching_loss": 0.015522783994674683,
      "grad_norm": 0.1097731664776802,
      "learning_rate": 4.4832890860431965e-06,
      "loss": 0.0165,
      "relation_loss": 0.000983536208514124,
      "selected_pairs": 26351.5,
      "state_query_norm": 41.87104721069336,
      "state_relation_loss": 7.636078298673965e-05,
      "step": 26130,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.000983536208514124,
      "weighted_relation_loss": 0.000983536208514124
    },
    {
      "action_query_norm": 39.63694725036621,
      "action_relation_loss": 0.0009138786990661174,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.452759588400376,
      "flow_matching_loss": 0.01518742935732007,
      "grad_norm": 0.08592668175697327,
      "learning_rate": 4.460505810416682e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009888707601930947,
      "selected_pairs": 26860.1,
      "state_query_norm": 41.89827995300293,
      "state_relation_loss": 7.499205676140264e-05,
      "step": 26140,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009888707601930947,
      "weighted_relation_loss": 0.0009888707601930947
    },
    {
      "action_query_norm": 39.66462326049805,
      "action_relation_loss": 0.0008696836302988231,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.462114125350794,
      "flow_matching_loss": 0.016160884220153094,
      "grad_norm": 0.11897945404052734,
      "learning_rate": 4.4377778695696984e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009403856995049864,
      "selected_pairs": 27406.5,
      "state_query_norm": 41.88684425354004,
      "state_relation_loss": 7.070207502692937e-05,
      "step": 26150,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0009403856995049864,
      "weighted_relation_loss": 0.0009403856995049864
    },
    {
      "action_query_norm": 39.68024520874023,
      "action_relation_loss": 0.0008778033079579473,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.471468662301216,
      "flow_matching_loss": 0.016207421943545342,
      "grad_norm": 0.08515817672014236,
      "learning_rate": 4.415105291118843e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009522601496428252,
      "selected_pairs": 27360.8,
      "state_query_norm": 41.871359252929686,
      "state_relation_loss": 7.445683149853721e-05,
      "step": 26160,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009522601496428252,
      "weighted_relation_loss": 0.0009522601496428252
    },
    {
      "action_query_norm": 39.634709167480466,
      "action_relation_loss": 0.0008998183708172292,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.48082319925164,
      "flow_matching_loss": 0.015034362766891718,
      "grad_norm": 0.09740619361400604,
      "learning_rate": 4.392488102613435e-06,
      "loss": 0.016,
      "relation_loss": 0.0009792749711778014,
      "selected_pairs": 27388.4,
      "state_query_norm": 41.88924903869629,
      "state_relation_loss": 7.945659963297658e-05,
      "step": 26170,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0009792749711778014,
      "weighted_relation_loss": 0.0009792749711778014
    },
    {
      "action_query_norm": 39.66842155456543,
      "action_relation_loss": 0.000884533382486552,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.490177736202057,
      "flow_matching_loss": 0.013994853757321835,
      "grad_norm": 0.11390700191259384,
      "learning_rate": 4.369926331535473e-06,
      "loss": 0.015,
      "relation_loss": 0.000958046643063426,
      "selected_pairs": 26666.3,
      "state_query_norm": 41.868129348754884,
      "state_relation_loss": 7.351324675255455e-05,
      "step": 26180,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.000958046643063426,
      "weighted_relation_loss": 0.000958046643063426
    },
    {
      "action_query_norm": 39.621965408325195,
      "action_relation_loss": 0.0009739350061863661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.49953227315248,
      "flow_matching_loss": 0.01628203596919775,
      "grad_norm": 0.0901283398270607,
      "learning_rate": 4.347420005299668e-06,
      "loss": 0.0173,
      "relation_loss": 0.0010522183263674378,
      "selected_pairs": 27031.0,
      "state_query_norm": 41.863019180297854,
      "state_relation_loss": 7.828331727068871e-05,
      "step": 26190,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0010522183263674378,
      "weighted_relation_loss": 0.0010522183263674378
    },
    {
      "action_query_norm": 39.677713012695314,
      "action_relation_loss": 0.000921225524507463,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.5088868101029,
      "flow_matching_loss": 0.01744939424097538,
      "grad_norm": 0.12504246830940247,
      "learning_rate": 4.324969151253317e-06,
      "loss": 0.0184,
      "relation_loss": 0.0009938185452483595,
      "selected_pairs": 27108.0,
      "state_query_norm": 41.892630386352536,
      "state_relation_loss": 7.259300909936428e-05,
      "step": 26200,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009938185452483595,
      "weighted_relation_loss": 0.0009938185452483595
    },
    {
      "action_query_norm": 39.61835861206055,
      "action_relation_loss": 0.0009563096391502768,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.51824134705332,
      "flow_matching_loss": 0.01682370537891984,
      "grad_norm": 0.08973018825054169,
      "learning_rate": 4.302573796676312e-06,
      "loss": 0.0179,
      "relation_loss": 0.0010315706604160368,
      "selected_pairs": 26549.1,
      "state_query_norm": 41.862711334228514,
      "state_relation_loss": 7.526103436248378e-05,
      "step": 26210,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0010315706604160368,
      "weighted_relation_loss": 0.0010315706604160368
    },
    {
      "action_query_norm": 39.65751647949219,
      "action_relation_loss": 0.0008721165242604912,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.52759588400374,
      "flow_matching_loss": 0.01523291040211916,
      "grad_norm": 0.08219482004642487,
      "learning_rate": 4.2802339687811385e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009452976926695555,
      "selected_pairs": 27013.8,
      "state_query_norm": 41.88423957824707,
      "state_relation_loss": 7.318116986425594e-05,
      "step": 26220,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009452976926695555,
      "weighted_relation_loss": 0.0009452976926695555
    },
    {
      "action_query_norm": 39.651680374145506,
      "action_relation_loss": 0.0008930219104513526,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.536950420954163,
      "flow_matching_loss": 0.015342861600220203,
      "grad_norm": 0.11061982065439224,
      "learning_rate": 4.257949694712771e-06,
      "loss": 0.0163,
      "relation_loss": 0.0009713805571664125,
      "selected_pairs": 26383.0,
      "state_query_norm": 41.87708053588867,
      "state_relation_loss": 7.83586292527616e-05,
      "step": 26230,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0009713805571664125,
      "weighted_relation_loss": 0.0009713805571664125
    },
    {
      "action_query_norm": 39.679342651367186,
      "action_relation_loss": 0.0010120365244802088,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.546304957904585,
      "flow_matching_loss": 0.0170130530372262,
      "grad_norm": 0.14725299179553986,
      "learning_rate": 4.235721001548726e-06,
      "loss": 0.0181,
      "relation_loss": 0.0010884838353376835,
      "selected_pairs": 26667.5,
      "state_query_norm": 41.89143295288086,
      "state_relation_loss": 7.644731886102818e-05,
      "step": 26240,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010884838353376835,
      "weighted_relation_loss": 0.0010884838353376835
    },
    {
      "action_query_norm": 39.66907005310058,
      "action_relation_loss": 0.0009508234914392232,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.555659494855004,
      "flow_matching_loss": 0.01670719189569354,
      "grad_norm": 0.09956305474042892,
      "learning_rate": 4.21354791629896e-06,
      "loss": 0.0177,
      "relation_loss": 0.0010228280269075185,
      "selected_pairs": 27389.7,
      "state_query_norm": 41.88086166381836,
      "state_relation_loss": 7.200454419944435e-05,
      "step": 26250,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0010228280269075185,
      "weighted_relation_loss": 0.0010228280269075185
    },
    {
      "action_query_norm": 39.602896499633786,
      "action_relation_loss": 0.0009381433541420847,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.565014031805426,
      "flow_matching_loss": 0.015620006155222655,
      "grad_norm": 0.10474415123462677,
      "learning_rate": 4.191430465905843e-06,
      "loss": 0.0166,
      "relation_loss": 0.001009661622811109,
      "selected_pairs": 26804.6,
      "state_query_norm": 41.86906814575195,
      "state_relation_loss": 7.151827376219443e-05,
      "step": 26260,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.001009661622811109,
      "weighted_relation_loss": 0.001009661622811109
    },
    {
      "action_query_norm": 39.6384693145752,
      "action_relation_loss": 0.0008933105738833547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.574368568755848,
      "flow_matching_loss": 0.017772225476801396,
      "grad_norm": 0.11404582858085632,
      "learning_rate": 4.169368677244184e-06,
      "loss": 0.0187,
      "relation_loss": 0.0009694931155536324,
      "selected_pairs": 26458.5,
      "state_query_norm": 41.8685604095459,
      "state_relation_loss": 7.61825307563413e-05,
      "step": 26270,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009694931155536324,
      "weighted_relation_loss": 0.0009694931155536324
    },
    {
      "action_query_norm": 39.66891021728516,
      "action_relation_loss": 0.0009136338776443154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.583723105706266,
      "flow_matching_loss": 0.014639153517782688,
      "grad_norm": 0.09769146889448166,
      "learning_rate": 4.147362577121122e-06,
      "loss": 0.0156,
      "relation_loss": 0.0009850973379798233,
      "selected_pairs": 26782.5,
      "state_query_norm": 41.90181655883789,
      "state_relation_loss": 7.146344723878429e-05,
      "step": 26280,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0009850973379798233,
      "weighted_relation_loss": 0.0009850973379798233
    },
    {
      "action_query_norm": 39.63126640319824,
      "action_relation_loss": 0.0008997811819426715,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.593077642656688,
      "flow_matching_loss": 0.01501066004857421,
      "grad_norm": 0.12632706761360168,
      "learning_rate": 4.125412192276157e-06,
      "loss": 0.016,
      "relation_loss": 0.0009769565891474485,
      "selected_pairs": 26708.4,
      "state_query_norm": 41.87737808227539,
      "state_relation_loss": 7.717540429439396e-05,
      "step": 26290,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0009769565891474485,
      "weighted_relation_loss": 0.0009769565891474485
    },
    {
      "action_query_norm": 39.673530960083006,
      "action_relation_loss": 0.0009003080602269619,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.60243217960711,
      "flow_matching_loss": 0.01684326184913516,
      "grad_norm": 0.10293737053871155,
      "learning_rate": 4.10351754938107e-06,
      "loss": 0.0178,
      "relation_loss": 0.000972754106624052,
      "selected_pairs": 27131.7,
      "state_query_norm": 41.884539413452146,
      "state_relation_loss": 7.244603366416414e-05,
      "step": 26300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.000972754106624052,
      "weighted_relation_loss": 0.000972754106624052
    },
    {
      "action_query_norm": 39.65223770141601,
      "action_relation_loss": 0.0009332058602012694,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.611786716557532,
      "flow_matching_loss": 0.015176938567310571,
      "grad_norm": 0.09837668389081955,
      "learning_rate": 4.081678675039913e-06,
      "loss": 0.0162,
      "relation_loss": 0.0010113772470504046,
      "selected_pairs": 27249.3,
      "state_query_norm": 41.86821327209473,
      "state_relation_loss": 7.817138393875211e-05,
      "step": 26310,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0010113772470504046,
      "weighted_relation_loss": 0.0010113772470504046
    },
    {
      "action_query_norm": 39.71415824890137,
      "action_relation_loss": 0.000913152750581503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.62114125350795,
      "flow_matching_loss": 0.018795057386159896,
      "grad_norm": 0.09635001420974731,
      "learning_rate": 4.0598955957889876e-06,
      "loss": 0.0198,
      "relation_loss": 0.000989520485745743,
      "selected_pairs": 27407.5,
      "state_query_norm": 41.875325775146486,
      "state_relation_loss": 7.636773880221881e-05,
      "step": 26320,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.000989520485745743,
      "weighted_relation_loss": 0.000989520485745743
    },
    {
      "action_query_norm": 39.58666725158692,
      "action_relation_loss": 0.0009488165029324591,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.630495790458372,
      "flow_matching_loss": 0.016310568992048503,
      "grad_norm": 0.11218909174203873,
      "learning_rate": 4.038168338096776e-06,
      "loss": 0.0173,
      "relation_loss": 0.001025289244716987,
      "selected_pairs": 27360.3,
      "state_query_norm": 41.864932250976565,
      "state_relation_loss": 7.647274396731518e-05,
      "step": 26330,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.001025289244716987,
      "weighted_relation_loss": 0.001025289244716987
    },
    {
      "action_query_norm": 39.64633407592773,
      "action_relation_loss": 0.0009436093328986317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.639850327408794,
      "flow_matching_loss": 0.015871758200228216,
      "grad_norm": 0.11773671209812164,
      "learning_rate": 4.016496928363944e-06,
      "loss": 0.0169,
      "relation_loss": 0.0010162199207115919,
      "selected_pairs": 26462.4,
      "state_query_norm": 41.87184867858887,
      "state_relation_loss": 7.261059072334319e-05,
      "step": 26340,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0010162199207115919,
      "weighted_relation_loss": 0.0010162199207115919
    },
    {
      "action_query_norm": 39.660831451416016,
      "action_relation_loss": 0.0009257888013962656,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.649204864359213,
      "flow_matching_loss": 0.01584779843688011,
      "grad_norm": 0.08738037943840027,
      "learning_rate": 3.994881392923317e-06,
      "loss": 0.0169,
      "relation_loss": 0.0010024469112977386,
      "selected_pairs": 27248.1,
      "state_query_norm": 41.88636245727539,
      "state_relation_loss": 7.665810408070683e-05,
      "step": 26350,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0010024469112977386,
      "weighted_relation_loss": 0.0010024469112977386
    },
    {
      "action_query_norm": 39.65416793823242,
      "action_relation_loss": 0.0009121297625824809,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.658559401309635,
      "flow_matching_loss": 0.017503482382744552,
      "grad_norm": 0.10593006759881973,
      "learning_rate": 3.9733217580397945e-06,
      "loss": 0.0185,
      "relation_loss": 0.0009864824416581542,
      "selected_pairs": 26503.0,
      "state_query_norm": 41.86671180725098,
      "state_relation_loss": 7.43526870792266e-05,
      "step": 26360,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009864824416581542,
      "weighted_relation_loss": 0.0009864824416581542
    },
    {
      "action_query_norm": 39.63783760070801,
      "action_relation_loss": 0.0009475375642068684,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.667913938260057,
      "flow_matching_loss": 0.01699923286214471,
      "grad_norm": 0.09209639579057693,
      "learning_rate": 3.951818049910366e-06,
      "loss": 0.018,
      "relation_loss": 0.0010220202500931918,
      "selected_pairs": 26573.6,
      "state_query_norm": 41.868142700195314,
      "state_relation_loss": 7.448268770531286e-05,
      "step": 26370,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0010220202500931918,
      "weighted_relation_loss": 0.0010220202500931918
    },
    {
      "action_query_norm": 39.61152648925781,
      "action_relation_loss": 0.0008581450500059872,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.67726847521048,
      "flow_matching_loss": 0.015715988352894783,
      "grad_norm": 0.10594681650400162,
      "learning_rate": 3.930370294664071e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009333116526249796,
      "selected_pairs": 26597.9,
      "state_query_norm": 41.87791748046875,
      "state_relation_loss": 7.5166597525822e-05,
      "step": 26380,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0009333116526249796,
      "weighted_relation_loss": 0.0009333116526249796
    },
    {
      "action_query_norm": 39.63604545593262,
      "action_relation_loss": 0.0009129472193308175,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.686623012160897,
      "flow_matching_loss": 0.014910059701651334,
      "grad_norm": 0.1386241465806961,
      "learning_rate": 3.908978518361938e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009871870919596405,
      "selected_pairs": 27296.2,
      "state_query_norm": 41.86071128845215,
      "state_relation_loss": 7.423986426147167e-05,
      "step": 26390,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0009871870919596405,
      "weighted_relation_loss": 0.0009871870919596405
    },
    {
      "action_query_norm": 39.64053382873535,
      "action_relation_loss": 0.0008568235556595027,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.69597754911132,
      "flow_matching_loss": 0.016865567583590747,
      "grad_norm": 0.13552334904670715,
      "learning_rate": 3.887642746997017e-06,
      "loss": 0.0178,
      "relation_loss": 0.0009291660448070615,
      "selected_pairs": 26890.6,
      "state_query_norm": 41.8894718170166,
      "state_relation_loss": 7.234249715111218e-05,
      "step": 26400,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009291660448070615,
      "weighted_relation_loss": 0.0009291660448070615
    },
    {
      "action_query_norm": 39.64152793884277,
      "action_relation_loss": 0.0009007192857097834,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.70533208606174,
      "flow_matching_loss": 0.016787741985172033,
      "grad_norm": 0.14047761261463165,
      "learning_rate": 3.866363006494255e-06,
      "loss": 0.0178,
      "relation_loss": 0.000975811161333695,
      "selected_pairs": 26556.2,
      "state_query_norm": 41.89957504272461,
      "state_relation_loss": 7.509187707910314e-05,
      "step": 26410,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.000975811161333695,
      "weighted_relation_loss": 0.000975811161333695
    },
    {
      "action_query_norm": 39.62889404296875,
      "action_relation_loss": 0.0009223563596606254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.71468662301216,
      "flow_matching_loss": 0.016630273871123792,
      "grad_norm": 0.14975497126579285,
      "learning_rate": 3.8451393227105725e-06,
      "loss": 0.0176,
      "relation_loss": 0.0009965064353309571,
      "selected_pairs": 27245.9,
      "state_query_norm": 41.892158126831056,
      "state_relation_loss": 7.4150077853119e-05,
      "step": 26420,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009965064353309571,
      "weighted_relation_loss": 0.0009965064353309571
    },
    {
      "action_query_norm": 39.674536514282224,
      "action_relation_loss": 0.0008848414581734687,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.72404115996258,
      "flow_matching_loss": 0.015522512421011925,
      "grad_norm": 0.09730160981416702,
      "learning_rate": 3.823971721434727e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009603614569641649,
      "selected_pairs": 26847.8,
      "state_query_norm": 41.864018249511716,
      "state_relation_loss": 7.552000461146236e-05,
      "step": 26430,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009603614569641649,
      "weighted_relation_loss": 0.0009603614569641649
    },
    {
      "action_query_norm": 39.64494094848633,
      "action_relation_loss": 0.0008877215092070401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.733395696913004,
      "flow_matching_loss": 0.015238524880260229,
      "grad_norm": 0.06760982424020767,
      "learning_rate": 3.8028602283873505e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009654023102484644,
      "selected_pairs": 26970.7,
      "state_query_norm": 41.86789512634277,
      "state_relation_loss": 7.768079522065818e-05,
      "step": 26440,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009654023102484644,
      "weighted_relation_loss": 0.0009654023102484644
    },
    {
      "action_query_norm": 39.6745418548584,
      "action_relation_loss": 0.000963884691009298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.742750233863426,
      "flow_matching_loss": 0.016990127600729467,
      "grad_norm": 0.10343816876411438,
      "learning_rate": 3.781804869220912e-06,
      "loss": 0.018,
      "relation_loss": 0.0010395081480965018,
      "selected_pairs": 26505.4,
      "state_query_norm": 41.89962654113769,
      "state_relation_loss": 7.562344617326744e-05,
      "step": 26450,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0010395081480965018,
      "weighted_relation_loss": 0.0010395081480965018
    },
    {
      "action_query_norm": 39.693871688842776,
      "action_relation_loss": 0.0009748719341587276,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.752104770813844,
      "flow_matching_loss": 0.015528245642781258,
      "grad_norm": 0.2775585651397705,
      "learning_rate": 3.760805669519646e-06,
      "loss": 0.0166,
      "relation_loss": 0.0010495693306438625,
      "selected_pairs": 26966.3,
      "state_query_norm": 41.88522605895996,
      "state_relation_loss": 7.469738266081549e-05,
      "step": 26460,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0010495693306438625,
      "weighted_relation_loss": 0.0010495693306438625
    },
    {
      "action_query_norm": 39.579816818237305,
      "action_relation_loss": 0.0009506957838311791,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.761459307764266,
      "flow_matching_loss": 0.0165098880417645,
      "grad_norm": 0.09173984080553055,
      "learning_rate": 3.7398626547995587e-06,
      "loss": 0.0175,
      "relation_loss": 0.0010227693594060839,
      "selected_pairs": 26209.0,
      "state_query_norm": 41.88097381591797,
      "state_relation_loss": 7.207357048173435e-05,
      "step": 26470,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0010227693594060839,
      "weighted_relation_loss": 0.0010227693594060839
    },
    {
      "action_query_norm": 39.62801628112793,
      "action_relation_loss": 0.0009204393834806978,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.770813844714688,
      "flow_matching_loss": 0.014862778317183255,
      "grad_norm": 0.0813010185956955,
      "learning_rate": 3.7189758505083995e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009919829957652838,
      "selected_pairs": 27201.9,
      "state_query_norm": 41.86747398376465,
      "state_relation_loss": 7.154359991545789e-05,
      "step": 26480,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009919829957652838,
      "weighted_relation_loss": 0.0009919829957652838
    },
    {
      "action_query_norm": 39.61808052062988,
      "action_relation_loss": 0.0009185533213894814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.780168381665106,
      "flow_matching_loss": 0.0168478443287313,
      "grad_norm": 0.09147733449935913,
      "learning_rate": 3.6981452820255837e-06,
      "loss": 0.0178,
      "relation_loss": 0.000991474074544385,
      "selected_pairs": 26734.9,
      "state_query_norm": 41.85606422424316,
      "state_relation_loss": 7.292075897566975e-05,
      "step": 26490,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.000991474074544385,
      "weighted_relation_loss": 0.000991474074544385
    },
    {
      "action_query_norm": 39.635884857177736,
      "action_relation_loss": 0.0009013767878059298,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.78952291861553,
      "flow_matching_loss": 0.01625341260805726,
      "grad_norm": 0.18466702103614807,
      "learning_rate": 3.6773709746622363e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009771408687811345,
      "selected_pairs": 26624.0,
      "state_query_norm": 41.87025489807129,
      "state_relation_loss": 7.57640867959708e-05,
      "step": 26500,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009771408687811345,
      "weighted_relation_loss": 0.0009771408687811345
    },
    {
      "action_query_norm": 39.626820755004886,
      "action_relation_loss": 0.0008928596682380885,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.79887745556595,
      "flow_matching_loss": 0.015010589640587569,
      "grad_norm": 0.09104015678167343,
      "learning_rate": 3.6566529536610995e-06,
      "loss": 0.016,
      "relation_loss": 0.0009637141483835876,
      "selected_pairs": 26500.9,
      "state_query_norm": 41.88348388671875,
      "state_relation_loss": 7.085449033183977e-05,
      "step": 26510,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009637141483835876,
      "weighted_relation_loss": 0.0009637141483835876
    },
    {
      "action_query_norm": 39.662906646728516,
      "action_relation_loss": 0.0008854836691170931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.80823199251637,
      "flow_matching_loss": 0.014466733392328023,
      "grad_norm": 0.10737787187099457,
      "learning_rate": 3.635991244196513e-06,
      "loss": 0.0154,
      "relation_loss": 0.0009598117845598608,
      "selected_pairs": 26987.8,
      "state_query_norm": 41.88660202026367,
      "state_relation_loss": 7.432811544276774e-05,
      "step": 26520,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009598117845598608,
      "weighted_relation_loss": 0.0009598117845598608
    },
    {
      "action_query_norm": 39.63901252746582,
      "action_relation_loss": 0.0009135630098171532,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.81758652946679,
      "flow_matching_loss": 0.01861895713955164,
      "grad_norm": 0.10467588901519775,
      "learning_rate": 3.615385871374405e-06,
      "loss": 0.0196,
      "relation_loss": 0.0009829101851209997,
      "selected_pairs": 26823.8,
      "state_query_norm": 41.89775924682617,
      "state_relation_loss": 6.934717239346356e-05,
      "step": 26530,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009829101851209997,
      "weighted_relation_loss": 0.0009829101851209997
    },
    {
      "action_query_norm": 39.667897415161136,
      "action_relation_loss": 0.0009255502955056727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.826941066417213,
      "flow_matching_loss": 0.017526093870401382,
      "grad_norm": 0.09814257174730301,
      "learning_rate": 3.5948368602322367e-06,
      "loss": 0.0185,
      "relation_loss": 0.0009978018642868847,
      "selected_pairs": 27078.5,
      "state_query_norm": 41.888078689575195,
      "state_relation_loss": 7.225157314678654e-05,
      "step": 26540,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0009978018642868847,
      "weighted_relation_loss": 0.0009978018642868847
    },
    {
      "action_query_norm": 39.684003829956055,
      "action_relation_loss": 0.0008825593104120344,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.836295603367635,
      "flow_matching_loss": 0.017084008175879718,
      "grad_norm": 0.0950678363442421,
      "learning_rate": 3.5743442357390187e-06,
      "loss": 0.018,
      "relation_loss": 0.0009544408007059246,
      "selected_pairs": 26732.4,
      "state_query_norm": 41.88298416137695,
      "state_relation_loss": 7.18814928404754e-05,
      "step": 26550,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009544408007059246,
      "weighted_relation_loss": 0.0009544408007059246
    },
    {
      "action_query_norm": 39.62660942077637,
      "action_relation_loss": 0.0009041894401889295,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.845650140318053,
      "flow_matching_loss": 0.016121591813862322,
      "grad_norm": 0.12108322978019714,
      "learning_rate": 3.5539080227951937e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009774712612852453,
      "selected_pairs": 27386.1,
      "state_query_norm": 41.87664985656738,
      "state_relation_loss": 7.328180072363466e-05,
      "step": 26560,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0009774712612852453,
      "weighted_relation_loss": 0.0009774712612852453
    },
    {
      "action_query_norm": 39.62976226806641,
      "action_relation_loss": 0.0009164384624455124,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.855004677268475,
      "flow_matching_loss": 0.015333004016429185,
      "grad_norm": 0.1555783897638321,
      "learning_rate": 3.533528246232709e-06,
      "loss": 0.0163,
      "relation_loss": 0.0009860316466074436,
      "selected_pairs": 26621.6,
      "state_query_norm": 41.892734146118165,
      "state_relation_loss": 6.959317797736731e-05,
      "step": 26570,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009860316466074436,
      "weighted_relation_loss": 0.0009860316466074436
    },
    {
      "action_query_norm": 39.62796173095703,
      "action_relation_loss": 0.0009345977916382254,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.864359214218897,
      "flow_matching_loss": 0.013997543603181839,
      "grad_norm": 0.13208851218223572,
      "learning_rate": 3.5132049308149117e-06,
      "loss": 0.015,
      "relation_loss": 0.001006208430044353,
      "selected_pairs": 27155.4,
      "state_query_norm": 41.889534378051756,
      "state_relation_loss": 7.161066023400053e-05,
      "step": 26580,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.001006208430044353,
      "weighted_relation_loss": 0.001006208430044353
    },
    {
      "action_query_norm": 39.667254257202146,
      "action_relation_loss": 0.0008905451570171863,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.873713751169316,
      "flow_matching_loss": 0.015600574016571046,
      "grad_norm": 0.11501195281744003,
      "learning_rate": 3.4929381012365425e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009646615362726152,
      "selected_pairs": 26755.4,
      "state_query_norm": 41.89408531188965,
      "state_relation_loss": 7.41163814382162e-05,
      "step": 26590,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009646615362726152,
      "weighted_relation_loss": 0.0009646615362726152
    },
    {
      "action_query_norm": 39.63995895385742,
      "action_relation_loss": 0.0008755947172176092,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.883068288119738,
      "flow_matching_loss": 0.015635400358587504,
      "grad_norm": 0.09641318768262863,
      "learning_rate": 3.472727782123697e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009515350160654634,
      "selected_pairs": 27486.0,
      "state_query_norm": 41.8782901763916,
      "state_relation_loss": 7.594029739266261e-05,
      "step": 26600,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0009515350160654634,
      "weighted_relation_loss": 0.0009515350160654634
    },
    {
      "action_query_norm": 39.65128746032715,
      "action_relation_loss": 0.0009704701835289598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.89242282507016,
      "flow_matching_loss": 0.015276941284537316,
      "grad_norm": 0.10688894242048264,
      "learning_rate": 3.4525739980338425e-06,
      "loss": 0.0163,
      "relation_loss": 0.0010413467884063722,
      "selected_pairs": 26988.9,
      "state_query_norm": 41.87557792663574,
      "state_relation_loss": 7.087661360856145e-05,
      "step": 26610,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0010413467884063722,
      "weighted_relation_loss": 0.0010413467884063722
    },
    {
      "action_query_norm": 39.61163063049317,
