{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.982857142857143,
  "global_step": 261,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "logps_train/chosen": -124.03406524658203,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -103.7885971069336,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.021984513849020004,
      "rewards_train/margins": 0.03294864948838949,
      "rewards_train/rejected": -0.010964135639369488,
      "step": 0
    },
    {
      "epoch": 0,
      "logps_train/chosen": -189.20074462890625,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -143.92283630371094,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.01742708310484886,
      "rewards_train/margins": -0.006499961018562317,
      "rewards_train/rejected": 0.02392704412341118,
      "step": 0
    },
    {
      "epoch": 0,
      "logps_train/chosen": -108.28680419921875,
      "logps_train/ref_chosen": -108.0,
      "logps_train/ref_rejected": -86.5,
      "logps_train/rejected": -86.16234588623047,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": -0.0033381874673068523,
      "rewards_train/margins": -0.008782926481217146,
      "rewards_train/rejected": 0.005444739013910294,
      "step": 0
    },
    {
      "epoch": 0,
      "logps_train/chosen": -159.1072235107422,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -137.1709747314453,
      "rewards_train/accuracies": 0.4375,
      "rewards_train/chosen": -0.1611127108335495,
      "rewards_train/margins": -0.18815578520298004,
      "rewards_train/rejected": 0.027043074369430542,
      "step": 0
    },
    {
      "epoch": 0.01,
      "learning_rate": 6.25e-06,
      "loss": 0.7329,
      "step": 1
    },
    {
      "epoch": 0.01,
      "logps_train/chosen": -152.52149963378906,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -145.44671630859375,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": -0.3556646704673767,
      "rewards_train/margins": -0.34302489552646875,
      "rewards_train/rejected": -0.012639774940907955,
      "step": 1
    },
    {
      "epoch": 0.01,
      "logps_train/chosen": -162.546875,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -115.53173065185547,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": -0.0027348892763257027,
      "rewards_train/margins": -0.006935393437743187,
      "rewards_train/rejected": 0.004200504161417484,
      "step": 1
    },
    {
      "epoch": 0.01,
      "logps_train/chosen": -166.0098876953125,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -137.22930908203125,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": -0.009777255356311798,
      "rewards_train/margins": 0.0004109693691134453,
      "rewards_train/rejected": -0.010188224725425243,
      "step": 1
    },
    {
      "epoch": 0.01,
      "logps_train/chosen": -136.130859375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -112.58856201171875,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.009960669092833996,
      "rewards_train/margins": 0.021551736630499363,
      "rewards_train/rejected": -0.011591067537665367,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.25e-05,
      "loss": 0.7722,
      "step": 2
    },
    {
      "epoch": 0.02,
      "logps_train/chosen": -124.05975341796875,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -109.16248321533203,
      "rewards_train/accuracies": 0.34375,
      "rewards_train/chosen": 0.018145941197872162,
      "rewards_train/margins": -0.004668369889259338,
      "rewards_train/rejected": 0.0228143110871315,
      "step": 2
    },
    {
      "epoch": 0.02,
      "logps_train/chosen": -165.04600524902344,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -138.3419952392578,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.011854463256895542,
      "rewards_train/margins": -0.011222056113183498,
      "rewards_train/rejected": 0.02307651937007904,
      "step": 2
    },
    {
      "epoch": 0.02,
      "logps_train/chosen": -142.4432373046875,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -123.91960144042969,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": 0.008557641878724098,
      "rewards_train/margins": 0.01167575130239129,
      "rewards_train/rejected": -0.0031181094236671925,
      "step": 2
    },
    {
      "epoch": 0.02,
      "logps_train/chosen": -165.10171508789062,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -143.83792114257812,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.02361709624528885,
      "rewards_train/margins": 0.017565627116709948,
      "rewards_train/rejected": 0.006051469128578901,
      "step": 2
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.6922,
      "step": 3
    },
    {
      "epoch": 0.03,
      "logps_train/chosen": -170.82505798339844,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -151.93087768554688,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": 0.013978248462080956,
      "rewards_train/margins": -0.005433313548564911,
      "rewards_train/rejected": 0.019411562010645866,
      "step": 3
    },
    {
      "epoch": 0.03,
      "logps_train/chosen": -139.39083862304688,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -129.07003784179688,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": 0.020486466586589813,
      "rewards_train/margins": 0.003638278692960739,
      "rewards_train/rejected": 0.016848187893629074,
      "step": 3
    },
    {
      "epoch": 0.03,
      "logps_train/chosen": -176.51419067382812,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -132.66586303710938,
      "rewards_train/accuracies": 0.4375,
      "rewards_train/chosen": 0.009908124804496765,
      "rewards_train/margins": -0.014521919190883636,
      "rewards_train/rejected": 0.0244300439953804,
      "step": 3
    },
    {
      "epoch": 0.03,
      "logps_train/chosen": -156.88722229003906,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -119.69668579101562,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": 0.019480425864458084,
      "rewards_train/margins": -0.0017685145139694214,
      "rewards_train/rejected": 0.021248940378427505,
      "step": 3
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.5e-05,
      "loss": 0.6956,
      "step": 4
    },
    {
      "epoch": 0.05,
      "logps_train/chosen": -150.92123413085938,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -117.88097381591797,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.034377966076135635,
      "rewards_train/margins": 0.017336303368210793,
      "rewards_train/rejected": 0.017041662707924843,
      "step": 4
    },
    {
      "epoch": 0.05,
      "logps_train/chosen": -180.184814453125,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -128.09719848632812,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.0534917414188385,
      "rewards_train/margins": 0.01155165582895279,
      "rewards_train/rejected": 0.04194008558988571,
      "step": 4
    },
    {
      "epoch": 0.05,
      "logps_train/chosen": -143.22642517089844,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -111.45314025878906,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.04542411118745804,
      "rewards_train/margins": 0.008365228772163391,
      "rewards_train/rejected": 0.03705888241529465,
      "step": 4
    },
    {
      "epoch": 0.05,
      "logps_train/chosen": -149.19285583496094,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -132.03118896484375,
      "rewards_train/accuracies": 0.40625,
      "rewards_train/chosen": 0.05766720697283745,
      "rewards_train/margins": 0.0012159161269664764,
      "rewards_train/rejected": 0.05645129084587097,
      "step": 4
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.125e-05,
      "loss": 0.6893,
      "step": 5
    },
    {
      "epoch": 0.06,
      "logps_train/chosen": -147.8840789794922,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -122.57456970214844,
      "rewards_train/accuracies": 0.40625,
      "rewards_train/chosen": 0.05192326009273529,
      "rewards_train/margins": -0.010029420256614685,
      "rewards_train/rejected": 0.061952680349349976,
      "step": 5
    },
    {
      "epoch": 0.06,
      "logps_train/chosen": -170.50418090820312,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -136.78695678710938,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.11383962631225586,
      "rewards_train/margins": 0.027300000190734863,
      "rewards_train/rejected": 0.086539626121521,
      "step": 5
    },
    {
      "epoch": 0.06,
      "logps_train/chosen": -155.12435913085938,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -116.24176025390625,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": 0.08434212952852249,
      "rewards_train/margins": 0.012327082455158234,
      "rewards_train/rejected": 0.07201504707336426,
      "step": 5
    },
    {
      "epoch": 0.06,
      "logps_train/chosen": -140.93975830078125,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -96.5,
      "logps_train/rejected": -96.0169677734375,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.09296239912509918,
      "rewards_train/margins": 0.042803697288036346,
      "rewards_train/rejected": 0.050158701837062836,
      "step": 5
    },
    {
      "epoch": 0.07,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.6848,
      "step": 6
    },
    {
      "epoch": 0.07,
      "logps_train/chosen": -140.3257293701172,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -110.98545837402344,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.1716260313987732,
      "rewards_train/margins": 0.07232072949409485,
      "rewards_train/rejected": 0.09930530190467834,
      "step": 6
    },
    {
      "epoch": 0.07,
      "logps_train/chosen": -154.02883911132812,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -129.6768798828125,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.20046155154705048,
      "rewards_train/margins": 0.06691007316112518,
      "rewards_train/rejected": 0.1335514783859253,
      "step": 6
    },
    {
      "epoch": 0.07,
      "logps_train/chosen": -148.28170776367188,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -124.03239440917969,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.17491726577281952,
      "rewards_train/margins": 0.04614986479282379,
      "rewards_train/rejected": 0.12876740097999573,
      "step": 6
    },
    {
      "epoch": 0.07,
      "logps_train/chosen": -147.0029296875,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -109.11970520019531,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.14702154695987701,
      "rewards_train/margins": 0.04444076120853424,
      "rewards_train/rejected": 0.10258078575134277,
      "step": 6
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.375e-05,
      "loss": 0.667,
      "step": 7
    },
    {
      "epoch": 0.08,
      "logps_train/chosen": -132.00901794433594,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -109.18241882324219,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.226222425699234,
      "rewards_train/margins": 0.0596490353345871,
      "rewards_train/rejected": 0.1665733903646469,
      "step": 7
    },
    {
      "epoch": 0.08,
      "logps_train/chosen": -176.80740356445312,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -171.0,
      "logps_train/rejected": -168.07736206054688,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.302267462015152,
      "rewards_train/margins": 0.03226765990257263,
      "rewards_train/rejected": 0.26999980211257935,
      "step": 7
    },
    {
      "epoch": 0.08,
      "logps_train/chosen": -161.03793334960938,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -107.19043731689453,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.2758948802947998,
      "rewards_train/margins": 0.10362982749938965,
      "rewards_train/rejected": 0.17226505279541016,
      "step": 7
    },
    {
      "epoch": 0.08,
      "logps_train/chosen": -165.82919311523438,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -142.60577392578125,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.31333446502685547,
      "rewards_train/margins": 0.09588496387004852,
      "rewards_train/rejected": 0.21744950115680695,
      "step": 7
    },
    {
      "epoch": 0.09,
      "learning_rate": 5e-05,
      "loss": 0.6623,
      "step": 8
    },
    {
      "epoch": 0.09,
      "logps_train/chosen": -156.38870239257812,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -145.6343231201172,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.423044353723526,
      "rewards_train/margins": 0.11528569459915161,
      "rewards_train/rejected": 0.3077586591243744,
      "step": 8
    },
    {
      "epoch": 0.09,
      "logps_train/chosen": -166.0334014892578,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -124.54536437988281,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.41580069065093994,
      "rewards_train/margins": 0.15119647979736328,
      "rewards_train/rejected": 0.26460421085357666,
      "step": 8
    },
    {
      "epoch": 0.09,
      "logps_train/chosen": -111.77043151855469,
      "logps_train/ref_chosen": -115.0,
      "logps_train/ref_rejected": -78.0,
      "logps_train/rejected": -76.39224243164062,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.2940497100353241,
      "rewards_train/margins": 0.1374729573726654,
      "rewards_train/rejected": 0.1565767526626587,
      "step": 8
    },
    {
      "epoch": 0.09,
      "logps_train/chosen": -148.72055053710938,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -107.92207336425781,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.42247551679611206,
      "rewards_train/margins": 0.2193703055381775,
      "rewards_train/rejected": 0.20310521125793457,
      "step": 8
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.99980726386944e-05,
      "loss": 0.6258,
      "step": 9
    },
    {
      "epoch": 0.1,
      "logps_train/chosen": -171.09103393554688,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -112.07714080810547,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.5837191343307495,
      "rewards_train/margins": 0.2688255310058594,
      "rewards_train/rejected": 0.31489360332489014,
      "step": 9
    },
    {
      "epoch": 0.1,
      "logps_train/chosen": -134.94949340820312,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -132.9979705810547,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": 0.4236791431903839,
      "rewards_train/margins": 0.02884700894355774,
      "rewards_train/rejected": 0.39483213424682617,
      "step": 9
    },
    {
      "epoch": 0.1,
      "logps_train/chosen": -154.36537170410156,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -127.57532501220703,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.5209816694259644,
      "rewards_train/margins": 0.10341629385948181,
      "rewards_train/rejected": 0.41756537556648254,
      "step": 9
    },
    {
      "epoch": 0.1,
      "logps_train/chosen": -149.526611328125,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -134.63084411621094,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.48503243923187256,
      "rewards_train/margins": 0.11374184489250183,
      "rewards_train/rejected": 0.3712905943393707,
      "step": 9
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.9992290851955325e-05,
      "loss": 0.6503,
      "step": 10
    },
    {
      "epoch": 0.11,
      "logps_train/chosen": -153.533935546875,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -126.86518859863281,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.5384516716003418,
      "rewards_train/margins": 0.16266614198684692,
      "rewards_train/rejected": 0.3757855296134949,
      "step": 10
    },
    {
      "epoch": 0.11,
      "logps_train/chosen": -167.98931884765625,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -144.03042602539062,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": 0.6893491744995117,
      "rewards_train/margins": 0.16035926342010498,
      "rewards_train/rejected": 0.5289899110794067,
      "step": 10
    },
    {
      "epoch": 0.11,
      "logps_train/chosen": -155.04034423828125,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -130.8607177734375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.6690115332603455,
      "rewards_train/margins": 0.2056686282157898,
      "rewards_train/rejected": 0.46334290504455566,
      "step": 10
    },
    {
      "epoch": 0.11,
      "logps_train/chosen": -130.7728271484375,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -109.23792266845703,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.6114521622657776,
      "rewards_train/margins": 0.13563531637191772,
      "rewards_train/rejected": 0.47581684589385986,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.998265553127013e-05,
      "loss": 0.6345,
      "step": 11
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -160.4620819091797,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -125.24628448486328,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.8416832685470581,
      "rewards_train/margins": 0.36943644285202026,
      "rewards_train/rejected": 0.47224682569503784,
      "step": 11
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -145.121826171875,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -132.90176391601562,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.646020233631134,
      "rewards_train/margins": 0.17330646514892578,
      "rewards_train/rejected": 0.47271376848220825,
      "step": 11
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -151.66098022460938,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -147.03488159179688,
      "rewards_train/accuracies": 0.5,
      "rewards_train/chosen": 0.7320475578308105,
      "rewards_train/margins": 0.1371973752975464,
      "rewards_train/rejected": 0.5948501825332642,
      "step": 11
    },
    {
      "epoch": 0.13,
      "logps_train/chosen": -142.59877014160156,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -114.5,
      "logps_train/rejected": -109.92486572265625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.6653189659118652,
      "rewards_train/margins": 0.20673072338104248,
      "rewards_train/rejected": 0.45858824253082275,
      "step": 11
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.996916816229837e-05,
      "loss": 0.6218,
      "step": 12
    },
    {
      "epoch": 0.14,
      "logps_train/chosen": -129.76406860351562,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -104.38932800292969,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.7065520882606506,
      "rewards_train/margins": 0.4883071482181549,
      "rewards_train/rejected": 0.21824494004249573,
      "step": 12
    },
    {
      "epoch": 0.14,
      "logps_train/chosen": -149.8843994140625,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -126.18021392822266,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": 0.9509168863296509,
      "rewards_train/margins": 0.3822464942932129,
      "rewards_train/rejected": 0.568670392036438,
      "step": 12
    },
    {
      "epoch": 0.14,
      "logps_train/chosen": -141.76194763183594,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -141.0,
      "logps_train/rejected": -134.5399169921875,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.6832780838012695,
      "rewards_train/margins": 0.04976832866668701,
      "rewards_train/rejected": 0.6335097551345825,
      "step": 12
    },
    {
      "epoch": 0.14,
      "logps_train/chosen": -135.6645965576172,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -115.1833724975586,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.6000440716743469,
      "rewards_train/margins": 0.09338122606277466,
      "rewards_train/rejected": 0.5066628456115723,
      "step": 12
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.995183082464269e-05,
      "loss": 0.6331,
      "step": 13
    },
    {
      "epoch": 0.15,
      "logps_train/chosen": -156.60064697265625,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -118.9285888671875,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 1.034368634223938,
      "rewards_train/margins": 0.5694149434566498,
      "rewards_train/rejected": 0.4649536907672882,
      "step": 13
    },
    {
      "epoch": 0.15,
      "logps_train/chosen": -173.77572631835938,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -163.0,
      "logps_train/rejected": -156.2445526123047,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.9569660425186157,
      "rewards_train/margins": 0.2895270586013794,
      "rewards_train/rejected": 0.6674389839172363,
      "step": 13
    },
    {
      "epoch": 0.15,
      "logps_train/chosen": -151.65478515625,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -115.70185852050781,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 1.0789567232131958,
      "rewards_train/margins": 0.48078322410583496,
      "rewards_train/rejected": 0.5981734991073608,
      "step": 13
    },
    {
      "epoch": 0.15,
      "logps_train/chosen": -110.07494354248047,
      "logps_train/ref_chosen": -118.5,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -97.74917602539062,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.8415290117263794,
      "rewards_train/margins": 0.2806065082550049,
      "rewards_train/rejected": 0.5609225034713745,
      "step": 13
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.9930646191528175e-05,
      "loss": 0.5759,
      "step": 14
    },
    {
      "epoch": 0.16,
      "logps_train/chosen": -177.0446319580078,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -176.0,
      "logps_train/rejected": -168.82325744628906,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 1.0113569498062134,
      "rewards_train/margins": 0.356975793838501,
      "rewards_train/rejected": 0.6543811559677124,
      "step": 14
    },
    {
      "epoch": 0.16,
      "logps_train/chosen": -151.45489501953125,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -136.52796936035156,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.8385694026947021,
      "rewards_train/margins": 0.13999921083450317,
      "rewards_train/rejected": 0.698570191860199,
      "step": 14
    },
    {
      "epoch": 0.16,
      "logps_train/chosen": -135.15379333496094,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -125.3636703491211,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.8960466980934143,
      "rewards_train/margins": 0.2092685103416443,
      "rewards_train/rejected": 0.68677818775177,
      "step": 14
    },
    {
      "epoch": 0.16,
      "logps_train/chosen": -146.58248901367188,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -120.76148223876953,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.9701685905456543,
      "rewards_train/margins": 0.3346949815750122,
      "rewards_train/rejected": 0.6354736089706421,
      "step": 14
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.9905617529390203e-05,
      "loss": 0.6141,
      "step": 15
    },
    {
      "epoch": 0.17,
      "logps_train/chosen": -105.37283325195312,
      "logps_train/ref_chosen": -112.5,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -107.2294921875,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.717697262763977,
      "rewards_train/margins": 0.21457242965698242,
      "rewards_train/rejected": 0.5031248331069946,
      "step": 15
    },
    {
      "epoch": 0.17,
      "logps_train/chosen": -140.62977600097656,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -97.55780792236328,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.7388781309127808,
      "rewards_train/margins": 0.35042065382003784,
      "rewards_train/rejected": 0.3884574770927429,
      "step": 15
    },
    {
      "epoch": 0.17,
      "logps_train/chosen": -150.58285522460938,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -99.97172546386719,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.8648346662521362,
      "rewards_train/margins": 0.408088356256485,
      "rewards_train/rejected": 0.45674630999565125,
      "step": 15
    },
    {
      "epoch": 0.17,
      "logps_train/chosen": -141.79176330566406,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -125.314453125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.8631088733673096,
      "rewards_train/margins": 0.32648766040802,
      "rewards_train/rejected": 0.5366212129592896,
      "step": 15
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.987674869737077e-05,
      "loss": 0.5859,
      "step": 16
    },
    {
      "epoch": 0.18,
      "logps_train/chosen": -136.44541931152344,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -105.30747985839844,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.7658092975616455,
      "rewards_train/margins": 0.42277851700782776,
      "rewards_train/rejected": 0.34303078055381775,
      "step": 16
    },
    {
      "epoch": 0.18,
      "logps_train/chosen": -142.18093872070312,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -115.7771987915039,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.7697968482971191,
      "rewards_train/margins": 0.11177456378936768,
      "rewards_train/rejected": 0.6580222845077515,
      "step": 16
    },
    {
      "epoch": 0.18,
      "logps_train/chosen": -154.24090576171875,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -126.07774353027344,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 1.0580410957336426,
      "rewards_train/margins": 0.3286939859390259,
      "rewards_train/rejected": 0.7293471097946167,
      "step": 16
    },
    {
      "epoch": 0.18,
      "logps_train/chosen": -152.1510467529297,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -104.72694396972656,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.752070426940918,
      "rewards_train/margins": 0.3003508448600769,
      "rewards_train/rejected": 0.45171958208084106,
      "step": 16
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.984404414672346e-05,
      "loss": 0.6126,
      "step": 17
    },
    {
      "epoch": 0.19,
      "logps_train/chosen": -142.31809997558594,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -126.29586791992188,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.7054945230484009,
      "rewards_train/margins": 0.10695618391036987,
      "rewards_train/rejected": 0.598538339138031,
      "step": 17
    },
    {
      "epoch": 0.19,
      "logps_train/chosen": -89.5583724975586,
      "logps_train/ref_chosen": -94.5,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -88.00601959228516,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": 0.49502912163734436,
      "rewards_train/margins": 0.10490813851356506,
      "rewards_train/rejected": 0.3901209831237793,
      "step": 17
    },
    {
      "epoch": 0.19,
      "logps_train/chosen": -141.4034881591797,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -98.5,
      "logps_train/rejected": -95.42372131347656,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.9821124076843262,
      "rewards_train/margins": 0.6551483273506165,
      "rewards_train/rejected": 0.3269640803337097,
      "step": 17
    },
    {
      "epoch": 0.19,
      "logps_train/chosen": -126.50776672363281,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -115.35845947265625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.8917524814605713,
      "rewards_train/margins": 0.34312617778778076,
      "rewards_train/rejected": 0.5486263036727905,
      "step": 17
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.980750892012711e-05,
      "loss": 0.616,
      "step": 18
    },
    {
      "epoch": 0.21,
      "logps_train/chosen": -161.98281860351562,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -129.00392150878906,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.9355075359344482,
      "rewards_train/margins": 0.31050950288772583,
      "rewards_train/rejected": 0.6249980330467224,
      "step": 18
    },
    {
      "epoch": 0.21,
      "logps_train/chosen": -109.2407455444336,
      "logps_train/ref_chosen": -115.5,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -87.75328063964844,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.634067952632904,
      "rewards_train/margins": 0.21584081649780273,
      "rewards_train/rejected": 0.4182271361351013,
      "step": 18
    },
    {
      "epoch": 0.21,
      "logps_train/chosen": -125.10116577148438,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -109.39839172363281,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.7695701122283936,
      "rewards_train/margins": 0.3268895447254181,
      "rewards_train/rejected": 0.44268056750297546,
      "step": 18
    },
    {
      "epoch": 0.21,
      "logps_train/chosen": -154.5267791748047,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -111.5,
      "logps_train/rejected": -105.82466125488281,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.8263745307922363,
      "rewards_train/margins": 0.2686058282852173,
      "rewards_train/rejected": 0.557768702507019,
      "step": 18
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.976714865090827e-05,
      "loss": 0.6399,
      "step": 19
    },
    {
      "epoch": 0.22,
      "logps_train/chosen": -170.05242919921875,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -133.59002685546875,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 1.0042303800582886,
      "rewards_train/margins": 0.6750620007514954,
      "rewards_train/rejected": 0.3291683793067932,
      "step": 19
    },
    {
      "epoch": 0.22,
      "logps_train/chosen": -138.31671142578125,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -93.0,
      "logps_train/rejected": -90.01516723632812,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.5841480493545532,
      "rewards_train/margins": 0.30216851830482483,
      "rewards_train/rejected": 0.2819795310497284,
      "step": 19
    },
    {
      "epoch": 0.22,
      "logps_train/chosen": -127.0712890625,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -87.40908813476562,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.6961421966552734,
      "rewards_train/margins": 0.25760751962661743,
      "rewards_train/rejected": 0.438534677028656,
      "step": 19
    },
    {
      "epoch": 0.22,
      "logps_train/chosen": -152.77476501464844,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -118.90397644042969,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.7926408648490906,
      "rewards_train/margins": 0.37669146060943604,
      "rewards_train/rejected": 0.41594940423965454,
      "step": 19
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.972296956217265e-05,
      "loss": 0.5838,
      "step": 20
    },
    {
      "epoch": 0.23,
      "logps_train/chosen": -148.97178649902344,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -102.91510009765625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.6525766849517822,
      "rewards_train/margins": 0.4128361642360687,
      "rewards_train/rejected": 0.2397405207157135,
      "step": 20
    },
    {
      "epoch": 0.23,
      "logps_train/chosen": -188.7393035888672,
      "logps_train/ref_chosen": -201.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -136.55224609375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 1.208785057067871,
      "rewards_train/margins": 1.0652238130569458,
      "rewards_train/rejected": 0.1435612440109253,
      "step": 20
    },
    {
      "epoch": 0.23,
      "logps_train/chosen": -136.43887329101562,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -106.124755859375,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.8320896625518799,
      "rewards_train/margins": 0.5457369983196259,
      "rewards_train/rejected": 0.28635266423225403,
      "step": 20
    },
    {
      "epoch": 0.23,
      "logps_train/chosen": -137.87374877929688,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -103.64527893066406,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.6388455629348755,
      "rewards_train/margins": 0.2835492193698883,
      "rewards_train/rejected": 0.3552963435649872,
      "step": 20
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.967497846584552e-05,
      "loss": 0.5307,
      "step": 21
    },
    {
      "epoch": 0.24,
      "logps_train/chosen": -127.54042053222656,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -93.16413116455078,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.4815409779548645,
      "rewards_train/margins": 0.36279795318841934,
      "rewards_train/rejected": 0.11874302476644516,
      "step": 21
    },
    {
      "epoch": 0.24,
      "logps_train/chosen": -145.63665771484375,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -127.99638366699219,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.3527398109436035,
      "rewards_train/margins": 0.19827693700790405,
      "rewards_train/rejected": 0.15446287393569946,
      "step": 21
    },
    {
      "epoch": 0.24,
      "logps_train/chosen": -143.23275756835938,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -99.59611511230469,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.8170070648193359,
      "rewards_train/margins": 0.7010326832532883,
      "rewards_train/rejected": 0.11597438156604767,
      "step": 21
    },
    {
      "epoch": 0.24,
      "logps_train/chosen": -134.1798095703125,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -103.10169982910156,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.8344590663909912,
      "rewards_train/margins": 0.4115239083766937,
      "rewards_train/rejected": 0.4229351580142975,
      "step": 21
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.962318276162148e-05,
      "loss": 0.5787,
      "step": 22
    },
    {
      "epoch": 0.25,
      "logps_train/chosen": -169.95230102539062,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -147.85079956054688,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 1.0285433530807495,
      "rewards_train/margins": 0.5781982243061066,
      "rewards_train/rejected": 0.45034512877464294,
      "step": 22
    },
    {
      "epoch": 0.25,
      "logps_train/chosen": -159.36582946777344,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -116.03822326660156,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.9459124803543091,
      "rewards_train/margins": 0.6589144468307495,
      "rewards_train/rejected": 0.28699803352355957,
      "step": 22
    },
    {
      "epoch": 0.25,
      "logps_train/chosen": -140.4551544189453,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -113.26671600341797,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.6787030696868896,
      "rewards_train/margins": 0.34983301162719727,
      "rewards_train/rejected": 0.3288700580596924,
      "step": 22
    },
    {
      "epoch": 0.25,
      "logps_train/chosen": -182.3084259033203,
      "logps_train/ref_chosen": -191.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -122.35601806640625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.8974780440330505,
      "rewards_train/margins": 0.5072007179260254,
      "rewards_train/rejected": 0.39027732610702515,
      "step": 22
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.9567590435823383e-05,
      "loss": 0.5552,
      "step": 23
    },
    {
      "epoch": 0.26,
      "logps_train/chosen": -123.2986068725586,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -114.5,
      "logps_train/rejected": -111.90111541748047,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.4740452170372009,
      "rewards_train/margins": 0.20946910977363586,
      "rewards_train/rejected": 0.26457610726356506,
      "step": 23
    },
    {
      "epoch": 0.26,
      "logps_train/chosen": -143.50405883789062,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -114.5,
      "logps_train/rejected": -110.6120376586914,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.8933444023132324,
      "rewards_train/margins": 0.49614936113357544,
      "rewards_train/rejected": 0.397195041179657,
      "step": 23
    },
    {
      "epoch": 0.26,
      "logps_train/chosen": -155.1194610595703,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -121.85575866699219,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 1.0275070667266846,
      "rewards_train/margins": 0.6742154359817505,
      "rewards_train/rejected": 0.3532916307449341,
      "step": 23
    },
    {
      "epoch": 0.26,
      "logps_train/chosen": -140.42703247070312,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -122.62283325195312,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": 0.627119779586792,
      "rewards_train/margins": 0.5090075731277466,
      "rewards_train/rejected": 0.11811220645904541,
      "step": 23
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.950821006017107e-05,
      "loss": 0.5787,
      "step": 24
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -136.32363891601562,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -101.22764587402344,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.6326753497123718,
      "rewards_train/margins": 0.4577834904193878,
      "rewards_train/rejected": 0.174891859292984,
      "step": 24
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -157.54037475585938,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -126.21077728271484,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.776432454586029,
      "rewards_train/margins": 0.4170653820037842,
      "rewards_train/rejected": 0.3593670725822449,
      "step": 24
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -107.36483001708984,
      "logps_train/ref_chosen": -112.5,
      "logps_train/ref_rejected": -83.0,
      "logps_train/rejected": -82.24800109863281,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.5118322372436523,
      "rewards_train/margins": 0.4391593262553215,
      "rewards_train/rejected": 0.07267291098833084,
      "step": 24
    },
    {
      "epoch": 0.27,
      "logps_train/chosen": -137.41708374023438,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -112.7099609375,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.5886627435684204,
      "rewards_train/margins": 0.302530437707901,
      "rewards_train/rejected": 0.2861323058605194,
      "step": 24
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.944505079045958e-05,
      "loss": 0.6016,
      "step": 25
    },
    {
      "epoch": 0.29,
      "logps_train/chosen": -157.17239379882812,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -100.15357971191406,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.8933084011077881,
      "rewards_train/margins": 0.4814213514328003,
      "rewards_train/rejected": 0.4118870496749878,
      "step": 25
    },
    {
      "epoch": 0.29,
      "logps_train/chosen": -128.82952880859375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -114.12107849121094,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.7554748058319092,
      "rewards_train/margins": 0.49004340171813965,
      "rewards_train/rejected": 0.26543140411376953,
      "step": 25
    },
    {
      "epoch": 0.29,
      "logps_train/chosen": -174.43142700195312,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -161.0,
      "logps_train/rejected": -156.95753479003906,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.6013880968093872,
      "rewards_train/margins": 0.15554097294807434,
      "rewards_train/rejected": 0.44584712386131287,
      "step": 25
    },
    {
      "epoch": 0.29,
      "logps_train/chosen": -179.6756591796875,
      "logps_train/ref_chosen": -190.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -136.37588500976562,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.9904413819313049,
      "rewards_train/margins": 0.6243195533752441,
      "rewards_train/rejected": 0.3661218285560608,
      "step": 25
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.9378122365147536e-05,
      "loss": 0.5856,
      "step": 26
    },
    {
      "epoch": 0.3,
      "logps_train/chosen": -160.62625122070312,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -112.54586029052734,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.568233847618103,
      "rewards_train/margins": 0.46276117116212845,
      "rewards_train/rejected": 0.10547267645597458,
      "step": 26
    },
    {
      "epoch": 0.3,
      "logps_train/chosen": -180.06460571289062,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -170.0,
      "logps_train/rejected": -165.44969177246094,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.8118870854377747,
      "rewards_train/margins": 0.4134972393512726,
      "rewards_train/rejected": 0.3983898460865021,
      "step": 26
    },
    {
      "epoch": 0.3,
      "logps_train/chosen": -165.16346740722656,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -125.14801025390625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.733361005783081,
      "rewards_train/margins": 0.4440605640411377,
      "rewards_train/rejected": 0.28930044174194336,
      "step": 26
    },
    {
      "epoch": 0.3,
      "logps_train/chosen": -149.80445861816406,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -172.0,
      "logps_train/rejected": -168.4322967529297,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.5533437728881836,
      "rewards_train/margins": 0.24852675199508667,
      "rewards_train/rejected": 0.3048170208930969,
      "step": 26
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.9307435103855507e-05,