      "action_relation_loss": 0.0009073043300304561,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.90177736202058,
      "flow_matching_loss": 0.016218674555420875,
      "grad_norm": 0.15417544543743134,
      "learning_rate": 3.432476773455712e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009767894574906678,
      "selected_pairs": 27061.7,
      "state_query_norm": 41.8619571685791,
      "state_relation_loss": 6.948513109819032e-05,
      "step": 26620,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009767894574906678,
      "weighted_relation_loss": 0.0009767894574906678
    },
    {
      "action_query_norm": 39.65698585510254,
      "action_relation_loss": 0.0008526358171366156,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.911131898971,
      "flow_matching_loss": 0.014957844279706478,
      "grad_norm": 0.0957469642162323,
      "learning_rate": 3.4124361328093378e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009282517421524972,
      "selected_pairs": 26921.1,
      "state_query_norm": 41.85503158569336,
      "state_relation_loss": 7.56159242882859e-05,
      "step": 26630,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009282517421524972,
      "weighted_relation_loss": 0.0009282517421524972
    },
    {
      "action_query_norm": 39.64010009765625,
      "action_relation_loss": 0.0009075729292817413,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.920486435921422,
      "flow_matching_loss": 0.017459241673350334,
      "grad_norm": 0.10330759733915329,
      "learning_rate": 3.392452100445975e-06,
      "loss": 0.0184,
      "relation_loss": 0.0009825169283431024,
      "selected_pairs": 26167.9,
      "state_query_norm": 41.87077941894531,
      "state_relation_loss": 7.494400269933977e-05,
      "step": 26640,
      "valid_action_samples": 229.2,
      "weighted_auxiliary_loss": 0.0009825169283431024,
      "weighted_relation_loss": 0.0009825169283431024
    },
    {
      "action_query_norm": 39.67158470153809,
      "action_relation_loss": 0.0009076482267118991,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.929840972871844,
      "flow_matching_loss": 0.01746275769546628,
      "grad_norm": 0.09948833286762238,
      "learning_rate": 3.3725247006481163e-06,
      "loss": 0.0184,
      "relation_loss": 0.0009797724662348628,
      "selected_pairs": 26805.8,
      "state_query_norm": 41.87612648010254,
      "state_relation_loss": 7.212424025055952e-05,
      "step": 26650,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009797724662348628,
      "weighted_relation_loss": 0.0009797724662348628
    },
    {
      "action_query_norm": 39.6728759765625,
      "action_relation_loss": 0.0009041184617672116,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.939195509822262,
      "flow_matching_loss": 0.01661729849874973,
      "grad_norm": 0.08960279077291489,
      "learning_rate": 3.3526539576294313e-06,
      "loss": 0.0176,
      "relation_loss": 0.000978225440485403,
      "selected_pairs": 26824.6,
      "state_query_norm": 41.87129669189453,
      "state_relation_loss": 7.410698526655324e-05,
      "step": 26660,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.000978225440485403,
      "weighted_relation_loss": 0.000978225440485403
    },
    {
      "action_query_norm": 39.620840835571286,
      "action_relation_loss": 0.0009067269507795572,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.948550046772684,
      "flow_matching_loss": 0.014280292857438325,
      "grad_norm": 0.08905833214521408,
      "learning_rate": 3.332839895534745e-06,
      "loss": 0.0153,
      "relation_loss": 0.0009776836261153222,
      "selected_pairs": 26559.0,
      "state_query_norm": 41.89307174682617,
      "state_relation_loss": 7.095667606336065e-05,
      "step": 26670,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0009776836261153222,
      "weighted_relation_loss": 0.0009776836261153222
    },
    {
      "action_query_norm": 39.657937622070314,
      "action_relation_loss": 0.0008603361318819224,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.957904583723106,
      "flow_matching_loss": 0.014845180511474609,
      "grad_norm": 0.08493076264858246,
      "learning_rate": 3.3130825384400155e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009302789811044931,
      "selected_pairs": 27244.6,
      "state_query_norm": 41.90447998046875,
      "state_relation_loss": 6.994283903623e-05,
      "step": 26680,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009302789811044931,
      "weighted_relation_loss": 0.0009302789811044931
    },
    {
      "action_query_norm": 39.67068748474121,
      "action_relation_loss": 0.0009061620919965207,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.96725912067353,
      "flow_matching_loss": 0.016867816727608443,
      "grad_norm": 0.13624078035354614,
      "learning_rate": 3.293381910352278e-06,
      "loss": 0.0178,
      "relation_loss": 0.0009787655901163817,
      "selected_pairs": 27035.9,
      "state_query_norm": 41.87907791137695,
      "state_relation_loss": 7.260351339937188e-05,
      "step": 26690,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009787655901163817,
      "weighted_relation_loss": 0.0009787655901163817
    },
    {
      "action_query_norm": 39.69578819274902,
      "action_relation_loss": 0.0008323817513883114,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.976613657623947,
      "flow_matching_loss": 0.016641771607100963,
      "grad_norm": 0.09162028878927231,
      "learning_rate": 3.273738035209678e-06,
      "loss": 0.0175,
      "relation_loss": 0.0009048223437275738,
      "selected_pairs": 26709.1,
      "state_query_norm": 41.85835723876953,
      "state_relation_loss": 7.244058579090052e-05,
      "step": 26700,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0009048223437275738,
      "weighted_relation_loss": 0.0009048223437275738
    },
    {
      "action_query_norm": 39.67858200073242,
      "action_relation_loss": 0.0008996942604426294,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.98596819457437,
      "flow_matching_loss": 0.016546560544520618,
      "grad_norm": 0.10807742178440094,
      "learning_rate": 3.254150936881356e-06,
      "loss": 0.0175,
      "relation_loss": 0.0009716256405226886,
      "selected_pairs": 26479.0,
      "state_query_norm": 41.883810424804686,
      "state_relation_loss": 7.193138226284646e-05,
      "step": 26710,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0009716256405226886,
      "weighted_relation_loss": 0.0009716256405226886
    },
    {
      "action_query_norm": 39.6477653503418,
      "action_relation_loss": 0.0008804543700534851,
      "auxiliary_loss_weight": 1.0,
      "epoch": 24.99532273152479,
      "flow_matching_loss": 0.015110890101641416,
      "grad_norm": 0.1785440891981125,
      "learning_rate": 3.234620639167496e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009531219373457133,
      "selected_pairs": 26271.5,
      "state_query_norm": 41.875718688964845,
      "state_relation_loss": 7.266756874741986e-05,
      "step": 26720,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0009531219373457133,
      "weighted_relation_loss": 0.0009531219373457133
    },
    {
      "action_query_norm": 39.66155014038086,
      "action_relation_loss": 0.0008767570194322616,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.00467726847521,
      "flow_matching_loss": 0.015231436491012574,
      "grad_norm": 0.08754817396402359,
      "learning_rate": 3.2151471657992482e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009545787121169269,
      "selected_pairs": 24424.8,
      "state_query_norm": 41.85646514892578,
      "state_relation_loss": 7.78216912294738e-05,
      "step": 26730,
      "valid_action_samples": 214.9,
      "weighted_auxiliary_loss": 0.0009545787121169269,
      "weighted_relation_loss": 0.0009545787121169269
    },
    {
      "action_query_norm": 39.66976699829102,
      "action_relation_loss": 0.0009082972013857215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.01403180542563,
      "flow_matching_loss": 0.01676566507667303,
      "grad_norm": 0.17227396368980408,
      "learning_rate": 3.195730540438718e-06,
      "loss": 0.0177,
      "relation_loss": 0.0009792527358513325,
      "selected_pairs": 26532.7,
      "state_query_norm": 41.8820743560791,
      "state_relation_loss": 7.095553155522794e-05,
      "step": 26740,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009792527358513325,
      "weighted_relation_loss": 0.0009792527358513325
    },
    {
      "action_query_norm": 39.635731887817386,
      "action_relation_loss": 0.0008856554166413843,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.023386342376053,
      "flow_matching_loss": 0.01817179685458541,
      "grad_norm": 0.14294065535068512,
      "learning_rate": 3.1763707866789336e-06,
      "loss": 0.0191,
      "relation_loss": 0.0009576460462994874,
      "selected_pairs": 26872.2,
      "state_query_norm": 41.881818771362305,
      "state_relation_loss": 7.199062820291146e-05,
      "step": 26750,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009576460462994874,
      "weighted_relation_loss": 0.0009576460462994874
    },
    {
      "action_query_norm": 39.672926330566405,
      "action_relation_loss": 0.00086411505471915,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.03274087932647,
      "flow_matching_loss": 0.015985178761184216,
      "grad_norm": 0.10434707254171371,
      "learning_rate": 3.1570679280438165e-06,
      "loss": 0.0169,
      "relation_loss": 0.000932595320045948,
      "selected_pairs": 27334.5,
      "state_query_norm": 41.89159240722656,
      "state_relation_loss": 6.848027187515982e-05,
      "step": 26760,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.000932595320045948,
      "weighted_relation_loss": 0.000932595320045948
    },
    {
      "action_query_norm": 39.636404037475586,
      "action_relation_loss": 0.000894047471228987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.042095416276894,
      "flow_matching_loss": 0.015429922100156545,
      "grad_norm": 0.08965949714183807,
      "learning_rate": 3.1378219879881767e-06,
      "loss": 0.0164,
      "relation_loss": 0.0009669731487520039,
      "selected_pairs": 26759.2,
      "state_query_norm": 41.892121124267575,
      "state_relation_loss": 7.292567752301694e-05,
      "step": 26770,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009669731487520039,
      "weighted_relation_loss": 0.0009669731487520039
    },
    {
      "action_query_norm": 39.674739074707034,
      "action_relation_loss": 0.0008988177287392318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.051449953227316,
      "flow_matching_loss": 0.01680777734145522,
      "grad_norm": 0.10252218693494797,
      "learning_rate": 3.11863298989763e-06,
      "loss": 0.0178,
      "relation_loss": 0.0009708246914669871,
      "selected_pairs": 27291.6,
      "state_query_norm": 41.857889556884764,
      "state_relation_loss": 7.200696418294684e-05,
      "step": 26780,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0009708246914669871,
      "weighted_relation_loss": 0.0009708246914669871
    },
    {
      "action_query_norm": 39.72322158813476,
      "action_relation_loss": 0.0008745296625420451,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.060804490177738,
      "flow_matching_loss": 0.015183998178690672,
      "grad_norm": 0.08947357535362244,
      "learning_rate": 3.0995009570886303e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009458144952077419,
      "selected_pairs": 27340.9,
      "state_query_norm": 41.893134689331056,
      "state_relation_loss": 7.128482830012217e-05,
      "step": 26790,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009458144952077419,
      "weighted_relation_loss": 0.0009458144952077419
    },
    {
      "action_query_norm": 39.64872398376465,
      "action_relation_loss": 0.0009031562658492476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.070159027128156,
      "flow_matching_loss": 0.01489031668752432,
      "grad_norm": 0.10361263900995255,
      "learning_rate": 3.0804259128083936e-06,
      "loss": 0.0159,
      "relation_loss": 0.000979761709459126,
      "selected_pairs": 27036.1,
      "state_query_norm": 41.87743873596192,
      "state_relation_loss": 7.660544433747418e-05,
      "step": 26800,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.000979761709459126,
      "weighted_relation_loss": 0.000979761709459126
    },
    {
      "action_query_norm": 39.66851387023926,
      "action_relation_loss": 0.0008662171196192503,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.079513564078578,
      "flow_matching_loss": 0.015887696482241153,
      "grad_norm": 0.1585412174463272,
      "learning_rate": 3.06140788023489e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009390266088303179,
      "selected_pairs": 27618.5,
      "state_query_norm": 41.882158279418945,
      "state_relation_loss": 7.280948557308875e-05,
      "step": 26810,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0009390266088303179,
      "weighted_relation_loss": 0.0009390266088303179
    },
    {
      "action_query_norm": 39.63195610046387,
      "action_relation_loss": 0.0009304075967520476,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.088868101029,
      "flow_matching_loss": 0.015803261194378138,
      "grad_norm": 0.11090075224637985,
      "learning_rate": 3.042446882476846e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010027514596004038,
      "selected_pairs": 27602.7,
      "state_query_norm": 41.88927116394043,
      "state_relation_loss": 7.234386939671821e-05,
      "step": 26820,
      "valid_action_samples": 235.4,
      "weighted_auxiliary_loss": 0.0010027514596004038,
      "weighted_relation_loss": 0.0010027514596004038
    },
    {
      "action_query_norm": 39.598168182373044,
      "action_relation_loss": 0.0008762929472140967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.09822263797942,
      "flow_matching_loss": 0.015683733951300384,
      "grad_norm": 0.09333687275648117,
      "learning_rate": 3.023542942573643e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009473476151470095,
      "selected_pairs": 26807.6,
      "state_query_norm": 41.846276092529294,
      "state_relation_loss": 7.105467811925336e-05,
      "step": 26830,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009473476151470095,
      "weighted_relation_loss": 0.0009473476151470095
    },
    {
      "action_query_norm": 39.675053405761716,
      "action_relation_loss": 0.0009077587921638042,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.10757717492984,
      "flow_matching_loss": 0.014762382954359055,
      "grad_norm": 0.09098543226718903,
      "learning_rate": 3.004696083495351e-06,
      "loss": 0.0157,
      "relation_loss": 0.0009782740846276284,
      "selected_pairs": 27126.7,
      "state_query_norm": 41.88989143371582,
      "state_relation_loss": 7.051529028103686e-05,
      "step": 26840,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009782740846276284,
      "weighted_relation_loss": 0.0009782740846276284
    },
    {
      "action_query_norm": 39.63624267578125,
      "action_relation_loss": 0.0009096520545426757,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.116931711880262,
      "flow_matching_loss": 0.016682927776128054,
      "grad_norm": 0.11164301633834839,
      "learning_rate": 2.98590632814269e-06,
      "loss": 0.0177,
      "relation_loss": 0.0009803646651562303,
      "selected_pairs": 27054.0,
      "state_query_norm": 41.86977958679199,
      "state_relation_loss": 7.071259788062889e-05,
      "step": 26850,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009803646651562303,
      "weighted_relation_loss": 0.0009803646651562303
    },
    {
      "action_query_norm": 39.62252616882324,
      "action_relation_loss": 0.0008820203540381045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.126286248830684,
      "flow_matching_loss": 0.016625765711069107,
      "grad_norm": 0.1166820079088211,
      "learning_rate": 2.9671736993469702e-06,
      "loss": 0.0176,
      "relation_loss": 0.0009561201964970678,
      "selected_pairs": 26206.7,
      "state_query_norm": 41.86268920898438,
      "state_relation_loss": 7.409984536934643e-05,
      "step": 26860,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0009561201964970678,
      "weighted_relation_loss": 0.0009561201964970678
    },
    {
      "action_query_norm": 39.66965560913086,
      "action_relation_loss": 0.0008639160136226564,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.135640785781103,
      "flow_matching_loss": 0.014878455363214017,
      "grad_norm": 0.10472092777490616,
      "learning_rate": 2.9484982198701216e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009362732118461281,
      "selected_pairs": 26733.7,
      "state_query_norm": 41.87066650390625,
      "state_relation_loss": 7.235720404423774e-05,
      "step": 26870,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009362732118461281,
      "weighted_relation_loss": 0.0009362732118461281
    },
    {
      "action_query_norm": 39.59205017089844,
      "action_relation_loss": 0.0009115985652897507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.144995322731525,
      "flow_matching_loss": 0.015622481610625983,
      "grad_norm": 0.12030352652072906,
      "learning_rate": 2.929879912404604e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009829982533119618,
      "selected_pairs": 27156.1,
      "state_query_norm": 41.873312377929686,
      "state_relation_loss": 7.139970184653067e-05,
      "step": 26880,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0009829982533119618,
      "weighted_relation_loss": 0.0009829982533119618
    },
    {
      "action_query_norm": 39.7108341217041,
      "action_relation_loss": 0.0009040143573656678,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.154349859681947,
      "flow_matching_loss": 0.016121668461710214,
      "grad_norm": 0.11107154190540314,
      "learning_rate": 2.9113187995734147e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009749117481987924,
      "selected_pairs": 27372.3,
      "state_query_norm": 41.877095794677736,
      "state_relation_loss": 7.08973835571669e-05,
      "step": 26890,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009749117481987924,
      "weighted_relation_loss": 0.0009749117481987924
    },
    {
      "action_query_norm": 39.703239822387694,
      "action_relation_loss": 0.0008911298587918282,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.163704396632365,
      "flow_matching_loss": 0.01443872731178999,
      "grad_norm": 0.14949214458465576,
      "learning_rate": 2.8928149039300524e-06,
      "loss": 0.0154,
      "relation_loss": 0.0009689067141152918,
      "selected_pairs": 27436.6,
      "state_query_norm": 41.86292724609375,
      "state_relation_loss": 7.77768531406764e-05,
      "step": 26900,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0009689067141152918,
      "weighted_relation_loss": 0.0009689067141152918
    },
    {
      "action_query_norm": 39.628395080566406,
      "action_relation_loss": 0.0009105970675591379,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.173058933582787,
      "flow_matching_loss": 0.014902983512729406,
      "grad_norm": 0.11117589473724365,
      "learning_rate": 2.8743682479584975e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009796895203180612,
      "selected_pairs": 27062.2,
      "state_query_norm": 41.856004333496095,
      "state_relation_loss": 6.90924447553698e-05,
      "step": 26910,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009796895203180612,
      "weighted_relation_loss": 0.0009796895203180612
    },
    {
      "action_query_norm": 39.673302459716794,
      "action_relation_loss": 0.0008760081778746098,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.18241347053321,
      "flow_matching_loss": 0.016226501576602458,
      "grad_norm": 0.09282805770635605,
      "learning_rate": 2.8559788540731824e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009488634299486875,
      "selected_pairs": 26963.9,
      "state_query_norm": 41.87430953979492,
      "state_relation_loss": 7.285525498446077e-05,
      "step": 26920,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009488634299486875,
      "weighted_relation_loss": 0.0009488634299486875
    },
    {
      "action_query_norm": 39.664811325073245,
      "action_relation_loss": 0.0009074640343897045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.19176800748363,
      "flow_matching_loss": 0.016191073786467315,
      "grad_norm": 0.10703232139348984,
      "learning_rate": 2.8376467446189493e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009787838847842068,
      "selected_pairs": 26874.1,
      "state_query_norm": 41.86488304138184,
      "state_relation_loss": 7.131985112209804e-05,
      "step": 26930,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009787838847842068,
      "weighted_relation_loss": 0.0009787838847842068
    },
    {
      "action_query_norm": 39.652221298217775,
      "action_relation_loss": 0.0009301560057792813,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.20112254443405,
      "flow_matching_loss": 0.01520815771073103,
      "grad_norm": 0.10461783409118652,
      "learning_rate": 2.8193719418710406e-06,
      "loss": 0.0162,
      "relation_loss": 0.0010070154385175556,
      "selected_pairs": 26698.5,
      "state_query_norm": 41.85964393615723,
      "state_relation_loss": 7.685943564865738e-05,
      "step": 26940,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0010070154385175556,
      "weighted_relation_loss": 0.0010070154385175556
    },
    {
      "action_query_norm": 39.650915908813474,
      "action_relation_loss": 0.0008903648937121034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.21047708138447,
      "flow_matching_loss": 0.01599861737340689,
      "grad_norm": 0.09567037969827652,
      "learning_rate": 2.8011544680350667e-06,
      "loss": 0.017,
      "relation_loss": 0.0009632275206968189,
      "selected_pairs": 26916.6,
      "state_query_norm": 41.85034065246582,
      "state_relation_loss": 7.286263353307731e-05,
      "step": 26950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009632275206968189,
      "weighted_relation_loss": 0.0009632275206968189
    },
    {
      "action_query_norm": 39.64620933532715,
      "action_relation_loss": 0.0008826684963423759,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.219831618334894,
      "flow_matching_loss": 0.01578133963048458,
      "grad_norm": 0.09879422187805176,
      "learning_rate": 2.7829943452469753e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009537226636894047,
      "selected_pairs": 26252.4,
      "state_query_norm": 41.87533760070801,
      "state_relation_loss": 7.105416298145429e-05,
      "step": 26960,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0009537226636894047,
      "weighted_relation_loss": 0.0009537226636894047
    },
    {
      "action_query_norm": 39.66032676696777,
      "action_relation_loss": 0.0009404433483723551,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.229186155285312,
      "flow_matching_loss": 0.01488957116380334,
      "grad_norm": 0.1166117861866951,
      "learning_rate": 2.7648915955730213e-06,
      "loss": 0.0159,
      "relation_loss": 0.0010152757458854467,
      "selected_pairs": 27243.6,
      "state_query_norm": 41.86651420593262,
      "state_relation_loss": 7.48323836887721e-05,
      "step": 26970,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0010152757458854467,
      "weighted_relation_loss": 0.0010152757458854467
    },
    {
      "action_query_norm": 39.684579086303714,
      "action_relation_loss": 0.0008744737831875681,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.238540692235734,
      "flow_matching_loss": 0.01447580400854349,
      "grad_norm": 0.08738978952169418,
      "learning_rate": 2.746846241009765e-06,
      "loss": 0.0154,
      "relation_loss": 0.000945954822236672,
      "selected_pairs": 27429.0,
      "state_query_norm": 41.8771255493164,
      "state_relation_loss": 7.148104268708266e-05,
      "step": 26980,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.000945954822236672,
      "weighted_relation_loss": 0.000945954822236672
    },
    {
      "action_query_norm": 39.68146896362305,
      "action_relation_loss": 0.0008807344362139702,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.247895229186156,
      "flow_matching_loss": 0.016253172513097526,
      "grad_norm": 0.08410736918449402,
      "learning_rate": 2.7288583034839945e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009525625617243349,
      "selected_pairs": 26765.5,
      "state_query_norm": 41.87154426574707,
      "state_relation_loss": 7.182813205872663e-05,
      "step": 26990,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009525625617243349,
      "weighted_relation_loss": 0.0009525625617243349
    },
    {
      "action_query_norm": 39.70161819458008,
      "action_relation_loss": 0.0008306110044941306,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.257249766136574,
      "flow_matching_loss": 0.016872160974889992,
      "grad_norm": 0.13074266910552979,
      "learning_rate": 2.7109278048527752e-06,
      "loss": 0.0178,
      "relation_loss": 0.00090193283976987,
      "selected_pairs": 27078.4,
      "state_query_norm": 41.86245918273926,
      "state_relation_loss": 7.132182981877122e-05,
      "step": 27000,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00090193283976987,
      "weighted_relation_loss": 0.00090193283976987
    },
    {
      "action_query_norm": 39.62060356140137,
      "action_relation_loss": 0.0009129073820076883,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.266604303086996,
      "flow_matching_loss": 0.014967519417405128,
      "grad_norm": 0.10547448694705963,
      "learning_rate": 2.6930547669033413e-06,
      "loss": 0.016,
      "relation_loss": 0.000985588616458699,
      "selected_pairs": 26503.1,
      "state_query_norm": 41.86387062072754,
      "state_relation_loss": 7.268123990797904e-05,
      "step": 27010,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.000985588616458699,
      "weighted_relation_loss": 0.000985588616458699
    },
    {
      "action_query_norm": 39.66051177978515,
      "action_relation_loss": 0.0008654964214656502,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.27595884003742,
      "flow_matching_loss": 0.0162355937063694,
      "grad_norm": 0.09975717216730118,
      "learning_rate": 2.675239211353109e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009407988691236824,
      "selected_pairs": 26916.3,
      "state_query_norm": 41.8670711517334,
      "state_relation_loss": 7.530243128712754e-05,
      "step": 27020,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009407988691236824,
      "weighted_relation_loss": 0.0009407988691236824
    },
    {
      "action_query_norm": 39.69667778015137,
      "action_relation_loss": 0.0008598733344115317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.28531337698784,
      "flow_matching_loss": 0.015129115525633096,
      "grad_norm": 0.08540557324886322,
      "learning_rate": 2.6574811598496786e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009290678892284632,
      "selected_pairs": 26648.8,
      "state_query_norm": 41.87453079223633,
      "state_relation_loss": 6.919456063769758e-05,
      "step": 27030,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0009290678892284632,
      "weighted_relation_loss": 0.0009290678892284632
    },
    {
      "action_query_norm": 39.66443977355957,
      "action_relation_loss": 0.0009044169040862471,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.29466791393826,
      "flow_matching_loss": 0.016274064406752588,
      "grad_norm": 0.14387640357017517,
      "learning_rate": 2.6397806339707456e-06,
      "loss": 0.0173,
      "relation_loss": 0.0009760221000760794,
      "selected_pairs": 26339.6,
      "state_query_norm": 41.85658988952637,
      "state_relation_loss": 7.160519671742805e-05,
      "step": 27040,
      "valid_action_samples": 230.0,
      "weighted_auxiliary_loss": 0.0009760221000760794,
      "weighted_relation_loss": 0.0009760221000760794
    },
    {
      "action_query_norm": 39.63493690490723,
      "action_relation_loss": 0.0008873185608536005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.30402245088868,
      "flow_matching_loss": 0.014859911706298589,
      "grad_norm": 0.11071638017892838,
      "learning_rate": 2.6221376552241217e-06,
      "loss": 0.0158,
      "relation_loss": 0.000959739403333515,
      "selected_pairs": 26583.7,
      "state_query_norm": 41.86204452514649,
      "state_relation_loss": 7.242082647280767e-05,
      "step": 27050,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.000959739403333515,
      "weighted_relation_loss": 0.000959739403333515
    },
    {
      "action_query_norm": 39.62471809387207,
      "action_relation_loss": 0.0008719245437532663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.313376987839103,
      "flow_matching_loss": 0.01449693962931633,
      "grad_norm": 0.11409670859575272,
      "learning_rate": 2.604552245047681e-06,
      "loss": 0.0154,
      "relation_loss": 0.0009501790686044842,
      "selected_pairs": 26685.0,
      "state_query_norm": 41.837389373779295,
      "state_relation_loss": 7.825452557881362e-05,
      "step": 27060,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0009501790686044842,
      "weighted_relation_loss": 0.0009501790686044842
    },
    {
      "action_query_norm": 39.60242614746094,
      "action_relation_loss": 0.0008674164768308401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.32273152478952,
      "flow_matching_loss": 0.01571193803101778,
      "grad_norm": 0.09680254757404327,
      "learning_rate": 2.587024424809359e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009404790529515594,
      "selected_pairs": 26525.5,
      "state_query_norm": 41.874203491210935,
      "state_relation_loss": 7.30625783035066e-05,
      "step": 27070,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009404790529515594,
      "weighted_relation_loss": 0.0009404790529515594
    },
    {
      "action_query_norm": 39.63177375793457,
      "action_relation_loss": 0.0008788893581368029,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.332086061739943,
      "flow_matching_loss": 0.014825086947530508,
      "grad_norm": 0.09659114480018616,
      "learning_rate": 2.5695542158071186e-06,
      "loss": 0.0158,
      "relation_loss": 0.000951351597905159,
      "selected_pairs": 26923.0,
      "state_query_norm": 41.87208557128906,
      "state_relation_loss": 7.246223321999423e-05,
      "step": 27080,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.000951351597905159,
      "weighted_relation_loss": 0.000951351597905159
    },
    {
      "action_query_norm": 39.63889350891113,
      "action_relation_loss": 0.0008934522164054215,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.341440598690365,
      "flow_matching_loss": 0.015243793651461602,
      "grad_norm": 0.12511977553367615,
      "learning_rate": 2.5521416392689068e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009656674461439252,
      "selected_pairs": 26053.7,
      "state_query_norm": 41.84023590087891,
      "state_relation_loss": 7.221523483167403e-05,
      "step": 27090,
      "valid_action_samples": 228.7,
      "weighted_auxiliary_loss": 0.0009656674461439252,
      "weighted_relation_loss": 0.0009656674461439252
    },
    {
      "action_query_norm": 39.636482620239256,
      "action_relation_loss": 0.0008744109713006764,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.350795135640787,
      "flow_matching_loss": 0.01490480313077569,
      "grad_norm": 0.12344551831483841,
      "learning_rate": 2.5347867163526384e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009449353034142405,
      "selected_pairs": 26646.4,
      "state_query_norm": 41.85891456604004,
      "state_relation_loss": 7.052435030345805e-05,
      "step": 27100,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0009449353034142405,
      "weighted_relation_loss": 0.0009449353034142405
    },
    {
      "action_query_norm": 39.591927719116214,
      "action_relation_loss": 0.0009374800371006131,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.360149672591206,
      "flow_matching_loss": 0.015745549555867912,
      "grad_norm": 0.1281605064868927,
      "learning_rate": 2.517489468146189e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010116680758073926,
      "selected_pairs": 26331.1,
      "state_query_norm": 41.856183242797854,
      "state_relation_loss": 7.418803506880068e-05,
      "step": 27110,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0010116680758073926,
      "weighted_relation_loss": 0.0010116680758073926
    },
    {
      "action_query_norm": 39.67952651977539,
      "action_relation_loss": 0.0008690603543072938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.369504209541628,
      "flow_matching_loss": 0.015978868026286362,
      "grad_norm": 0.09804314374923706,
      "learning_rate": 2.500249915667341e-06,
      "loss": 0.0169,
      "relation_loss": 0.0009418723173439502,
      "selected_pairs": 27032.1,
      "state_query_norm": 41.882388305664065,
      "state_relation_loss": 7.281196376425214e-05,
      "step": 27120,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009418723173439502,
      "weighted_relation_loss": 0.0009418723173439502
    },
    {
      "action_query_norm": 39.68598861694336,
      "action_relation_loss": 0.0008520279603544623,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.37885874649205,
      "flow_matching_loss": 0.014847141969949008,
      "grad_norm": 0.09879640489816666,
      "learning_rate": 2.4830680798637817e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009270097885746509,
      "selected_pairs": 26896.3,
      "state_query_norm": 41.86321487426758,
      "state_relation_loss": 7.498181948903949e-05,
      "step": 27130,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009270097885746509,
      "weighted_relation_loss": 0.0009270097885746509
    },
    {
      "action_query_norm": 39.65447654724121,
      "action_relation_loss": 0.0008446457621175795,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.388213283442468,
      "flow_matching_loss": 0.016398245003074408,
      "grad_norm": 0.10420732200145721,
      "learning_rate": 2.465943981613056e-06,
      "loss": 0.0173,
      "relation_loss": 0.0009206237271428108,
      "selected_pairs": 27339.1,
      "state_query_norm": 41.85956649780273,
      "state_relation_loss": 7.597797084599734e-05,
      "step": 27140,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009206237271428108,
      "weighted_relation_loss": 0.0009206237271428108
    },
    {
      "action_query_norm": 39.63589363098144,
      "action_relation_loss": 0.0009097616537474095,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.39756782039289,
      "flow_matching_loss": 0.016306533198803663,
      "grad_norm": 0.07832937687635422,
      "learning_rate": 2.448877641722569e-06,
      "loss": 0.0173,
      "relation_loss": 0.000982873881002888,
      "selected_pairs": 27184.9,
      "state_query_norm": 41.85996627807617,
      "state_relation_loss": 7.311224071600008e-05,
      "step": 27150,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.000982873881002888,
      "weighted_relation_loss": 0.000982873881002888
    },
    {
      "action_query_norm": 39.620063400268556,
      "action_relation_loss": 0.0009005152911413461,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.406922357343312,
      "flow_matching_loss": 0.016319817584007978,
      "grad_norm": 0.08860436826944351,
      "learning_rate": 2.431869080929522e-06,
      "loss": 0.0173,
      "relation_loss": 0.0009681441239081323,
      "selected_pairs": 27335.9,
      "state_query_norm": 41.896328353881835,
      "state_relation_loss": 6.762882330804131e-05,
      "step": 27160,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009681441239081323,
      "weighted_relation_loss": 0.0009681441239081323
    },
    {
      "action_query_norm": 39.70271110534668,
      "action_relation_loss": 0.0008449080749414861,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.416276894293734,
      "flow_matching_loss": 0.015531440824270248,
      "grad_norm": 0.11172015219926834,
      "learning_rate": 2.4149183199009216e-06,
      "loss": 0.0164,
      "relation_loss": 0.0009183083719108253,
      "selected_pairs": 26875.0,
      "state_query_norm": 41.88652648925781,
      "state_relation_loss": 7.34002940589562e-05,
      "step": 27170,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009183083719108253,
      "weighted_relation_loss": 0.0009183083719108253
    },
    {
      "action_query_norm": 39.677804183959964,
      "action_relation_loss": 0.0008552788116503507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.425631431244152,
      "flow_matching_loss": 0.01516117425635457,
      "grad_norm": 0.10669843852519989,
      "learning_rate": 2.3980253792335427e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009276543278247118,
      "selected_pairs": 26785.0,
      "state_query_norm": 41.8777156829834,
      "state_relation_loss": 7.237550889840349e-05,
      "step": 27180,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0009276543278247118,
      "weighted_relation_loss": 0.0009276543278247118
    },
    {
      "action_query_norm": 39.63109855651855,
      "action_relation_loss": 0.0008859847381245345,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.434985968194574,
      "flow_matching_loss": 0.01513103088364005,
      "grad_norm": 0.10770182311534882,
      "learning_rate": 2.381190279453899e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009588059387169778,
      "selected_pairs": 26754.7,
      "state_query_norm": 41.86743659973145,
      "state_relation_loss": 7.282118749571965e-05,
      "step": 27190,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009588059387169778,
      "weighted_relation_loss": 0.0009588059387169778
    },
    {
      "action_query_norm": 39.59057960510254,
      "action_relation_loss": 0.0009075893205590547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.444340505144996,
      "flow_matching_loss": 0.015265236515551806,
      "grad_norm": 0.11849770694971085,
      "learning_rate": 2.364413041018232e-06,
      "loss": 0.0162,
      "relation_loss": 0.000980768323643133,
      "selected_pairs": 26662.0,
      "state_query_norm": 41.8578598022461,
      "state_relation_loss": 7.317901217902545e-05,
      "step": 27200,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.000980768323643133,
      "weighted_relation_loss": 0.000980768323643133
    },
    {
      "action_query_norm": 39.65291404724121,
      "action_relation_loss": 0.0009054139547515661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.453695042095415,
      "flow_matching_loss": 0.01563961962237954,
      "grad_norm": 0.07489535957574844,
      "learning_rate": 2.347693684312463e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009751463483553379,
      "selected_pairs": 27317.7,
      "state_query_norm": 41.870671081542966,
      "state_relation_loss": 6.973238378122914e-05,
      "step": 27210,
      "valid_action_samples": 234.2,
      "weighted_auxiliary_loss": 0.0009751463483553379,
      "weighted_relation_loss": 0.0009751463483553379
    },
    {
      "action_query_norm": 39.64082641601563,
      "action_relation_loss": 0.0009272228344343603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.463049579045837,
      "flow_matching_loss": 0.01566577944904566,
      "grad_norm": 0.17110523581504822,
      "learning_rate": 2.331032229652186e-06,
      "loss": 0.0167,
      "relation_loss": 0.000997875991743058,
      "selected_pairs": 26672.8,
      "state_query_norm": 41.86819267272949,
      "state_relation_loss": 7.065316167427227e-05,
      "step": 27220,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.000997875991743058,
      "weighted_relation_loss": 0.000997875991743058
    },
    {
      "action_query_norm": 39.67392768859863,
      "action_relation_loss": 0.0008624342910479755,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.47240411599626,
      "flow_matching_loss": 0.015261368732899428,
      "grad_norm": 0.11797086149454117,
      "learning_rate": 2.314428697282628e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009325773047748953,
      "selected_pairs": 26569.9,
      "state_query_norm": 41.87619171142578,
      "state_relation_loss": 7.014302391326055e-05,
      "step": 27230,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0009325773047748953,
      "weighted_relation_loss": 0.0009325773047748953
    },
    {
      "action_query_norm": 39.652821731567386,
      "action_relation_loss": 0.0008764160156715661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.48175865294668,
      "flow_matching_loss": 0.015349532198160886,
      "grad_norm": 0.07861553132534027,
      "learning_rate": 2.297883107378673e-06,
      "loss": 0.0163,
      "relation_loss": 0.0009480942448135465,
      "selected_pairs": 26504.0,
      "state_query_norm": 41.86081237792969,
      "state_relation_loss": 7.167823350755498e-05,
      "step": 27240,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009480942448135465,
      "weighted_relation_loss": 0.0009480942448135465
    },
    {
      "action_query_norm": 39.69519119262695,
      "action_relation_loss": 0.000853626528987661,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.4911131898971,
      "flow_matching_loss": 0.0151871919631958,
      "grad_norm": 0.10516238212585449,
      "learning_rate": 2.2813954800447513e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009239516104571522,
      "selected_pairs": 26848.6,
      "state_query_norm": 41.85221710205078,
      "state_relation_loss": 7.032508801785298e-05,
      "step": 27250,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009239516104571522,
      "weighted_relation_loss": 0.0009239516104571522
    },
    {
      "action_query_norm": 39.65634422302246,
      "action_relation_loss": 0.0009134530846495181,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.50046772684752,
      "flow_matching_loss": 0.01500543737784028,
      "grad_norm": 0.09587015956640244,
      "learning_rate": 2.2649658353148974e-06,
      "loss": 0.016,
      "relation_loss": 0.0009818730992265045,
      "selected_pairs": 27200.8,
      "state_query_norm": 41.87159652709961,
      "state_relation_loss": 6.842000511824154e-05,
      "step": 27260,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009818730992265045,
      "weighted_relation_loss": 0.0009818730992265045
    },
    {
      "action_query_norm": 39.694990921020505,
      "action_relation_loss": 0.0008649024064652621,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.509822263797943,
      "flow_matching_loss": 0.01576734883710742,
      "grad_norm": 0.09971209615468979,
      "learning_rate": 2.2485941931526645e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009405908058397472,
      "selected_pairs": 26897.3,
      "state_query_norm": 41.84257202148437,
      "state_relation_loss": 7.56884015572723e-05,
      "step": 27270,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009405908058397472,
      "weighted_relation_loss": 0.0009405908058397472
    },
    {
      "action_query_norm": 39.609021759033205,
      "action_relation_loss": 0.0008435948751866817,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.51917680074836,
      "flow_matching_loss": 0.015543664153665305,
      "grad_norm": 0.09790739417076111,
      "learning_rate": 2.232280573451151e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009170795616228134,
      "selected_pairs": 27030.8,
      "state_query_norm": 41.861336135864256,
      "state_relation_loss": 7.348468570853584e-05,
      "step": 27280,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009170795616228134,
      "weighted_relation_loss": 0.0009170795616228134
    },
    {
      "action_query_norm": 39.629740142822264,
      "action_relation_loss": 0.0008997193130198866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.528531337698784,
      "flow_matching_loss": 0.01642386568710208,
      "grad_norm": 0.0911327451467514,
      "learning_rate": 2.216024996032945e-06,
      "loss": 0.0174,
      "relation_loss": 0.0009706980898045003,
      "selected_pairs": 26465.1,
      "state_query_norm": 41.85860824584961,
      "state_relation_loss": 7.097877314663493e-05,
      "step": 27290,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009706980898045003,
      "weighted_relation_loss": 0.0009706980898045003
    },
    {
      "action_query_norm": 39.67556915283203,
      "action_relation_loss": 0.0009200059226714075,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.537885874649206,
      "flow_matching_loss": 0.015232015587389468,
      "grad_norm": 0.10417116433382034,
      "learning_rate": 2.1998274806501074e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009918140480294825,
      "selected_pairs": 27131.8,
      "state_query_norm": 41.882342147827146,
      "state_relation_loss": 7.18081253580749e-05,
      "step": 27300,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009918140480294825,
      "weighted_relation_loss": 0.0009918140480294825
    },
    {
      "action_query_norm": 39.676065826416014,
      "action_relation_loss": 0.0009512531745713204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.547240411599624,
      "flow_matching_loss": 0.01653039949014783,
      "grad_norm": 0.12729565799236298,
      "learning_rate": 2.183688046984139e-06,
      "loss": 0.0176,
      "relation_loss": 0.0010279839916620404,
      "selected_pairs": 26712.3,
      "state_query_norm": 41.84691276550293,
      "state_relation_loss": 7.673081054235809e-05,
      "step": 27310,
      "valid_action_samples": 231.6,
      "weighted_auxiliary_loss": 0.0010279839916620404,
      "weighted_relation_loss": 0.0010279839916620404
    },
    {
      "action_query_norm": 39.59800148010254,
      "action_relation_loss": 0.0009108429017942399,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.556594948550046,
      "flow_matching_loss": 0.015563104674220085,
      "grad_norm": 0.08977604657411575,
      "learning_rate": 2.167606714645981e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009823839529417456,
      "selected_pairs": 27203.6,
      "state_query_norm": 41.85697898864746,
      "state_relation_loss": 7.15410387783777e-05,
      "step": 27320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009823839529417456,
      "weighted_relation_loss": 0.0009823839529417456
    },
    {
      "action_query_norm": 39.63368644714355,
      "action_relation_loss": 0.0009472113568335771,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.565949485500468,
      "flow_matching_loss": 0.015935658756643534,
      "grad_norm": 0.09520276635885239,
      "learning_rate": 2.151583503175958e-06,
      "loss": 0.017,
      "relation_loss": 0.0010196468268986792,
      "selected_pairs": 26394.5,
      "state_query_norm": 41.845940017700194,
      "state_relation_loss": 7.243546424433589e-05,
      "step": 27330,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0010196468268986792,
      "weighted_relation_loss": 0.0010196468268986792
    },
    {
      "action_query_norm": 39.641251754760745,
      "action_relation_loss": 0.0008997399476356804,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.57530402245089,
      "flow_matching_loss": 0.014688229095190764,
      "grad_norm": 0.0894111841917038,
      "learning_rate": 2.1356184320437957e-06,
      "loss": 0.0157,
      "relation_loss": 0.0009711995196994394,
      "selected_pairs": 26857.3,
      "state_query_norm": 41.87835006713867,
      "state_relation_loss": 7.145957060856744e-05,
      "step": 27340,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009711995196994394,
      "weighted_relation_loss": 0.0009711995196994394
    },
    {
      "action_query_norm": 39.65560188293457,
      "action_relation_loss": 0.0008830546285025775,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.58465855940131,
      "flow_matching_loss": 0.014936944842338562,
      "grad_norm": 0.11431188881397247,
      "learning_rate": 2.119711520648554e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009588051412720233,
      "selected_pairs": 26735.5,
      "state_query_norm": 41.852850341796874,
      "state_relation_loss": 7.57505084038712e-05,
      "step": 27350,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009588051412720233,
      "weighted_relation_loss": 0.0009588051412720233
    },
    {
      "action_query_norm": 39.657711791992185,
      "action_relation_loss": 0.0009419203619472682,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.59401309635173,
      "flow_matching_loss": 0.015811200998723508,
      "grad_norm": 0.11635052412748337,
      "learning_rate": 2.103862788318628e-06,
      "loss": 0.0168,
      "relation_loss": 0.0010098945698700846,
      "selected_pairs": 26667.0,
      "state_query_norm": 41.87105255126953,
      "state_relation_loss": 6.797420574002899e-05,
      "step": 27360,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0010098945698700846,
      "weighted_relation_loss": 0.0010098945698700846
    },
    {
      "action_query_norm": 39.67958106994629,
      "action_relation_loss": 0.0009050215245224536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.603367633302152,
      "flow_matching_loss": 0.015931091643869875,
      "grad_norm": 0.09099660813808441,
      "learning_rate": 2.088072254311729e-06,
      "loss": 0.0169,
      "relation_loss": 0.0009802551881875842,
      "selected_pairs": 27383.4,
      "state_query_norm": 41.86248092651367,
      "state_relation_loss": 7.523365602537523e-05,
      "step": 27370,
      "valid_action_samples": 234.5,
      "weighted_auxiliary_loss": 0.0009802551881875842,
      "weighted_relation_loss": 0.0009802551881875842
    },
    {
      "action_query_norm": 39.66944541931152,
      "action_relation_loss": 0.0008848227909766138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.61272217025257,
      "flow_matching_loss": 0.014991457667201757,
      "grad_norm": 0.09033077210187912,
      "learning_rate": 2.0723399378148435e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009558122488670051,
      "selected_pairs": 27645.2,
      "state_query_norm": 41.86572456359863,
      "state_relation_loss": 7.098947462509386e-05,
      "step": 27380,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0009558122488670051,
      "weighted_relation_loss": 0.0009558122488670051
    },
    {
      "action_query_norm": 39.64140701293945,
      "action_relation_loss": 0.0008343969180714339,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.622076707202993,
      "flow_matching_loss": 0.015578180644661188,
      "grad_norm": 0.10999361425638199,
      "learning_rate": 2.0566658579442065e-06,
      "loss": 0.0165,
      "relation_loss": 0.000903679383918643,
      "selected_pairs": 27403.9,
      "state_query_norm": 41.86255569458008,
      "state_relation_loss": 6.928246839379426e-05,
      "step": 27390,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.000903679383918643,
      "weighted_relation_loss": 0.000903679383918643
    },
    {
      "action_query_norm": 39.650146865844725,
      "action_relation_loss": 0.0008784470381215214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.631431244153415,
      "flow_matching_loss": 0.01565159484744072,
      "grad_norm": 0.09087327122688293,
      "learning_rate": 2.0410500337453176e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009531685092952103,
      "selected_pairs": 26668.3,
      "state_query_norm": 41.845045471191405,
      "state_relation_loss": 7.472148499800823e-05,
      "step": 27400,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0009531685092952103,
      "weighted_relation_loss": 0.0009531685092952103
    },
    {
      "action_query_norm": 39.64921073913574,
      "action_relation_loss": 0.0008489838102832437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.640785781103837,
      "flow_matching_loss": 0.014871441200375557,
      "grad_norm": 0.08309406042098999,
      "learning_rate": 2.0254924841928645e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009176837804261595,
      "selected_pairs": 26539.0,
      "state_query_norm": 41.8598331451416,
      "state_relation_loss": 6.869997014291584e-05,
      "step": 27410,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009176837804261595,
      "weighted_relation_loss": 0.0009176837804261595
    },
    {
      "action_query_norm": 39.63262100219727,
      "action_relation_loss": 0.000878970930352807,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.650140318054255,
      "flow_matching_loss": 0.014432639349251986,
      "grad_norm": 0.08076729625463486,
      "learning_rate": 2.009993228190754e-06,
      "loss": 0.0154,
      "relation_loss": 0.0009479155123699456,
      "selected_pairs": 26538.6,
      "state_query_norm": 41.88066940307617,
      "state_relation_loss": 6.894458711030893e-05,
      "step": 27420,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009479155123699456,
      "weighted_relation_loss": 0.0009479155123699456
    },
    {
      "action_query_norm": 39.66122398376465,
      "action_relation_loss": 0.0008379022765439004,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.659494855004677,
      "flow_matching_loss": 0.015476338192820549,
      "grad_norm": 0.11476708948612213,
      "learning_rate": 1.9945522845720323e-06,
      "loss": 0.0164,
      "relation_loss": 0.00091223637573421,
      "selected_pairs": 25618.6,
      "state_query_norm": 41.848490142822264,
      "state_relation_loss": 7.433410828525667e-05,
      "step": 27430,
      "valid_action_samples": 226.8,
      "weighted_auxiliary_loss": 0.00091223637573421,
      "weighted_relation_loss": 0.00091223637573421
    },
    {
      "action_query_norm": 39.645969009399415,
      "action_relation_loss": 0.0009383212716784328,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.6688493919551,
      "flow_matching_loss": 0.015275403019040823,
      "grad_norm": 0.11779464036226273,
      "learning_rate": 1.9791696720988963e-06,
      "loss": 0.0163,
      "relation_loss": 0.001009344996418804,
      "selected_pairs": 27127.9,
      "state_query_norm": 41.87330017089844,
      "state_relation_loss": 7.10237323801266e-05,
      "step": 27440,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.001009344996418804,
      "weighted_relation_loss": 0.001009344996418804
    },
    {
      "action_query_norm": 39.595926666259764,
      "action_relation_loss": 0.0009243338368833065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.678203928905518,
      "flow_matching_loss": 0.0158309874124825,
      "grad_norm": 0.09237007051706314,
      "learning_rate": 1.9638454094626833e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009960134630091488,
      "selected_pairs": 26978.1,
      "state_query_norm": 41.87054634094238,
      "state_relation_loss": 7.16796268534381e-05,
      "step": 27450,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009960134630091488,
      "weighted_relation_loss": 0.0009960134630091488
    },
    {
      "action_query_norm": 39.64953422546387,
      "action_relation_loss": 0.0008700514794327318,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.68755846585594,
      "flow_matching_loss": 0.015229786187410355,
      "grad_norm": 0.09847939014434814,
      "learning_rate": 1.9485795152838104e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009424391260836274,
      "selected_pairs": 26769.2,
      "state_query_norm": 41.8762451171875,
      "state_relation_loss": 7.238764883368276e-05,
      "step": 27460,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009424391260836274,
      "weighted_relation_loss": 0.0009424391260836274
    },
    {
      "action_query_norm": 39.664567184448245,
      "action_relation_loss": 0.0008823135984130203,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.69691300280636,
      "flow_matching_loss": 0.014080640207976104,
      "grad_norm": 0.09309322386980057,
      "learning_rate": 1.933372008111778e-06,
      "loss": 0.015,
      "relation_loss": 0.0009530590323265642,
      "selected_pairs": 27462.2,
      "state_query_norm": 41.85639190673828,
      "state_relation_loss": 7.07454397343099e-05,
      "step": 27470,
      "valid_action_samples": 234.8,
      "weighted_auxiliary_loss": 0.0009530590323265642,
      "weighted_relation_loss": 0.0009530590323265642
    },
    {
      "action_query_norm": 39.624019622802734,
      "action_relation_loss": 0.0009205288370139897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.706267539756784,
      "flow_matching_loss": 0.015841183811426164,
      "grad_norm": 0.11046949028968811,
      "learning_rate": 1.918222906425143e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009948793100193142,
      "selected_pairs": 26966.5,
      "state_query_norm": 41.83461494445801,
      "state_relation_loss": 7.435048610204831e-05,
      "step": 27480,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009948793100193142,
      "weighted_relation_loss": 0.0009948793100193142
    },
    {
      "action_query_norm": 39.679673767089845,
      "action_relation_loss": 0.0008602423244155943,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.715622076707202,
      "flow_matching_loss": 0.01610702695325017,
      "grad_norm": 0.0833875834941864,
      "learning_rate": 1.903132228631488e-06,
      "loss": 0.017,
      "relation_loss": 0.0009341054712422193,
      "selected_pairs": 26807.4,
      "state_query_norm": 41.85378456115723,
      "state_relation_loss": 7.386315046460367e-05,
      "step": 27490,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009341054712422193,
      "weighted_relation_loss": 0.0009341054712422193
    },
    {
      "action_query_norm": 39.62633743286133,
      "action_relation_loss": 0.0008801022020634263,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.724976613657624,
      "flow_matching_loss": 0.016260913386940957,
      "grad_norm": 0.10682828724384308,
      "learning_rate": 1.8880999930674215e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009531303949188441,
      "selected_pairs": 26761.1,
      "state_query_norm": 41.847007751464844,
      "state_relation_loss": 7.302820013137535e-05,
      "step": 27500,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009531303949188441,
      "weighted_relation_loss": 0.0009531303949188441
    },
    {
      "action_query_norm": 39.59763069152832,
      "action_relation_loss": 0.0008928640629164875,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.734331150608046,
      "flow_matching_loss": 0.015385664440691471,
      "grad_norm": 0.08811097592115402,
      "learning_rate": 1.8731262179985166e-06,
      "loss": 0.0164,
      "relation_loss": 0.0009693590342067182,
      "selected_pairs": 27240.0,
      "state_query_norm": 41.84300231933594,
      "state_relation_loss": 7.649496255908161e-05,
      "step": 27510,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009693590342067182,
      "weighted_relation_loss": 0.0009693590342067182
    },
    {
      "action_query_norm": 39.63476982116699,
      "action_relation_loss": 0.0008868142555002123,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.743685687558465,
      "flow_matching_loss": 0.013812353648245335,
      "grad_norm": 0.0710466131567955,
      "learning_rate": 1.8582109216193244e-06,
      "loss": 0.0148,
      "relation_loss": 0.0009574519412126392,
      "selected_pairs": 26778.9,
      "state_query_norm": 41.85900230407715,
      "state_relation_loss": 7.063767407089472e-05,
      "step": 27520,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0009574519412126392,
      "weighted_relation_loss": 0.0009574519412126392
    },
    {
      "action_query_norm": 39.59937057495117,
      "action_relation_loss": 0.000870872673112899,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.753040224508887,
      "flow_matching_loss": 0.01604750584810972,
      "grad_norm": 0.11102977395057678,
      "learning_rate": 1.8433541220533368e-06,
      "loss": 0.017,
      "relation_loss": 0.0009399853763170541,
      "selected_pairs": 26755.4,
      "state_query_norm": 41.865814971923825,
      "state_relation_loss": 6.911268756084609e-05,
      "step": 27530,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009399853763170541,
      "weighted_relation_loss": 0.0009399853763170541
    },
    {
      "action_query_norm": 39.66381721496582,
      "action_relation_loss": 0.0008597702137194574,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.76239476145931,
      "flow_matching_loss": 0.015227428171783686,
      "grad_norm": 0.1338522732257843,
      "learning_rate": 1.8285558373529577e-06,
      "loss": 0.0162,
      "relation_loss": 0.0009342769510112703,
      "selected_pairs": 27057.7,
      "state_query_norm": 41.85731163024902,
      "state_relation_loss": 7.450673583662138e-05,
      "step": 27540,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009342769510112703,
      "weighted_relation_loss": 0.0009342769510112703
    },
    {
      "action_query_norm": 39.636803436279294,
      "action_relation_loss": 0.0008775088761467486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.77174929840973,
      "flow_matching_loss": 0.015162258315831423,
      "grad_norm": 0.09125950187444687,
      "learning_rate": 1.8138160854995145e-06,
      "loss": 0.0161,
      "relation_loss": 0.000947779807029292,
      "selected_pairs": 27013.4,
      "state_query_norm": 41.845244598388675,
      "state_relation_loss": 7.027092287898995e-05,
      "step": 27550,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.000947779807029292,
      "weighted_relation_loss": 0.000947779807029292
    },
    {
      "action_query_norm": 39.63600044250488,
      "action_relation_loss": 0.0008939493563957512,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.78110383536015,
      "flow_matching_loss": 0.016080754343420267,
      "grad_norm": 0.1073039323091507,
      "learning_rate": 1.7991348844031864e-06,
      "loss": 0.0171,
      "relation_loss": 0.0009693166997749358,
      "selected_pairs": 26875.3,
      "state_query_norm": 41.85712776184082,
      "state_relation_loss": 7.536734265158884e-05,
      "step": 27560,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009693166997749358,
      "weighted_relation_loss": 0.0009693166997749358
    },
    {
      "action_query_norm": 39.65760269165039,
      "action_relation_loss": 0.0008796849171631039,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.79045837231057,
      "flow_matching_loss": 0.014804805535823107,
      "grad_norm": 0.09061452001333237,
      "learning_rate": 1.78451225190302e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009531625604722649,
      "selected_pairs": 26593.5,
      "state_query_norm": 41.8589900970459,
      "state_relation_loss": 7.347764112637378e-05,
      "step": 27570,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0009531625604722649,
      "weighted_relation_loss": 0.0009531625604722649
    },
    {
      "action_query_norm": 39.63074150085449,
      "action_relation_loss": 0.0008544107549823821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.799812909260993,
      "flow_matching_loss": 0.014930321276187897,
      "grad_norm": 0.09854493290185928,
      "learning_rate": 1.7699482057668914e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009271881775930524,
      "selected_pairs": 26303.4,
      "state_query_norm": 41.86456108093262,
      "state_relation_loss": 7.277742624864914e-05,
      "step": 27580,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0009271881775930524,
      "weighted_relation_loss": 0.0009271881775930524
    },
    {
      "action_query_norm": 39.650032424926756,
      "action_relation_loss": 0.0008591135148890317,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.80916744621141,
      "flow_matching_loss": 0.015123195201158523,
      "grad_norm": 0.07982396334409714,