      "loss": 0.6072,
      "step": 27
    },
    {
      "epoch": 0.31,
      "logps_train/chosen": -140.88067626953125,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -111.68878173828125,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.4423023760318756,
      "rewards_train/margins": 0.2377922534942627,
      "rewards_train/rejected": 0.20451012253761292,
      "step": 27
    },
    {
      "epoch": 0.31,
      "logps_train/chosen": -127.85845947265625,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -104.07398986816406,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.3409123420715332,
      "rewards_train/margins": 0.4412309601902962,
      "rewards_train/rejected": -0.10031861811876297,
      "step": 27
    },
    {
      "epoch": 0.31,
      "logps_train/chosen": -124.84896087646484,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -100.35652160644531,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.36427372694015503,
      "rewards_train/margins": 0.3226316273212433,
      "rewards_train/rejected": 0.04164209961891174,
      "step": 27
    },
    {
      "epoch": 0.31,
      "logps_train/chosen": -147.45672607421875,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -114.71719360351562,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.6011910438537598,
      "rewards_train/margins": 0.5735943987965584,
      "rewards_train/rejected": 0.027596645057201385,
      "step": 27
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.923299990577488e-05,
      "loss": 0.617,
      "step": 28
    },
    {
      "epoch": 0.32,
      "logps_train/chosen": -169.756103515625,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -124.39460754394531,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.40368735790252686,
      "rewards_train/margins": 0.5618972331285477,
      "rewards_train/rejected": -0.1582098752260208,
      "step": 28
    },
    {
      "epoch": 0.32,
      "logps_train/chosen": -123.26806640625,
      "logps_train/ref_chosen": -127.0,
      "logps_train/ref_rejected": -90.5,
      "logps_train/rejected": -93.16129302978516,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.394532173871994,
      "rewards_train/margins": 0.6385668218135834,
      "rewards_train/rejected": -0.24403464794158936,
      "step": 28
    },
    {
      "epoch": 0.32,
      "logps_train/chosen": -162.7403106689453,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -112.55574035644531,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.8793867826461792,
      "rewards_train/margins": 1.0994132608175278,
      "rewards_train/rejected": -0.22002647817134857,
      "step": 28
    },
    {
      "epoch": 0.32,
      "logps_train/chosen": -108.07917785644531,
      "logps_train/ref_chosen": -112.5,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -100.18183135986328,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.4343189299106598,
      "rewards_train/margins": 0.35391509532928467,
      "rewards_train/rejected": 0.08040383458137512,
      "step": 28
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.9154828247987275e-05,
      "loss": 0.5549,
      "step": 29
    },
    {
      "epoch": 0.33,
      "logps_train/chosen": -152.16064453125,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -132.60389709472656,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.7116701602935791,
      "rewards_train/margins": 0.6406141370534897,
      "rewards_train/rejected": 0.07105602324008942,
      "step": 29
    },
    {
      "epoch": 0.33,
      "logps_train/chosen": -150.0108642578125,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -124.35476684570312,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.49451807141304016,
      "rewards_train/margins": 0.45050277933478355,
      "rewards_train/rejected": 0.04401529207825661,
      "step": 29
    },
    {
      "epoch": 0.33,
      "logps_train/chosen": -139.51673889160156,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -123.01943969726562,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.6350452899932861,
      "rewards_train/margins": 0.6675560399889946,
      "rewards_train/rejected": -0.032510749995708466,
      "step": 29
    },
    {
      "epoch": 0.33,
      "logps_train/chosen": -151.99301147460938,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -137.28713989257812,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.34091004729270935,
      "rewards_train/margins": 0.29519854485988617,
      "rewards_train/rejected": 0.04571150243282318,
      "step": 29
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.907293218369499e-05,
      "loss": 0.5685,
      "step": 30
    },
    {
      "epoch": 0.34,
      "logps_train/chosen": -171.75759887695312,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -101.5311508178711,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.736739993095398,
      "rewards_train/margins": 0.8000111803412437,
      "rewards_train/rejected": -0.0632711872458458,
      "step": 30
    },
    {
      "epoch": 0.34,
      "logps_train/chosen": -150.02069091796875,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -146.40304565429688,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.6478447914123535,
      "rewards_train/margins": 0.6080716103315353,
      "rewards_train/rejected": 0.039773181080818176,
      "step": 30
    },
    {
      "epoch": 0.34,
      "logps_train/chosen": -138.49688720703125,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -118.42485809326172,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.20158207416534424,
      "rewards_train/margins": 0.424081951379776,
      "rewards_train/rejected": -0.22249987721443176,
      "step": 30
    },
    {
      "epoch": 0.34,
      "logps_train/chosen": -153.21910095214844,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -131.95118713378906,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.5112931728363037,
      "rewards_train/margins": 0.5308257788419724,
      "rewards_train/rejected": -0.01953260600566864,
      "step": 30
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.898732434036244e-05,
      "loss": 0.5647,
      "step": 31
    },
    {
      "epoch": 0.35,
      "logps_train/chosen": -110.5118408203125,
      "logps_train/ref_chosen": -111.0,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -105.46853637695312,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.026843346655368805,
      "rewards_train/margins": 0.23717240244150162,
      "rewards_train/rejected": -0.2103290557861328,
      "step": 31
    },
    {
      "epoch": 0.35,
      "logps_train/chosen": -170.99050903320312,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -158.0,
      "logps_train/rejected": -157.1632080078125,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.5195033550262451,
      "rewards_train/margins": 0.4441733881831169,
      "rewards_train/rejected": 0.0753299668431282,
      "step": 31
    },
    {
      "epoch": 0.35,
      "logps_train/chosen": -159.28607177734375,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -138.23997497558594,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.43447864055633545,
      "rewards_train/margins": 0.6162890195846558,
      "rewards_train/rejected": -0.1818103790283203,
      "step": 31
    },
    {
      "epoch": 0.35,
      "logps_train/chosen": -134.58811950683594,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -132.0167236328125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.36775100231170654,
      "rewards_train/margins": 0.45528803020715714,
      "rewards_train/rejected": -0.08753702789545059,
      "step": 31
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.889801791776921e-05,
      "loss": 0.6137,
      "step": 32
    },
    {
      "epoch": 0.37,
      "logps_train/chosen": -135.58267211914062,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -115.02655029296875,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.2093113362789154,
      "rewards_train/margins": 0.3869669735431671,
      "rewards_train/rejected": -0.1776556372642517,
      "step": 32
    },
    {
      "epoch": 0.37,
      "logps_train/chosen": -169.59982299804688,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -144.22299194335938,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.26462656259536743,
      "rewards_train/margins": 0.3123156949877739,
      "rewards_train/rejected": -0.047689132392406464,
      "step": 32
    },
    {
      "epoch": 0.37,
      "logps_train/chosen": -146.15602111816406,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -115.34562683105469,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.4857647120952606,
      "rewards_train/margins": 0.7841946184635162,
      "rewards_train/rejected": -0.2984299063682556,
      "step": 32
    },
    {
      "epoch": 0.37,
      "logps_train/chosen": -139.1509552001953,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -151.7754364013672,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.40150654315948486,
      "rewards_train/margins": 0.8612766265869141,
      "rewards_train/rejected": -0.4597700834274292,
      "step": 32
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.880502668597475e-05,
      "loss": 0.6015,
      "step": 33
    },
    {
      "epoch": 0.38,
      "logps_train/chosen": -164.56207275390625,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -149.7305145263672,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.6187927722930908,
      "rewards_train/margins": 0.6844227313995361,
      "rewards_train/rejected": -0.06562995910644531,
      "step": 33
    },
    {
      "epoch": 0.38,
      "logps_train/chosen": -157.14566040039062,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -111.04812622070312,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.4366801977157593,
      "rewards_train/margins": 0.8013320863246918,
      "rewards_train/rejected": -0.3646518886089325,
      "step": 33
    },
    {
      "epoch": 0.38,
      "logps_train/chosen": -159.72222900390625,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -166.0,
      "logps_train/rejected": -166.43585205078125,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.3263116478919983,
      "rewards_train/margins": 0.37236351892352104,
      "rewards_train/rejected": -0.04605187103152275,
      "step": 33
    },
    {
      "epoch": 0.38,
      "logps_train/chosen": -149.86741638183594,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -103.66566467285156,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.577516496181488,
      "rewards_train/margins": 1.0174227356910706,
      "rewards_train/rejected": -0.4399062395095825,
      "step": 33
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.870836498319523e-05,
      "loss": 0.5305,
      "step": 34
    },
    {
      "epoch": 0.39,
      "logps_train/chosen": -161.78330993652344,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -130.53228759765625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.26034095883369446,
      "rewards_train/margins": 0.532904863357544,
      "rewards_train/rejected": -0.2725639045238495,
      "step": 34
    },
    {
      "epoch": 0.39,
      "logps_train/chosen": -167.86843872070312,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -139.84494018554688,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.5324914455413818,
      "rewards_train/margins": 0.7851487994194031,
      "rewards_train/rejected": -0.25265735387802124,
      "step": 34
    },
    {
      "epoch": 0.39,
      "logps_train/chosen": -147.22430419921875,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -123.09727478027344,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.3912162780761719,
      "rewards_train/margins": 0.3929597958922386,
      "rewards_train/rejected": -0.001743517816066742,
      "step": 34
    },
    {
      "epoch": 0.39,
      "logps_train/chosen": -147.94500732421875,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -137.91192626953125,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.15667131543159485,
      "rewards_train/margins": 0.4005984365940094,
      "rewards_train/rejected": -0.24392712116241455,
      "step": 34
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.86080477135927e-05,
      "loss": 0.5872,
      "step": 35
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -149.14828491210938,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -131.72792053222656,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.3583168387413025,
      "rewards_train/margins": 0.661382257938385,
      "rewards_train/rejected": -0.3030654191970825,
      "step": 35
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -173.38485717773438,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -130.51498413085938,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.5212795734405518,
      "rewards_train/margins": 0.8276603519916534,
      "rewards_train/rejected": -0.3063807785511017,
      "step": 35
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -153.11077880859375,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -115.9320068359375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.3146057426929474,
      "rewards_train/margins": 0.8578799068927765,
      "rewards_train/rejected": -0.5432741641998291,
      "step": 35
    },
    {
      "epoch": 0.4,
      "logps_train/chosen": -172.79013061523438,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -155.0,
      "logps_train/rejected": -157.91104125976562,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.24540139734745026,
      "rewards_train/margins": 0.5402161329984665,
      "rewards_train/rejected": -0.29481473565101624,
      "step": 35
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.850409034497704e-05,
      "loss": 0.5427,
      "step": 36
    },
    {
      "epoch": 0.41,
      "logps_train/chosen": -147.08950805664062,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -132.83718872070312,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.3379243314266205,
      "rewards_train/margins": 0.9198374450206757,
      "rewards_train/rejected": -0.5819131135940552,
      "step": 36
    },
    {
      "epoch": 0.41,
      "logps_train/chosen": -161.61907958984375,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -127.36285400390625,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.3235408067703247,
      "rewards_train/margins": 1.1210931539535522,
      "rewards_train/rejected": -0.7975523471832275,
      "step": 36
    },
    {
      "epoch": 0.41,
      "logps_train/chosen": -131.70205688476562,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -119.81352996826172,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.042293842881917953,
      "rewards_train/margins": 0.5976700149476528,
      "rewards_train/rejected": -0.5553761720657349,
      "step": 36
    },
    {
      "epoch": 0.41,
      "logps_train/chosen": -125.20751953125,
      "logps_train/ref_chosen": -126.0,
      "logps_train/ref_rejected": -93.5,
      "logps_train/rejected": -101.25257873535156,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.11919025331735611,
      "rewards_train/margins": 0.896144725382328,
      "rewards_train/rejected": -0.7769544720649719,
      "step": 36
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.839650890642104e-05,
      "loss": 0.5169,
      "step": 37
    },
    {
      "epoch": 0.42,
      "logps_train/chosen": -121.39013671875,
      "logps_train/ref_chosen": -120.0,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -102.62464141845703,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.1606450080871582,
      "rewards_train/margins": 0.42508548498153687,
      "rewards_train/rejected": -0.5857304930686951,
      "step": 37
    },
    {
      "epoch": 0.42,
      "logps_train/chosen": -125.78103637695312,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -111.85375213623047,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.0980682522058487,
      "rewards_train/margins": 0.46781833469867706,
      "rewards_train/rejected": -0.36975008249282837,
      "step": 37
    },
    {
      "epoch": 0.42,
      "logps_train/chosen": -145.85195922851562,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -128.85755920410156,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.18382151424884796,
      "rewards_train/margins": 0.7501192837953568,
      "rewards_train/rejected": -0.5662977695465088,
      "step": 37
    },
    {
      "epoch": 0.42,
      "logps_train/chosen": -120.22520446777344,
      "logps_train/ref_chosen": -122.0,
      "logps_train/ref_rejected": -88.0,
      "logps_train/rejected": -94.60220336914062,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.18724526464939117,
      "rewards_train/margins": 0.8372600227594376,
      "rewards_train/rejected": -0.6500147581100464,
      "step": 37
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.828531998578885e-05,
      "loss": 0.5599,
      "step": 38
    },
    {
      "epoch": 0.43,
      "logps_train/chosen": -182.186279296875,
      "logps_train/ref_chosen": -186.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -157.68670654296875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.349829763174057,
      "rewards_train/margins": 0.827485203742981,
      "rewards_train/rejected": -0.47765544056892395,
      "step": 38
    },
    {
      "epoch": 0.43,
      "logps_train/chosen": -148.2696075439453,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -132.3610382080078,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": 0.011784523725509644,
      "rewards_train/margins": 0.44945111870765686,
      "rewards_train/rejected": -0.4376665949821472,
      "step": 38
    },
    {
      "epoch": 0.43,
      "logps_train/chosen": -148.4195556640625,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -126.39881134033203,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.09119848161935806,
      "rewards_train/margins": 0.628833569586277,
      "rewards_train/rejected": -0.537635087966919,
      "step": 38
    },
    {
      "epoch": 0.43,
      "logps_train/chosen": -154.6384735107422,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -131.3975067138672,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": 0.4014357924461365,
      "rewards_train/margins": 0.9915284514427185,
      "rewards_train/rejected": -0.590092658996582,
      "step": 38
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.8170540727178326e-05,
      "loss": 0.5291,
      "step": 39
    },
    {
      "epoch": 0.45,
      "logps_train/chosen": -151.14736938476562,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -128.07752990722656,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.21706223487854004,
      "rewards_train/margins": 0.8728617429733276,
      "rewards_train/rejected": -0.6557995080947876,
      "step": 39
    },
    {
      "epoch": 0.45,
      "logps_train/chosen": -175.7981719970703,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -152.55126953125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.25172606110572815,
      "rewards_train/margins": 1.094938188791275,
      "rewards_train/rejected": -0.8432121276855469,
      "step": 39
    },
    {
      "epoch": 0.45,
      "logps_train/chosen": -142.94915771484375,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -126.44970703125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.24932152032852173,
      "rewards_train/margins": 0.8067926168441772,
      "rewards_train/rejected": -0.5574710965156555,
      "step": 39
    },
    {
      "epoch": 0.45,
      "logps_train/chosen": -175.91030883789062,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -129.6051025390625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.29153841733932495,
      "rewards_train/margins": 0.695236325263977,
      "rewards_train/rejected": -0.4036979079246521,
      "step": 39
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.805218882827761e-05,
      "loss": 0.5147,
      "step": 40
    },
    {
      "epoch": 0.46,
      "logps_train/chosen": -139.74484252929688,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -110.22254180908203,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.10924912989139557,
      "rewards_train/margins": 0.5274574607610703,
      "rewards_train/rejected": -0.6367065906524658,
      "step": 40
    },
    {
      "epoch": 0.46,
      "logps_train/chosen": -160.38890075683594,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -146.0457000732422,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": -0.15876291692256927,
      "rewards_train/margins": 0.4254945069551468,
      "rewards_train/rejected": -0.5842574238777161,
      "step": 40
    },
    {
      "epoch": 0.46,
      "logps_train/chosen": -153.27549743652344,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -135.46401977539062,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.12430575489997864,
      "rewards_train/margins": 0.5440479218959808,
      "rewards_train/rejected": -0.4197421669960022,
      "step": 40
    },
    {
      "epoch": 0.46,
      "logps_train/chosen": -176.70468139648438,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -124.0498275756836,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.16253966093063354,
      "rewards_train/margins": 0.6728933453559875,
      "rewards_train/rejected": -0.510353684425354,
      "step": 40
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.793028253763633e-05,
      "loss": 0.5968,
      "step": 41
    },
    {
      "epoch": 0.47,
      "logps_train/chosen": -171.22650146484375,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -146.68588256835938,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.2716864347457886,
      "rewards_train/margins": 0.5631259977817535,
      "rewards_train/rejected": -0.29143956303596497,
      "step": 41
    },
    {
      "epoch": 0.47,
      "logps_train/chosen": -125.95775604248047,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -99.5,
      "logps_train/rejected": -107.48272705078125,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.14445701241493225,
      "rewards_train/margins": 0.6441479623317719,
      "rewards_train/rejected": -0.7886049747467041,
      "step": 41
    },
    {
      "epoch": 0.47,
      "logps_train/chosen": -126.57373046875,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -81.5,
      "logps_train/rejected": -87.17964935302734,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.19477389752864838,
      "rewards_train/margins": 0.7732878178358078,
      "rewards_train/rejected": -0.5785139203071594,
      "step": 41
    },
    {
      "epoch": 0.47,
      "logps_train/chosen": -148.91766357421875,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -88.5,
      "logps_train/rejected": -95.6549072265625,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.44352418184280396,
      "rewards_train/margins": 0.24266964197158813,
      "rewards_train/rejected": -0.6861938238143921,
      "step": 41
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.780484065185188e-05,
      "loss": 0.6466,
      "step": 42
    },
    {
      "epoch": 0.48,
      "logps_train/chosen": -115.9168930053711,
      "logps_train/ref_chosen": -114.0,
      "logps_train/ref_rejected": -79.5,
      "logps_train/rejected": -87.66254425048828,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.21981437504291534,
      "rewards_train/margins": 0.599864050745964,
      "rewards_train/rejected": -0.8196784257888794,
      "step": 42
    },
    {
      "epoch": 0.48,
      "logps_train/chosen": -184.0168914794922,
      "logps_train/ref_chosen": -186.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -145.45132446289062,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.2588578462600708,
      "rewards_train/margins": 0.7897325158119202,
      "rewards_train/rejected": -0.5308746695518494,
      "step": 42
    },
    {
      "epoch": 0.48,
      "logps_train/chosen": -124.75608825683594,
      "logps_train/ref_chosen": -125.5,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -101.11264038085938,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.04958619177341461,
      "rewards_train/margins": 0.9850691705942154,
      "rewards_train/rejected": -0.9354829788208008,
      "step": 42
    },
    {
      "epoch": 0.48,
      "logps_train/chosen": -163.1077880859375,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -129.63858032226562,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.006554238498210907,
      "rewards_train/margins": 0.6211441680788994,
      "rewards_train/rejected": -0.6145899295806885,
      "step": 42
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.767588251267121e-05,
      "loss": 0.5648,
      "step": 43
    },
    {
      "epoch": 0.49,
      "logps_train/chosen": -151.1636962890625,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -93.0,
      "logps_train/rejected": -102.26740264892578,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.17907707393169403,
      "rewards_train/margins": 1.0967354625463486,
      "rewards_train/rejected": -0.9176583886146545,
      "step": 43
    },
    {
      "epoch": 0.49,
      "logps_train/chosen": -154.32647705078125,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -118.2194595336914,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.08923967182636261,
      "rewards_train/margins": 0.5269937962293625,
      "rewards_train/rejected": -0.6162334680557251,
      "step": 43
    },
    {
      "epoch": 0.49,
      "logps_train/chosen": -132.7676544189453,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -116.9314193725586,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": -0.3089926540851593,
      "rewards_train/margins": 0.21969613432884216,
      "rewards_train/rejected": -0.5286887884140015,
      "step": 43
    },
    {
      "epoch": 0.49,
      "logps_train/chosen": -158.78457641601562,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -104.1181640625,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.05724158138036728,
      "rewards_train/margins": 0.9021327868103981,
      "rewards_train/rejected": -0.9593743681907654,
      "step": 43
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.754342800400852e-05,
      "loss": 0.5647,
      "step": 44
    },
    {
      "epoch": 0.5,
      "logps_train/chosen": -145.93472290039062,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -123.93000030517578,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.3344568610191345,
      "rewards_train/margins": 0.7847806811332703,
      "rewards_train/rejected": -0.45032382011413574,
      "step": 44
    },
    {
      "epoch": 0.5,
      "logps_train/chosen": -140.81419372558594,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -109.78305053710938,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.007155388593673706,
      "rewards_train/margins": 0.5099476277828217,
      "rewards_train/rejected": -0.502792239189148,
      "step": 44
    },
    {
      "epoch": 0.5,
      "logps_train/chosen": -151.26019287109375,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -121.91029357910156,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.026507869362831116,
      "rewards_train/margins": 0.7881536334753036,
      "rewards_train/rejected": -0.8146615028381348,
      "step": 44
    },
    {
      "epoch": 0.5,
      "logps_train/chosen": -157.45396423339844,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -120.49565887451172,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.14910393953323364,
      "rewards_train/margins": 0.5894901752471924,
      "rewards_train/rejected": -0.44038623571395874,
      "step": 44
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.7407497548879384e-05,
      "loss": 0.5792,
      "step": 45
    },
    {
      "epoch": 0.51,
      "logps_train/chosen": -113.65975952148438,
      "logps_train/ref_chosen": -113.5,
      "logps_train/ref_rejected": -70.5,
      "logps_train/rejected": -76.96720886230469,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.006112739443778992,
      "rewards_train/margins": 0.6379715949296951,
      "rewards_train/rejected": -0.6440843343734741,
      "step": 45
    },
    {
      "epoch": 0.51,
      "logps_train/chosen": -154.4318389892578,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -128.56004333496094,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.1915566474199295,
      "rewards_train/margins": 0.7697056084871292,
      "rewards_train/rejected": -0.5781489610671997,
      "step": 45
    },
    {
      "epoch": 0.51,
      "logps_train/chosen": -164.68267822265625,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -124.21025848388672,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.11348249763250351,
      "rewards_train/margins": 0.9802481904625893,
      "rewards_train/rejected": -1.0937306880950928,
      "step": 45
    },
    {
      "epoch": 0.51,
      "logps_train/chosen": -148.53305053710938,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -122.53600311279297,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.3209740221500397,
      "rewards_train/margins": 0.7358295619487762,
      "rewards_train/rejected": -0.4148555397987366,
      "step": 45
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.726811210625176e-05,
      "loss": 0.5427,
      "step": 46
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -169.46038818359375,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -117.86236572265625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.38960567116737366,
      "rewards_train/margins": 0.9380492866039276,
      "rewards_train/rejected": -0.548443615436554,
      "step": 46
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -109.06082916259766,
      "logps_train/ref_chosen": -110.0,
      "logps_train/ref_rejected": -78.5,
      "logps_train/rejected": -86.27485656738281,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.10861498862504959,
      "rewards_train/margins": 0.8819746449589729,
      "rewards_train/rejected": -0.7733596563339233,
      "step": 46
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -149.625732421875,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -111.72176361083984,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.3020744025707245,
      "rewards_train/margins": 1.0752268731594086,
      "rewards_train/rejected": -0.7731524705886841,
      "step": 46
    },
    {
      "epoch": 0.53,
      "logps_train/chosen": -160.92486572265625,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -140.57608032226562,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.2807552218437195,
      "rewards_train/margins": 0.5618008971214294,
      "rewards_train/rejected": -0.28104567527770996,
      "step": 46
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.7125293167814345e-05,
      "loss": 0.5141,
      "step": 47
    },
    {
      "epoch": 0.54,
      "logps_train/chosen": -174.048828125,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -150.9656219482422,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.28046780824661255,
      "rewards_train/margins": 0.7596479058265686,
      "rewards_train/rejected": -0.47918009757995605,
      "step": 47
    },
    {
      "epoch": 0.54,
      "logps_train/chosen": -134.04562377929688,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -144.9838409423828,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.08473768830299377,
      "rewards_train/margins": 0.6593497097492218,
      "rewards_train/rejected": -0.7440873980522156,
      "step": 47
    },
    {
      "epoch": 0.54,
      "logps_train/chosen": -153.0111083984375,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -124.59916687011719,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.3758174479007721,
      "rewards_train/margins": 0.956462949514389,
      "rewards_train/rejected": -1.3322803974151611,
      "step": 47
    },
    {
      "epoch": 0.54,
      "logps_train/chosen": -139.15069580078125,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -112.43024444580078,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.30602389574050903,
      "rewards_train/margins": 0.9506352543830872,
      "rewards_train/rejected": -0.6446113586425781,
      "step": 47
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.697906275466279e-05,
      "loss": 0.5201,
      "step": 48
    },
    {
      "epoch": 0.55,
      "logps_train/chosen": -129.7398681640625,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -114.27020263671875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.321838915348053,
      "rewards_train/margins": 0.39028793573379517,
      "rewards_train/rejected": -0.7121268510818481,
      "step": 48
    },
    {
      "epoch": 0.55,
      "logps_train/chosen": -170.0889434814453,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -177.0,
      "logps_train/rejected": -180.01104736328125,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.4118082821369171,
      "rewards_train/margins": 0.7699443101882935,
      "rewards_train/rejected": -0.35813602805137634,
      "step": 48
    },
    {
      "epoch": 0.55,
      "logps_train/chosen": -166.91763305664062,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -129.64866638183594,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.2073570191860199,
      "rewards_train/margins": 0.8548417389392853,
      "rewards_train/rejected": -0.6474847197532654,
      "step": 48
    },
    {
      "epoch": 0.55,
      "logps_train/chosen": -160.1295166015625,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -138.43392944335938,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.05831342190504074,
      "rewards_train/margins": 0.5032433345913887,
      "rewards_train/rejected": -0.5615567564964294,
      "step": 48
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.68294434139043e-05,
      "loss": 0.6017,
      "step": 49
    },
    {
      "epoch": 0.56,
      "logps_train/chosen": -149.0851593017578,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -115.5216064453125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.5192182064056396,
      "rewards_train/margins": 1.6972576379776,
      "rewards_train/rejected": -1.1780394315719604,
      "step": 49
    },
    {
      "epoch": 0.56,
      "logps_train/chosen": -125.34487915039062,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -108.78245544433594,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.09363719820976257,
      "rewards_train/margins": 0.7584063112735748,
      "rewards_train/rejected": -0.6647691130638123,
      "step": 49
    },
    {
      "epoch": 0.56,
      "logps_train/chosen": -141.67967224121094,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -124.4493408203125,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": 0.10556776821613312,
      "rewards_train/margins": 0.5667611807584763,
      "rewards_train/rejected": -0.46119341254234314,
      "step": 49
    },
    {
      "epoch": 0.56,
      "logps_train/chosen": -124.61006164550781,
      "logps_train/ref_chosen": -125.0,
      "logps_train/ref_rejected": -88.0,
      "logps_train/rejected": -95.22357177734375,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.030595704913139343,
      "rewards_train/margins": 0.7752193659543991,
      "rewards_train/rejected": -0.7446236610412598,
      "step": 49
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.667645821518111e-05,
      "loss": 0.4851,
      "step": 50
    },
    {
      "epoch": 0.57,
      "logps_train/chosen": -176.50062561035156,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -155.60519409179688,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.17474240064620972,
      "rewards_train/margins": 0.5643627047538757,
      "rewards_train/rejected": -0.389620304107666,
      "step": 50
    },
    {
      "epoch": 0.57,
      "logps_train/chosen": -165.77471923828125,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -145.21810913085938,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.4006544351577759,
      "rewards_train/margins": 0.8921915590763092,
      "rewards_train/rejected": -0.4915371239185333,
      "step": 50
    },
    {
      "epoch": 0.57,
      "logps_train/chosen": -142.41026306152344,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -121.59886169433594,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.2132701277732849,
      "rewards_train/margins": 0.9618284106254578,
      "rewards_train/rejected": -0.7485582828521729,
      "step": 50
    },
    {
      "epoch": 0.57,
      "logps_train/chosen": -150.48184204101562,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -128.40724182128906,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.04735386371612549,
      "rewards_train/margins": 0.7860215902328491,
      "rewards_train/rejected": -0.8333754539489746,
      "step": 50
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.65201307471134e-05,
      "loss": 0.5391,
      "step": 51
    },
    {
      "epoch": 0.58,
      "logps_train/chosen": -158.06637573242188,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -139.3981170654297,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": 0.11431728303432465,
      "rewards_train/margins": 0.3441133350133896,
      "rewards_train/rejected": -0.22979605197906494,
      "step": 51
    },
    {
      "epoch": 0.58,
      "logps_train/chosen": -132.14430236816406,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -121.54273986816406,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.13733020424842834,
      "rewards_train/margins": 1.0312997996807098,
      "rewards_train/rejected": -1.1686300039291382,
      "step": 51
    },
    {
      "epoch": 0.58,
      "logps_train/chosen": -145.66217041015625,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -144.03201293945312,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.26250535249710083,
      "rewards_train/margins": 0.5208224654197693,
      "rewards_train/rejected": -0.7833278179168701,
      "step": 51
    },
    {
      "epoch": 0.58,
      "logps_train/chosen": -169.46548461914062,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -150.265869140625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.2237648218870163,
      "rewards_train/margins": 0.9571881741285324,
      "rewards_train/rejected": -0.7334233522415161,
      "step": 51
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.6360485113662216e-05,
      "loss": 0.6187,
      "step": 52
    },
    {
      "epoch": 0.59,
      "logps_train/chosen": -163.33822631835938,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -145.1938018798828,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.24283689260482788,
      "rewards_train/margins": 0.8202248811721802,
      "rewards_train/rejected": -0.5773879885673523,
      "step": 52
    },
    {
      "epoch": 0.59,
      "logps_train/chosen": -181.57000732421875,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -133.94961547851562,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.14456185698509216,
      "rewards_train/margins": 0.8153055608272552,
      "rewards_train/rejected": -0.6707437038421631,
      "step": 52
    },
    {
      "epoch": 0.59,
      "logps_train/chosen": -160.7414093017578,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -131.62620544433594,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.27234378457069397,
      "rewards_train/margins": 1.1264202892780304,
      "rewards_train/rejected": -0.8540765047073364,
      "step": 52
    },
    {
      "epoch": 0.59,
      "logps_train/chosen": -129.25555419921875,
      "logps_train/ref_chosen": -127.5,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -129.04898071289062,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": -0.16788795590400696,
      "rewards_train/margins": 0.3479476869106293,
      "rewards_train/rejected": -0.5158356428146362,
      "step": 52
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.6197545930412874e-05,
      "loss": 0.5332,
      "step": 53
    },
    {
      "epoch": 0.61,
      "logps_train/chosen": -153.34988403320312,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -160.0,
      "logps_train/rejected": -164.3514862060547,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": -0.12678614258766174,
      "rewards_train/margins": 0.33057892322540283,
      "rewards_train/rejected": -0.4573650658130646,
      "step": 53
    },
    {
      "epoch": 0.61,
      "logps_train/chosen": -162.29542541503906,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -130.75132751464844,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.25033989548683167,
      "rewards_train/margins": 1.0030124485492706,
      "rewards_train/rejected": -0.752672553062439,
      "step": 53
    },
    {
      "epoch": 0.61,
      "logps_train/chosen": -138.71487426757812,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -91.0,
      "logps_train/rejected": -101.21212768554688,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.04726358875632286,
      "rewards_train/margins": 1.0392774604260921,
      "rewards_train/rejected": -0.9920138716697693,
      "step": 53
    },
    {
      "epoch": 0.61,
      "logps_train/chosen": -159.54571533203125,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -129.88558959960938,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.34285280108451843,
      "rewards_train/margins": 0.4222678244113922,
      "rewards_train/rejected": -0.7651206254959106,
      "step": 53