      "learning_rate": 1.7554427636914895e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009379981609527021,
      "selected_pairs": 26986.1,
      "state_query_norm": 41.848741912841795,
      "state_relation_loss": 7.888464315328748e-05,
      "step": 27590,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009379981609527021,
      "weighted_relation_loss": 0.0009379981609527021
    },
    {
      "action_query_norm": 39.644547653198245,
      "action_relation_loss": 0.0008610630640760065,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.818521983161833,
      "flow_matching_loss": 0.015835592336952687,
      "grad_norm": 0.07844182848930359,
      "learning_rate": 1.7409959433022871e-06,
      "loss": 0.0168,
      "relation_loss": 0.000930514675565064,
      "selected_pairs": 26508.6,
      "state_query_norm": 41.8557788848877,
      "state_relation_loss": 6.945161294424906e-05,
      "step": 27600,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.000930514675565064,
      "weighted_relation_loss": 0.000930514675565064
    },
    {
      "action_query_norm": 39.62813262939453,
      "action_relation_loss": 0.0008668927242979407,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.827876520112255,
      "flow_matching_loss": 0.01572551382705569,
      "grad_norm": 0.0957050696015358,
      "learning_rate": 1.7266077621535481e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009374840359669179,
      "selected_pairs": 27127.3,
      "state_query_norm": 41.86926460266113,
      "state_relation_loss": 7.059132476570085e-05,
      "step": 27610,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009374840359669179,
      "weighted_relation_loss": 0.0009374840359669179
    },
    {
      "action_query_norm": 39.64097213745117,
      "action_relation_loss": 0.0008653523691464216,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.837231057062674,
      "flow_matching_loss": 0.016092150006443262,
      "grad_norm": 0.1046275645494461,
      "learning_rate": 1.7122782377282597e-06,
      "loss": 0.017,
      "relation_loss": 0.0009388277132529766,
      "selected_pairs": 27129.8,
      "state_query_norm": 41.845429229736325,
      "state_relation_loss": 7.347533173742704e-05,
      "step": 27620,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009388277132529766,
      "weighted_relation_loss": 0.0009388277132529766
    },
    {
      "action_query_norm": 39.65033111572266,
      "action_relation_loss": 0.0008494809211697429,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.846585594013096,
      "flow_matching_loss": 0.01686357231810689,
      "grad_norm": 0.11251793056726456,
      "learning_rate": 1.6980073874381497e-06,
      "loss": 0.0178,
      "relation_loss": 0.0009232921700458973,
      "selected_pairs": 26411.4,
      "state_query_norm": 41.87282600402832,
      "state_relation_loss": 7.381126779364422e-05,
      "step": 27630,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0009232921700458973,
      "weighted_relation_loss": 0.0009232921700458973
    },
    {
      "action_query_norm": 39.6106761932373,
      "action_relation_loss": 0.0008825307770166547,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.855940130963518,
      "flow_matching_loss": 0.015554792154580355,
      "grad_norm": 0.09145613014698029,
      "learning_rate": 1.6837952286236413e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009548465837724507,
      "selected_pairs": 26317.5,
      "state_query_norm": 41.875094985961915,
      "state_relation_loss": 7.231580166262574e-05,
      "step": 27640,
      "valid_action_samples": 229.9,
      "weighted_auxiliary_loss": 0.0009548465837724507,
      "weighted_relation_loss": 0.0009548465837724507
    },
    {
      "action_query_norm": 39.63366622924805,
      "action_relation_loss": 0.0008772836707066744,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.86529466791394,
      "flow_matching_loss": 0.015171672496944667,
      "grad_norm": 0.10583841800689697,
      "learning_rate": 1.6696417785538488e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009469404001720249,
      "selected_pairs": 27108.6,
      "state_query_norm": 41.86784477233887,
      "state_relation_loss": 6.965672801015899e-05,
      "step": 27650,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009469404001720249,
      "weighted_relation_loss": 0.0009469404001720249
    },
    {
      "action_query_norm": 39.632316970825194,
      "action_relation_loss": 0.000929276691749692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.874649204864358,
      "flow_matching_loss": 0.01720346650108695,
      "grad_norm": 0.13874897360801697,
      "learning_rate": 1.6555470544265538e-06,
      "loss": 0.0182,
      "relation_loss": 0.0009994113352149725,
      "selected_pairs": 26638.3,
      "state_query_norm": 41.87184219360351,
      "state_relation_loss": 7.013463728071656e-05,
      "step": 27660,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0009994113352149725,
      "weighted_relation_loss": 0.0009994113352149725
    },
    {
      "action_query_norm": 39.66423721313477,
      "action_relation_loss": 0.00089947241358459,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.88400374181478,
      "flow_matching_loss": 0.015112551301717759,
      "grad_norm": 0.0914393961429596,
      "learning_rate": 1.6415110733681738e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009696025226730853,
      "selected_pairs": 26759.9,
      "state_query_norm": 41.872409439086915,
      "state_relation_loss": 7.013011309027206e-05,
      "step": 27670,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009696025226730853,
      "weighted_relation_loss": 0.0009696025226730853
    },
    {
      "action_query_norm": 39.64276809692383,
      "action_relation_loss": 0.0008363544708117843,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.893358278765202,
      "flow_matching_loss": 0.01460843812674284,
      "grad_norm": 0.11525722593069077,
      "learning_rate": 1.6275338524337436e-06,
      "loss": 0.0155,
      "relation_loss": 0.0009080503310542554,
      "selected_pairs": 27058.3,
      "state_query_norm": 41.86710548400879,
      "state_relation_loss": 7.16958638804499e-05,
      "step": 27680,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009080503310542554,
      "weighted_relation_loss": 0.0009080503310542554
    },
    {
      "action_query_norm": 39.675923919677736,
      "action_relation_loss": 0.000878742552595213,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.90271281571562,
      "flow_matching_loss": 0.016994745563715698,
      "grad_norm": 0.10822414606809616,
      "learning_rate": 1.6136154086069056e-06,
      "loss": 0.0179,
      "relation_loss": 0.0009545244451146573,
      "selected_pairs": 26735.9,
      "state_query_norm": 41.854943466186526,
      "state_relation_loss": 7.57819005229976e-05,
      "step": 27690,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009545244451146573,
      "weighted_relation_loss": 0.0009545244451146573
    },
    {
      "action_query_norm": 39.66478233337402,
      "action_relation_loss": 0.0008913610188756138,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.912067352666043,
      "flow_matching_loss": 0.015670129097998142,
      "grad_norm": 0.10163560509681702,
      "learning_rate": 1.5997557587998868e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009649583895225078,
      "selected_pairs": 26480.1,
      "state_query_norm": 41.83958511352539,
      "state_relation_loss": 7.359737101069186e-05,
      "step": 27700,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0009649583895225078,
      "weighted_relation_loss": 0.0009649583895225078
    },
    {
      "action_query_norm": 39.623261260986325,
      "action_relation_loss": 0.0008867593540344387,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.921421889616465,
      "flow_matching_loss": 0.014668061304837466,
      "grad_norm": 0.0888809859752655,
      "learning_rate": 1.5859549198534606e-06,
      "loss": 0.0156,
      "relation_loss": 0.0009552025876473635,
      "selected_pairs": 27104.8,
      "state_query_norm": 41.86666564941406,
      "state_relation_loss": 6.844324016128667e-05,
      "step": 27710,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009552025876473635,
      "weighted_relation_loss": 0.0009552025876473635
    },
    {
      "action_query_norm": 39.671471786499026,
      "action_relation_loss": 0.0008571277372539044,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.930776426566887,
      "flow_matching_loss": 0.017400580737739802,
      "grad_norm": 0.09813526272773743,
      "learning_rate": 1.572212908536963e-06,
      "loss": 0.0183,
      "relation_loss": 0.000929281156277284,
      "selected_pairs": 26603.9,
      "state_query_norm": 41.871245574951175,
      "state_relation_loss": 7.21534204785712e-05,
      "step": 27720,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.000929281156277284,
      "weighted_relation_loss": 0.000929281156277284
    },
    {
      "action_query_norm": 39.61890068054199,
      "action_relation_loss": 0.000878565973835066,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.940130963517305,
      "flow_matching_loss": 0.017087058071047066,
      "grad_norm": 0.15929186344146729,
      "learning_rate": 1.5585297415482202e-06,
      "loss": 0.018,
      "relation_loss": 0.0009502932836767286,
      "selected_pairs": 27251.4,
      "state_query_norm": 41.864806747436525,
      "state_relation_loss": 7.17273145710351e-05,
      "step": 27730,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009502932836767286,
      "weighted_relation_loss": 0.0009502932836767286
    },
    {
      "action_query_norm": 39.64575271606445,
      "action_relation_loss": 0.0008693408046383411,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.949485500467727,
      "flow_matching_loss": 0.015589412488043308,
      "grad_norm": 0.21668295562267303,
      "learning_rate": 1.5449054355135717e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009393284562975168,
      "selected_pairs": 27147.6,
      "state_query_norm": 41.88109855651855,
      "state_relation_loss": 6.99876545695588e-05,
      "step": 27740,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0009393284562975168,
      "weighted_relation_loss": 0.0009393284562975168
    },
    {
      "action_query_norm": 39.64314727783203,
      "action_relation_loss": 0.0008723663282580674,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.95884003741815,
      "flow_matching_loss": 0.014360692631453276,
      "grad_norm": 0.09921425580978394,
      "learning_rate": 1.5313400069878415e-06,
      "loss": 0.0153,
      "relation_loss": 0.0009405023360159248,
      "selected_pairs": 26940.9,
      "state_query_norm": 41.862810134887695,
      "state_relation_loss": 6.813600593886804e-05,
      "step": 27750,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009405023360159248,
      "weighted_relation_loss": 0.0009405023360159248
    },
    {
      "action_query_norm": 39.628662490844725,
      "action_relation_loss": 0.0008646920847240836,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.968194574368567,
      "flow_matching_loss": 0.015212933346629143,
      "grad_norm": 0.12480832636356354,
      "learning_rate": 1.5178334724542887e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009355812566354871,
      "selected_pairs": 26800.9,
      "state_query_norm": 41.86572036743164,
      "state_relation_loss": 7.088917518558447e-05,
      "step": 27760,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009355812566354871,
      "weighted_relation_loss": 0.0009355812566354871
    },
    {
      "action_query_norm": 39.658139038085935,
      "action_relation_loss": 0.0009185982635244727,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.97754911131899,
      "flow_matching_loss": 0.015230578929185867,
      "grad_norm": 0.07714258879423141,
      "learning_rate": 1.504385848324641e-06,
      "loss": 0.0162,
      "relation_loss": 0.000993308750912547,
      "selected_pairs": 26500.4,
      "state_query_norm": 41.85286636352539,
      "state_relation_loss": 7.471048957086168e-05,
      "step": 27770,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.000993308750912547,
      "weighted_relation_loss": 0.000993308750912547
    },
    {
      "action_query_norm": 39.59041404724121,
      "action_relation_loss": 0.0009219754254445433,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.98690364826941,
      "flow_matching_loss": 0.015897790249437094,
      "grad_norm": 0.0895533338189125,
      "learning_rate": 1.490997150939033e-06,
      "loss": 0.0169,
      "relation_loss": 0.000997581594856456,
      "selected_pairs": 26393.8,
      "state_query_norm": 41.85474548339844,
      "state_relation_loss": 7.560615485999734e-05,
      "step": 27780,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.000997581594856456,
      "weighted_relation_loss": 0.000997581594856456
    },
    {
      "action_query_norm": 39.66511268615723,
      "action_relation_loss": 0.0008643127046525478,
      "auxiliary_loss_weight": 1.0,
      "epoch": 25.996258185219833,
      "flow_matching_loss": 0.016111088823527096,
      "grad_norm": 0.11016852408647537,
      "learning_rate": 1.4776673965659793e-06,
      "loss": 0.017,
      "relation_loss": 0.0009339999814983457,
      "selected_pairs": 26226.8,
      "state_query_norm": 41.86539192199707,
      "state_relation_loss": 6.968727939238306e-05,
      "step": 27790,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0009339999814983457,
      "weighted_relation_loss": 0.0009339999814983457
    },
    {
      "action_query_norm": 39.671189880371095,
      "action_relation_loss": 0.0008307294745463877,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.00561272217025,
      "flow_matching_loss": 0.015901742223650216,
      "grad_norm": 0.18303467333316803,
      "learning_rate": 1.4643966014023957e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009026828338392079,
      "selected_pairs": 23525.5,
      "state_query_norm": 41.84772834777832,
      "state_relation_loss": 7.195335783762857e-05,
      "step": 27800,
      "valid_action_samples": 210.5,
      "weighted_auxiliary_loss": 0.0009026828338392079,
      "weighted_relation_loss": 0.0009026828338392079
    },
    {
      "action_query_norm": 39.641668701171874,
      "action_relation_loss": 0.0008543546253349632,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.014967259120674,
      "flow_matching_loss": 0.015909876395016908,
      "grad_norm": 0.10138076543807983,
      "learning_rate": 1.4511847815735503e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009284150029998273,
      "selected_pairs": 26433.4,
      "state_query_norm": 41.853329467773435,
      "state_relation_loss": 7.406036856991705e-05,
      "step": 27810,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0009284150029998273,
      "weighted_relation_loss": 0.0009284150029998273
    },
    {
      "action_query_norm": 39.65014762878418,
      "action_relation_loss": 0.0008514391898643226,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.024321796071096,
      "flow_matching_loss": 0.014551912806928158,
      "grad_norm": 0.09941010177135468,
      "learning_rate": 1.438031953133051e-06,
      "loss": 0.0155,
      "relation_loss": 0.0009211133234202862,
      "selected_pairs": 26899.3,
      "state_query_norm": 41.83898239135742,
      "state_relation_loss": 6.967413173697423e-05,
      "step": 27820,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009211133234202862,
      "weighted_relation_loss": 0.0009211133234202862
    },
    {
      "action_query_norm": 39.63155288696289,
      "action_relation_loss": 0.0009215996309649199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.033676333021514,
      "flow_matching_loss": 0.014775276742875576,
      "grad_norm": 0.1735236495733261,
      "learning_rate": 1.4249381320628197e-06,
      "loss": 0.0158,
      "relation_loss": 0.000990403932519257,
      "selected_pairs": 26736.4,
      "state_query_norm": 41.87614707946777,
      "state_relation_loss": 6.880430591991172e-05,
      "step": 27830,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.000990403932519257,
      "weighted_relation_loss": 0.000990403932519257
    },
    {
      "action_query_norm": 39.65272903442383,
      "action_relation_loss": 0.0008518761140294373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.043030869971936,
      "flow_matching_loss": 0.0160762133076787,
      "grad_norm": 0.08106189966201782,
      "learning_rate": 1.4119033342730902e-06,
      "loss": 0.017,
      "relation_loss": 0.000923769100336358,
      "selected_pairs": 26393.2,
      "state_query_norm": 41.856695556640624,
      "state_relation_loss": 7.189298630692065e-05,
      "step": 27840,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.000923769100336358,
      "weighted_relation_loss": 0.000923769100336358
    },
    {
      "action_query_norm": 39.67107963562012,
      "action_relation_loss": 0.0008820830145850778,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.052385406922358,
      "flow_matching_loss": 0.01572790751233697,
      "grad_norm": 0.07313233613967896,
      "learning_rate": 1.3989275756023656e-06,
      "loss": 0.0167,
      "relation_loss": 0.0009507169714197516,
      "selected_pairs": 26730.6,
      "state_query_norm": 41.86863708496094,
      "state_relation_loss": 6.863396847620607e-05,
      "step": 27850,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009507169714197516,
      "weighted_relation_loss": 0.0009507169714197516
    },
    {
      "action_query_norm": 39.63654747009277,
      "action_relation_loss": 0.0008568030782043934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.061739943872777,
      "flow_matching_loss": 0.01511557586491108,
      "grad_norm": 0.10427594929933548,
      "learning_rate": 1.386010871817417e-06,
      "loss": 0.016,
      "relation_loss": 0.0009286393644288183,
      "selected_pairs": 26758.3,
      "state_query_norm": 41.85126571655273,
      "state_relation_loss": 7.183627822087147e-05,
      "step": 27860,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009286393644288183,
      "weighted_relation_loss": 0.0009286393644288183
    },
    {
      "action_query_norm": 39.587542343139646,
      "action_relation_loss": 0.0008484998717904091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.0710944808232,
      "flow_matching_loss": 0.014786520786583424,
      "grad_norm": 0.12413239479064941,
      "learning_rate": 1.3731532386132616e-06,
      "loss": 0.0157,
      "relation_loss": 0.0009162176284007728,
      "selected_pairs": 27431.0,
      "state_query_norm": 41.85679130554199,
      "state_relation_loss": 6.771775879315101e-05,
      "step": 27870,
      "valid_action_samples": 234.7,
      "weighted_auxiliary_loss": 0.0009162176284007728,
      "weighted_relation_loss": 0.0009162176284007728
    },
    {
      "action_query_norm": 39.66048355102539,
      "action_relation_loss": 0.0008492009714245796,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.08044901777362,
      "flow_matching_loss": 0.017072334606200456,
      "grad_norm": 0.08408541232347488,
      "learning_rate": 1.360354691613136e-06,
      "loss": 0.018,
      "relation_loss": 0.0009173251746688038,
      "selected_pairs": 26596.1,
      "state_query_norm": 41.872418594360354,
      "state_relation_loss": 6.812420397181996e-05,
      "step": 27880,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0009173251746688038,
      "weighted_relation_loss": 0.0009173251746688038
    },
    {
      "action_query_norm": 39.617078399658205,
      "action_relation_loss": 0.000849212339380756,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.089803554724043,
      "flow_matching_loss": 0.014406866114586591,
      "grad_norm": 0.1510850340127945,
      "learning_rate": 1.3476152463684776e-06,
      "loss": 0.0153,
      "relation_loss": 0.0009196002210956067,
      "selected_pairs": 27646.8,
      "state_query_norm": 41.85567092895508,
      "state_relation_loss": 7.038788353384007e-05,
      "step": 27890,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.0009196002210956067,
      "weighted_relation_loss": 0.0009196002210956067
    },
    {
      "action_query_norm": 39.61057395935059,
      "action_relation_loss": 0.0009268367430195212,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.09915809167446,
      "flow_matching_loss": 0.01618626797571778,
      "grad_norm": 0.11414042860269547,
      "learning_rate": 1.3349349183589155e-06,
      "loss": 0.0172,
      "relation_loss": 0.0009944937017280608,
      "selected_pairs": 25914.5,
      "state_query_norm": 41.84916877746582,
      "state_relation_loss": 6.765696234651841e-05,
      "step": 27900,
      "valid_action_samples": 228.1,
      "weighted_auxiliary_loss": 0.0009944937017280608,
      "weighted_relation_loss": 0.0009944937017280608
    },
    {
      "action_query_norm": 39.65551986694336,
      "action_relation_loss": 0.0008603898459114134,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.108512628624883,
      "flow_matching_loss": 0.015998300258070232,
      "grad_norm": 0.08244314789772034,
      "learning_rate": 1.3223137229922356e-06,
      "loss": 0.0169,
      "relation_loss": 0.0009317194751929491,
      "selected_pairs": 26860.2,
      "state_query_norm": 41.85655670166015,
      "state_relation_loss": 7.132962782634422e-05,
      "step": 27910,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009317194751929491,
      "weighted_relation_loss": 0.0009317194751929491
    },
    {
      "action_query_norm": 39.65742568969726,
      "action_relation_loss": 0.00087785305804573,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.117867165575305,
      "flow_matching_loss": 0.01586518231779337,
      "grad_norm": 0.15784093737602234,
      "learning_rate": 1.3097516756043981e-06,
      "loss": 0.0168,
      "relation_loss": 0.0009488581621553749,
      "selected_pairs": 27218.2,
      "state_query_norm": 41.856984329223636,
      "state_relation_loss": 7.100511429598555e-05,
      "step": 27920,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0009488581621553749,
      "weighted_relation_loss": 0.0009488581621553749
    },
    {
      "action_query_norm": 39.65466842651367,
      "action_relation_loss": 0.0008358316321391612,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.127221702525723,
      "flow_matching_loss": 0.015420082956552505,
      "grad_norm": 0.09950180351734161,
      "learning_rate": 1.29724879145946e-06,
      "loss": 0.0163,
      "relation_loss": 0.0009094262379221619,
      "selected_pairs": 27204.3,
      "state_query_norm": 41.836819076538085,
      "state_relation_loss": 7.359461160376668e-05,
      "step": 27930,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009094262379221619,
      "weighted_relation_loss": 0.0009094262379221619
    },
    {
      "action_query_norm": 39.624995422363284,
      "action_relation_loss": 0.0008424563857261091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.136576239476145,
      "flow_matching_loss": 0.014700467884540557,
      "grad_norm": 0.10571230947971344,
      "learning_rate": 1.284805085749613e-06,
      "loss": 0.0156,
      "relation_loss": 0.0009123660507611931,
      "selected_pairs": 26619.8,
      "state_query_norm": 41.86155548095703,
      "state_relation_loss": 6.990966212470085e-05,
      "step": 27940,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009123660507611931,
      "weighted_relation_loss": 0.0009123660507611931
    },
    {
      "action_query_norm": 39.67131004333496,
      "action_relation_loss": 0.0008962326857727021,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.145930776426567,
      "flow_matching_loss": 0.015458303783088923,
      "grad_norm": 0.09730827063322067,
      "learning_rate": 1.2724205735951288e-06,
      "loss": 0.0164,
      "relation_loss": 0.0009658164344727993,
      "selected_pairs": 26919.6,
      "state_query_norm": 41.847573852539064,
      "state_relation_loss": 6.958373996894807e-05,
      "step": 27950,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009658164344727993,
      "weighted_relation_loss": 0.0009658164344727993
    },
    {
      "action_query_norm": 39.59475250244141,
      "action_relation_loss": 0.0008637998544145375,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.15528531337699,
      "flow_matching_loss": 0.015176588576287032,
      "grad_norm": 0.08301448822021484,
      "learning_rate": 1.2600952700443591e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009323568374384195,
      "selected_pairs": 26902.0,
      "state_query_norm": 41.86464233398438,
      "state_relation_loss": 6.855697647552006e-05,
      "step": 27960,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009323568374384195,
      "weighted_relation_loss": 0.0009323568374384195
    },
    {
      "action_query_norm": 39.6956069946289,
      "action_relation_loss": 0.0008345316979102791,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.164639850327408,
      "flow_matching_loss": 0.014527595881372691,
      "grad_norm": 0.13715435564517975,
      "learning_rate": 1.247829190073707e-06,
      "loss": 0.0154,
      "relation_loss": 0.0009067285864148289,
      "selected_pairs": 26850.0,
      "state_query_norm": 41.854491806030275,
      "state_relation_loss": 7.219688341137953e-05,
      "step": 27970,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009067285864148289,
      "weighted_relation_loss": 0.0009067285864148289
    },
    {
      "action_query_norm": 39.69060935974121,
      "action_relation_loss": 0.0008406914537772536,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.17399438727783,
      "flow_matching_loss": 0.014404419530183076,
      "grad_norm": 0.15358372032642365,
      "learning_rate": 1.2356223485876173e-06,
      "loss": 0.0153,
      "relation_loss": 0.0009130763064604253,
      "selected_pairs": 26825.6,
      "state_query_norm": 41.856774520874026,
      "state_relation_loss": 7.23848570487462e-05,
      "step": 27980,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009130763064604253,
      "weighted_relation_loss": 0.0009130763064604253
    },
    {
      "action_query_norm": 39.647560119628906,
      "action_relation_loss": 0.0008775901515036821,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.18334892422825,
      "flow_matching_loss": 0.014211690425872803,
      "grad_norm": 0.09495893120765686,
      "learning_rate": 1.2234747604185526e-06,
      "loss": 0.0152,
      "relation_loss": 0.0009497534018009901,
      "selected_pairs": 26924.3,
      "state_query_norm": 41.85004997253418,
      "state_relation_loss": 7.216325102490372e-05,
      "step": 27990,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009497534018009901,
      "weighted_relation_loss": 0.0009497534018009901
    },
    {
      "action_query_norm": 39.66967544555664,
      "action_relation_loss": 0.0008830831851810217,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.19270346117867,
      "flow_matching_loss": 0.015585193876177073,
      "grad_norm": 0.09034153819084167,
      "learning_rate": 1.2113864403269836e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009509807277936488,
      "selected_pairs": 26739.9,
      "state_query_norm": 41.87400741577149,
      "state_relation_loss": 6.789752624172251e-05,
      "step": 28000,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009509807277936488,
      "weighted_relation_loss": 0.0009509807277936488
    },
    {
      "action_query_norm": 39.64529151916504,
      "action_relation_loss": 0.0008814377710223198,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.202057998129092,
      "flow_matching_loss": 0.0166507875546813,
      "grad_norm": 0.11764965951442719,
      "learning_rate": 1.1993574030013555e-06,
      "loss": 0.0176,
      "relation_loss": 0.0009492719371337443,
      "selected_pairs": 26754.3,
      "state_query_norm": 41.85911750793457,
      "state_relation_loss": 6.783417411497794e-05,
      "step": 28010,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009492719371337443,
      "weighted_relation_loss": 0.0009492719371337443
    },
    {
      "action_query_norm": 39.640329360961914,
      "action_relation_loss": 0.0008595449733547866,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.211412535079514,
      "flow_matching_loss": 0.01556482519954443,
      "grad_norm": 0.08017983287572861,
      "learning_rate": 1.1873876630580816e-06,
      "loss": 0.0165,
      "relation_loss": 0.0009351032203994691,
      "selected_pairs": 26880.9,
      "state_query_norm": 41.846350860595706,
      "state_relation_loss": 7.555825941381044e-05,
      "step": 28020,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009351032203994691,
      "weighted_relation_loss": 0.0009351032203994691
    },
    {
      "action_query_norm": 39.61856460571289,
      "action_relation_loss": 0.0009100047522224486,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.220767072029936,
      "flow_matching_loss": 0.016118446178734303,
      "grad_norm": 0.08878940343856812,
      "learning_rate": 1.1754772350415278e-06,
      "loss": 0.0171,
      "relation_loss": 0.000981049076654017,
      "selected_pairs": 27549.8,
      "state_query_norm": 41.84383697509766,
      "state_relation_loss": 7.104432806954719e-05,
      "step": 28030,
      "valid_action_samples": 235.2,
      "weighted_auxiliary_loss": 0.000981049076654017,
      "weighted_relation_loss": 0.000981049076654017
    },
    {
      "action_query_norm": 39.63968391418457,
      "action_relation_loss": 0.0008695249620359391,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.230121608980355,
      "flow_matching_loss": 0.01592717319726944,
      "grad_norm": 0.08942247927188873,
      "learning_rate": 1.163626133423984e-06,
      "loss": 0.0169,
      "relation_loss": 0.0009397595946211368,
      "selected_pairs": 26825.9,
      "state_query_norm": 41.871531295776364,
      "state_relation_loss": 7.023463294899556e-05,
      "step": 28040,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009397595946211368,
      "weighted_relation_loss": 0.0009397595946211368
    },
    {
      "action_query_norm": 39.63047142028809,
      "action_relation_loss": 0.0008975798089522868,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.239476145930777,
      "flow_matching_loss": 0.014986842311918736,
      "grad_norm": 0.09622978419065475,
      "learning_rate": 1.1518343726056591e-06,
      "loss": 0.016,
      "relation_loss": 0.0009656812297180295,
      "selected_pairs": 26512.2,
      "state_query_norm": 41.856902313232425,
      "state_relation_loss": 6.8101410579402e-05,
      "step": 28050,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009656812297180295,
      "weighted_relation_loss": 0.0009656812297180295
    },
    {
      "action_query_norm": 39.66448783874512,
      "action_relation_loss": 0.0008604363712947816,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.2488306828812,
      "flow_matching_loss": 0.015189112536609173,
      "grad_norm": 0.09148899465799332,
      "learning_rate": 1.1401019669146474e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009352448105346411,
      "selected_pairs": 27363.8,
      "state_query_norm": 41.8690486907959,
      "state_relation_loss": 7.480843996745534e-05,
      "step": 28060,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009352448105346411,
      "weighted_relation_loss": 0.0009352448105346411
    },
    {
      "action_query_norm": 39.69604759216308,
      "action_relation_loss": 0.0008672711788676679,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.258185219831617,
      "flow_matching_loss": 0.01513543538749218,
      "grad_norm": 0.13560405373573303,
      "learning_rate": 1.128428930606934e-06,
      "loss": 0.0161,
      "relation_loss": 0.0009359629650134594,
      "selected_pairs": 27109.1,
      "state_query_norm": 41.87206192016602,
      "state_relation_loss": 6.869178760098294e-05,
      "step": 28070,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009359629650134594,
      "weighted_relation_loss": 0.0009359629650134594
    },
    {
      "action_query_norm": 39.6533390045166,