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.6031338320779534e-05,
      "loss": 0.6483,
      "step": 54
    },
    {
      "epoch": 0.62,
      "logps_train/chosen": -173.44290161132812,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -133.28878784179688,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.12924429774284363,
      "rewards_train/margins": 0.9989194571971893,
      "rewards_train/rejected": -0.8696751594543457,
      "step": 54
    },
    {
      "epoch": 0.62,
      "logps_train/chosen": -148.29855346679688,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -115.97592163085938,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.009714528918266296,
      "rewards_train/margins": 0.9764523059129715,
      "rewards_train/rejected": -0.9861668348312378,
      "step": 54
    },
    {
      "epoch": 0.62,
      "logps_train/chosen": -163.42010498046875,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -132.7950439453125,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.16990332305431366,
      "rewards_train/margins": 1.1158150881528854,
      "rewards_train/rejected": -0.9459117650985718,
      "step": 54
    },
    {
      "epoch": 0.62,
      "logps_train/chosen": -169.2532958984375,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -126.75450134277344,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.2445676475763321,
      "rewards_train/margins": 0.5689687579870224,
      "rewards_train/rejected": -0.8135364055633545,
      "step": 54
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.586188791213143e-05,
      "loss": 0.5139,
      "step": 55
    },
    {
      "epoch": 0.63,
      "logps_train/chosen": -128.13194274902344,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -120.18585205078125,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.3469837009906769,
      "rewards_train/margins": 1.0525586307048798,
      "rewards_train/rejected": -1.3995423316955566,
      "step": 55
    },
    {
      "epoch": 0.63,
      "logps_train/chosen": -137.48464965820312,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -109.7799072265625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.32219475507736206,
      "rewards_train/margins": 0.4786474108695984,
      "rewards_train/rejected": -0.8008421659469604,
      "step": 55
    },
    {
      "epoch": 0.63,
      "logps_train/chosen": -155.82305908203125,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -140.88726806640625,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.3568549156188965,
      "rewards_train/margins": 1.018384873867035,
      "rewards_train/rejected": -0.6615299582481384,
      "step": 55
    },
    {
      "epoch": 0.63,
      "logps_train/chosen": -137.8286590576172,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -92.5,
      "logps_train/rejected": -101.40814208984375,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.3273978531360626,
      "rewards_train/margins": 0.5737676322460175,
      "rewards_train/rejected": -0.9011654853820801,
      "step": 55
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.568922083184144e-05,
      "loss": 0.5308,
      "step": 56
    },
    {
      "epoch": 0.64,
      "logps_train/chosen": -201.31527709960938,
      "logps_train/ref_chosen": -204.0,
      "logps_train/ref_rejected": -166.0,
      "logps_train/rejected": -172.15939331054688,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.23741737008094788,
      "rewards_train/margins": 0.8567750751972198,
      "rewards_train/rejected": -0.619357705116272,
      "step": 56
    },
    {
      "epoch": 0.64,
      "logps_train/chosen": -173.6513671875,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -154.0,
      "logps_train/rejected": -163.62249755859375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.11093807220458984,
      "rewards_train/margins": 1.0794376134872437,
      "rewards_train/rejected": -0.9684995412826538,
      "step": 56
    },
    {
      "epoch": 0.64,
      "logps_train/chosen": -170.75677490234375,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -160.66265869140625,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": -0.1634213924407959,
      "rewards_train/margins": 0.5591435432434082,
      "rewards_train/rejected": -0.7225649356842041,
      "step": 56
    },
    {
      "epoch": 0.64,
      "logps_train/chosen": -176.91485595703125,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -158.0,
      "logps_train/rejected": -165.61184692382812,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": -0.07683837413787842,
      "rewards_train/margins": 0.6816108822822571,
      "rewards_train/rejected": -0.7584492564201355,
      "step": 56
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.5513363703257496e-05,
      "loss": 0.5886,
      "step": 57
    },
    {
      "epoch": 0.65,
      "logps_train/chosen": -147.32388305664062,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -118.71624755859375,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": -0.22408804297447205,
      "rewards_train/margins": 0.9452424347400665,
      "rewards_train/rejected": -1.1693304777145386,
      "step": 57
    },
    {
      "epoch": 0.65,
      "logps_train/chosen": -149.79339599609375,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -138.41961669921875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.2672303020954132,
      "rewards_train/margins": 0.8092033565044403,
      "rewards_train/rejected": -1.0764336585998535,
      "step": 57
    },
    {
      "epoch": 0.65,
      "logps_train/chosen": -108.96585083007812,
      "logps_train/ref_chosen": -104.5,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -111.37809753417969,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.4704621434211731,
      "rewards_train/margins": 0.4902119040489197,
      "rewards_train/rejected": -0.9606740474700928,
      "step": 57
    },
    {
      "epoch": 0.65,
      "logps_train/chosen": -176.698486328125,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -154.18609619140625,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": 0.010303527116775513,
      "rewards_train/margins": 0.8293034136295319,
      "rewards_train/rejected": -0.8189998865127563,
      "step": 57
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.533434364159761e-05,
      "loss": 0.5843,
      "step": 58
    },
    {
      "epoch": 0.66,
      "logps_train/chosen": -136.48675537109375,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -142.8067626953125,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.18441638350486755,
      "rewards_train/margins": 0.39918962121009827,
      "rewards_train/rejected": -0.5836060047149658,
      "step": 58
    },
    {
      "epoch": 0.66,
      "logps_train/chosen": -184.3641357421875,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -153.0880584716797,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.08172591775655746,
      "rewards_train/margins": 0.6115528717637062,
      "rewards_train/rejected": -0.6932787895202637,
      "step": 58
    },
    {
      "epoch": 0.66,
      "logps_train/chosen": -175.23277282714844,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -124.69281005859375,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.07252401113510132,
      "rewards_train/margins": 0.7904370427131653,
      "rewards_train/rejected": -0.717913031578064,
      "step": 58
    },
    {
      "epoch": 0.66,
      "logps_train/chosen": -144.44598388671875,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -113.03366088867188,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.09381596744060516,
      "rewards_train/margins": 1.031522586941719,
      "rewards_train/rejected": -1.1253385543823242,
      "step": 58
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.515218824976895e-05,
      "loss": 0.5473,
      "step": 59
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -135.3798065185547,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -105.5,
      "logps_train/rejected": -120.29751586914062,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.3549787104129791,
      "rewards_train/margins": 1.1142815053462982,
      "rewards_train/rejected": -1.4692602157592773,
      "step": 59
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -140.24740600585938,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -105.5,
      "logps_train/rejected": -117.92471313476562,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.19534514844417572,
      "rewards_train/margins": 1.0404373556375504,
      "rewards_train/rejected": -1.235782504081726,
      "step": 59
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -170.62718200683594,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -133.84637451171875,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": 0.15193051099777222,
      "rewards_train/margins": 1.3069776892662048,
      "rewards_train/rejected": -1.1550471782684326,
      "step": 59
    },
    {
      "epoch": 0.67,
      "logps_train/chosen": -171.68701171875,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -140.77157592773438,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.14379817247390747,
      "rewards_train/margins": 0.9716156721115112,
      "rewards_train/rejected": -0.8278174996376038,
      "step": 59
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.496692561411182e-05,
      "loss": 0.4245,
      "step": 60
    },
    {
      "epoch": 0.69,
      "logps_train/chosen": -145.77418518066406,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -127.12965393066406,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.3304464817047119,
      "rewards_train/margins": 0.8365238904953003,
      "rewards_train/rejected": -1.1669703722000122,
      "step": 60
    },
    {
      "epoch": 0.69,
      "logps_train/chosen": -173.94161987304688,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -132.8108673095703,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.3185763955116272,
      "rewards_train/margins": 0.8773545622825623,
      "rewards_train/rejected": -1.1959309577941895,
      "step": 60
    },
    {
      "epoch": 0.69,
      "logps_train/chosen": -190.071044921875,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -159.0,
      "logps_train/rejected": -168.48336791992188,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.10808217525482178,
      "rewards_train/margins": 0.8708211779594421,
      "rewards_train/rejected": -0.9789033532142639,
      "step": 60
    },
    {
      "epoch": 0.69,
      "logps_train/chosen": -121.50714111328125,
      "logps_train/ref_chosen": -116.5,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -117.76773071289062,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.5074518322944641,
      "rewards_train/margins": 1.0132667422294617,
      "rewards_train/rejected": -1.5207185745239258,
      "step": 60
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.477858430006906e-05,
      "loss": 0.5009,
      "step": 61
    },
    {
      "epoch": 0.7,
      "logps_train/chosen": -148.4321746826172,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -105.5,
      "logps_train/rejected": -122.11833953857422,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.6903851628303528,
      "rewards_train/margins": 0.9600227475166321,
      "rewards_train/rejected": -1.6504079103469849,
      "step": 61
    },
    {
      "epoch": 0.7,
      "logps_train/chosen": -167.27960205078125,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -140.42367553710938,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.3914361596107483,
      "rewards_train/margins": 0.738186776638031,
      "rewards_train/rejected": -1.1296229362487793,
      "step": 61
    },
    {
      "epoch": 0.7,
      "logps_train/chosen": -175.978271484375,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -129.72894287109375,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.16933763027191162,
      "rewards_train/margins": 1.182854413986206,
      "rewards_train/rejected": -1.3521920442581177,
      "step": 61
    },
    {
      "epoch": 0.7,
      "logps_train/chosen": -153.08892822265625,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -137.57305908203125,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.2504955530166626,
      "rewards_train/margins": 0.9743890762329102,
      "rewards_train/rejected": -1.2248846292495728,
      "step": 61
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.458719334778153e-05,
      "loss": 0.5142,
      "step": 62
    },
    {
      "epoch": 0.71,
      "logps_train/chosen": -143.49911499023438,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -156.21580505371094,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": -0.044589295983314514,
      "rewards_train/margins": 0.8543341010808945,
      "rewards_train/rejected": -0.898923397064209,
      "step": 62
    },
    {
      "epoch": 0.71,
      "logps_train/chosen": -156.5689239501953,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -125.41449737548828,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.5022292137145996,
      "rewards_train/margins": 0.6355087757110596,
      "rewards_train/rejected": -1.1377379894256592,
      "step": 62
    },
    {
      "epoch": 0.71,
      "logps_train/chosen": -201.266357421875,
      "logps_train/ref_chosen": -196.0,
      "logps_train/ref_rejected": -159.0,
      "logps_train/rejected": -173.8751220703125,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.43591195344924927,
      "rewards_train/margins": 1.0374639630317688,
      "rewards_train/rejected": -1.473375916481018,
      "step": 62
    },
    {
      "epoch": 0.71,
      "logps_train/chosen": -149.0179901123047,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -170.43136596679688,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -1.208733081817627,
      "rewards_train/margins": 0.9972443580627441,
      "rewards_train/rejected": -2.205977439880371,
      "step": 62
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.43927822676105e-05,
      "loss": 0.5598,
      "step": 63
    },
    {
      "epoch": 0.72,
      "logps_train/chosen": -126.84237670898438,
      "logps_train/ref_chosen": -117.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -115.09578704833984,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.9945892691612244,
      "rewards_train/margins": 0.3996577858924866,
      "rewards_train/rejected": -1.394247055053711,
      "step": 63
    },
    {
      "epoch": 0.72,
      "logps_train/chosen": -142.31930541992188,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -130.88372802734375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.607906699180603,
      "rewards_train/margins": 0.7457003593444824,
      "rewards_train/rejected": -1.3536070585250854,
      "step": 63
    },
    {
      "epoch": 0.72,
      "logps_train/chosen": -124.56082916259766,
      "logps_train/ref_chosen": -115.5,
      "logps_train/ref_rejected": -95.5,
      "logps_train/rejected": -110.76518249511719,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.923465371131897,
      "rewards_train/margins": 0.618140697479248,
      "rewards_train/rejected": -1.541606068611145,
      "step": 63
    },
    {
      "epoch": 0.72,
      "logps_train/chosen": -179.1411895751953,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -157.0,
      "logps_train/rejected": -171.82464599609375,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.33872923254966736,
      "rewards_train/margins": 1.1219574511051178,
      "rewards_train/rejected": -1.4606866836547852,
      "step": 63
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.419538103558742e-05,
      "loss": 0.5848,
      "step": 64
    },
    {
      "epoch": 0.73,
      "logps_train/chosen": -168.5423583984375,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -148.78189086914062,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.5333375334739685,
      "rewards_train/margins": 0.8630149960517883,
      "rewards_train/rejected": -1.3963525295257568,
      "step": 64
    },
    {
      "epoch": 0.73,
      "logps_train/chosen": -140.47198486328125,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -124.3873291015625,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -1.0393857955932617,
      "rewards_train/margins": 0.7289131879806519,
      "rewards_train/rejected": -1.7682989835739136,
      "step": 64
    },
    {
      "epoch": 0.73,
      "logps_train/chosen": -174.19720458984375,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -125.15985870361328,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.5634703040122986,
      "rewards_train/margins": 1.0730233788490295,
      "rewards_train/rejected": -1.6364936828613281,
      "step": 64
    },
    {
      "epoch": 0.73,
      "logps_train/chosen": -160.61813354492188,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -143.4553680419922,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.6700178384780884,
      "rewards_train/margins": 0.5729802846908569,
      "rewards_train/rejected": -1.2429981231689453,
      "step": 64
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.3995020088792e-05,
      "loss": 0.5701,
      "step": 65
    },
    {
      "epoch": 0.74,
      "logps_train/chosen": -188.96270751953125,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -148.6373291015625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -1.192512035369873,
      "rewards_train/margins": 0.6878228187561035,
      "rewards_train/rejected": -1.8803348541259766,
      "step": 65
    },
    {
      "epoch": 0.74,
      "logps_train/chosen": -173.8163299560547,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -141.50582885742188,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.7177664041519165,
      "rewards_train/margins": 0.6095980405807495,
      "rewards_train/rejected": -1.327364444732666,
      "step": 65
    },
    {
      "epoch": 0.74,
      "logps_train/chosen": -156.86846923828125,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -131.81948852539062,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.514484167098999,
      "rewards_train/margins": 1.3581879138946533,
      "rewards_train/rejected": -1.8726720809936523,
      "step": 65
    },
    {
      "epoch": 0.74,
      "logps_train/chosen": -153.33334350585938,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -140.69818115234375,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.8610811233520508,
      "rewards_train/margins": 0.6778771877288818,
      "rewards_train/rejected": -1.5389583110809326,
      "step": 65
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.379173032065912e-05,
      "loss": 0.571,
      "step": 66
    },
    {
      "epoch": 0.75,
      "logps_train/chosen": -123.39452362060547,
      "logps_train/ref_chosen": -111.5,
      "logps_train/ref_rejected": -91.0,
      "logps_train/rejected": -110.48931121826172,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -1.208837628364563,
      "rewards_train/margins": 0.7541805505752563,
      "rewards_train/rejected": -1.9630181789398193,
      "step": 66
    },
    {
      "epoch": 0.75,
      "logps_train/chosen": -149.31439208984375,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -124.77264404296875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.7231384515762329,
      "rewards_train/margins": 1.0160391330718994,
      "rewards_train/rejected": -1.7391775846481323,
      "step": 66
    },
    {
      "epoch": 0.75,
      "logps_train/chosen": -189.39576721191406,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -165.7838134765625,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.8770769834518433,
      "rewards_train/margins": 0.7735692262649536,
      "rewards_train/rejected": -1.6506462097167969,
      "step": 66
    },
    {
      "epoch": 0.75,
      "logps_train/chosen": -172.92323303222656,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -151.0,
      "logps_train/rejected": -169.48965454101562,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.6337293982505798,
      "rewards_train/margins": 1.203711450099945,
      "rewards_train/rejected": -1.837440848350525,
      "step": 66
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.358554307621541e-05,
      "loss": 0.519,
      "step": 67
    },
    {
      "epoch": 0.77,
      "logps_train/chosen": -106.08445739746094,
      "logps_train/ref_chosen": -95.5,
      "logps_train/ref_rejected": -84.5,
      "logps_train/rejected": -104.09698486328125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -1.045597791671753,
      "rewards_train/margins": 0.9242256879806519,
      "rewards_train/rejected": -1.9698234796524048,
      "step": 67
    },
    {
      "epoch": 0.77,
      "logps_train/chosen": -157.91793823242188,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -139.233154296875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -1.0456013679504395,
      "rewards_train/margins": 1.0714645385742188,
      "rewards_train/rejected": -2.117065906524658,
      "step": 67
    },
    {
      "epoch": 0.77,
      "logps_train/chosen": -191.77606201171875,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -159.0693359375,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.44342678785324097,
      "rewards_train/margins": 1.00628000497818,
      "rewards_train/rejected": -1.449706792831421,
      "step": 67
    },
    {
      "epoch": 0.77,
      "logps_train/chosen": -136.33993530273438,
      "logps_train/ref_chosen": -125.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -121.84136199951172,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -1.1351652145385742,
      "rewards_train/margins": 0.6773405075073242,
      "rewards_train/rejected": -1.8125057220458984,
      "step": 67
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.337649014724621e-05,
      "loss": 0.5137,
      "step": 68
    },
    {
      "epoch": 0.78,
      "logps_train/chosen": -163.60760498046875,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -143.341064453125,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.8888216018676758,
      "rewards_train/margins": 0.49487829208374023,
      "rewards_train/rejected": -1.383699893951416,
      "step": 68
    },
    {
      "epoch": 0.78,
      "logps_train/chosen": -172.0780029296875,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -118.73085021972656,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.7458871603012085,
      "rewards_train/margins": 0.9197756052017212,
      "rewards_train/rejected": -1.6656627655029297,
      "step": 68
    },
    {
      "epoch": 0.78,
      "logps_train/chosen": -135.83432006835938,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -148.3021697998047,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.9420253038406372,
      "rewards_train/margins": 0.8683685064315796,
      "rewards_train/rejected": -1.8103938102722168,
      "step": 68
    },
    {
      "epoch": 0.78,
      "logps_train/chosen": -156.50640869140625,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -100.5,
      "logps_train/rejected": -119.85636138916016,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -1.0331604480743408,
      "rewards_train/margins": 0.8918311595916748,
      "rewards_train/rejected": -1.9249916076660156,
      "step": 68
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.31646037673936e-05,
      "loss": 0.6014,
      "step": 69
    },
    {
      "epoch": 0.79,
      "logps_train/chosen": -109.36697387695312,
      "logps_train/ref_chosen": -99.0,
      "logps_train/ref_rejected": -69.5,
      "logps_train/rejected": -90.85383605957031,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -1.0384798049926758,
      "rewards_train/margins": 1.0995402336120605,
      "rewards_train/rejected": -2.1380200386047363,
      "step": 69
    },
    {
      "epoch": 0.79,
      "logps_train/chosen": -164.5358428955078,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -157.33303833007812,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -1.1899994611740112,
      "rewards_train/margins": 0.6788498163223267,
      "rewards_train/rejected": -1.868849277496338,
      "step": 69
    },
    {
      "epoch": 0.79,
      "logps_train/chosen": -172.2188720703125,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -148.18170166015625,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.46329355239868164,
      "rewards_train/margins": 1.061615228652954,
      "rewards_train/rejected": -1.5249087810516357,
      "step": 69
    },
    {
      "epoch": 0.79,
      "logps_train/chosen": -157.74252319335938,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -144.2315673828125,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.8678563833236694,
      "rewards_train/margins": 0.7282001972198486,
      "rewards_train/rejected": -1.596056580543518,
      "step": 69
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.2949916607186357e-05,
      "loss": 0.5289,
      "step": 70
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -151.2384033203125,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -136.65538024902344,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.9337508678436279,
      "rewards_train/margins": 0.9835443496704102,
      "rewards_train/rejected": -1.917295217514038,
      "step": 70
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -164.0899658203125,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -119.64618682861328,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.9898558855056763,
      "rewards_train/margins": 1.0959542989730835,
      "rewards_train/rejected": -2.0858101844787598,
      "step": 70
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -158.28138732910156,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -87.5,
      "logps_train/rejected": -107.51951599121094,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.7542130947113037,
      "rewards_train/margins": 1.254964828491211,
      "rewards_train/rejected": -2.0091779232025146,
      "step": 70
    },
    {
      "epoch": 0.8,
      "logps_train/chosen": -158.1058349609375,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -135.90130615234375,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -1.465502142906189,
      "rewards_train/margins": 0.7091010808944702,
      "rewards_train/rejected": -2.174603223800659,
      "step": 70
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.273246176900252e-05,
      "loss": 0.5182,
      "step": 71
    },
    {
      "epoch": 0.81,
      "logps_train/chosen": -180.446533203125,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -157.4761962890625,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.7675294876098633,
      "rewards_train/margins": 0.56602942943573,
      "rewards_train/rejected": -1.3335589170455933,
      "step": 71
    },
    {
      "epoch": 0.81,
      "logps_train/chosen": -166.73056030273438,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -143.46798706054688,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.6558671593666077,
      "rewards_train/margins": 1.4000133872032166,
      "rewards_train/rejected": -2.055880546569824,
      "step": 71
    },
    {
      "epoch": 0.81,
      "logps_train/chosen": -188.02003479003906,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -167.78900146484375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.7188003063201904,
      "rewards_train/margins": 1.1408617496490479,
      "rewards_train/rejected": -1.8596620559692383,
      "step": 71
    },
    {
      "epoch": 0.81,
      "logps_train/chosen": -144.71981811523438,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -128.44407653808594,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -1.008699893951416,
      "rewards_train/margins": 0.8348284959793091,
      "rewards_train/rejected": -1.843528389930725,
      "step": 71
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.251227278196536e-05,
      "loss": 0.4895,
      "step": 72
    },
    {
      "epoch": 0.82,
      "logps_train/chosen": -132.26991271972656,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -113.90031433105469,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.6027978658676147,
      "rewards_train/margins": 1.070241093635559,
      "rewards_train/rejected": -1.6730389595031738,
      "step": 72
    },
    {
      "epoch": 0.82,
      "logps_train/chosen": -156.4181671142578,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -129.92678833007812,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.6840531826019287,
      "rewards_train/margins": 0.7702467441558838,
      "rewards_train/rejected": -1.4542999267578125,
      "step": 72
    },
    {
      "epoch": 0.82,
      "logps_train/chosen": -137.28082275390625,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -150.2662353515625,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.9548407793045044,
      "rewards_train/margins": 0.8424848318099976,
      "rewards_train/rejected": -1.797325611114502,
      "step": 72
    },
    {
      "epoch": 0.82,
      "logps_train/chosen": -159.76388549804688,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -127.27428436279297,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.3386008143424988,
      "rewards_train/margins": 1.3924408555030823,
      "rewards_train/rejected": -1.731041669845581,
      "step": 72
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.228938359677354e-05,
      "loss": 0.4813,
      "step": 73
    },
    {
      "epoch": 0.83,
      "logps_train/chosen": -158.0514373779297,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -126.70331573486328,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.8166680335998535,
      "rewards_train/margins": 0.6750500202178955,
      "rewards_train/rejected": -1.491718053817749,
      "step": 73
    },
    {
      "epoch": 0.83,
      "logps_train/chosen": -167.51153564453125,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -154.432861328125,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.33699363470077515,
      "rewards_train/margins": 0.9434502720832825,
      "rewards_train/rejected": -1.2804439067840576,
      "step": 73
    },
    {
      "epoch": 0.83,
      "logps_train/chosen": -197.72607421875,
      "logps_train/ref_chosen": -196.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -160.88992309570312,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.23745089769363403,
      "rewards_train/margins": 0.9035444855690002,
      "rewards_train/rejected": -1.1409953832626343,
      "step": 73
    },
    {
      "epoch": 0.83,
      "logps_train/chosen": -149.27923583984375,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -128.9501953125,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.7999448180198669,
      "rewards_train/margins": 1.0314993262290955,
      "rewards_train/rejected": -1.8314441442489624,
      "step": 73
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.206382858046636e-05,
      "loss": 0.5071,
      "step": 74
    },
    {
      "epoch": 0.85,
      "logps_train/chosen": -176.96607971191406,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -150.25148010253906,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.6526632308959961,
      "rewards_train/margins": 0.5287351608276367,
      "rewards_train/rejected": -1.1813983917236328,
      "step": 74
    },
    {
      "epoch": 0.85,
      "logps_train/chosen": -202.90512084960938,
      "logps_train/ref_chosen": -201.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -153.5736083984375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.19715136289596558,
      "rewards_train/margins": 0.9527884125709534,
      "rewards_train/rejected": -1.149939775466919,
      "step": 74
    },
    {
      "epoch": 0.85,
      "logps_train/chosen": -149.58419799804688,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -133.56790161132812,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.5854701995849609,
      "rewards_train/margins": 1.0398268699645996,
      "rewards_train/rejected": -1.6252970695495605,
      "step": 74
    },
    {
      "epoch": 0.85,
      "logps_train/chosen": -132.50787353515625,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -119.30223083496094,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.8568595051765442,
      "rewards_train/margins": 0.9427974820137024,
      "rewards_train/rejected": -1.7996569871902466,
      "step": 74
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.1835642511124656e-05,
      "loss": 0.5643,
      "step": 75
    },
    {
      "epoch": 0.86,
      "logps_train/chosen": -158.19993591308594,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -128.560546875,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.4262433648109436,
      "rewards_train/margins": 1.189382255077362,
      "rewards_train/rejected": -1.6156256198883057,
      "step": 75
    },
    {
      "epoch": 0.86,
      "logps_train/chosen": -188.24725341796875,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -144.20211791992188,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.1420358419418335,
      "rewards_train/margins": 0.9777122735977173,
      "rewards_train/rejected": -1.1197481155395508,
      "step": 75
    },
    {
      "epoch": 0.86,
      "logps_train/chosen": -137.14773559570312,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -112.96627807617188,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.8436791896820068,
      "rewards_train/margins": 0.7864446640014648,
      "rewards_train/rejected": -1.6301238536834717,
      "step": 75
    },
    {
      "epoch": 0.86,
      "logps_train/chosen": -174.13377380371094,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -139.75009155273438,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.310349702835083,
      "rewards_train/margins": 1.0647573471069336,
      "rewards_train/rejected": -1.3751070499420166,
      "step": 75
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.160486057250849e-05,
      "loss": 0.4979,
      "step": 76
    },
    {
      "epoch": 0.87,
      "logps_train/chosen": -142.59698486328125,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -129.82534790039062,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.3968072235584259,
      "rewards_train/margins": 0.733596533536911,
      "rewards_train/rejected": -1.130403757095337,
      "step": 76
    },
    {
      "epoch": 0.87,
      "logps_train/chosen": -161.75323486328125,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -125.47662353515625,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": -1.1422175168991089,
      "rewards_train/margins": 0.11257338523864746,
      "rewards_train/rejected": -1.2547909021377563,
      "step": 76
    },
    {
      "epoch": 0.87,
      "logps_train/chosen": -178.9916229248047,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -132.39141845703125,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.31464019417762756,
      "rewards_train/margins": 0.715956062078476,
      "rewards_train/rejected": -1.0305962562561035,
      "step": 76
    },
    {
      "epoch": 0.87,
      "logps_train/chosen": -152.24346923828125,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -127.48640441894531,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.2509094476699829,
      "rewards_train/margins": 1.095436692237854,
      "rewards_train/rejected": -1.346346139907837,
      "step": 76
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.137151834863213e-05,
      "loss": 0.6958,
      "step": 77
    },
    {
      "epoch": 0.88,
      "logps_train/chosen": -148.580078125,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -157.56333923339844,
      "rewards_train/accuracies": 0.625,
      "rewards_train/chosen": -0.48818454146385193,
      "rewards_train/margins": 0.4652206003665924,
      "rewards_train/rejected": -0.9534051418304443,
      "step": 77
    },
    {
      "epoch": 0.88,
      "logps_train/chosen": -166.72164916992188,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -147.3184356689453,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.47060221433639526,
      "rewards_train/margins": 0.5743276476860046,
      "rewards_train/rejected": -1.0449298620224,
      "step": 77
    },
    {
      "epoch": 0.88,
      "logps_train/chosen": -143.37298583984375,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -130.98110961914062,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.5442638397216797,
      "rewards_train/margins": 1.0199611186981201,
      "rewards_train/rejected": -1.5642249584197998,
      "step": 77
    },
    {
      "epoch": 0.88,
      "logps_train/chosen": -155.43408203125,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -129.516845703125,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.3463389575481415,
      "rewards_train/margins": 0.9465561211109161,
      "rewards_train/rejected": -1.2928950786590576,
      "step": 77
    },
    {
      "epoch": 0.89,
      "learning_rate": 4.1135651818277445e-05,
      "loss": 0.5414,
      "step": 78
    },
    {
      "epoch": 0.89,
      "logps_train/chosen": -154.96820068359375,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -132.89260864257812,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.20873412489891052,
      "rewards_train/margins": 1.1599203646183014,
      "rewards_train/rejected": -1.368654489517212,
      "step": 78
    },
    {
      "epoch": 0.89,
      "logps_train/chosen": -144.24415588378906,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -130.75668334960938,
      "rewards_train/accuracies": 0.53125,
      "rewards_train/chosen": -0.38398557901382446,
      "rewards_train/margins": 0.327619731426239,
      "rewards_train/rejected": -0.7116053104400635,
      "step": 78
    },
    {
      "epoch": 0.89,
      "logps_train/chosen": -139.5198974609375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -116.1756591796875,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.28363025188446045,
      "rewards_train/margins": 0.8827880620956421,
      "rewards_train/rejected": -1.1664183139801025,
      "step": 78
    },
    {
      "epoch": 0.89,
      "logps_train/chosen": -180.26397705078125,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -149.57452392578125,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.1939750611782074,
      "rewards_train/margins": 1.027538686990738,
      "rewards_train/rejected": -1.2215137481689453,
      "step": 78
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.089729734944634e-05,
      "loss": 0.544,
      "step": 79
    },
    {
      "epoch": 0.9,
      "logps_train/chosen": -143.99681091308594,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -153.2803497314453,
      "rewards_train/accuracies": 0.5625,
      "rewards_train/chosen": -0.678196370601654,
      "rewards_train/margins": 0.220004141330719,
      "rewards_train/rejected": -0.898200511932373,
      "step": 79
    },
    {
      "epoch": 0.9,
      "logps_train/chosen": -158.53390502929688,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -146.01658630371094,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.03825424611568451,
      "rewards_train/margins": 0.9192156940698624,
      "rewards_train/rejected": -0.9574699401855469,
      "step": 79
    },
    {
      "epoch": 0.9,
      "logps_train/chosen": -156.74154663085938,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -150.85284423828125,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.48094165325164795,
      "rewards_train/margins": 0.6565214395523071,
      "rewards_train/rejected": -1.137463092803955,
      "step": 79
    },
    {
      "epoch": 0.9,
      "logps_train/chosen": -183.3664093017578,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -156.0,
      "logps_train/rejected": -164.84950256347656,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.10578218847513199,
      "rewards_train/margins": 0.8176443204283714,
      "rewards_train/rejected": -0.9234265089035034,
      "step": 79
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.065649169375324e-05,
      "loss": 0.6275,
      "step": 80
    },
    {
      "epoch": 0.91,
      "logps_train/chosen": -161.34866333007812,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -130.3014678955078,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.00708414614200592,
      "rewards_train/margins": 1.3033236116170883,
      "rewards_train/rejected": -1.3104077577590942,
      "step": 80
    },
    {
      "epoch": 0.91,
      "logps_train/chosen": -196.28811645507812,
      "logps_train/ref_chosen": -194.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -164.7366943359375,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.1925809383392334,
      "rewards_train/margins": 1.0608729124069214,
      "rewards_train/rejected": -1.2534538507461548,
      "step": 80
    },
    {
      "epoch": 0.91,
      "logps_train/chosen": -143.91098022460938,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -137.1011962890625,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.027426350861787796,
      "rewards_train/margins": 0.7220006696879864,
      "rewards_train/rejected": -0.7494270205497742,
      "step": 80
    },
    {
      "epoch": 0.91,
      "logps_train/chosen": -186.50070190429688,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -138.23883056640625,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.17532047629356384,
      "rewards_train/margins": 1.30330428481102,
      "rewards_train/rejected": -1.127983808517456,