      "action_relation_loss": 0.0008426542684901505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.26753975678204,
      "flow_matching_loss": 0.014872606750577689,
      "grad_norm": 0.10542087256908417,
      "learning_rate": 1.1168152778663621e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009171813144348562,
      "selected_pairs": 27081.4,
      "state_query_norm": 41.861813354492185,
      "state_relation_loss": 7.452704958268441e-05,
      "step": 28080,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0009171813144348562,
      "weighted_relation_loss": 0.0009171813144348562
    },
    {
      "action_query_norm": 39.61381301879883,
      "action_relation_loss": 0.0008431705355178565,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.27689429373246,
      "flow_matching_loss": 0.015693057794123888,
      "grad_norm": 0.07800383120775223,
      "learning_rate": 1.10526102280461e-06,
      "loss": 0.0166,
      "relation_loss": 0.0009113254607655108,
      "selected_pairs": 27477.3,
      "state_query_norm": 41.871102523803714,
      "state_relation_loss": 6.815492088207975e-05,
      "step": 28090,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0009113254607655108,
      "weighted_relation_loss": 0.0009113254607655108
    },
    {
      "action_query_norm": 39.66087532043457,
      "action_relation_loss": 0.0008374777680728585,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.28624883068288,
      "flow_matching_loss": 0.014293713681399822,
      "grad_norm": 0.08786031603813171,
      "learning_rate": 1.0937661794611864e-06,
      "loss": 0.0152,
      "relation_loss": 0.000909712037537247,
      "selected_pairs": 27060.9,
      "state_query_norm": 41.84513473510742,
      "state_relation_loss": 7.223427601275035e-05,
      "step": 28100,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.000909712037537247,
      "weighted_relation_loss": 0.000909712037537247
    },
    {
      "action_query_norm": 39.622351837158206,
      "action_relation_loss": 0.0008731721027288586,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.2956033676333,
      "flow_matching_loss": 0.014851337950676679,
      "grad_norm": 0.10759598761796951,
      "learning_rate": 1.082330761803413e-06,
      "loss": 0.0158,
      "relation_loss": 0.0009459921333473176,
      "selected_pairs": 26294.4,
      "state_query_norm": 41.84781455993652,
      "state_relation_loss": 7.282002770807594e-05,
      "step": 28110,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0009459921333473176,
      "weighted_relation_loss": 0.0009459921333473176
    },
    {
      "action_query_norm": 39.677435302734374,
      "action_relation_loss": 0.0008178750227671117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.304957904583723,
      "flow_matching_loss": 0.014224985986948014,
      "grad_norm": 0.08979330211877823,
      "learning_rate": 1.0709547837263966e-06,
      "loss": 0.0151,
      "relation_loss": 0.0008921051456127315,
      "selected_pairs": 26966.2,
      "state_query_norm": 41.84813232421875,
      "state_relation_loss": 7.423011920764111e-05,
      "step": 28120,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0008921051456127315,
      "weighted_relation_loss": 0.0008921051456127315
    },
    {
      "action_query_norm": 39.65572166442871,
      "action_relation_loss": 0.0008790568972472101,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.314312441534145,
      "flow_matching_loss": 0.014284979458898306,
      "grad_norm": 0.13869735598564148,
      "learning_rate": 1.059638259053025e-06,
      "loss": 0.0152,
      "relation_loss": 0.0009453231527004391,
      "selected_pairs": 27038.1,
      "state_query_norm": 41.87980194091797,
      "state_relation_loss": 6.626624344789889e-05,
      "step": 28130,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009453231527004391,
      "weighted_relation_loss": 0.0009453231527004391
    },
    {
      "action_query_norm": 39.681895065307614,
      "action_relation_loss": 0.0008490706328302622,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.323666978484564,
      "flow_matching_loss": 0.014573099091649055,
      "grad_norm": 0.11810046434402466,
      "learning_rate": 1.0483812015339434e-06,
      "loss": 0.0155,
      "relation_loss": 0.0009217168029863387,
      "selected_pairs": 26833.2,
      "state_query_norm": 41.84616584777832,
      "state_relation_loss": 7.264616142492742e-05,
      "step": 28140,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009217168029863387,
      "weighted_relation_loss": 0.0009217168029863387
    },
    {
      "action_query_norm": 39.63842239379883,
      "action_relation_loss": 0.0009157643769867718,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.333021515434986,
      "flow_matching_loss": 0.014194931276142598,
      "grad_norm": 0.10267619788646698,
      "learning_rate": 1.037183624847543e-06,
      "loss": 0.0152,
      "relation_loss": 0.0009849480877164752,
      "selected_pairs": 27104.8,
      "state_query_norm": 41.8600772857666,
      "state_relation_loss": 6.91837260092143e-05,
      "step": 28150,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009849480877164752,
      "weighted_relation_loss": 0.0009849480877164752
    },
    {
      "action_query_norm": 39.67832450866699,
      "action_relation_loss": 0.0008650039439089596,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.342376052385408,
      "flow_matching_loss": 0.014920077007263898,
      "grad_norm": 0.09799753874540329,
      "learning_rate": 1.0260455425999238e-06,
      "loss": 0.0159,
      "relation_loss": 0.0009372060303576291,
      "selected_pairs": 26944.4,
      "state_query_norm": 41.862876892089844,
      "state_relation_loss": 7.220207335194573e-05,
      "step": 28160,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009372060303576291,
      "weighted_relation_loss": 0.0009372060303576291
    },
    {
      "action_query_norm": 39.621507263183595,
      "action_relation_loss": 0.0008602386922575533,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.351730589335826,
      "flow_matching_loss": 0.01461428627371788,
      "grad_norm": 0.08904361724853516,
      "learning_rate": 1.0149669683249208e-06,
      "loss": 0.0155,
      "relation_loss": 0.0009273039177060127,
      "selected_pairs": 27103.4,
      "state_query_norm": 41.85674705505371,
      "state_relation_loss": 6.706522035528905e-05,
      "step": 28170,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009273039177060127,
      "weighted_relation_loss": 0.0009273039177060127
    },
    {
      "action_query_norm": 39.620048904418944,
      "action_relation_loss": 0.0008738175558391959,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.36108512628625,
      "flow_matching_loss": 0.013734692335128784,
      "grad_norm": 0.08777312934398651,
      "learning_rate": 1.003947915484027e-06,
      "loss": 0.0147,
      "relation_loss": 0.0009437339671421796,
      "selected_pairs": 26802.4,
      "state_query_norm": 41.8517765045166,
      "state_relation_loss": 6.991640402702615e-05,
      "step": 28180,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009437339671421796,
      "weighted_relation_loss": 0.0009437339671421796
    },
    {
      "action_query_norm": 39.65504837036133,
      "action_relation_loss": 0.0008506299345754087,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.37043966323667,
      "flow_matching_loss": 0.014786608703434467,
      "grad_norm": 0.3004353940486908,
      "learning_rate": 9.92988397466449e-07,
      "loss": 0.0157,
      "relation_loss": 0.0009206948336213828,
      "selected_pairs": 27061.0,
      "state_query_norm": 41.840018844604494,
      "state_relation_loss": 7.006490268395283e-05,
      "step": 28190,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009206948336213828,
      "weighted_relation_loss": 0.0009206948336213828
    },
    {
      "action_query_norm": 39.620906066894534,
      "action_relation_loss": 0.0008606294461060316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.379794200187092,
      "flow_matching_loss": 0.015064599737524987,
      "grad_norm": 0.11347852647304535,
      "learning_rate": 9.820884275890286e-07,
      "loss": 0.016,
      "relation_loss": 0.0009329744148999453,
      "selected_pairs": 26301.3,
      "state_query_norm": 41.85024757385254,
      "state_relation_loss": 7.234497825265862e-05,
      "step": 28200,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0009329744148999453,
      "weighted_relation_loss": 0.0009329744148999453
    },
    {
      "action_query_norm": 39.65857429504395,
      "action_relation_loss": 0.000884477753425017,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.38914873713751,
      "flow_matching_loss": 0.016242151521146296,
      "grad_norm": 0.10881073027849197,
      "learning_rate": 9.712480190962546e-07,
      "loss": 0.0172,
      "relation_loss": 0.0009585649007931351,
      "selected_pairs": 26941.7,
      "state_query_norm": 41.83742408752441,
      "state_relation_loss": 7.408714518533088e-05,
      "step": 28210,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009585649007931351,
      "weighted_relation_loss": 0.0009585649007931351
    },
    {
      "action_query_norm": 39.625186920166016,
      "action_relation_loss": 0.0009920631011482329,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.398503274087933,
      "flow_matching_loss": 0.015293512772768737,
      "grad_norm": 0.0860452800989151,
      "learning_rate": 9.604671851602464e-07,
      "loss": 0.0164,
      "relation_loss": 0.0010627392271999269,
      "selected_pairs": 26439.2,
      "state_query_norm": 41.85536842346191,
      "state_relation_loss": 7.067614387779031e-05,
      "step": 28220,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.0010627392271999269,
      "weighted_relation_loss": 0.0010627392271999269
    },
    {
      "action_query_norm": 39.65043563842774,
      "action_relation_loss": 0.0008641115622594952,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.407857811038355,
      "flow_matching_loss": 0.014095180947333575,
      "grad_norm": 0.07280628383159637,
      "learning_rate": 9.497459388807306e-07,
      "loss": 0.015,
      "relation_loss": 0.0009322577796410769,
      "selected_pairs": 26755.4,
      "state_query_norm": 41.86665496826172,
      "state_relation_loss": 6.814623411628417e-05,
      "step": 28230,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009322577796410769,
      "weighted_relation_loss": 0.0009322577796410769
    },
    {
      "action_query_norm": 39.65539360046387,
      "action_relation_loss": 0.0008373555901926011,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.417212347988773,
      "flow_matching_loss": 0.015960930567234754,
      "grad_norm": 0.08253970742225647,
      "learning_rate": 9.390842932850364e-07,
      "loss": 0.0169,
      "relation_loss": 0.00090967922587879,
      "selected_pairs": 26965.5,
      "state_query_norm": 41.87668952941895,
      "state_relation_loss": 7.232363386719954e-05,
      "step": 28240,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.00090967922587879,
      "weighted_relation_loss": 0.00090967922587879
    },
    {
      "action_query_norm": 39.65853309631348,
      "action_relation_loss": 0.0008930868352763355,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.426566884939195,
      "flow_matching_loss": 0.014145549759268761,
      "grad_norm": 0.09158233553171158,
      "learning_rate": 9.28482261328073e-07,
      "loss": 0.0151,
      "relation_loss": 0.000963734369724989,
      "selected_pairs": 26763.9,
      "state_query_norm": 41.863480758666995,
      "state_relation_loss": 7.064752571750432e-05,
      "step": 28250,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.000963734369724989,
      "weighted_relation_loss": 0.000963734369724989
    },
    {
      "action_query_norm": 39.64287910461426,
      "action_relation_loss": 0.000862046517431736,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.435921421889617,
      "flow_matching_loss": 0.015917772985994817,
      "grad_norm": 0.07654521614313126,
      "learning_rate": 9.179398558923025e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009346037812065333,
      "selected_pairs": 26596.3,
      "state_query_norm": 41.854994201660155,
      "state_relation_loss": 7.255726450239308e-05,
      "step": 28260,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0009346037812065333,
      "weighted_relation_loss": 0.0009346037812065333
    },
    {
      "action_query_norm": 39.62461204528809,
      "action_relation_loss": 0.0008573054510634392,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.44527595884004,
      "flow_matching_loss": 0.015303488820791245,
      "grad_norm": 0.08835297077894211,
      "learning_rate": 9.074570897877388e-07,
      "loss": 0.0162,
      "relation_loss": 0.0009281546343117952,
      "selected_pairs": 26827.6,
      "state_query_norm": 41.85005187988281,
      "state_relation_loss": 7.084919125190935e-05,
      "step": 28270,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009281546343117952,
      "weighted_relation_loss": 0.0009281546343117952
    },
    {
      "action_query_norm": 39.65593490600586,
      "action_relation_loss": 0.000859598774695769,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.454630495790457,
      "flow_matching_loss": 0.014934796653687955,
      "grad_norm": 0.08803530782461166,
      "learning_rate": 8.970339757519375e-07,
      "loss": 0.0159,
      "relation_loss": 0.0009293978684581817,
      "selected_pairs": 26453.5,
      "state_query_norm": 41.86092185974121,
      "state_relation_loss": 6.979909740039148e-05,
      "step": 28280,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009293978684581817,
      "weighted_relation_loss": 0.0009293978684581817
    },
    {
      "action_query_norm": 39.63332176208496,
      "action_relation_loss": 0.0008670614275615663,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.46398503274088,
      "flow_matching_loss": 0.014295704569667578,
      "grad_norm": 0.06928049027919769,
      "learning_rate": 8.866705264499619e-07,
      "loss": 0.0152,
      "relation_loss": 0.0009424159012269229,
      "selected_pairs": 27339.7,
      "state_query_norm": 41.84821243286133,
      "state_relation_loss": 7.535447584814393e-05,
      "step": 28290,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009424159012269229,
      "weighted_relation_loss": 0.0009424159012269229
    },
    {
      "action_query_norm": 39.651140213012695,
      "action_relation_loss": 0.0008796587237156928,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.4733395696913,
      "flow_matching_loss": 0.014647044241428375,
      "grad_norm": 0.08378918468952179,
      "learning_rate": 8.763667544743836e-07,
      "loss": 0.0156,
      "relation_loss": 0.0009505440481007099,
      "selected_pairs": 26894.4,
      "state_query_norm": 41.8590591430664,
      "state_relation_loss": 7.088531274348497e-05,
      "step": 28300,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009505440481007099,
      "weighted_relation_loss": 0.0009505440481007099
    },
    {
      "action_query_norm": 39.64626541137695,
      "action_relation_loss": 0.0008831059793010354,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.48269410664172,
      "flow_matching_loss": 0.017618186306208372,
      "grad_norm": 0.0958213284611702,
      "learning_rate": 8.661226723452543e-07,
      "loss": 0.0186,
      "relation_loss": 0.0009545958833768964,
      "selected_pairs": 27079.4,
      "state_query_norm": 41.846435546875,
      "state_relation_loss": 7.148990625864826e-05,
      "step": 28310,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0009545958833768964,
      "weighted_relation_loss": 0.0009545958833768964
    },
    {
      "action_query_norm": 39.63633842468262,
      "action_relation_loss": 0.0008625829825177789,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.492048643592142,
      "flow_matching_loss": 0.01526644816622138,
      "grad_norm": 0.09374122321605682,
      "learning_rate": 8.559382925101001e-07,
      "loss": 0.0162,
      "relation_loss": 0.0009359119751024992,
      "selected_pairs": 26578.0,
      "state_query_norm": 41.867415237426755,
      "state_relation_loss": 7.332898821914569e-05,
      "step": 28320,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0009359119751024992,
      "weighted_relation_loss": 0.0009359119751024992
    },
    {
      "action_query_norm": 39.64849624633789,
      "action_relation_loss": 0.0008505463425535709,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.501403180542564,
      "flow_matching_loss": 0.014172057900577784,
      "grad_norm": 0.06494192779064178,
      "learning_rate": 8.458136273438943e-07,
      "loss": 0.0151,
      "relation_loss": 0.0009216780541464687,
      "selected_pairs": 27505.6,
      "state_query_norm": 41.84809455871582,
      "state_relation_loss": 7.1131726144813e-05,
      "step": 28330,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0009216780541464687,
      "weighted_relation_loss": 0.0009216780541464687
    },
    {
      "action_query_norm": 39.649146270751956,
      "action_relation_loss": 0.0008816609566565603,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.510757717492986,
      "flow_matching_loss": 0.015117313526570798,
      "grad_norm": 0.10477404296398163,
      "learning_rate": 8.357486891490795e-07,
      "loss": 0.0161,
      "relation_loss": 0.000954103167168796,
      "selected_pairs": 26677.5,
      "state_query_norm": 41.86533737182617,
      "state_relation_loss": 7.244222651934252e-05,
      "step": 28340,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.000954103167168796,
      "weighted_relation_loss": 0.000954103167168796
    },
    {
      "action_query_norm": 39.599341583251956,
      "action_relation_loss": 0.0009296165313571692,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.520112254443404,
      "flow_matching_loss": 0.014730715844780206,
      "grad_norm": 0.08477769792079926,
      "learning_rate": 8.257434901554895e-07,
      "loss": 0.0157,
      "relation_loss": 0.0010026923497207462,
      "selected_pairs": 26203.8,
      "state_query_norm": 41.849786376953126,
      "state_relation_loss": 7.307581981876864e-05,
      "step": 28350,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0010026923497207462,
      "weighted_relation_loss": 0.0010026923497207462
    },
    {
      "action_query_norm": 39.69462356567383,
      "action_relation_loss": 0.0008565088559407741,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.529466791393826,
      "flow_matching_loss": 0.01594727449119091,
      "grad_norm": 0.12130893021821976,
      "learning_rate": 8.157980425203937e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009278109646402299,
      "selected_pairs": 27811.4,
      "state_query_norm": 41.867859268188475,
      "state_relation_loss": 7.130210324248764e-05,
      "step": 28360,
      "valid_action_samples": 236.3,
      "weighted_auxiliary_loss": 0.0009278109646402299,
      "weighted_relation_loss": 0.0009278109646402299
    },
    {
      "action_query_norm": 39.62821655273437,
      "action_relation_loss": 0.0008688224421348423,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.53882132834425,
      "flow_matching_loss": 0.015807518269866705,
      "grad_norm": 0.09250479191541672,
      "learning_rate": 8.059123583284367e-07,
      "loss": 0.0168,
      "relation_loss": 0.0009431079728528857,
      "selected_pairs": 26949.1,
      "state_query_norm": 41.854378890991214,
      "state_relation_loss": 7.428551907651126e-05,
      "step": 28370,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009431079728528857,
      "weighted_relation_loss": 0.0009431079728528857
    },
    {
      "action_query_norm": 39.62423973083496,
      "action_relation_loss": 0.0008855259628035128,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.548175865294667,
      "flow_matching_loss": 0.01740448083728552,
      "grad_norm": 0.08908013254404068,
      "learning_rate": 7.960864495916654e-07,
      "loss": 0.0184,
      "relation_loss": 0.0009579392964951694,
      "selected_pairs": 26763.2,
      "state_query_norm": 41.86465187072754,
      "state_relation_loss": 7.241333441925235e-05,
      "step": 28380,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009579392964951694,
      "weighted_relation_loss": 0.0009579392964951694
    },
    {
      "action_query_norm": 39.62983932495117,
      "action_relation_loss": 0.0008963454514741897,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.55753040224509,
      "flow_matching_loss": 0.014210263453423976,
      "grad_norm": 0.12814395129680634,
      "learning_rate": 7.863203282494847e-07,
      "loss": 0.0152,
      "relation_loss": 0.0009690660517662764,
      "selected_pairs": 26762.7,
      "state_query_norm": 41.87068061828613,
      "state_relation_loss": 7.272059556271416e-05,
      "step": 28390,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009690660517662764,
      "weighted_relation_loss": 0.0009690660517662764
    },
    {
      "action_query_norm": 39.62518577575683,
      "action_relation_loss": 0.0009293579554650932,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.56688493919551,
      "flow_matching_loss": 0.017333152517676354,
      "grad_norm": 0.08074310421943665,
      "learning_rate": 7.766140061686522e-07,
      "loss": 0.0183,
      "relation_loss": 0.0009999521891586483,
      "selected_pairs": 27036.6,
      "state_query_norm": 41.85542106628418,
      "state_relation_loss": 7.059422750899103e-05,
      "step": 28400,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009999521891586483,
      "weighted_relation_loss": 0.0009999521891586483
    },
    {
      "action_query_norm": 39.651794815063475,
      "action_relation_loss": 0.0008401985047385097,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.57623947614593,
      "flow_matching_loss": 0.01641394291073084,
      "grad_norm": 0.09967577457427979,
      "learning_rate": 7.669674951432615e-07,
      "loss": 0.0173,
      "relation_loss": 0.0009092452237382531,
      "selected_pairs": 26948.5,
      "state_query_norm": 41.85815315246582,
      "state_relation_loss": 6.90467182721477e-05,
      "step": 28410,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009092452237382531,
      "weighted_relation_loss": 0.0009092452237382531
    },
    {
      "action_query_norm": 39.61768493652344,
      "action_relation_loss": 0.0008287169912364334,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.58559401309635,
      "flow_matching_loss": 0.014203791785985232,
      "grad_norm": 0.08436927944421768,
      "learning_rate": 7.573808068947363e-07,
      "loss": 0.0151,
      "relation_loss": 0.000897643924690783,
      "selected_pairs": 26869.9,
      "state_query_norm": 41.853887939453124,
      "state_relation_loss": 6.892693745612633e-05,
      "step": 28420,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.000897643924690783,
      "weighted_relation_loss": 0.000897643924690783
    },
    {
      "action_query_norm": 39.63897552490234,
      "action_relation_loss": 0.0008718652476090938,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.594948550046773,
      "flow_matching_loss": 0.016299405414611102,
      "grad_norm": 0.09243573248386383,
      "learning_rate": 7.478539530718087e-07,
      "loss": 0.0172,
      "relation_loss": 0.00094425497809425,
      "selected_pairs": 26874.9,
      "state_query_norm": 41.84778366088867,
      "state_relation_loss": 7.238973412313499e-05,
      "step": 28430,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.00094425497809425,
      "weighted_relation_loss": 0.00094425497809425
    },
    {
      "action_query_norm": 39.65989456176758,
      "action_relation_loss": 0.0008717112534213812,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.604303086997195,
      "flow_matching_loss": 0.016100508999079464,
      "grad_norm": 0.08880136907100677,
      "learning_rate": 7.383869452504965e-07,
      "loss": 0.017,
      "relation_loss": 0.0009447218151763081,
      "selected_pairs": 26556.0,
      "state_query_norm": 41.84941864013672,
      "state_relation_loss": 7.301056939468253e-05,
      "step": 28440,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0009447218151763081,
      "weighted_relation_loss": 0.0009447218151763081
    },
    {
      "action_query_norm": 39.637155532836914,
      "action_relation_loss": 0.0008585487608797848,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.613657623947613,
      "flow_matching_loss": 0.014721298683434724,
      "grad_norm": 0.07581348717212677,
      "learning_rate": 7.289797949341204e-07,
      "loss": 0.0157,
      "relation_loss": 0.0009297193551901728,
      "selected_pairs": 26809.8,
      "state_query_norm": 41.86197700500488,
      "state_relation_loss": 7.117059285519645e-05,
      "step": 28450,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009297193551901728,
      "weighted_relation_loss": 0.0009297193551901728
    },
    {
      "action_query_norm": 39.61488723754883,
      "action_relation_loss": 0.0008725251886062324,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.623012160898035,
      "flow_matching_loss": 0.014550598990172147,
      "grad_norm": 0.11681336909532547,
      "learning_rate": 7.196325135532423e-07,
      "loss": 0.0155,
      "relation_loss": 0.0009425251395441592,
      "selected_pairs": 27050.9,
      "state_query_norm": 41.8538875579834,
      "state_relation_loss": 6.999995130172466e-05,
      "step": 28460,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009425251395441592,
      "weighted_relation_loss": 0.0009425251395441592
    },
    {
      "action_query_norm": 39.60507583618164,
      "action_relation_loss": 0.0008365374698769302,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.632366697848457,
      "flow_matching_loss": 0.016119442880153656,
      "grad_norm": 0.17258374392986298,
      "learning_rate": 7.10345112465699e-07,
      "loss": 0.017,
      "relation_loss": 0.0009088669379707426,
      "selected_pairs": 26548.4,
      "state_query_norm": 41.85109252929688,
      "state_relation_loss": 7.23294768249616e-05,
      "step": 28470,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0009088669379707426,
      "weighted_relation_loss": 0.0009088669379707426
    },
    {
      "action_query_norm": 39.63541450500488,
      "action_relation_loss": 0.0008816694898996502,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.641721234798876,
      "flow_matching_loss": 0.01553791994228959,
      "grad_norm": 0.08059713989496231,
      "learning_rate": 7.011176029565525e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009516961639747023,
      "selected_pairs": 26919.3,
      "state_query_norm": 41.84547004699707,
      "state_relation_loss": 7.002667989581823e-05,
      "step": 28480,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009516961639747023,
      "weighted_relation_loss": 0.0009516961639747023
    },
    {
      "action_query_norm": 39.611444091796876,
      "action_relation_loss": 0.0008705836837179959,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.651075771749298,
      "flow_matching_loss": 0.013305032067000866,
      "grad_norm": 0.11130866408348083,
      "learning_rate": 6.919499962381004e-07,
      "loss": 0.0142,
      "relation_loss": 0.0009425159601960331,
      "selected_pairs": 26856.5,
      "state_query_norm": 41.84097785949707,
      "state_relation_loss": 7.193226483650505e-05,
      "step": 28490,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009425159601960331,
      "weighted_relation_loss": 0.0009425159601960331
    },
    {
      "action_query_norm": 39.66086921691895,
      "action_relation_loss": 0.0008481691998895258,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.66043030869972,
      "flow_matching_loss": 0.015903167985379697,
      "grad_norm": 0.10743676126003265,
      "learning_rate": 6.828423034498488e-07,
      "loss": 0.0168,
      "relation_loss": 0.0009194192010909319,
      "selected_pairs": 26417.8,
      "state_query_norm": 41.83880195617676,
      "state_relation_loss": 7.125000483938493e-05,
      "step": 28500,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0009194192010909319,
      "weighted_relation_loss": 0.0009194192010909319
    },
    {
      "action_query_norm": 39.63527488708496,
      "action_relation_loss": 0.0008623810310382396,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.669784845650142,
      "flow_matching_loss": 0.01621002536267042,
      "grad_norm": 0.08317774534225464,
      "learning_rate": 6.737945356585007e-07,
      "loss": 0.0171,
      "relation_loss": 0.0009319498552940786,
      "selected_pairs": 26731.4,
      "state_query_norm": 41.864274978637695,
      "state_relation_loss": 6.956881261430681e-05,
      "step": 28510,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009319498552940786,
      "weighted_relation_loss": 0.0009319498552940786
    },
    {
      "action_query_norm": 39.59367370605469,
      "action_relation_loss": 0.0008769971784204245,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.67913938260056,
      "flow_matching_loss": 0.015061859786510468,
      "grad_norm": 0.0660277009010315,
      "learning_rate": 6.648067038579508e-07,
      "loss": 0.016,
      "relation_loss": 0.0009478008607402444,
      "selected_pairs": 26295.4,
      "state_query_norm": 41.849493408203124,
      "state_relation_loss": 7.080368741299025e-05,
      "step": 28520,
      "valid_action_samples": 229.8,
      "weighted_auxiliary_loss": 0.0009478008607402444,
      "weighted_relation_loss": 0.0009478008607402444
    },
    {
      "action_query_norm": 39.61347618103027,
      "action_relation_loss": 0.0008652270713355392,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.688493919550982,
      "flow_matching_loss": 0.015895376354455946,
      "grad_norm": 0.10710233449935913,
      "learning_rate": 6.558788189692578e-07,
      "loss": 0.0168,
      "relation_loss": 0.0009363599005155265,
      "selected_pairs": 26465.6,
      "state_query_norm": 41.855908203125,
      "state_relation_loss": 7.113282772479579e-05,
      "step": 28530,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009363599005155265,
      "weighted_relation_loss": 0.0009363599005155265
    },
    {
      "action_query_norm": 39.65139884948731,
      "action_relation_loss": 0.0008889519434887916,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.697848456501404,
      "flow_matching_loss": 0.016475363355129957,
      "grad_norm": 0.09638884663581848,
      "learning_rate": 6.470108918406492e-07,
      "loss": 0.0174,
      "relation_loss": 0.0009636924020014703,
      "selected_pairs": 26969.8,
      "state_query_norm": 41.84925422668457,
      "state_relation_loss": 7.47404650610406e-05,
      "step": 28540,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009636924020014703,
      "weighted_relation_loss": 0.0009636924020014703
    },
    {
      "action_query_norm": 39.65304946899414,
      "action_relation_loss": 0.0008760451048146934,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.707202993451823,
      "flow_matching_loss": 0.015940498746931552,
      "grad_norm": 0.10035654157400131,
      "learning_rate": 6.382029332474893e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009462155168876052,
      "selected_pairs": 26464.8,
      "state_query_norm": 41.85585060119629,
      "state_relation_loss": 7.017040697974153e-05,
      "step": 28550,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009462155168876052,
      "weighted_relation_loss": 0.0009462155168876052
    },
    {
      "action_query_norm": 39.60987930297851,
      "action_relation_loss": 0.0009567569533828646,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.716557530402245,
      "flow_matching_loss": 0.015921305399388075,
      "grad_norm": 0.10026781260967255,
      "learning_rate": 6.294549538922778e-07,
      "loss": 0.017,
      "relation_loss": 0.0010307552816811948,
      "selected_pairs": 27220.1,
      "state_query_norm": 41.8381160736084,
      "state_relation_loss": 7.399833048111759e-05,
      "step": 28560,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0010307552816811948,