      "step": 80
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.041327198075838e-05,
      "loss": 0.4788,
      "step": 81
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -128.11949157714844,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -107.81585693359375,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.03633883595466614,
      "rewards_train/margins": 1.1293288171291351,
      "rewards_train/rejected": -1.1656676530838013,
      "step": 81
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -165.39813232421875,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -141.9514617919922,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.31048035621643066,
      "rewards_train/margins": 1.1549919247627258,
      "rewards_train/rejected": -0.8445115685462952,
      "step": 81
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -126.2926025390625,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -113.79829406738281,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.19713124632835388,
      "rewards_train/margins": 0.9133626520633698,
      "rewards_train/rejected": -1.1104938983917236,
      "step": 81
    },
    {
      "epoch": 0.93,
      "logps_train/chosen": -187.25137329101562,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -142.0508575439453,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": -0.020840942859649658,
      "rewards_train/margins": 1.0018226504325867,
      "rewards_train/rejected": -1.0226635932922363,
      "step": 81
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.016767571224284e-05,
      "loss": 0.4871,
      "step": 82
    },
    {
      "epoch": 0.94,
      "logps_train/chosen": -136.57688903808594,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -91.5,
      "logps_train/rejected": -102.67213439941406,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.5452377796173096,
      "rewards_train/margins": 0.577444314956665,
      "rewards_train/rejected": -1.1226820945739746,
      "step": 82
    },
    {
      "epoch": 0.94,
      "logps_train/chosen": -162.8528289794922,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -123.56939697265625,
      "rewards_train/accuracies": 0.59375,
      "rewards_train/chosen": -0.4730757772922516,
      "rewards_train/margins": 0.4601331651210785,
      "rewards_train/rejected": -0.9332089424133301,
      "step": 82
    },
    {
      "epoch": 0.94,
      "logps_train/chosen": -152.0038299560547,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -127.97988891601562,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.0006765499711036682,
      "rewards_train/margins": 1.3138957843184471,
      "rewards_train/rejected": -1.3145723342895508,
      "step": 82
    },
    {
      "epoch": 0.94,
      "logps_train/chosen": -141.26974487304688,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -132.22300720214844,
      "rewards_train/accuracies": 0.46875,
      "rewards_train/chosen": -0.452718585729599,
      "rewards_train/margins": 0.32729652523994446,
      "rewards_train/rejected": -0.7800151109695435,
      "step": 82
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.991974075642621e-05,
      "loss": 0.5842,
      "step": 83
    },
    {
      "epoch": 0.95,
      "logps_train/chosen": -144.98745727539062,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -131.68789672851562,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.15765535831451416,
      "rewards_train/margins": 1.060597538948059,
      "rewards_train/rejected": -1.2182528972625732,
      "step": 83
    },
    {
      "epoch": 0.95,
      "logps_train/chosen": -159.57968139648438,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -131.57052612304688,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.22476573288440704,
      "rewards_train/margins": 1.0797476023435593,
      "rewards_train/rejected": -1.3045133352279663,
      "step": 83
    },
    {
      "epoch": 0.95,
      "logps_train/chosen": -138.2335968017578,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -133.75772094726562,
      "rewards_train/accuracies": 0.8125,
      "rewards_train/chosen": -0.14119389653205872,
      "rewards_train/margins": 1.2634479701519012,
      "rewards_train/rejected": -1.40464186668396,
      "step": 83
    },
    {
      "epoch": 0.95,
      "logps_train/chosen": -145.84381103515625,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -147.16595458984375,
      "rewards_train/accuracies": 0.65625,
      "rewards_train/chosen": -0.2642488479614258,
      "rewards_train/margins": 0.4790076017379761,
      "rewards_train/rejected": -0.7432564496994019,
      "step": 83
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.96695053421277e-05,
      "loss": 0.4781,
      "step": 84
    },
    {
      "epoch": 0.96,
      "logps_train/chosen": -183.54966735839844,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -143.7237548828125,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.6588726043701172,
      "rewards_train/margins": 0.7602804899215698,
      "rewards_train/rejected": -1.419153094291687,
      "step": 84
    },
    {
      "epoch": 0.96,
      "logps_train/chosen": -118.30413055419922,
      "logps_train/ref_chosen": -113.5,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -104.0281982421875,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": -0.46830376982688904,
      "rewards_train/margins": 0.7527767717838287,
      "rewards_train/rejected": -1.2210805416107178,
      "step": 84
    },
    {
      "epoch": 0.96,
      "logps_train/chosen": -142.31927490234375,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -121.17881774902344,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.44811493158340454,
      "rewards_train/margins": 0.5898844599723816,
      "rewards_train/rejected": -1.0379993915557861,
      "step": 84
    },
    {
      "epoch": 0.96,
      "logps_train/chosen": -162.7830047607422,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -154.27557373046875,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.023749448359012604,
      "rewards_train/margins": 0.9083372130990028,
      "rewards_train/rejected": -0.8845877647399902,
      "step": 84
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.941700805287168e-05,
      "loss": 0.5488,
      "step": 85
    },
    {
      "epoch": 0.97,
      "logps_train/chosen": -145.57501220703125,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -146.69969177246094,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.0080266073346138,
      "rewards_train/margins": 0.8595626428723335,
      "rewards_train/rejected": -0.8515360355377197,
      "step": 85
    },
    {
      "epoch": 0.97,
      "logps_train/chosen": -157.25,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -141.33470153808594,
      "rewards_train/accuracies": 0.78125,
      "rewards_train/chosen": 0.25292864441871643,
      "rewards_train/margins": 0.8953824937343597,
      "rewards_train/rejected": -0.6424538493156433,
      "step": 85
    },
    {
      "epoch": 0.97,
      "logps_train/chosen": -156.912109375,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -123.60111999511719,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": 0.2099611759185791,
      "rewards_train/margins": 0.9257863759994507,
      "rewards_train/rejected": -0.7158252000808716,
      "step": 85
    },
    {
      "epoch": 0.97,
      "logps_train/chosen": -148.09971618652344,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -98.0,
      "logps_train/rejected": -110.16642761230469,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.032508380711078644,
      "rewards_train/margins": 1.2118464782834053,
      "rewards_train/rejected": -1.1793380975723267,
      "step": 85
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.916228782093857e-05,
      "loss": 0.4959,
      "step": 86
    },
    {
      "epoch": 0.98,
      "logps_train/chosen": -134.76730346679688,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -116.19819641113281,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": -0.11325030773878098,
      "rewards_train/margins": 0.5541764572262764,
      "rewards_train/rejected": -0.6674267649650574,
      "step": 86
    },
    {
      "epoch": 0.98,
      "logps_train/chosen": -177.84475708007812,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -145.89642333984375,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": 0.3474574685096741,
      "rewards_train/margins": 1.1841707825660706,
      "rewards_train/rejected": -0.8367133140563965,
      "step": 86
    },
    {
      "epoch": 0.98,
      "logps_train/chosen": -162.54910278320312,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -128.0146484375,
      "rewards_train/accuracies": 0.75,
      "rewards_train/chosen": 0.29553040862083435,
      "rewards_train/margins": 1.3141840398311615,
      "rewards_train/rejected": -1.0186536312103271,
      "step": 86
    },
    {
      "epoch": 0.98,
      "logps_train/chosen": -159.21815490722656,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -137.9909210205078,
      "rewards_train/accuracies": 0.6875,
      "rewards_train/chosen": 0.2795528173446655,
      "rewards_train/margins": 0.6489570438861847,
      "rewards_train/rejected": -0.36940422654151917,
      "step": 86
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.890538392136188e-05,
      "loss": 0.5114,
      "step": 87
    },
    {
      "epoch": 0.99,
      "logps_train/chosen": -156.0184326171875,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -113.9042739868164,
      "rewards_train/accuracies": 0.71875,
      "rewards_train/chosen": -0.060827575623989105,
      "rewards_train/margins": 0.9319678917527199,
      "rewards_train/rejected": -0.992795467376709,
      "step": 87
    },
    {
      "epoch": 0.99,
      "logps_train/chosen": -153.46629333496094,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -147.35433959960938,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": 0.40214991569519043,
      "rewards_train/margins": 1.7021600008010864,
      "rewards_train/rejected": -1.300010085105896,
      "step": 87
    },
    {
      "epoch": 0.99,
      "logps_train/chosen": -137.97434997558594,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -142.22134399414062,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": 0.3973763883113861,
      "rewards_train/margins": 1.8008342683315277,
      "rewards_train/rejected": -1.4034578800201416,
      "step": 87
    },
    {
      "epoch": 0.99,
      "logps_train/chosen": -149.7021026611328,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -150.05059814453125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.41885271668434143,
      "rewards_train/margins": 2.0571157038211823,
      "rewards_train/rejected": -1.6382629871368408,
      "step": 87
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.8646335965872414e-05,
      "loss": 0.3438,
      "step": 88
    },
    {
      "epoch": 1.01,
      "logps_train/chosen": -133.6719207763672,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -138.8072509765625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.3787682056427002,
      "rewards_train/margins": 1.741170048713684,
      "rewards_train/rejected": -1.3624018430709839,
      "step": 88
    },
    {
      "epoch": 1.01,
      "logps_train/chosen": -175.55409240722656,
      "logps_train/ref_chosen": -185.0,
      "logps_train/ref_rejected": -160.0,
      "logps_train/rejected": -170.5850067138672,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 0.9660758972167969,
      "rewards_train/margins": 2.0684245824813843,
      "rewards_train/rejected": -1.1023486852645874,
      "step": 88
    },
    {
      "epoch": 1.01,
      "logps_train/chosen": -130.45492553710938,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -122.48587799072266,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.4400542378425598,
      "rewards_train/margins": 2.027607262134552,
      "rewards_train/rejected": -1.5875530242919922,
      "step": 88
    },
    {
      "epoch": 1.01,
      "logps_train/chosen": -139.69839477539062,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -145.28656005859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.48426252603530884,
      "rewards_train/margins": 2.070731222629547,
      "rewards_train/rejected": -1.5864686965942383,
      "step": 88
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.838518389679065e-05,
      "loss": 0.223,
      "step": 89
    },
    {
      "epoch": 1.02,
      "logps_train/chosen": -135.4624786376953,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -144.20643615722656,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.40640440583229065,
      "rewards_train/margins": 1.82263645529747,
      "rewards_train/rejected": -2.2290408611297607,
      "step": 89
    },
    {
      "epoch": 1.02,
      "logps_train/chosen": -151.22457885742188,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -158.74038696289062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 1.1002472639083862,
      "rewards_train/margins": 2.5626660585403442,
      "rewards_train/rejected": -1.462418794631958,
      "step": 89
    },
    {
      "epoch": 1.02,
      "logps_train/chosen": -152.0464324951172,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -123.42019653320312,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 1.0309046506881714,
      "rewards_train/margins": 2.756957173347473,
      "rewards_train/rejected": -1.7260525226593018,
      "step": 89
    },
    {
      "epoch": 1.02,
      "logps_train/chosen": -142.5914764404297,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -130.45376586914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 1.0303058624267578,
      "rewards_train/margins": 2.694628357887268,
      "rewards_train/rejected": -1.6643224954605103,
      "step": 89
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.812196798086799e-05,
      "loss": 0.2086,
      "step": 90
    },
    {
      "epoch": 1.03,
      "logps_train/chosen": -122.85346221923828,
      "logps_train/ref_chosen": -127.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -120.95262908935547,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.44590410590171814,
      "rewards_train/margins": 2.2182177007198334,
      "rewards_train/rejected": -1.7723135948181152,
      "step": 90
    },
    {
      "epoch": 1.03,
      "logps_train/chosen": -135.0833282470703,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -148.67274475097656,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.5127615928649902,
      "rewards_train/margins": 1.8486876487731934,
      "rewards_train/rejected": -1.3359260559082031,
      "step": 90
    },
    {
      "epoch": 1.03,
      "logps_train/chosen": -127.56909942626953,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -137.51315307617188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.8834019899368286,
      "rewards_train/margins": 2.3031264543533325,
      "rewards_train/rejected": -1.419724464416504,
      "step": 90
    },
    {
      "epoch": 1.03,
      "logps_train/chosen": -157.03421020507812,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -134.88729858398438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.7317352294921875,
      "rewards_train/margins": 2.5376535654067993,
      "rewards_train/rejected": -1.8059183359146118,
      "step": 90
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.785672880307817e-05,
      "loss": 0.2009,
      "step": 91
    },
    {
      "epoch": 1.04,
      "logps_train/chosen": -175.03244018554688,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -156.64846801757812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.8252712488174438,
      "rewards_train/margins": 2.3022282123565674,
      "rewards_train/rejected": -1.4769569635391235,
      "step": 91
    },
    {
      "epoch": 1.04,
      "logps_train/chosen": -134.37451171875,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -126.92540740966797,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.2630377411842346,
      "rewards_train/margins": 1.791417419910431,
      "rewards_train/rejected": -1.5283796787261963,
      "step": 91
    },
    {
      "epoch": 1.04,
      "logps_train/chosen": -122.10877227783203,
      "logps_train/ref_chosen": -121.0,
      "logps_train/ref_rejected": -92.5,
      "logps_train/rejected": -109.1966552734375,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.07845546305179596,
      "rewards_train/margins": 1.5968745797872543,
      "rewards_train/rejected": -1.6753300428390503,
      "step": 91
    },
    {
      "epoch": 1.04,
      "logps_train/chosen": -175.49851989746094,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -151.86138916015625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.5926284790039062,
      "rewards_train/margins": 2.2545496225357056,
      "rewards_train/rejected": -1.6619211435317993,
      "step": 91
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.7589507260359404e-05,
      "loss": 0.2275,
      "step": 92
    },
    {
      "epoch": 1.05,
      "logps_train/chosen": -149.17735290527344,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -150.0,
      "logps_train/rejected": -168.39157104492188,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.5334359407424927,
      "rewards_train/margins": 2.4423190355300903,
      "rewards_train/rejected": -1.9088830947875977,
      "step": 92
    },
    {
      "epoch": 1.05,
      "logps_train/chosen": -128.9286346435547,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -155.26699829101562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.3554763197898865,
      "rewards_train/margins": 2.472897946834564,
      "rewards_train/rejected": -2.1174216270446777,
      "step": 92
    },
    {
      "epoch": 1.05,
      "logps_train/chosen": -126.43780517578125,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -122.29857635498047,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.2506530284881592,
      "rewards_train/margins": 2.3148858547210693,
      "rewards_train/rejected": -2.06423282623291,
      "step": 92
    },
    {
      "epoch": 1.05,
      "logps_train/chosen": -155.54661560058594,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -121.3519515991211,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 0.8316656351089478,
      "rewards_train/margins": 2.5136386156082153,
      "rewards_train/rejected": -1.6819729804992676,
      "step": 92
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.732034455530863e-05,
      "loss": 0.1687,
      "step": 93
    },
    {
      "epoch": 1.06,
      "logps_train/chosen": -180.7027587890625,
      "logps_train/ref_chosen": -190.0,
      "logps_train/ref_rejected": -171.0,
      "logps_train/rejected": -185.62432861328125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.9221908450126648,
      "rewards_train/margins": 2.366423189640045,
      "rewards_train/rejected": -1.4442323446273804,
      "step": 93
    },
    {
      "epoch": 1.06,
      "logps_train/chosen": -174.7982940673828,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -162.9814453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.7939980030059814,
      "rewards_train/margins": 2.73032546043396,
      "rewards_train/rejected": -1.9363274574279785,
      "step": 93
    },
    {
      "epoch": 1.06,
      "logps_train/chosen": -133.31781005859375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -116.48727416992188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.2709043323993683,
      "rewards_train/margins": 2.5305700600147247,
      "rewards_train/rejected": -2.2596657276153564,
      "step": 93
    },
    {
      "epoch": 1.06,
      "logps_train/chosen": -156.94393920898438,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -139.75140380859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.225722536444664,
      "rewards_train/margins": 2.1422575563192368,
      "rewards_train/rejected": -1.9165350198745728,
      "step": 93
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.704928218982845e-05,
      "loss": 0.1829,
      "step": 94
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -139.14205932617188,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -131.75674438476562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.3425329923629761,
      "rewards_train/margins": 2.1845154762268066,
      "rewards_train/rejected": -1.8419824838638306,
      "step": 94
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -145.99407958984375,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -156.46421813964844,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.3601621985435486,
      "rewards_train/margins": 2.0587562918663025,