      "weighted_relation_loss": 0.0010307552816811948
    },
    {
      "action_query_norm": 39.621521377563475,
      "action_relation_loss": 0.0008981679857242852,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.725912067352667,
      "flow_matching_loss": 0.015466335788369178,
      "grad_norm": 0.09115804731845856,
      "learning_rate": 6.207669644046344e-07,
      "loss": 0.0164,
      "relation_loss": 0.0009681867551989854,
      "selected_pairs": 26270.6,
      "state_query_norm": 41.82695350646973,
      "state_relation_loss": 7.001876147114672e-05,
      "step": 28570,
      "valid_action_samples": 229.7,
      "weighted_auxiliary_loss": 0.0009681867551989854,
      "weighted_relation_loss": 0.0009681867551989854
    },
    {
      "action_query_norm": 39.643828582763675,
      "action_relation_loss": 0.0008794799796305597,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.73526660430309,
      "flow_matching_loss": 0.015679371543228626,
      "grad_norm": 0.08175988495349884,
      "learning_rate": 6.121389753412865e-07,
      "loss": 0.0166,
      "relation_loss": 0.000954264507163316,
      "selected_pairs": 26530.8,
      "state_query_norm": 41.84668731689453,
      "state_relation_loss": 7.47845206205966e-05,
      "step": 28580,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.000954264507163316,
      "weighted_relation_loss": 0.000954264507163316
    },
    {
      "action_query_norm": 39.67546005249024,
      "action_relation_loss": 0.0008655015437398106,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.744621141253507,
      "flow_matching_loss": 0.015400486718863248,
      "grad_norm": 0.09697435051202774,
      "learning_rate": 6.035709971860592e-07,
      "loss": 0.0163,
      "relation_loss": 0.0009376461151987314,
      "selected_pairs": 27202.7,
      "state_query_norm": 41.846879959106445,
      "state_relation_loss": 7.214456054498442e-05,
      "step": 28590,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009376461151987314,
      "weighted_relation_loss": 0.0009376461151987314
    },
    {
      "action_query_norm": 39.558179473876955,
      "action_relation_loss": 0.0009153033723123371,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.75397567820393,
      "flow_matching_loss": 0.015526272635906935,
      "grad_norm": 0.11490985006093979,
      "learning_rate": 5.950630403498469e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009892854723148048,
      "selected_pairs": 26524.3,
      "state_query_norm": 41.85024757385254,
      "state_relation_loss": 7.398209563689307e-05,
      "step": 28600,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009892854723148048,
      "weighted_relation_loss": 0.0009892854723148048
    },
    {
      "action_query_norm": 39.682421112060545,
      "action_relation_loss": 0.000839847099268809,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.76333021515435,
      "flow_matching_loss": 0.014521317742764949,
      "grad_norm": 0.08321060240268707,
      "learning_rate": 5.866151151706189e-07,
      "loss": 0.0154,
      "relation_loss": 0.0009118522750213743,
      "selected_pairs": 26920.2,
      "state_query_norm": 41.86498870849609,
      "state_relation_loss": 7.200516993179917e-05,
      "step": 28610,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009118522750213743,
      "weighted_relation_loss": 0.0009118522750213743
    },
    {
      "action_query_norm": 39.680715560913086,
      "action_relation_loss": 0.0008313695318065584,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.77268475210477,
      "flow_matching_loss": 0.01560564422979951,
      "grad_norm": 0.1358676701784134,
      "learning_rate": 5.782272319134086e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009012854541651904,
      "selected_pairs": 26525.9,
      "state_query_norm": 41.861296844482425,
      "state_relation_loss": 6.99159147188766e-05,
      "step": 28620,
      "valid_action_samples": 230.8,
      "weighted_auxiliary_loss": 0.0009012854541651904,
      "weighted_relation_loss": 0.0009012854541651904
    },
    {
      "action_query_norm": 39.712248992919925,
      "action_relation_loss": 0.0008278743014670908,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.78203928905519,
      "flow_matching_loss": 0.01627481058239937,
      "grad_norm": 0.101941779255867,
      "learning_rate": 5.698994007702796e-07,
      "loss": 0.0172,
      "relation_loss": 0.0009002725419122726,
      "selected_pairs": 27614.9,
      "state_query_norm": 41.86453475952148,
      "state_relation_loss": 7.2398242627969e-05,
      "step": 28630,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0009002725419122726,
      "weighted_relation_loss": 0.0009002725419122726
    },
    {
      "action_query_norm": 39.654079818725585,
      "action_relation_loss": 0.0008184782462194562,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.791393826005613,
      "flow_matching_loss": 0.016553377173840998,
      "grad_norm": 0.10081907361745834,
      "learning_rate": 5.616316318603321e-07,
      "loss": 0.0174,
      "relation_loss": 0.0008900835702661425,
      "selected_pairs": 26989.4,
      "state_query_norm": 41.85095558166504,
      "state_relation_loss": 7.16053182259202e-05,
      "step": 28640,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0008900835702661425,
      "weighted_relation_loss": 0.0008900835702661425
    },
    {
      "action_query_norm": 39.63445053100586,
      "action_relation_loss": 0.0008496764348819852,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.800748362956035,
      "flow_matching_loss": 0.016653108410537243,
      "grad_norm": 0.11198677867650986,
      "learning_rate": 5.534239352296799e-07,
      "loss": 0.0176,
      "relation_loss": 0.000919535115826875,
      "selected_pairs": 27045.7,
      "state_query_norm": 41.8620719909668,
      "state_relation_loss": 6.985867585171946e-05,
      "step": 28650,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.000919535115826875,
      "weighted_relation_loss": 0.000919535115826875
    },
    {
      "action_query_norm": 39.666648864746094,
      "action_relation_loss": 0.000844495085766539,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.810102899906454,
      "flow_matching_loss": 0.01602595569565892,
      "grad_norm": 0.09269803762435913,
      "learning_rate": 5.452763208514621e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009155944630037993,
      "selected_pairs": 27084.3,
      "state_query_norm": 41.847874069213866,
      "state_relation_loss": 7.109939033398405e-05,
      "step": 28660,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0009155944630037993,
      "weighted_relation_loss": 0.0009155944630037993
    },
    {
      "action_query_norm": 39.617350006103514,
      "action_relation_loss": 0.0008450622379314154,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.819457436856876,
      "flow_matching_loss": 0.014945136290043593,
      "grad_norm": 0.1221344992518425,
      "learning_rate": 5.371887986257873e-07,
      "loss": 0.0159,
      "relation_loss": 0.0009162595204543322,
      "selected_pairs": 26415.9,
      "state_query_norm": 41.85838851928711,
      "state_relation_loss": 7.11972803401295e-05,
      "step": 28670,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0009162595204543322,
      "weighted_relation_loss": 0.0009162595204543322
    },
    {
      "action_query_norm": 39.644013214111325,
      "action_relation_loss": 0.0008534702297765762,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.828811973807298,
      "flow_matching_loss": 0.014925552997738123,
      "grad_norm": 0.15266533195972443,
      "learning_rate": 5.291613783797611e-07,
      "loss": 0.0158,
      "relation_loss": 0.0009216948063112795,
      "selected_pairs": 26874.0,
      "state_query_norm": 41.852695083618165,
      "state_relation_loss": 6.822456234658602e-05,
      "step": 28680,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009216948063112795,
      "weighted_relation_loss": 0.0009216948063112795
    },
    {
      "action_query_norm": 39.657915115356445,
      "action_relation_loss": 0.000854128273203969,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.838166510757716,
      "flow_matching_loss": 0.01569548072293401,
      "grad_norm": 0.09750299900770187,
      "learning_rate": 5.211940698674534e-07,
      "loss": 0.0166,
      "relation_loss": 0.0009283618477638811,
      "selected_pairs": 27107.9,
      "state_query_norm": 41.84534339904785,
      "state_relation_loss": 7.423359347740189e-05,
      "step": 28690,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009283618477638811,
      "weighted_relation_loss": 0.0009283618477638811
    },
    {
      "action_query_norm": 39.67698593139649,
      "action_relation_loss": 0.0008543217205442488,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.84752104770814,
      "flow_matching_loss": 0.015253886766731738,
      "grad_norm": 0.09625180065631866,
      "learning_rate": 5.132868827698978e-07,
      "loss": 0.0162,
      "relation_loss": 0.0009277714067138732,
      "selected_pairs": 27614.3,
      "state_query_norm": 41.87360153198242,
      "state_relation_loss": 7.344969199039043e-05,
      "step": 28700,
      "valid_action_samples": 235.5,
      "weighted_auxiliary_loss": 0.0009277714067138732,
      "weighted_relation_loss": 0.0009277714067138732
    },
    {
      "action_query_norm": 39.609663772583005,
      "action_relation_loss": 0.0008569508965592832,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.85687558465856,
      "flow_matching_loss": 0.016040606331080197,
      "grad_norm": 0.09162753820419312,
      "learning_rate": 5.054398266950755e-07,
      "loss": 0.017,
      "relation_loss": 0.0009279041201807559,
      "selected_pairs": 26969.0,
      "state_query_norm": 41.843343353271486,
      "state_relation_loss": 7.095323198882397e-05,
      "step": 28710,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009279041201807559,
      "weighted_relation_loss": 0.0009279041201807559
    },
    {
      "action_query_norm": 39.60528755187988,
      "action_relation_loss": 0.000888454366941005,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.86623012160898,
      "flow_matching_loss": 0.016941096913069487,
      "grad_norm": 0.12506866455078125,
      "learning_rate": 4.976529111778872e-07,
      "loss": 0.0179,
      "relation_loss": 0.0009580962243489922,
      "selected_pairs": 27363.5,
      "state_query_norm": 41.850123977661134,
      "state_relation_loss": 6.964185085962526e-05,
      "step": 28720,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009580962243489922,
      "weighted_relation_loss": 0.0009580962243489922
    },
    {
      "action_query_norm": 39.651970291137694,
      "action_relation_loss": 0.0008434828021563589,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.8755846585594,
      "flow_matching_loss": 0.015463265310972928,
      "grad_norm": 0.10136889666318893,
      "learning_rate": 4.899261456801862e-07,
      "loss": 0.0164,
      "relation_loss": 0.0009149033925496042,
      "selected_pairs": 27175.2,
      "state_query_norm": 41.84763946533203,
      "state_relation_loss": 7.14205831172876e-05,
      "step": 28730,
      "valid_action_samples": 233.6,
      "weighted_auxiliary_loss": 0.0009149033925496042,
      "weighted_relation_loss": 0.0009149033925496042
    },
    {
      "action_query_norm": 39.613430404663085,
      "action_relation_loss": 0.0008602927497122437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.884939195509823,
      "flow_matching_loss": 0.01533218901604414,
      "grad_norm": 0.07782738655805588,
      "learning_rate": 4.822595395907126e-07,
      "loss": 0.0163,
      "relation_loss": 0.0009331360284704715,
      "selected_pairs": 26594.9,
      "state_query_norm": 41.85956954956055,
      "state_relation_loss": 7.284327657544054e-05,
      "step": 28740,
      "valid_action_samples": 231.1,
      "weighted_auxiliary_loss": 0.0009331360284704715,
      "weighted_relation_loss": 0.0009331360284704715
    },
    {
      "action_query_norm": 39.64050178527832,
      "action_relation_loss": 0.0008821388997603208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.894293732460245,
      "flow_matching_loss": 0.015251823607832194,
      "grad_norm": 0.11454382538795471,
      "learning_rate": 4.7465310222510886e-07,
      "loss": 0.0162,
      "relation_loss": 0.0009544356376864016,
      "selected_pairs": 27334.8,
      "state_query_norm": 41.861137771606444,
      "state_relation_loss": 7.229673501569777e-05,
      "step": 28750,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009544356376864016,
      "weighted_relation_loss": 0.0009544356376864016
    },
    {
      "action_query_norm": 39.64859733581543,
      "action_relation_loss": 0.0008658050559461117,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.903648269410663,
      "flow_matching_loss": 0.016264089941978456,
      "grad_norm": 0.0940447747707367,
      "learning_rate": 4.6710684282593175e-07,
      "loss": 0.0172,
      "relation_loss": 0.0009361579664982856,
      "selected_pairs": 26942.0,
      "state_query_norm": 41.863331985473636,
      "state_relation_loss": 7.035290655039716e-05,
      "step": 28760,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009361579664982856,
      "weighted_relation_loss": 0.0009361579664982856
    },
    {
      "action_query_norm": 39.57572021484375,
      "action_relation_loss": 0.0008925489790271967,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.913002806361085,
      "flow_matching_loss": 0.015184399392455816,
      "grad_norm": 0.09794878214597702,
      "learning_rate": 4.5962077056257993e-07,
      "loss": 0.0161,
      "relation_loss": 0.0009582395316101611,
      "selected_pairs": 26663.3,
      "state_query_norm": 41.85313949584961,
      "state_relation_loss": 6.569055367435794e-05,
      "step": 28770,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0009582395316101611,
      "weighted_relation_loss": 0.0009582395316101611
    },
    {
      "action_query_norm": 39.64562225341797,
      "action_relation_loss": 0.0008510694839060307,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.922357343311507,
      "flow_matching_loss": 0.013933141436427831,
      "grad_norm": 0.11147879809141159,
      "learning_rate": 4.521948945313492e-07,
      "loss": 0.0149,
      "relation_loss": 0.000924852560274303,
      "selected_pairs": 26896.0,
      "state_query_norm": 41.86099281311035,
      "state_relation_loss": 7.378307564067654e-05,
      "step": 28780,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.000924852560274303,
      "weighted_relation_loss": 0.000924852560274303
    },
    {
      "action_query_norm": 39.617773818969724,
      "action_relation_loss": 0.0008323312737047672,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.931711880261926,
      "flow_matching_loss": 0.016706989239901306,
      "grad_norm": 0.0878731831908226,
      "learning_rate": 4.4482922375537195e-07,
      "loss": 0.0176,
      "relation_loss": 0.0009035772352945059,
      "selected_pairs": 25845.2,
      "state_query_norm": 41.857351303100586,
      "state_relation_loss": 7.124595358618536e-05,
      "step": 28790,
      "valid_action_samples": 227.8,
      "weighted_auxiliary_loss": 0.0009035772352945059,
      "weighted_relation_loss": 0.0009035772352945059
    },
    {
      "action_query_norm": 39.647891235351565,
      "action_relation_loss": 0.0008373683143872767,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.941066417212348,
      "flow_matching_loss": 0.015252165589481592,
      "grad_norm": 0.0681026503443718,
      "learning_rate": 4.375237671846333e-07,
      "loss": 0.0162,
      "relation_loss": 0.0009082240809220821,
      "selected_pairs": 26368.6,
      "state_query_norm": 41.85227394104004,
      "state_relation_loss": 7.085576798999682e-05,
      "step": 28800,
      "valid_action_samples": 230.1,
      "weighted_auxiliary_loss": 0.0009082240809220821,
      "weighted_relation_loss": 0.0009082240809220821
    },
    {
      "action_query_norm": 39.60274772644043,
      "action_relation_loss": 0.0009287880617193878,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.95042095416277,
      "flow_matching_loss": 0.014959146734327078,
      "grad_norm": 0.08130199462175369,
      "learning_rate": 4.302785336959547e-07,
      "loss": 0.016,
      "relation_loss": 0.0009991355007514357,
      "selected_pairs": 26246.3,
      "state_query_norm": 41.857316207885745,
      "state_relation_loss": 7.034744339762256e-05,
      "step": 28810,
      "valid_action_samples": 229.6,
      "weighted_auxiliary_loss": 0.0009991355007514357,
      "weighted_relation_loss": 0.0009991355007514357
    },
    {
      "action_query_norm": 39.61853523254395,
      "action_relation_loss": 0.0008948162838350981,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.95977549111319,
      "flow_matching_loss": 0.0159721152856946,
      "grad_norm": 0.10887188464403152,
      "learning_rate": 4.2309353209297744e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009645728103350848,
      "selected_pairs": 26515.1,
      "state_query_norm": 41.845768737792966,
      "state_relation_loss": 6.975652722758241e-05,
      "step": 28820,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009645728103350848,
      "weighted_relation_loss": 0.0009645728103350848
    },
    {
      "action_query_norm": 39.65040550231934,
      "action_relation_loss": 0.0009129009384196251,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.96913002806361,
      "flow_matching_loss": 0.014719068072736263,
      "grad_norm": 0.06934136152267456,
      "learning_rate": 4.159687711061566e-07,
      "loss": 0.0157,
      "relation_loss": 0.0009819723840337248,
      "selected_pairs": 27667.7,
      "state_query_norm": 41.86290168762207,
      "state_relation_loss": 6.907144052092917e-05,
      "step": 28830,
      "valid_action_samples": 235.7,
      "weighted_auxiliary_loss": 0.0009819723840337248,
      "weighted_relation_loss": 0.0009819723840337248
    },
    {
      "action_query_norm": 39.606472396850585,
      "action_relation_loss": 0.000895658292574808,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.978484565014032,
      "flow_matching_loss": 0.016638503037393094,
      "grad_norm": 0.1057848259806633,
      "learning_rate": 4.0890425939275055e-07,
      "loss": 0.0176,
      "relation_loss": 0.0009658967435825616,
      "selected_pairs": 27366.7,
      "state_query_norm": 41.85597114562988,
      "state_relation_loss": 7.023846192169003e-05,
      "step": 28840,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.0009658967435825616,
      "weighted_relation_loss": 0.0009658967435825616
    },
    {
      "action_query_norm": 39.633188247680664,
      "action_relation_loss": 0.0008428691304288805,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.987839101964454,
      "flow_matching_loss": 0.014765645936131477,
      "grad_norm": 0.10350771993398666,
      "learning_rate": 4.0190000553679827e-07,
      "loss": 0.0157,
      "relation_loss": 0.000913438742281869,
      "selected_pairs": 27371.7,
      "state_query_norm": 41.85002326965332,
      "state_relation_loss": 7.05695969372755e-05,
      "step": 28850,
      "valid_action_samples": 234.4,
      "weighted_auxiliary_loss": 0.000913438742281869,
      "weighted_relation_loss": 0.000913438742281869
    },
    {
      "action_query_norm": 39.6237735748291,
      "action_relation_loss": 0.0008506122103426605,
      "auxiliary_loss_weight": 1.0,
      "epoch": 26.997193638914872,
      "flow_matching_loss": 0.0156971862539649,
      "grad_norm": 0.08564760535955429,
      "learning_rate": 3.9495601804913627e-07,
      "loss": 0.0166,
      "relation_loss": 0.0009198930056300014,
      "selected_pairs": 26738.6,
      "state_query_norm": 41.8658203125,
      "state_relation_loss": 6.928079346835148e-05,
      "step": 28860,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009198930056300014,
      "weighted_relation_loss": 0.0009198930056300014
    },
    {
      "action_query_norm": 39.607208251953125,
      "action_relation_loss": 0.0010124846128746867,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.006548175865294,
      "flow_matching_loss": 0.0162100313231349,
      "grad_norm": 0.09093144536018372,
      "learning_rate": 3.880723053673652e-07,
      "loss": 0.0173,
      "relation_loss": 0.0010826096346136182,
      "selected_pairs": 24389.7,
      "state_query_norm": 41.873185348510745,
      "state_relation_loss": 7.012500900600571e-05,
      "step": 28870,
      "valid_action_samples": 214.5,
      "weighted_auxiliary_loss": 0.0010826096346136182,
      "weighted_relation_loss": 0.0010826096346136182
    },
    {
      "action_query_norm": 39.57945938110352,
      "action_relation_loss": 0.000843101175269112,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.015902712815716,
      "flow_matching_loss": 0.016119916271418334,
      "grad_norm": 0.11721689999103546,
      "learning_rate": 3.812488758558386e-07,
      "loss": 0.017,
      "relation_loss": 0.0009131827799137681,
      "selected_pairs": 26616.7,
      "state_query_norm": 41.856432342529295,
      "state_relation_loss": 7.008160682744346e-05,
      "step": 28880,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009131827799137681,
      "weighted_relation_loss": 0.0009131827799137681
    },
    {
      "action_query_norm": 39.622490310668944,
      "action_relation_loss": 0.0008496502821799367,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.02525724976614,
      "flow_matching_loss": 0.015567156206816434,
      "grad_norm": 0.08853636682033539,
      "learning_rate": 3.744857378056743e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009171613317448645,
      "selected_pairs": 25756.0,
      "state_query_norm": 41.86892318725586,
      "state_relation_loss": 6.751104156137444e-05,
      "step": 28890,
      "valid_action_samples": 227.4,
      "weighted_auxiliary_loss": 0.0009171613317448645,
      "weighted_relation_loss": 0.0009171613317448645
    },
    {
      "action_query_norm": 39.565089797973634,
      "action_relation_loss": 0.000880575564224273,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.034611786716557,
      "flow_matching_loss": 0.01555020771920681,
      "grad_norm": 0.11238008737564087,
      "learning_rate": 3.677828994347154e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009487355418968946,
      "selected_pairs": 27298.1,
      "state_query_norm": 41.854632186889646,
      "state_relation_loss": 6.815998822276014e-05,
      "step": 28900,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0009487355418968946,
      "weighted_relation_loss": 0.0009487355418968946
    },
    {
      "action_query_norm": 39.618696212768555,
      "action_relation_loss": 0.0009029376378748566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.04396632366698,
      "flow_matching_loss": 0.015320939477533102,
      "grad_norm": 0.11105699837207794,
      "learning_rate": 3.61140368887547e-07,
      "loss": 0.0163,
      "relation_loss": 0.0009715760068502277,
      "selected_pairs": 26803.4,
      "state_query_norm": 41.85441551208496,
      "state_relation_loss": 6.863836315460504e-05,
      "step": 28910,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009715760068502277,
      "weighted_relation_loss": 0.0009715760068502277
    },
    {
      "action_query_norm": 39.5911319732666,
      "action_relation_loss": 0.0008557288674637675,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.0533208606174,
      "flow_matching_loss": 0.015790384262800217,
      "grad_norm": 0.11104793101549149,
      "learning_rate": 3.545581542354681e-07,
      "loss": 0.0167,
      "relation_loss": 0.0009312421665526926,
      "selected_pairs": 26695.0,
      "state_query_norm": 41.84425773620605,
      "state_relation_loss": 7.551330199930816e-05,
      "step": 28920,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0009312421665526926,
      "weighted_relation_loss": 0.0009312421665526926
    },
    {
      "action_query_norm": 39.64312591552734,
      "action_relation_loss": 0.000824382237624377,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.06267539756782,
      "flow_matching_loss": 0.015399938821792603,
      "grad_norm": 0.13268788158893585,
      "learning_rate": 3.480362634764922e-07,
      "loss": 0.0163,
      "relation_loss": 0.0008953353739343583,
      "selected_pairs": 26777.4,
      "state_query_norm": 41.87816734313965,
      "state_relation_loss": 7.095312903402374e-05,
      "step": 28930,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0008953353739343583,
      "weighted_relation_loss": 0.0008953353739343583
    },
    {
      "action_query_norm": 39.657481002807614,
      "action_relation_loss": 0.0008742203761357814,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.07202993451824,
      "flow_matching_loss": 0.014639804512262345,
      "grad_norm": 0.13304057717323303,
      "learning_rate": 3.4157470453533014e-07,
      "loss": 0.0156,
      "relation_loss": 0.0009459529188461602,
      "selected_pairs": 26833.5,
      "state_query_norm": 41.84819984436035,
      "state_relation_loss": 7.173254853114486e-05,
      "step": 28940,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009459529188461602,
      "weighted_relation_loss": 0.0009459529188461602
    },
    {
      "action_query_norm": 39.67115516662598,
      "action_relation_loss": 0.00081224421155639,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.081384471468663,
      "flow_matching_loss": 0.015637591667473315,
      "grad_norm": 0.09683747589588165,
      "learning_rate": 3.3517348526339034e-07,
      "loss": 0.0165,
      "relation_loss": 0.0008855599910020829,
      "selected_pairs": 26879.9,
      "state_query_norm": 41.85437850952148,
      "state_relation_loss": 7.331579254241661e-05,
      "step": 28950,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0008855599910020829,
      "weighted_relation_loss": 0.0008855599910020829
    },
    {
      "action_query_norm": 39.59283790588379,
      "action_relation_loss": 0.0008800231851637363,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.09073900841908,
      "flow_matching_loss": 0.016410048492252825,
      "grad_norm": 0.11160697042942047,
      "learning_rate": 3.288326134387454e-07,
      "loss": 0.0174,
      "relation_loss": 0.0009457184467464685,
      "selected_pairs": 26700.1,
      "state_query_norm": 41.84064979553223,
      "state_relation_loss": 6.56952692224877e-05,
      "step": 28960,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0009457184467464685,
      "weighted_relation_loss": 0.0009457184467464685
    },
    {
      "action_query_norm": 39.6741828918457,
      "action_relation_loss": 0.000852268235757947,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.100093545369504,
      "flow_matching_loss": 0.016320077423006295,
      "grad_norm": 0.19257400929927826,
      "learning_rate": 3.225520967661655e-07,
      "loss": 0.0172,
      "relation_loss": 0.0009215157071594149,
      "selected_pairs": 25933.8,
      "state_query_norm": 41.837272262573244,
      "state_relation_loss": 6.924746121512726e-05,
      "step": 28970,
      "valid_action_samples": 228.2,
      "weighted_auxiliary_loss": 0.0009215157071594149,
      "weighted_relation_loss": 0.0009215157071594149
    },
    {
      "action_query_norm": 39.666018676757815,
      "action_relation_loss": 0.0008441108802799135,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.109448082319926,
      "flow_matching_loss": 0.014138293359428645,
      "grad_norm": 0.1303272545337677,
      "learning_rate": 3.163319428770628e-07,
      "loss": 0.0151,
      "relation_loss": 0.0009180887835100293,
      "selected_pairs": 26861.3,
      "state_query_norm": 41.85596618652344,
      "state_relation_loss": 7.397790614049882e-05,
      "step": 28980,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009180887835100293,
      "weighted_relation_loss": 0.0009180887835100293
    },
    {
      "action_query_norm": 39.606885528564455,
      "action_relation_loss": 0.00090159383835271,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.118802619270348,
      "flow_matching_loss": 0.016538347210735084,
      "grad_norm": 0.09424149990081787,
      "learning_rate": 3.1017215932951374e-07,
      "loss": 0.0175,
      "relation_loss": 0.0009742496360559017,
      "selected_pairs": 25896.5,
      "state_query_norm": 41.85257835388184,
      "state_relation_loss": 7.265578169608488e-05,
      "step": 28990,
      "valid_action_samples": 228.0,
      "weighted_auxiliary_loss": 0.0009742496360559017,
      "weighted_relation_loss": 0.0009742496360559017
    },
    {
      "action_query_norm": 39.66734161376953,
      "action_relation_loss": 0.0008227377315051853,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.128157156220766,
      "flow_matching_loss": 0.017078576888889075,
      "grad_norm": 0.09331122040748596,
      "learning_rate": 3.040727536082366e-07,
      "loss": 0.018,
      "relation_loss": 0.0008952799427788705,
      "selected_pairs": 27011.8,
      "state_query_norm": 41.85029678344726,
      "state_relation_loss": 7.254221236507874e-05,
      "step": 29000,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0008952799427788705,
      "weighted_relation_loss": 0.0008952799427788705
    },
    {
      "action_query_norm": 39.66279754638672,
      "action_relation_loss": 0.0008633801247924566,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.137511693171188,
      "flow_matching_loss": 0.01329552149400115,
      "grad_norm": 0.08900583535432816,
      "learning_rate": 2.980337331245864e-07,
      "loss": 0.0142,
      "relation_loss": 0.0009340395336039364,
      "selected_pairs": 26690.9,
      "state_query_norm": 41.85728302001953,
      "state_relation_loss": 7.065939935273491e-05,
      "step": 29010,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0009340395336039364,
      "weighted_relation_loss": 0.0009340395336039364
    },
    {
      "action_query_norm": 39.65528411865235,
      "action_relation_loss": 0.000858654105104506,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.14686623012161,
      "flow_matching_loss": 0.01570185236632824,
      "grad_norm": 0.10876559466123581,
      "learning_rate": 2.9205510521653213e-07,
      "loss": 0.0166,
      "relation_loss": 0.0009327861655037851,
      "selected_pairs": 26415.2,
      "state_query_norm": 41.846191024780275,
      "state_relation_loss": 7.413206622004509e-05,
      "step": 29020,
      "valid_action_samples": 230.3,
      "weighted_auxiliary_loss": 0.0009327861655037851,
      "weighted_relation_loss": 0.0009327861655037851
    },
    {
      "action_query_norm": 39.67303810119629,
      "action_relation_loss": 0.0008853790699504316,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.15622076707203,
      "flow_matching_loss": 0.016328035946935417,
      "grad_norm": 0.12854470312595367,
      "learning_rate": 2.86136877148685e-07,
      "loss": 0.0173,
      "relation_loss": 0.0009587106993421912,
      "selected_pairs": 27251.0,
      "state_query_norm": 41.84789962768555,
      "state_relation_loss": 7.333163157454691e-05,
      "step": 29030,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009587106993421912,
      "weighted_relation_loss": 0.0009587106993421912
    },
    {
      "action_query_norm": 39.66117858886719,
      "action_relation_loss": 0.000892601883970201,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.16557530402245,
      "flow_matching_loss": 0.01514557022601366,
      "grad_norm": 0.13169743120670319,
      "learning_rate": 2.8027905611223704e-07,
      "loss": 0.0161,
      "relation_loss": 0.0009635671391151846,
      "selected_pairs": 27244.7,
      "state_query_norm": 41.85586929321289,
      "state_relation_loss": 7.096526969689876e-05,
      "step": 29040,
      "valid_action_samples": 233.9,
      "weighted_auxiliary_loss": 0.0009635671391151846,
      "weighted_relation_loss": 0.0009635671391151846
    },
    {
      "action_query_norm": 39.65888671875,
      "action_relation_loss": 0.000805976358242333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.174929840972872,
      "flow_matching_loss": 0.015442149434238672,
      "grad_norm": 0.09091610461473465,
      "learning_rate": 2.7448164922500573e-07,
      "loss": 0.0163,
      "relation_loss": 0.0008761935867369175,
      "selected_pairs": 26881.1,
      "state_query_norm": 41.8560546875,
      "state_relation_loss": 7.021722303761635e-05,
      "step": 29050,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0008761935867369175,
      "weighted_relation_loss": 0.0008761935867369175
    },
    {
      "action_query_norm": 39.647233200073245,
      "action_relation_loss": 0.0008505293691996485,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.184284377923294,
      "flow_matching_loss": 0.015892752166837453,
      "grad_norm": 0.10529579222202301,
      "learning_rate": 2.687446635313784e-07,
      "loss": 0.0168,
      "relation_loss": 0.000922801491105929,
      "selected_pairs": 26736.4,
      "state_query_norm": 41.843739700317386,
      "state_relation_loss": 7.227212699945085e-05,
      "step": 29060,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.000922801491105929,
      "weighted_relation_loss": 0.000922801491105929
    },
    {
      "action_query_norm": 39.6738166809082,
      "action_relation_loss": 0.0008286795229651034,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.193638914873713,
      "flow_matching_loss": 0.01581536130979657,
      "grad_norm": 0.07520264387130737,
      "learning_rate": 2.630681060023343e-07,
      "loss": 0.0167,
      "relation_loss": 0.000901704007992521,
      "selected_pairs": 26973.5,
      "state_query_norm": 41.83982162475586,
      "state_relation_loss": 7.302447338588535e-05,
      "step": 29070,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.000901704007992521,
      "weighted_relation_loss": 0.000901704007992521
    },
    {
      "action_query_norm": 39.694832992553714,
      "action_relation_loss": 0.0008491465006954968,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.202993451824135,
      "flow_matching_loss": 0.01597609808668494,
      "grad_norm": 0.09564127773046494,
      "learning_rate": 2.5745198353542833e-07,
      "loss": 0.0169,
      "relation_loss": 0.0009239064645953476,
      "selected_pairs": 27291.9,
      "state_query_norm": 41.8461727142334,
      "state_relation_loss": 7.475997917936184e-05,
      "step": 29080,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0009239064645953476,
      "weighted_relation_loss": 0.0009239064645953476
    },
    {
      "action_query_norm": 39.57882575988769,
      "action_relation_loss": 0.0008577435975894332,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.212347988774557,
      "flow_matching_loss": 0.013651513960212469,
      "grad_norm": 0.08981341868638992,
      "learning_rate": 2.518963029547794e-07,
      "loss": 0.0146,
      "relation_loss": 0.0009271429036743939,
      "selected_pairs": 26821.9,
      "state_query_norm": 41.84629936218262,
      "state_relation_loss": 6.93993119057268e-05,
      "step": 29090,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009271429036743939,
      "weighted_relation_loss": 0.0009271429036743939
    },
    {
      "action_query_norm": 39.68475875854492,
      "action_relation_loss": 0.0008189171843696385,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.221702525724975,
      "flow_matching_loss": 0.01399768739938736,
      "grad_norm": 0.10282093286514282,
      "learning_rate": 2.464010710110598e-07,
      "loss": 0.0149,
      "relation_loss": 0.0008881885674782098,
      "selected_pairs": 26685.8,
      "state_query_norm": 41.859025955200195,
      "state_relation_loss": 6.927138128958177e-05,
      "step": 29100,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0008881885674782098,
      "weighted_relation_loss": 0.0008881885674782098
    },
    {
      "action_query_norm": 39.615610885620114,
      "action_relation_loss": 0.0008808488550130278,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.231057062675397,
      "flow_matching_loss": 0.014886588882654905,
      "grad_norm": 0.10758261382579803,
      "learning_rate": 2.4096629438150054e-07,
      "loss": 0.0158,
      "relation_loss": 0.0009536414989270269,
      "selected_pairs": 27080.2,
      "state_query_norm": 41.842438507080075,
      "state_relation_loss": 7.279265992110595e-05,
      "step": 29110,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.0009536414989270269,
      "weighted_relation_loss": 0.0009536414989270269
    },
    {
      "action_query_norm": 39.659010314941405,
      "action_relation_loss": 0.0008738217933569103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.24041159962582,
      "flow_matching_loss": 0.016296081710606813,
      "grad_norm": 0.14663560688495636,
      "learning_rate": 2.3559197966985802e-07,
      "loss": 0.0172,
      "relation_loss": 0.0009434471779968589,
      "selected_pairs": 26675.5,
      "state_query_norm": 41.861503219604494,
      "state_relation_loss": 6.962538318475709e-05,
      "step": 29120,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0009434471779968589,
      "weighted_relation_loss": 0.0009434471779968589
    },
    {
      "action_query_norm": 39.675604248046874,
      "action_relation_loss": 0.0008299152017571032,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.24976613657624,
      "flow_matching_loss": 0.014728987775743007,
      "grad_norm": 0.08032910525798798,
      "learning_rate": 2.3027813340644188e-07,
      "loss": 0.0156,
      "relation_loss": 0.000903541297884658,
      "selected_pairs": 27268.6,
      "state_query_norm": 41.84622459411621,
      "state_relation_loss": 7.362610558629968e-05,
      "step": 29130,
      "valid_action_samples": 234.0,
      "weighted_auxiliary_loss": 0.000903541297884658,
      "weighted_relation_loss": 0.000903541297884658
    },
    {
      "action_query_norm": 39.68092994689941,
      "action_relation_loss": 0.0008622008608654141,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.25912067352666,
      "flow_matching_loss": 0.016641626227647068,
      "grad_norm": 0.0883960947394371,
      "learning_rate": 2.2502476204807055e-07,
      "loss": 0.0176,
      "relation_loss": 0.0009298541001044214,
      "selected_pairs": 27477.1,
      "state_query_norm": 41.86327209472656,
      "state_relation_loss": 6.765323305444327e-05,
      "step": 29140,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0009298541001044214,
      "weighted_relation_loss": 0.0009298541001044214
    },
    {
      "action_query_norm": 39.574425506591794,
      "action_relation_loss": 0.000865565543062985,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.26847521047708,
      "flow_matching_loss": 0.014705355558544398,
      "grad_norm": 0.14670561254024506,
      "learning_rate": 2.1983187197808786e-07,
      "loss": 0.0156,
      "relation_loss": 0.0009314214170444757,
      "selected_pairs": 26501.6,
      "state_query_norm": 41.84529571533203,
      "state_relation_loss": 6.585586561413947e-05,
      "step": 29150,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.0009314214170444757,
      "weighted_relation_loss": 0.0009314214170444757
    },
    {
      "action_query_norm": 39.6411636352539,
      "action_relation_loss": 0.0008655435813125223,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.277829747427504,
      "flow_matching_loss": 0.01391857974231243,
      "grad_norm": 0.0681997686624527,
      "learning_rate": 2.1469946950634644e-07,
      "loss": 0.0149,
      "relation_loss": 0.0009372431377414614,
      "selected_pairs": 26661.5,
      "state_query_norm": 41.850059509277344,
      "state_relation_loss": 7.169956370489672e-05,
      "step": 29160,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0009372431377414614,
      "weighted_relation_loss": 0.0009372431377414614
    },
    {
      "action_query_norm": 39.641218185424805,
      "action_relation_loss": 0.0008497086993884296,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.287184284377922,
      "flow_matching_loss": 0.017031923960894348,
      "grad_norm": 0.08207210898399353,
      "learning_rate": 2.096275608691911e-07,
      "loss": 0.018,
      "relation_loss": 0.0009235502977389843,
      "selected_pairs": 26831.2,
      "state_query_norm": 41.84820289611817,
      "state_relation_loss": 7.384159907815046e-05,
      "step": 29170,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009235502977389843,
      "weighted_relation_loss": 0.0009235502977389843
    },
    {
      "action_query_norm": 39.63405990600586,
      "action_relation_loss": 0.0008971785253379494,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.296538821328344,
      "flow_matching_loss": 0.01438880180940032,
      "grad_norm": 0.11226276308298111,
      "learning_rate": 2.046161522294754e-07,
      "loss": 0.0154,
      "relation_loss": 0.0009703618532512337,
      "selected_pairs": 26462.4,
      "state_query_norm": 41.84394874572754,
      "state_relation_loss": 7.318331918213517e-05,
      "step": 29180,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009703618532512337,
      "weighted_relation_loss": 0.0009703618532512337
    },
    {
      "action_query_norm": 39.66330986022949,
      "action_relation_loss": 0.0008654047560412437,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.305893358278766,
      "flow_matching_loss": 0.013835979718714952,
      "grad_norm": 0.0852721631526947,
      "learning_rate": 1.9966524967653944e-07,
      "loss": 0.0148,
      "relation_loss": 0.00093944157124497,
      "selected_pairs": 27080.7,
      "state_query_norm": 41.85333366394043,
      "state_relation_loss": 7.403682902804576e-05,
      "step": 29190,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.00093944157124497,
      "weighted_relation_loss": 0.00093944157124497
    },
    {
      "action_query_norm": 39.640267562866214,
      "action_relation_loss": 0.000853920984081924,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.315247895229184,
      "flow_matching_loss": 0.015187699627131224,
      "grad_norm": 0.09343796968460083,
      "learning_rate": 1.9477485922618222e-07,
      "loss": 0.0161,
      "relation_loss": 0.0009258176200091839,
      "selected_pairs": 26930.0,
      "state_query_norm": 41.84671287536621,
      "state_relation_loss": 7.18966344720684e-05,
      "step": 29200,
      "valid_action_samples": 232.5,
      "weighted_auxiliary_loss": 0.0009258176200091839,
      "weighted_relation_loss": 0.0009258176200091839
    },
    {
      "action_query_norm": 39.65196685791015,
      "action_relation_loss": 0.0008891630626749247,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.324602432179606,
      "flow_matching_loss": 0.01457139365375042,
      "grad_norm": 0.12616023421287537,
      "learning_rate": 1.899449868207004e-07,
      "loss": 0.0155,
      "relation_loss": 0.0009604884078726172,
      "selected_pairs": 27219.7,
      "state_query_norm": 41.85857887268067,
      "state_relation_loss": 7.132534665288404e-05,
      "step": 29210,
      "valid_action_samples": 233.8,
      "weighted_auxiliary_loss": 0.0009604884078726172,
      "weighted_relation_loss": 0.0009604884078726172
    },
    {
      "action_query_norm": 39.64900932312012,
      "action_relation_loss": 0.0008483732992317528,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.33395696913003,
      "flow_matching_loss": 0.014520261622965335,
      "grad_norm": 0.12579520046710968,
      "learning_rate": 1.851756383288439e-07,
      "loss": 0.0154,
      "relation_loss": 0.0009177293279208243,
      "selected_pairs": 27020.5,
      "state_query_norm": 41.832845306396486,
      "state_relation_loss": 6.935602177691181e-05,
      "step": 29220,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009177293279208243,
      "weighted_relation_loss": 0.0009177293279208243
    },
    {
      "action_query_norm": 39.66847381591797,
      "action_relation_loss": 0.0008438663906417787,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.34331150608045,
      "flow_matching_loss": 0.014684491697698832,
      "grad_norm": 0.16181586682796478,
      "learning_rate": 1.8046681954581035e-07,
      "loss": 0.0156,
      "relation_loss": 0.0009149046614766121,
      "selected_pairs": 26618.3,
      "state_query_norm": 41.854754257202146,
      "state_relation_loss": 7.103826392267365e-05,
      "step": 29230,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0009149046614766121,
      "weighted_relation_loss": 0.0009149046614766121
    },
    {
      "action_query_norm": 39.63698310852051,
      "action_relation_loss": 0.0008669713162817061,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.35266604303087,
      "flow_matching_loss": 0.014429999236017465,
      "grad_norm": 0.07899461686611176,
      "learning_rate": 1.7581853619327294e-07,
      "loss": 0.0154,
      "relation_loss": 0.0009380473289638758,
      "selected_pairs": 26754.3,
      "state_query_norm": 41.85325584411621,
      "state_relation_loss": 7.107599885785021e-05,
      "step": 29240,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009380473289638758,
      "weighted_relation_loss": 0.0009380473289638758
    },
    {
      "action_query_norm": 39.627618408203126,
      "action_relation_loss": 0.0008468830317724496,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.36202057998129,
      "flow_matching_loss": 0.014383348543196917,
      "grad_norm": 0.08362231403589249,
      "learning_rate": 1.7123079391932472e-07,
      "loss": 0.0153,
      "relation_loss": 0.0009142011404037476,
      "selected_pairs": 27107.3,
      "state_query_norm": 41.85996551513672,
      "state_relation_loss": 6.731811590725556e-05,
      "step": 29250,
      "valid_action_samples": 233.3,
      "weighted_auxiliary_loss": 0.0009142011404037476,
      "weighted_relation_loss": 0.0009142011404037476
    },
    {
      "action_query_norm": 39.6520076751709,
      "action_relation_loss": 0.0008367626171093435,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.371375116931713,
      "flow_matching_loss": 0.014747098367661238,
      "grad_norm": 0.11078572273254395,
      "learning_rate": 1.6670359829850657e-07,
      "loss": 0.0157,
      "relation_loss": 0.0009046309802215547,
      "selected_pairs": 26942.3,
      "state_query_norm": 41.864244842529295,
      "state_relation_loss": 6.786835147067904e-05,
      "step": 29260,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009046309802215547,
      "weighted_relation_loss": 0.0009046309802215547
    },
    {
      "action_query_norm": 39.6389820098877,
      "action_relation_loss": 0.0008343318884726613,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.38072965388213,
      "flow_matching_loss": 0.014352597761899233,
      "grad_norm": 0.08320055902004242,
      "learning_rate": 1.6223695483179037e-07,
      "loss": 0.0153,
      "relation_loss": 0.0009060001058969647,
      "selected_pairs": 26880.1,
      "state_query_norm": 41.83717727661133,
      "state_relation_loss": 7.166821524151601e-05,
      "step": 29270,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009060001058969647,
      "weighted_relation_loss": 0.0009060001058969647
    },
    {
      "action_query_norm": 39.58853416442871,
      "action_relation_loss": 0.0008598899119533598,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.390084190832553,
      "flow_matching_loss": 0.016755907516926526,
      "grad_norm": 0.12825927138328552,
      "learning_rate": 1.5783086894656795e-07,
      "loss": 0.0177,
      "relation_loss": 0.0009316578041762113,
      "selected_pairs": 26976.1,
      "state_query_norm": 41.86686477661133,
      "state_relation_loss": 7.176790822995826e-05,
      "step": 29280,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009316578041762113,
      "weighted_relation_loss": 0.0009316578041762113
    },
    {
      "action_query_norm": 39.66921157836914,
      "action_relation_loss": 0.0008357336046174169,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.399438727782975,
      "flow_matching_loss": 0.015837958548218012,
      "grad_norm": 0.10234210640192032,
      "learning_rate": 1.5348534599665121e-07,
      "loss": 0.0167,
      "relation_loss": 0.000903996912529692,
      "selected_pairs": 26027.0,
      "state_query_norm": 41.858926391601564,
      "state_relation_loss": 6.826330791227519e-05,
      "step": 29290,
      "valid_action_samples": 228.6,
      "weighted_auxiliary_loss": 0.000903996912529692,
      "weighted_relation_loss": 0.000903996912529692
    },
    {
      "action_query_norm": 39.62960700988769,
      "action_relation_loss": 0.0008569037774577737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.408793264733397,
      "flow_matching_loss": 0.01463030818849802,
      "grad_norm": 0.11018048971891403,
      "learning_rate": 1.4920039126225527e-07,
      "loss": 0.0156,
      "relation_loss": 0.0009282489714678377,
      "selected_pairs": 27033.8,
      "state_query_norm": 41.858206558227536,
      "state_relation_loss": 7.13451903720852e-05,
      "step": 29300,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009282489714678377,
      "weighted_relation_loss": 0.0009282489714678377
    },
    {
      "action_query_norm": 39.65233306884765,
      "action_relation_loss": 0.0008511380234267563,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.418147801683816,
      "flow_matching_loss": 0.015383929572999477,
      "grad_norm": 0.07342327386140823,
      "learning_rate": 1.449760099500097e-07,
      "loss": 0.0163,
      "relation_loss": 0.0009229626622982324,
      "selected_pairs": 26390.8,
      "state_query_norm": 41.84255027770996,
      "state_relation_loss": 7.182464396464639e-05,
      "step": 29310,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0009229626622982324,
      "weighted_relation_loss": 0.0009229626622982324
    },
    {
      "action_query_norm": 39.60939979553223,
      "action_relation_loss": 0.0008481741999275983,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.427502338634238,
      "flow_matching_loss": 0.01607105853036046,
      "grad_norm": 0.06622724235057831,
      "learning_rate": 1.4081220719293075e-07,
      "loss": 0.017,
      "relation_loss": 0.0009161336987745017,
      "selected_pairs": 27198.1,
      "state_query_norm": 41.85150985717773,
      "state_relation_loss": 6.79594959365204e-05,
      "step": 29320,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009161336987745017,
      "weighted_relation_loss": 0.0009161336987745017
    },
    {
      "action_query_norm": 39.67853698730469,
      "action_relation_loss": 0.0008715158968698233,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.43685687558466,
      "flow_matching_loss": 0.013775940611958503,
      "grad_norm": 0.14725252985954285,
      "learning_rate": 1.3670898805043797e-07,
      "loss": 0.0147,
      "relation_loss": 0.0009421199676580728,
      "selected_pairs": 26898.8,
      "state_query_norm": 41.853010177612305,
      "state_relation_loss": 7.06040609657066e-05,
      "step": 29330,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009421199676580728,
      "weighted_relation_loss": 0.0009421199676580728
    },
    {
      "action_query_norm": 39.59255828857422,
      "action_relation_loss": 0.0008800457988400012,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.446211412535078,
      "flow_matching_loss": 0.015071311686187982,
      "grad_norm": 0.0809757336974144,
      "learning_rate": 1.326663575083209e-07,
      "loss": 0.016,
      "relation_loss": 0.0009496493672486395,
      "selected_pairs": 26940.5,
      "state_query_norm": 41.861502456665036,
      "state_relation_loss": 6.960355312912725e-05,
      "step": 29340,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009496493672486395,
      "weighted_relation_loss": 0.0009496493672486395
    },
    {
      "action_query_norm": 39.645718002319335,
      "action_relation_loss": 0.0008547610137611628,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.4555659494855,
      "flow_matching_loss": 0.016361195780336856,
      "grad_norm": 0.15148910880088806,
      "learning_rate": 1.2868432047876688e-07,
      "loss": 0.0173,
      "relation_loss": 0.0009270464885048568,
      "selected_pairs": 26902.1,
      "state_query_norm": 41.856890487670896,
      "state_relation_loss": 7.228546492115129e-05,
      "step": 29350,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009270464885048568,
      "weighted_relation_loss": 0.0009270464885048568
    },
    {
      "action_query_norm": 39.660007095336915,
      "action_relation_loss": 0.0008300901390612125,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.464920486435922,
      "flow_matching_loss": 0.0161772508174181,
      "grad_norm": 0.09311172366142273,
      "learning_rate": 1.2476288180032213e-07,
      "loss": 0.0171,
      "relation_loss": 0.0009025175822898746,
      "selected_pairs": 27131.4,
      "state_query_norm": 41.85389518737793,
      "state_relation_loss": 7.24274417734705e-05,
      "step": 29360,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009025175822898746,
      "weighted_relation_loss": 0.0009025175822898746
    },
    {
      "action_query_norm": 39.66616973876953,
      "action_relation_loss": 0.000885999680031091,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.474275023386344,
      "flow_matching_loss": 0.015010762307792903,
      "grad_norm": 0.10504524409770966,
      "learning_rate": 1.209020462379029e-07,
      "loss": 0.016,
      "relation_loss": 0.0009588993911165744,
      "selected_pairs": 26826.4,
      "state_query_norm": 41.85594863891602,
      "state_relation_loss": 7.289971399586648e-05,
      "step": 29370,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009588993911165744,
      "weighted_relation_loss": 0.0009588993911165744
    },
    {
      "action_query_norm": 39.65252151489258,
      "action_relation_loss": 0.0008764220809098333,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.483629560336762,
      "flow_matching_loss": 0.014586366899311542,
      "grad_norm": 0.08333509415388107,
      "learning_rate": 1.1710181848278435e-07,
      "loss": 0.0155,
      "relation_loss": 0.0009506510861683637,
      "selected_pairs": 27408.3,
      "state_query_norm": 41.83324851989746,
      "state_relation_loss": 7.422899652738124e-05,
      "step": 29380,
      "valid_action_samples": 234.6,
      "weighted_auxiliary_loss": 0.0009506510861683637,
      "weighted_relation_loss": 0.0009506510861683637
    },
    {
      "action_query_norm": 39.62266616821289,
      "action_relation_loss": 0.000868211273336783,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.492984097287184,
      "flow_matching_loss": 0.014341125171631575,
      "grad_norm": 0.14386554062366486,
      "learning_rate": 1.133622031526116e-07,
      "loss": 0.0153,
      "relation_loss": 0.0009381846059113741,
      "selected_pairs": 26802.8,
      "state_query_norm": 41.849874114990236,
      "state_relation_loss": 6.997333475737833e-05,
      "step": 29390,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009381846059113741,
      "weighted_relation_loss": 0.0009381846059113741
    },
    {
      "action_query_norm": 39.666205978393556,
      "action_relation_loss": 0.0008312012185342609,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.502338634237606,
      "flow_matching_loss": 0.01408311827108264,
      "grad_norm": 0.09412187337875366,
      "learning_rate": 1.0968320479136095e-07,
      "loss": 0.015,
      "relation_loss": 0.0009015340590849518,
      "selected_pairs": 26826.5,
      "state_query_norm": 41.858267593383786,
      "state_relation_loss": 7.03328383679036e-05,
      "step": 29400,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009015340590849518,
      "weighted_relation_loss": 0.0009015340590849518
    },
    {
      "action_query_norm": 39.666493606567386,
      "action_relation_loss": 0.0008952971256803721,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.511693171188025,
      "flow_matching_loss": 0.015503215789794921,
      "grad_norm": 0.14797315001487732,
      "learning_rate": 1.0606482786936767e-07,
      "loss": 0.0165,
      "relation_loss": 0.0009639647149015218,
      "selected_pairs": 26880.6,
      "state_query_norm": 41.86504135131836,
      "state_relation_loss": 6.866758776595816e-05,
      "step": 29410,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009639647149015218,
      "weighted_relation_loss": 0.0009639647149015218
    },
    {
      "action_query_norm": 39.64656410217285,
      "action_relation_loss": 0.0008491952321492136,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.521047708138447,
      "flow_matching_loss": 0.016496119555085897,
      "grad_norm": 0.07807393372058868,
      "learning_rate": 1.0250707678329808e-07,
      "loss": 0.0174,
      "relation_loss": 0.0009207117022015154,
      "selected_pairs": 26209.2,
      "state_query_norm": 41.84652862548828,
      "state_relation_loss": 7.151647187129129e-05,
      "step": 29420,
      "valid_action_samples": 229.4,
      "weighted_auxiliary_loss": 0.0009207117022015154,
      "weighted_relation_loss": 0.0009207117022015154
    },
    {
      "action_query_norm": 39.624993896484376,
      "action_relation_loss": 0.000848688097903505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.53040224508887,
      "flow_matching_loss": 0.015793965384364127,
      "grad_norm": 0.08809670805931091,
      "learning_rate": 9.900995585615525e-08,
      "loss": 0.0167,
      "relation_loss": 0.0009195182414259761,
      "selected_pairs": 26810.7,
      "state_query_norm": 41.86383399963379,
      "state_relation_loss": 7.083013879309874e-05,
      "step": 29430,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009195182414259761,
      "weighted_relation_loss": 0.0009195182414259761
    },
    {
      "action_query_norm": 39.62343902587891,
      "action_relation_loss": 0.0008787794038653373,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.53975678203929,
      "flow_matching_loss": 0.01645902534946799,
      "grad_norm": 0.08659353107213974,
      "learning_rate": 9.55734693372734e-08,
      "loss": 0.0174,
      "relation_loss": 0.0009509954776149243,
      "selected_pairs": 26461.8,
      "state_query_norm": 41.84008255004883,
      "state_relation_loss": 7.221607665996999e-05,
      "step": 29440,
      "valid_action_samples": 230.5,
      "weighted_auxiliary_loss": 0.0009509954776149243,
      "weighted_relation_loss": 0.0009509954776149243
    },
    {
      "action_query_norm": 39.599002075195315,
      "action_relation_loss": 0.00086666748393327,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.54911131898971,
      "flow_matching_loss": 0.01455099955201149,
      "grad_norm": 0.09451967477798462,
      "learning_rate": 9.219762140231236e-08,
      "loss": 0.0155,
      "relation_loss": 0.0009348879626486451,
      "selected_pairs": 26781.2,
      "state_query_norm": 41.85598182678223,
      "state_relation_loss": 6.822047871537506e-05,
      "step": 29450,
      "valid_action_samples": 231.9,
      "weighted_auxiliary_loss": 0.0009348879626486451,
      "weighted_relation_loss": 0.0009348879626486451
    },
    {
      "action_query_norm": 39.60804138183594,
      "action_relation_loss": 0.0008573288097977639,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.55846585594013,
      "flow_matching_loss": 0.01510022496804595,
      "grad_norm": 0.23753505945205688,
      "learning_rate": 8.888241615322978e-08,
      "loss": 0.016,
      "relation_loss": 0.0009245611901860685,
      "selected_pairs": 27061.8,
      "state_query_norm": 41.851264572143556,
      "state_relation_loss": 6.723235928802751e-05,
      "step": 29460,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009245611901860685,
      "weighted_relation_loss": 0.0009245611901860685
    },
    {
      "action_query_norm": 39.605418014526364,
      "action_relation_loss": 0.0008687550143804401,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.567820392890553,
      "flow_matching_loss": 0.015685200039297343,
      "grad_norm": 0.08105095475912094,
      "learning_rate": 8.562785761833114e-08,
      "loss": 0.0166,
      "relation_loss": 0.0009433106402866543,
      "selected_pairs": 26577.7,
      "state_query_norm": 41.84338264465332,
      "state_relation_loss": 7.455562008544803e-05,
      "step": 29470,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0009433106402866543,
      "weighted_relation_loss": 0.0009433106402866543
    },
    {
      "action_query_norm": 39.63513259887695,
      "action_relation_loss": 0.0008447539468761534,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.57717492984097,
      "flow_matching_loss": 0.0138771609403193,
      "grad_norm": 0.08492527157068253,
      "learning_rate": 8.243394975219753e-08,
      "loss": 0.0148,
      "relation_loss": 0.0009151354315690696,
      "selected_pairs": 27336.7,
      "state_query_norm": 41.85592041015625,
      "state_relation_loss": 7.03814879670972e-05,
      "step": 29480,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.0009151354315690696,
      "weighted_relation_loss": 0.0009151354315690696
    },
    {
      "action_query_norm": 39.63514175415039,
      "action_relation_loss": 0.0008833684551063925,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.586529466791394,
      "flow_matching_loss": 0.016149234492331743,
      "grad_norm": 0.10559122264385223,
      "learning_rate": 7.930069643573568e-08,
      "loss": 0.0171,
      "relation_loss": 0.000957246741745621,
      "selected_pairs": 27339.0,
      "state_query_norm": 41.86624946594238,
      "state_relation_loss": 7.387827936327085e-05,
      "step": 29490,
      "valid_action_samples": 234.3,
      "weighted_auxiliary_loss": 0.000957246741745621,
      "weighted_relation_loss": 0.000957246741745621
    },
    {
      "action_query_norm": 39.591957092285156,
      "action_relation_loss": 0.0008927281887736171,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.595884003741816,
      "flow_matching_loss": 0.015482277330011129,
      "grad_norm": 0.10240428149700165,
      "learning_rate": 7.622810147614456e-08,
      "loss": 0.0164,
      "relation_loss": 0.0009626518876757473,
      "selected_pairs": 26743.2,
      "state_query_norm": 41.85231323242188,
      "state_relation_loss": 6.992369235376827e-05,
      "step": 29500,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009626518876757473,
      "weighted_relation_loss": 0.0009626518876757473
    },
    {
      "action_query_norm": 39.62583541870117,
      "action_relation_loss": 0.0008546864555682987,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.605238540692234,
      "flow_matching_loss": 0.014725399110466242,
      "grad_norm": 0.10355565696954727,
      "learning_rate": 7.321616860690995e-08,
      "loss": 0.0157,
      "relation_loss": 0.0009256846562493593,
      "selected_pairs": 27046.2,
      "state_query_norm": 41.846356582641604,
      "state_relation_loss": 7.099819376890081e-05,
      "step": 29510,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009256846562493593,
      "weighted_relation_loss": 0.0009256846562493593
    },
    {
      "action_query_norm": 39.56326026916504,
      "action_relation_loss": 0.0008764915924984962,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.614593077642656,
      "flow_matching_loss": 0.016764580551534892,
      "grad_norm": 0.08486569672822952,
      "learning_rate": 7.026490148782095e-08,
      "loss": 0.0177,
      "relation_loss": 0.0009459393739234656,
      "selected_pairs": 26688.2,
      "state_query_norm": 41.8345386505127,
      "state_relation_loss": 6.944778069737368e-05,
      "step": 29520,
      "valid_action_samples": 231.5,
      "weighted_auxiliary_loss": 0.0009459393739234656,
      "weighted_relation_loss": 0.0009459393739234656
    },
    {
      "action_query_norm": 39.65273513793945,
      "action_relation_loss": 0.000861974322469905,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.623947614593078,
      "flow_matching_loss": 0.015152902342379093,
      "grad_norm": 0.12199165672063828,
      "learning_rate": 6.737430370494236e-08,
      "loss": 0.0161,
      "relation_loss": 0.0009362037584651261,
      "selected_pairs": 27292.7,
      "state_query_norm": 41.83314933776855,
      "state_relation_loss": 7.422944545396604e-05,
      "step": 29530,
      "valid_action_samples": 234.1,
      "weighted_auxiliary_loss": 0.0009362037584651261,
      "weighted_relation_loss": 0.0009362037584651261
    },
    {
      "action_query_norm": 39.65992965698242,
      "action_relation_loss": 0.0009047744853887707,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.6333021515435,
      "flow_matching_loss": 0.015477272123098374,
      "grad_norm": 0.10470261424779892,
      "learning_rate": 6.454437877062569e-08,
      "loss": 0.0165,
      "relation_loss": 0.000974648114060983,
      "selected_pairs": 26433.9,
      "state_query_norm": 41.857229232788086,
      "state_relation_loss": 6.987362939980812e-05,
      "step": 29540,
      "valid_action_samples": 230.4,
      "weighted_auxiliary_loss": 0.000974648114060983,
      "weighted_relation_loss": 0.000974648114060983
    },
    {
      "action_query_norm": 39.61976509094238,
      "action_relation_loss": 0.0008609130221884697,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.64265668849392,
      "flow_matching_loss": 0.015050254389643669,
      "grad_norm": 0.08810573071241379,
      "learning_rate": 6.177513012349256e-08,
      "loss": 0.016,
      "relation_loss": 0.0009274281852412969,
      "selected_pairs": 26987.9,
      "state_query_norm": 41.8682975769043,
      "state_relation_loss": 6.651517178397625e-05,
      "step": 29550,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009274281852412969,
      "weighted_relation_loss": 0.0009274281852412969
    },
    {
      "action_query_norm": 39.58647155761719,
      "action_relation_loss": 0.000863615027628839,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.65201122544434,
      "flow_matching_loss": 0.014547010883688927,
      "grad_norm": 0.17014074325561523,
      "learning_rate": 5.9066561128445775e-08,
      "loss": 0.0155,
      "relation_loss": 0.0009314619121141732,
      "selected_pairs": 27199.1,
      "state_query_norm": 41.83814468383789,
      "state_relation_loss": 6.784687502658926e-05,
      "step": 29560,
      "valid_action_samples": 233.7,
      "weighted_auxiliary_loss": 0.0009314619121141732,
      "weighted_relation_loss": 0.0009314619121141732
    },
    {
      "action_query_norm": 39.64404182434082,
      "action_relation_loss": 0.0008655427489429712,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.661365762394762,
      "flow_matching_loss": 0.01571812815964222,
      "grad_norm": 0.08683671802282333,
      "learning_rate": 5.6418675076641556e-08,
      "loss": 0.0167,
      "relation_loss": 0.0009392385720275343,
      "selected_pairs": 26895.2,
      "state_query_norm": 41.84986686706543,
      "state_relation_loss": 7.369582453975454e-05,
      "step": 29570,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009392385720275343,
      "weighted_relation_loss": 0.0009392385720275343
    },
    {
      "action_query_norm": 39.70482482910156,
      "action_relation_loss": 0.0008371080795768649,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.67072029934518,
      "flow_matching_loss": 0.0156644226051867,
      "grad_norm": 0.1395239233970642,
      "learning_rate": 5.383147518552845e-08,
      "loss": 0.0166,
      "relation_loss": 0.0009044790058396757,
      "selected_pairs": 27481.5,
      "state_query_norm": 41.860877990722656,
      "state_relation_loss": 6.737092990078963e-05,
      "step": 29580,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0009044790058396757,
      "weighted_relation_loss": 0.0009044790058396757
    },
    {
      "action_query_norm": 39.678015518188474,
      "action_relation_loss": 0.0008065496687777341,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.680074836295603,
      "flow_matching_loss": 0.015479712188243866,
      "grad_norm": 0.08576733618974686,
      "learning_rate": 5.1304964598786196e-08,
      "loss": 0.0164,
      "relation_loss": 0.0008772657485678792,
      "selected_pairs": 26572.2,
      "state_query_norm": 41.865835189819336,
      "state_relation_loss": 7.071608852129429e-05,
      "step": 29590,
      "valid_action_samples": 231.0,
      "weighted_auxiliary_loss": 0.0008772657485678792,
      "weighted_relation_loss": 0.0008772657485678792
    },
    {
      "action_query_norm": 39.616733932495116,
      "action_relation_loss": 0.000843379384605214,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.689429373246025,
      "flow_matching_loss": 0.014942145720124245,
      "grad_norm": 0.16197171807289124,
      "learning_rate": 4.883914638636467e-08,
      "loss": 0.0159,
      "relation_loss": 0.0009137841581832618,
      "selected_pairs": 26477.0,
      "state_query_norm": 41.84239120483399,
      "state_relation_loss": 7.04047706676647e-05,
      "step": 29600,
      "valid_action_samples": 230.6,
      "weighted_auxiliary_loss": 0.0009137841581832618,
      "weighted_relation_loss": 0.0009137841581832618
    },
    {
      "action_query_norm": 39.64512825012207,
      "action_relation_loss": 0.000838751852279529,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.698783910196447,
      "flow_matching_loss": 0.014470091369003058,
      "grad_norm": 0.0706259086728096,
      "learning_rate": 4.643402354446713e-08,
      "loss": 0.0154,
      "relation_loss": 0.000911760434973985,
      "selected_pairs": 27080.7,
      "state_query_norm": 41.847465133666994,
      "state_relation_loss": 7.300857942027505e-05,
      "step": 29610,
      "valid_action_samples": 233.2,
      "weighted_auxiliary_loss": 0.000911760434973985,
      "weighted_relation_loss": 0.000911760434973985
    },
    {
      "action_query_norm": 39.67167282104492,
      "action_relation_loss": 0.0008611553173977881,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.708138447146865,
      "flow_matching_loss": 0.015753974672406912,
      "grad_norm": 0.08091506361961365,
      "learning_rate": 4.4089598995544766e-08,
      "loss": 0.0167,
      "relation_loss": 0.0009315399976912886,
      "selected_pairs": 27015.6,
      "state_query_norm": 41.86689376831055,
      "state_relation_loss": 7.038467592792585e-05,
      "step": 29620,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009315399976912886,
      "weighted_relation_loss": 0.0009315399976912886
    },
    {
      "action_query_norm": 39.64918212890625,
      "action_relation_loss": 0.0008662849781103432,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.717492984097287,
      "flow_matching_loss": 0.014016858953982591,
      "grad_norm": 0.0652320384979248,
      "learning_rate": 4.180587558829663e-08,
      "loss": 0.015,
      "relation_loss": 0.0009370954183395952,
      "selected_pairs": 26874.9,
      "state_query_norm": 41.851545333862305,
      "state_relation_loss": 7.081044532242231e-05,
      "step": 29630,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009370954183395952,
      "weighted_relation_loss": 0.0009370954183395952
    },
    {
      "action_query_norm": 39.614312744140626,
      "action_relation_loss": 0.00088321243529208,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.72684752104771,
      "flow_matching_loss": 0.016678205877542495,
      "grad_norm": 0.1498451828956604,
      "learning_rate": 3.958285609765855e-08,
      "loss": 0.0176,
      "relation_loss": 0.0009540494298562408,
      "selected_pairs": 26825.2,
      "state_query_norm": 41.846757125854495,
      "state_relation_loss": 7.083700729708653e-05,
      "step": 29640,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009540494298562408,
      "weighted_relation_loss": 0.0009540494298562408
    },
    {
      "action_query_norm": 39.634564590454104,
      "action_relation_loss": 0.000851284246891737,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.736202057998128,
      "flow_matching_loss": 0.014570844266563654,
      "grad_norm": 0.1709519475698471,
      "learning_rate": 3.74205432248087e-08,
      "loss": 0.0155,
      "relation_loss": 0.0009244524640962482,
      "selected_pairs": 26736.3,
      "state_query_norm": 41.84908676147461,
      "state_relation_loss": 7.316822448046877e-05,
      "step": 29650,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009244524640962482,
      "weighted_relation_loss": 0.0009244524640962482
    },
    {
      "action_query_norm": 39.57811737060547,
      "action_relation_loss": 0.0008990652975626289,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.74555659494855,
      "flow_matching_loss": 0.01508805649355054,
      "grad_norm": 0.08841537684202194,
      "learning_rate": 3.531893959716204e-08,
      "loss": 0.0161,
      "relation_loss": 0.0009653025073930622,
      "selected_pairs": 26763.2,
      "state_query_norm": 41.85414848327637,
      "state_relation_loss": 6.623720364586915e-05,
      "step": 29660,
      "valid_action_samples": 231.8,
      "weighted_auxiliary_loss": 0.0009653025073930622,
      "weighted_relation_loss": 0.0009653025073930622
    },
    {
      "action_query_norm": 39.647884368896484,
      "action_relation_loss": 0.0008681022329255939,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.75491113189897,
      "flow_matching_loss": 0.0160292686894536,
      "grad_norm": 0.08832099288702011,
      "learning_rate": 3.327804776837029e-08,
      "loss": 0.017,
      "relation_loss": 0.0009402430907357485,
      "selected_pairs": 26829.4,
      "state_query_norm": 41.84179801940918,
      "state_relation_loss": 7.214085344457999e-05,
      "step": 29670,
      "valid_action_samples": 232.1,
      "weighted_auxiliary_loss": 0.0009402430907357485,
      "weighted_relation_loss": 0.0009402430907357485
    },
    {
      "action_query_norm": 39.621561431884764,
      "action_relation_loss": 0.0008725812309421599,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.764265668849394,
      "flow_matching_loss": 0.015045323222875596,
      "grad_norm": 0.06847404688596725,
      "learning_rate": 3.129787021829977e-08,
      "loss": 0.016,
      "relation_loss": 0.0009442263166420162,
      "selected_pairs": 27131.7,
      "state_query_norm": 41.838446044921874,
      "state_relation_loss": 7.164508133428171e-05,
      "step": 29680,
      "valid_action_samples": 233.4,
      "weighted_auxiliary_loss": 0.0009442263166420162,
      "weighted_relation_loss": 0.0009442263166420162
    },
    {
      "action_query_norm": 39.61129913330078,
      "action_relation_loss": 0.0008519151364453137,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.773620205799812,
      "flow_matching_loss": 0.01480162413790822,
      "grad_norm": 0.08069384843111038,
      "learning_rate": 2.9378409353059133e-08,
      "loss": 0.0157,
      "relation_loss": 0.0009206706774421037,
      "selected_pairs": 26986.1,
      "state_query_norm": 41.834969711303714,
      "state_relation_loss": 6.875554790894967e-05,
      "step": 29690,
      "valid_action_samples": 232.8,
      "weighted_auxiliary_loss": 0.0009206706774421037,
      "weighted_relation_loss": 0.0009206706774421037
    },
    {
      "action_query_norm": 39.62871475219727,
      "action_relation_loss": 0.0008637337712571024,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.782974742750234,
      "flow_matching_loss": 0.016518817096948624,
      "grad_norm": 0.09933112561702728,
      "learning_rate": 2.7519667504971593e-08,
      "loss": 0.0175,
      "relation_loss": 0.0009325064951553941,
      "selected_pairs": 27147.4,
      "state_query_norm": 41.848645782470705,
      "state_relation_loss": 6.877271371195093e-05,
      "step": 29700,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0009325064951553941,
      "weighted_relation_loss": 0.0009325064951553941
    },
    {
      "action_query_norm": 39.660556411743165,
      "action_relation_loss": 0.0008130431990139187,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.792329279700656,
      "flow_matching_loss": 0.015256292466074229,
      "grad_norm": 0.12410236150026321,
      "learning_rate": 2.572164693258605e-08,
      "loss": 0.0161,
      "relation_loss": 0.0008864462666679174,
      "selected_pairs": 26616.7,
      "state_query_norm": 41.85203552246094,
      "state_relation_loss": 7.340307492995635e-05,
      "step": 29710,
      "valid_action_samples": 231.2,
      "weighted_auxiliary_loss": 0.0008864462666679174,
      "weighted_relation_loss": 0.0008864462666679174
    },
    {
      "action_query_norm": 39.62894630432129,
      "action_relation_loss": 0.0008882765017915517,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.801683816651074,
      "flow_matching_loss": 0.014444481395184994,
      "grad_norm": 0.10753320902585983,
      "learning_rate": 2.3984349820665996e-08,
      "loss": 0.0154,
      "relation_loss": 0.0009588894376065582,
      "selected_pairs": 26397.5,
      "state_query_norm": 41.84396629333496,
      "state_relation_loss": 7.061293836159166e-05,
      "step": 29720,
      "valid_action_samples": 230.2,
      "weighted_auxiliary_loss": 0.0009588894376065582,
      "weighted_relation_loss": 0.0009588894376065582
    },
    {
      "action_query_norm": 39.63367347717285,
      "action_relation_loss": 0.0008561695518437773,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.811038353601496,
      "flow_matching_loss": 0.01600963417440653,
      "grad_norm": 0.11287222057580948,
      "learning_rate": 2.2307778280189485e-08,
      "loss": 0.0169,
      "relation_loss": 0.0009254613658413291,
      "selected_pairs": 26730.5,
      "state_query_norm": 41.8569694519043,
      "state_relation_loss": 6.929182418389246e-05,
      "step": 29730,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009254613658413291,
      "weighted_relation_loss": 0.0009254613658413291
    },
    {
      "action_query_norm": 39.62178268432617,
      "action_relation_loss": 0.0008541085640899837,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.82039289055192,
      "flow_matching_loss": 0.01480189897119999,
      "grad_norm": 0.11003366857767105,
      "learning_rate": 2.0691934348354704e-08,
      "loss": 0.0157,
      "relation_loss": 0.0009253724361769855,
      "selected_pairs": 26941.9,
      "state_query_norm": 41.85945930480957,
      "state_relation_loss": 7.126388081815093e-05,
      "step": 29740,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.0009253724361769855,
      "weighted_relation_loss": 0.0009253724361769855
    },
    {
      "action_query_norm": 39.63847122192383,
      "action_relation_loss": 0.0008585271600168199,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.82974742750234,
      "flow_matching_loss": 0.014421493280678988,
      "grad_norm": 0.08478638529777527,
      "learning_rate": 1.9136819988557763e-08,
      "loss": 0.0154,
      "relation_loss": 0.0009316222334746271,
      "selected_pairs": 27041.0,
      "state_query_norm": 41.85948867797852,
      "state_relation_loss": 7.309507491299882e-05,
      "step": 29750,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009316222334746271,
      "weighted_relation_loss": 0.0009316222334746271
    },
    {
      "action_query_norm": 39.63339424133301,
      "action_relation_loss": 0.0008690226823091507,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.83910196445276,
      "flow_matching_loss": 0.015069208294153213,
      "grad_norm": 0.10173647850751877,
      "learning_rate": 1.7642437090414908e-08,
      "loss": 0.016,
      "relation_loss": 0.0009408701385837049,
      "selected_pairs": 26848.6,
      "state_query_norm": 41.85227012634277,
      "state_relation_loss": 7.184746573329903e-05,
      "step": 29760,
      "valid_action_samples": 232.2,
      "weighted_auxiliary_loss": 0.0009408701385837049,
      "weighted_relation_loss": 0.0009408701385837049
    },
    {
      "action_query_norm": 39.652421188354495,
      "action_relation_loss": 0.0009144962939899414,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.84845650140318,
      "flow_matching_loss": 0.014981803204864263,
      "grad_norm": 0.07910740375518799,
      "learning_rate": 1.6208787469734755e-08,
      "loss": 0.016,
      "relation_loss": 0.0009879628429189324,
      "selected_pairs": 27035.4,
      "state_query_norm": 41.85365867614746,
      "state_relation_loss": 7.346654674620368e-05,
      "step": 29770,
      "valid_action_samples": 233.0,
      "weighted_auxiliary_loss": 0.0009879628429189324,
      "weighted_relation_loss": 0.0009879628429189324
    },
    {
      "action_query_norm": 39.643026733398436,
      "action_relation_loss": 0.000881643983302638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.857811038353603,
      "flow_matching_loss": 0.013920236099511384,
      "grad_norm": 0.09410411864519119,
      "learning_rate": 1.483587286854604e-08,
      "loss": 0.0149,
      "relation_loss": 0.000951907568378374,
      "selected_pairs": 26507.4,
      "state_query_norm": 41.86166687011719,
      "state_relation_loss": 7.026357488939538e-05,
      "step": 29780,
      "valid_action_samples": 230.7,
      "weighted_auxiliary_loss": 0.000951907568378374,
      "weighted_relation_loss": 0.000951907568378374
    },
    {
      "action_query_norm": 39.67456703186035,
      "action_relation_loss": 0.0008671715040691197,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.86716557530402,
      "flow_matching_loss": 0.015284329652786255,
      "grad_norm": 0.08867348730564117,
      "learning_rate": 1.3523694955064336e-08,
      "loss": 0.0162,
      "relation_loss": 0.0009359122370369733,
      "selected_pairs": 26895.5,
      "state_query_norm": 41.865135192871094,
      "state_relation_loss": 6.8740722417715e-05,
      "step": 29790,
      "valid_action_samples": 232.4,
      "weighted_auxiliary_loss": 0.0009359122370369733,
      "weighted_relation_loss": 0.0009359122370369733
    },
    {
      "action_query_norm": 39.666203308105466,
      "action_relation_loss": 0.0008578385750297457,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.876520112254443,
      "flow_matching_loss": 0.014960995130240917,
      "grad_norm": 0.07866515964269638,
      "learning_rate": 1.2272255323708681e-08,
      "loss": 0.0159,
      "relation_loss": 0.0009289056295529007,
      "selected_pairs": 27479.5,
      "state_query_norm": 41.84806022644043,
      "state_relation_loss": 7.106706543709151e-05,
      "step": 29800,
      "valid_action_samples": 234.9,
      "weighted_auxiliary_loss": 0.0009289056295529007,
      "weighted_relation_loss": 0.0009289056295529007
    },
    {
      "action_query_norm": 39.677376937866214,
      "action_relation_loss": 0.0008546525437850505,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.885874649204865,
      "flow_matching_loss": 0.01654194453731179,
      "grad_norm": 0.09327807277441025,
      "learning_rate": 1.1081555495096042e-08,
      "loss": 0.0175,
      "relation_loss": 0.0009335612412542105,
      "selected_pairs": 27056.0,
      "state_query_norm": 41.838003158569336,
      "state_relation_loss": 7.89086967415642e-05,
      "step": 29810,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009335612412542105,
      "weighted_relation_loss": 0.0009335612412542105
    },
    {
      "action_query_norm": 39.64237937927246,
      "action_relation_loss": 0.0008157853619195521,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.895229186155284,
      "flow_matching_loss": 0.015451095439493657,
      "grad_norm": 0.11760267615318298,
      "learning_rate": 9.951596916041306e-09,
      "loss": 0.0163,
      "relation_loss": 0.0008847849327139557,
      "selected_pairs": 26649.5,
      "state_query_norm": 41.85592880249023,
      "state_relation_loss": 6.8999575523776e-05,
      "step": 29820,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.0008847849327139557,
      "weighted_relation_loss": 0.0008847849327139557
    },
    {
      "action_query_norm": 39.63098373413086,
      "action_relation_loss": 0.0008562477654777467,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.904583723105706,
      "flow_matching_loss": 0.01528979754075408,
      "grad_norm": 0.08113061636686325,
      "learning_rate": 8.882380959551739e-09,
      "loss": 0.0162,
      "relation_loss": 0.000926700554555282,
      "selected_pairs": 26952.6,
      "state_query_norm": 41.8695011138916,
      "state_relation_loss": 7.045278834993951e-05,
      "step": 29830,
      "valid_action_samples": 232.6,
      "weighted_auxiliary_loss": 0.000926700554555282,
      "weighted_relation_loss": 0.000926700554555282
    },
    {
      "action_query_norm": 39.63846702575684,
      "action_relation_loss": 0.0008715051342733204,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.913938260056128,
      "flow_matching_loss": 0.014811729732900857,
      "grad_norm": 0.08482725918292999,
      "learning_rate": 7.873908924821428e-09,
      "loss": 0.0158,
      "relation_loss": 0.0009444007242564112,
      "selected_pairs": 26874.9,
      "state_query_norm": 41.85272102355957,
      "state_relation_loss": 7.289558052434587e-05,
      "step": 29840,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009444007242564112,
      "weighted_relation_loss": 0.0009444007242564112
    },
    {
      "action_query_norm": 39.6165283203125,
      "action_relation_loss": 0.000840972556034103,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.92329279700655,
      "flow_matching_loss": 0.01616855785250664,
      "grad_norm": 0.11037667840719223,
      "learning_rate": 6.926182037242379e-09,
      "loss": 0.0171,
      "relation_loss": 0.0009101361210923642,
      "selected_pairs": 27057.3,
      "state_query_norm": 41.8645378112793,
      "state_relation_loss": 6.916356578585691e-05,
      "step": 29850,
      "valid_action_samples": 233.1,
      "weighted_auxiliary_loss": 0.0009101361210923642,
      "weighted_relation_loss": 0.0009101361210923642
    },
    {
      "action_query_norm": 39.6601921081543,
      "action_relation_loss": 0.0008955532219260931,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.932647333956968,
      "flow_matching_loss": 0.01494816467165947,
      "grad_norm": 0.09159873425960541,
      "learning_rate": 6.039201448387876e-09,
      "loss": 0.0159,
      "relation_loss": 0.0009660164767410606,
      "selected_pairs": 27502.6,
      "state_query_norm": 41.85438003540039,
      "state_relation_loss": 7.04632511769887e-05,
      "step": 29860,
      "valid_action_samples": 235.0,
      "weighted_auxiliary_loss": 0.0009660164767410606,
      "weighted_relation_loss": 0.0009660164767410606
    },
    {
      "action_query_norm": 39.62078285217285,
      "action_relation_loss": 0.0009351086395327002,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.94200187090739,
      "flow_matching_loss": 0.01761757079511881,
      "grad_norm": 0.09976695477962494,
      "learning_rate": 5.212968236029125e-09,
      "loss": 0.0186,
      "relation_loss": 0.001006076339399442,
      "selected_pairs": 27643.9,
      "state_query_norm": 41.85299415588379,
      "state_relation_loss": 7.09677035047207e-05,
      "step": 29870,
      "valid_action_samples": 235.6,
      "weighted_auxiliary_loss": 0.001006076339399442,
      "weighted_relation_loss": 0.001006076339399442
    },
    {
      "action_query_norm": 39.70389175415039,
      "action_relation_loss": 0.0008270522346720099,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.951356407857812,
      "flow_matching_loss": 0.01584635376930237,
      "grad_norm": 0.08771960437297821,
      "learning_rate": 4.447483404118602e-09,
      "loss": 0.0167,
      "relation_loss": 0.000899383082287386,
      "selected_pairs": 26883.5,
      "state_query_norm": 41.865726470947266,
      "state_relation_loss": 7.233085707412102e-05,
      "step": 29880,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.000899383082287386,
      "weighted_relation_loss": 0.000899383082287386
    },
    {
      "action_query_norm": 39.610940551757814,
      "action_relation_loss": 0.0009054286987520754,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.96071094480823,
      "flow_matching_loss": 0.016668033506721257,
      "grad_norm": 0.13972485065460205,
      "learning_rate": 3.742747882784503e-09,
      "loss": 0.0176,
      "relation_loss": 0.0009726423886604607,
      "selected_pairs": 26183.9,
      "state_query_norm": 41.854460906982425,
      "state_relation_loss": 6.721369391016196e-05,
      "step": 29890,
      "valid_action_samples": 229.3,
      "weighted_auxiliary_loss": 0.0009726423886604607,
      "weighted_relation_loss": 0.0009726423886604607
    },
    {
      "action_query_norm": 39.64356117248535,
      "action_relation_loss": 0.0008356747857760638,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.970065481758652,
      "flow_matching_loss": 0.014556619711220265,
      "grad_norm": 0.12543411552906036,
      "learning_rate": 3.0987625283473987e-09,
      "loss": 0.0155,
      "relation_loss": 0.0009032710047904402,
      "selected_pairs": 26968.3,
      "state_query_norm": 41.856480026245116,
      "state_relation_loss": 6.759620882803574e-05,
      "step": 29900,
      "valid_action_samples": 232.7,
      "weighted_auxiliary_loss": 0.0009032710047904402,
      "weighted_relation_loss": 0.0009032710047904402
    },
    {
      "action_query_norm": 39.64381713867188,
      "action_relation_loss": 0.0009198758925776929,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.979420018709074,
      "flow_matching_loss": 0.015446959342807532,
      "grad_norm": 0.16947174072265625,
      "learning_rate": 2.515528123320232e-09,
      "loss": 0.0164,
      "relation_loss": 0.000987293035723269,
      "selected_pairs": 26639.2,
      "state_query_norm": 41.874443435668944,
      "state_relation_loss": 6.741715733369347e-05,
      "step": 29910,
      "valid_action_samples": 231.3,
      "weighted_auxiliary_loss": 0.000987293035723269,
      "weighted_relation_loss": 0.000987293035723269
    },
    {
      "action_query_norm": 39.66566009521485,
      "action_relation_loss": 0.000839479872956872,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.988774555659496,
      "flow_matching_loss": 0.014700146205723286,
      "grad_norm": 0.0628650039434433,
      "learning_rate": 1.9930453763750133e-09,
      "loss": 0.0156,
      "relation_loss": 0.0009076531685423106,
      "selected_pairs": 26734.2,
      "state_query_norm": 41.8576774597168,
      "state_relation_loss": 6.817330213380047e-05,
      "step": 29920,
      "valid_action_samples": 231.7,
      "weighted_auxiliary_loss": 0.0009076531685423106,
      "weighted_relation_loss": 0.0009076531685423106
    },
    {
      "action_query_norm": 39.619694519042966,
      "action_relation_loss": 0.0008887511503417045,
      "auxiliary_loss_weight": 1.0,
      "epoch": 27.998129092609915,
      "flow_matching_loss": 0.015662228874862193,
      "grad_norm": 0.14480628073215485,
      "learning_rate": 1.5313149223872281e-09,
      "loss": 0.0166,
      "relation_loss": 0.0009564513282384723,
      "selected_pairs": 26664.9,
      "state_query_norm": 41.84622573852539,
      "state_relation_loss": 6.77001771691721e-05,
      "step": 29930,
      "valid_action_samples": 231.4,
      "weighted_auxiliary_loss": 0.0009564513282384723,
      "weighted_relation_loss": 0.0009564513282384723
    },
    {
      "action_query_norm": 39.627939987182614,
      "action_relation_loss": 0.0008861621899995953,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.007483629560337,
      "flow_matching_loss": 0.014446358475834132,
      "grad_norm": 0.12763653695583344,
      "learning_rate": 1.1303373224025305e-09,
      "loss": 0.0154,
      "relation_loss": 0.0009577607095707207,
      "selected_pairs": 24608.7,
      "state_query_norm": 41.840902709960936,
      "state_relation_loss": 7.159853048506193e-05,
      "step": 29940,
      "valid_action_samples": 215.5,
      "weighted_auxiliary_loss": 0.0009577607095707207,
      "weighted_relation_loss": 0.0009577607095707207
    },
    {
      "action_query_norm": 39.626326370239255,
      "action_relation_loss": 0.0008008444332517683,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.01683816651076,
      "flow_matching_loss": 0.014643988758325576,
      "grad_norm": 0.07169917225837708,
      "learning_rate": 7.901130636367438e-10,
      "loss": 0.0155,
      "relation_loss": 0.0008710240828804672,
      "selected_pairs": 26558.4,
      "state_query_norm": 41.85442848205567,
      "state_relation_loss": 7.017965835984797e-05,
      "step": 29950,
      "valid_action_samples": 230.9,
      "weighted_auxiliary_loss": 0.0008710240828804672,
      "weighted_relation_loss": 0.0008710240828804672
    },
    {
      "action_query_norm": 39.6061336517334,
      "action_relation_loss": 0.000836722640087828,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.026192703461177,
      "flow_matching_loss": 0.01591390063986182,
      "grad_norm": 0.0737992376089096,
      "learning_rate": 5.106425595036158e-10,
      "loss": 0.0168,
      "relation_loss": 0.0009071920998394489,
      "selected_pairs": 26233.1,
      "state_query_norm": 41.86643905639649,
      "state_relation_loss": 7.046945647743996e-05,
      "step": 29960,
      "valid_action_samples": 229.5,
      "weighted_auxiliary_loss": 0.0009071920998394489,
      "weighted_relation_loss": 0.0009071920998394489
    },
    {
      "action_query_norm": 39.63439826965332,
      "action_relation_loss": 0.000853474740870297,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.0355472404116,
      "flow_matching_loss": 0.014647952653467656,
      "grad_norm": 0.10678240656852722,
      "learning_rate": 2.9192614958706286e-10,
      "loss": 0.0156,
      "relation_loss": 0.0009215975936967879,
      "selected_pairs": 26875.3,
      "state_query_norm": 41.853287887573245,
      "state_relation_loss": 6.812285027990583e-05,
      "step": 29970,
      "valid_action_samples": 232.3,
      "weighted_auxiliary_loss": 0.0009215975936967879,
      "weighted_relation_loss": 0.0009215975936967879
    },
    {
      "action_query_norm": 39.68527336120606,
      "action_relation_loss": 0.0008674254000652582,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.04490177736202,
      "flow_matching_loss": 0.015663078241050243,
      "grad_norm": 0.13001056015491486,
      "learning_rate": 1.3396409964117064e-10,
      "loss": 0.0166,
      "relation_loss": 0.0009343850775621832,
      "selected_pairs": 26808.6,
      "state_query_norm": 41.8635311126709,
      "state_relation_loss": 6.695968259009532e-05,
      "step": 29980,
      "valid_action_samples": 232.0,
      "weighted_auxiliary_loss": 0.0009343850775621832,
      "weighted_relation_loss": 0.0009343850775621832
    },
    {
      "action_query_norm": 39.639875411987305,
      "action_relation_loss": 0.0008806628233287483,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.054256314312443,
      "flow_matching_loss": 0.014712597895413636,
      "grad_norm": 0.09217794239521027,
      "learning_rate": 3.6756601606846574e-11,
      "loss": 0.0157,
      "relation_loss": 0.0009520060324575752,
      "selected_pairs": 27151.3,
      "state_query_norm": 41.848148345947266,
      "state_relation_loss": 7.134320876502897e-05,
      "step": 29990,
      "valid_action_samples": 233.5,
      "weighted_auxiliary_loss": 0.0009520060324575752,
      "weighted_relation_loss": 0.0009520060324575752
    },
    {
      "action_query_norm": 39.63685760498047,
      "action_relation_loss": 0.0009216151724103838,
      "auxiliary_loss_weight": 1.0,
      "epoch": 28.06361085126286,
      "flow_matching_loss": 0.015820432733744384,
      "grad_norm": 0.09764866530895233,
      "learning_rate": 3.0377360626943076e-13,
      "loss": 0.0168,
      "relation_loss": 0.0009941779193468393,
      "selected_pairs": 27014.0,
      "state_query_norm": 41.86003227233887,
      "state_relation_loss": 7.2562738932902e-05,
      "step": 30000,
      "valid_action_samples": 232.9,
      "weighted_auxiliary_loss": 0.0009941779193468393,
      "weighted_relation_loss": 0.0009941779193468393
    },
    {
      "epoch": 28.06361085126286,
      "step": 30000,
      "total_flos": 0.0,
      "train_loss": 0.043716008874773976,
      "train_runtime": 60060.3548,
      "train_samples_per_second": 127.871,
      "train_steps_per_second": 0.499
    }
  ],
  "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
}