      "rewards_train/rejected": -1.698594093322754,
      "step": 94
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -125.1644287109375,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -98.5,
      "logps_train/rejected": -120.51481628417969,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.10804449021816254,
      "rewards_train/margins": 2.082010880112648,
      "rewards_train/rejected": -2.1900553703308105,
      "step": 94
    },
    {
      "epoch": 1.07,
      "logps_train/chosen": -156.272216796875,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -144.20132446289062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.561450183391571,
      "rewards_train/margins": 2.499405324459076,
      "rewards_train/rejected": -1.9379551410675049,
      "step": 94
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.677636195872802e-05,
      "loss": 0.2071,
      "step": 95
    },
    {
      "epoch": 1.09,
      "logps_train/chosen": -165.90792846679688,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -166.0856170654297,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.7180935144424438,
      "rewards_train/margins": 2.9360300302505493,
      "rewards_train/rejected": -2.2179365158081055,
      "step": 95
    },
    {
      "epoch": 1.09,
      "logps_train/chosen": -143.52719116210938,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -130.509521484375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.600570797920227,
      "rewards_train/margins": 2.178702235221863,
      "rewards_train/rejected": -2.77927303314209,
      "step": 95
    },
    {
      "epoch": 1.09,
      "logps_train/chosen": -168.9429931640625,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -131.9114532470703,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 0.6327519416809082,
      "rewards_train/margins": 3.1334948539733887,
      "rewards_train/rejected": -2.5007429122924805,
      "step": 95
    },
    {
      "epoch": 1.09,
      "logps_train/chosen": -161.4586944580078,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -135.01678466796875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.04224124550819397,
      "rewards_train/margins": 2.7115966379642487,
      "rewards_train/rejected": -2.6693553924560547,
      "step": 95
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.6501625943278805e-05,
      "loss": 0.1683,
      "step": 96
    },
    {
      "epoch": 1.1,
      "logps_train/chosen": -166.95816040039062,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -168.36917114257812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.10113706439733505,
      "rewards_train/margins": 2.489418126642704,
      "rewards_train/rejected": -2.590555191040039,
      "step": 96
    },
    {
      "epoch": 1.1,
      "logps_train/chosen": -159.73220825195312,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -146.8890380859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": 0.12797528505325317,
      "rewards_train/margins": 2.4644392132759094,
      "rewards_train/rejected": -2.3364639282226562,
      "step": 96
    },
    {
      "epoch": 1.1,
      "logps_train/chosen": -141.54803466796875,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -137.75836181640625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.1825377345085144,
      "rewards_train/margins": 2.6527701020240784,
      "rewards_train/rejected": -2.8353078365325928,
      "step": 96
    },
    {
      "epoch": 1.1,
      "logps_train/chosen": -167.2300262451172,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -154.9856414794922,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.023921027779579163,
      "rewards_train/margins": 2.4235100895166397,
      "rewards_train/rejected": -2.3995890617370605,
      "step": 96
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.622511650472601e-05,
      "loss": 0.1993,
      "step": 97
    },
    {
      "epoch": 1.11,
      "logps_train/chosen": -145.75787353515625,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -143.81246948242188,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": 0.23278886079788208,
      "rewards_train/margins": 2.254021942615509,
      "rewards_train/rejected": -2.021233081817627,
      "step": 97
    },
    {
      "epoch": 1.11,
      "logps_train/chosen": -134.80609130859375,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -98.0,
      "logps_train/rejected": -121.690185546875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.004826478660106659,
      "rewards_train/margins": 2.3503986075520515,
      "rewards_train/rejected": -2.355225086212158,
      "step": 97
    },
    {
      "epoch": 1.11,
      "logps_train/chosen": -122.42782592773438,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -128.15924072265625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.4224705398082733,
      "rewards_train/margins": 2.235249310731888,
      "rewards_train/rejected": -2.657719850540161,
      "step": 97
    },
    {
      "epoch": 1.11,
      "logps_train/chosen": -164.96200561523438,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -157.0484619140625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": 0.057143136858940125,
      "rewards_train/margins": 2.3765157908201218,
      "rewards_train/rejected": -2.3193726539611816,
      "step": 97
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.5946876277757066e-05,
      "loss": 0.2219,
      "step": 98
    },
    {
      "epoch": 1.12,
      "logps_train/chosen": -157.20069885253906,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -158.66604614257812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.008192479610443115,
      "rewards_train/margins": 2.5320449471473694,
      "rewards_train/rejected": -2.5402374267578125,
      "step": 98
    },
    {
      "epoch": 1.12,
      "logps_train/chosen": -168.69036865234375,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -165.33319091796875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.05536573380231857,
      "rewards_train/margins": 2.6265848353505135,
      "rewards_train/rejected": -2.681950569152832,
      "step": 98
    },
    {
      "epoch": 1.12,
      "logps_train/chosen": -150.9794464111328,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -150.70811462402344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.5467734336853027,
      "rewards_train/margins": 2.44610857963562,
      "rewards_train/rejected": -2.992882013320923,
      "step": 98
    },
    {
      "epoch": 1.12,
      "logps_train/chosen": -137.99195861816406,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -140.39208984375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.13567104935646057,
      "rewards_train/margins": 2.4768291413784027,
      "rewards_train/rejected": -2.6125001907348633,
      "step": 98
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.5666948163927716e-05,
      "loss": 0.1803,
      "step": 99
    },
    {
      "epoch": 1.13,
      "logps_train/chosen": -146.2452392578125,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -150.16941833496094,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.3603140115737915,
      "rewards_train/margins": 2.619420647621155,
      "rewards_train/rejected": -2.9797346591949463,
      "step": 99
    },
    {
      "epoch": 1.13,
      "logps_train/chosen": -126.06419372558594,
      "logps_train/ref_chosen": -121.0,
      "logps_train/ref_rejected": -86.0,
      "logps_train/rejected": -114.22348022460938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.49977877736091614,
      "rewards_train/margins": 2.319640129804611,
      "rewards_train/rejected": -2.8194189071655273,
      "step": 99
    },
    {
      "epoch": 1.13,
      "logps_train/chosen": -132.54383850097656,
      "logps_train/ref_chosen": -125.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -168.75698852539062,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.7930556535720825,
      "rewards_train/margins": 2.4375261068344116,
      "rewards_train/rejected": -3.230581760406494,
      "step": 99
    },
    {
      "epoch": 1.13,
      "logps_train/chosen": -130.33595275878906,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -93.5,
      "logps_train/rejected": -121.05641174316406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6355488896369934,
      "rewards_train/margins": 2.140502631664276,
      "rewards_train/rejected": -2.7760515213012695,
      "step": 99
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.5385375325047166e-05,
      "loss": 0.2181,
      "step": 100
    },
    {
      "epoch": 1.14,
      "logps_train/chosen": -173.3467254638672,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -161.84597778320312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.3217822313308716,
      "rewards_train/margins": 2.759886145591736,
      "rewards_train/rejected": -3.0816683769226074,
      "step": 100
    },
    {
      "epoch": 1.14,
      "logps_train/chosen": -180.46633911132812,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -174.64852905273438,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.31440621614456177,
      "rewards_train/margins": 2.7816969752311707,
      "rewards_train/rejected": -3.0961031913757324,
      "step": 100
    },
    {
      "epoch": 1.14,
      "logps_train/chosen": -149.21096801757812,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -135.708251953125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.562260091304779,
      "rewards_train/margins": 2.4379597306251526,
      "rewards_train/rejected": -3.0002198219299316,
      "step": 100
    },
    {
      "epoch": 1.14,
      "logps_train/chosen": -165.3883819580078,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -147.68844604492188,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.2641317844390869,
      "rewards_train/margins": 2.305885076522827,
      "rewards_train/rejected": -2.570016860961914,
      "step": 100
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.510220117652297e-05,
      "loss": 0.1737,
      "step": 101
    },
    {
      "epoch": 1.15,
      "logps_train/chosen": -153.82940673828125,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -131.58132934570312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.41795065999031067,
      "rewards_train/margins": 3.12226340174675,
      "rewards_train/rejected": -3.5402140617370605,
      "step": 101
    },
    {
      "epoch": 1.15,
      "logps_train/chosen": -149.13304138183594,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -142.9083251953125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.5003165006637573,
      "rewards_train/margins": 2.4736016988754272,
      "rewards_train/rejected": -2.9739181995391846,
      "step": 101
    },
    {
      "epoch": 1.15,
      "logps_train/chosen": -185.08547973632812,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -164.0,
      "logps_train/rejected": -201.46383666992188,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.43471941351890564,
      "rewards_train/margins": 3.3151794970035553,
      "rewards_train/rejected": -3.749898910522461,
      "step": 101
    },
    {
      "epoch": 1.15,
      "logps_train/chosen": -143.16180419921875,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -152.1904754638672,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8441107273101807,
      "rewards_train/margins": 2.4427101612091064,
      "rewards_train/rejected": -3.286820888519287,
      "step": 101
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.481746938066684e-05,
      "loss": 0.1572,
      "step": 102
    },
    {
      "epoch": 1.17,
      "logps_train/chosen": -169.8740234375,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -151.0,
      "logps_train/rejected": -182.55886840820312,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -0.9510747790336609,
      "rewards_train/margins": 2.172243893146515,
      "rewards_train/rejected": -3.123318672180176,
      "step": 102
    },
    {
      "epoch": 1.17,
      "logps_train/chosen": -140.3857421875,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -151.04632568359375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9004732966423035,
      "rewards_train/margins": 2.5984941124916077,
      "rewards_train/rejected": -3.498967409133911,
      "step": 102
    },
    {
      "epoch": 1.17,
      "logps_train/chosen": -180.0685577392578,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -165.80479431152344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9504103660583496,
      "rewards_train/margins": 3.016885280609131,
      "rewards_train/rejected": -3.9672956466674805,
      "step": 102
    },
    {
      "epoch": 1.17,
      "logps_train/chosen": -159.4775390625,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -159.30438232421875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7955068349838257,
      "rewards_train/margins": 2.654363751411438,
      "rewards_train/rejected": -3.4498705863952637,
      "step": 102
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.4531223839962453e-05,
      "loss": 0.1811,
      "step": 103
    },
    {
      "epoch": 1.18,
      "logps_train/chosen": -126.63336181640625,
      "logps_train/ref_chosen": -117.5,
      "logps_train/ref_rejected": -92.0,
      "logps_train/rejected": -131.54600524902344,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.9199284911155701,
      "rewards_train/margins": 3.021292269229889,
      "rewards_train/rejected": -3.941220760345459,
      "step": 103
    },
    {
      "epoch": 1.18,
      "logps_train/chosen": -138.7056121826172,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -143.95974731445312,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.7718613147735596,
      "rewards_train/margins": 3.3114187717437744,
      "rewards_train/rejected": -4.083280086517334,
      "step": 103
    },
    {
      "epoch": 1.18,
      "logps_train/chosen": -170.283447265625,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -173.4869384765625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.28478938341140747,
      "rewards_train/margins": 3.222693145275116,
      "rewards_train/rejected": -3.5074825286865234,
      "step": 103
    },
    {
      "epoch": 1.18,
      "logps_train/chosen": -186.765380859375,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -184.8590850830078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7643318176269531,
      "rewards_train/margins": 2.9727487564086914,
      "rewards_train/rejected": -3.7370805740356445,
      "step": 103
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.4243508690296135e-05,
      "loss": 0.1504,
      "step": 104
    },
    {
      "epoch": 1.19,
      "logps_train/chosen": -132.2598876953125,
      "logps_train/ref_chosen": -122.5,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -142.23316955566406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9834104180335999,
      "rewards_train/margins": 2.9381489157676697,
      "rewards_train/rejected": -3.9215593338012695,
      "step": 104
    },
    {
      "epoch": 1.19,
      "logps_train/chosen": -133.6953887939453,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -161.70277404785156,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.7267659902572632,
      "rewards_train/margins": 2.867143750190735,
      "rewards_train/rejected": -3.593909740447998,
      "step": 104
    },
    {
      "epoch": 1.19,
      "logps_train/chosen": -148.66311645507812,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -100.5,
      "logps_train/rejected": -142.30751037597656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7122358083724976,
      "rewards_train/margins": 3.461971879005432,
      "rewards_train/rejected": -4.17420768737793,
      "step": 104
    },
    {
      "epoch": 1.19,
      "logps_train/chosen": -169.8518829345703,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -159.99441528320312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9660478830337524,
      "rewards_train/margins": 2.7869099378585815,
      "rewards_train/rejected": -3.752957820892334,
      "step": 104
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.39543682941516e-05,
      "loss": 0.1721,
      "step": 105
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -163.95489501953125,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -164.74134826660156,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6238088607788086,
      "rewards_train/margins": 2.583432197570801,
      "rewards_train/rejected": -4.207241058349609,
      "step": 105
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -170.38116455078125,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -171.26705932617188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8710801005363464,
      "rewards_train/margins": 3.070517122745514,
      "rewards_train/rejected": -3.9415972232818604,
      "step": 105
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -171.02854919433594,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -151.22903442382812,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.7450417280197144,
      "rewards_train/margins": 2.1690722703933716,
      "rewards_train/rejected": -3.914113998413086,
      "step": 105
    },
    {
      "epoch": 1.2,
      "logps_train/chosen": -162.40505981445312,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -158.7891845703125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.560684084892273,
      "rewards_train/margins": 2.450362801551819,
      "rewards_train/rejected": -4.011046886444092,
      "step": 105
    },
    {
      "epoch": 1.21,
      "learning_rate": 3.366384723376977e-05,
      "loss": 0.2371,
      "step": 106
    },
    {
      "epoch": 1.21,
      "logps_train/chosen": -119.2367935180664,
      "logps_train/ref_chosen": -99.5,
      "logps_train/ref_rejected": -99.5,
      "logps_train/rejected": -139.24002075195312,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.9458470344543457,
      "rewards_train/margins": 2.0191712379455566,
      "rewards_train/rejected": -3.9650182723999023,
      "step": 106
    },
    {
      "epoch": 1.21,
      "logps_train/chosen": -144.74407958984375,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -153.19857788085938,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4341243505477905,
      "rewards_train/margins": 2.543986439704895,
      "rewards_train/rejected": -3.9781107902526855,
      "step": 106
    },
    {
      "epoch": 1.21,
      "logps_train/chosen": -190.86489868164062,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -181.99560546875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8407862186431885,
      "rewards_train/margins": 3.3185408115386963,
      "rewards_train/rejected": -4.159327030181885,
      "step": 106
    },
    {
      "epoch": 1.21,
      "logps_train/chosen": -173.3730926513672,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -179.78939819335938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8697314262390137,
      "rewards_train/margins": 3.73850679397583,
      "rewards_train/rejected": -4.608238220214844,
      "step": 106
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.3371990304274656e-05,
      "loss": 0.1757,
      "step": 107
    },
    {
      "epoch": 1.22,
      "logps_train/chosen": -186.10415649414062,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -159.8262939453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9052414894104004,
      "rewards_train/margins": 3.8516058921813965,
      "rewards_train/rejected": -4.756847381591797,
      "step": 107
    },
    {
      "epoch": 1.22,
      "logps_train/chosen": -146.3466339111328,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -157.24111938476562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.206477165222168,
      "rewards_train/margins": 2.8152427673339844,
      "rewards_train/rejected": -4.021719932556152,
      "step": 107
    },
    {
      "epoch": 1.22,
      "logps_train/chosen": -172.4299774169922,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -178.18833923339844,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.0049124956130981,
      "rewards_train/margins": 2.5578664541244507,
      "rewards_train/rejected": -3.562778949737549,
      "step": 107
    },
    {
      "epoch": 1.22,
      "logps_train/chosen": -148.16546630859375,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -114.5,
      "logps_train/rejected": -157.4647216796875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6409605741500854,
      "rewards_train/margins": 2.648480534553528,
      "rewards_train/rejected": -4.289441108703613,
      "step": 107
    },
    {
      "epoch": 1.23,
      "learning_rate": 3.3078842506766484e-05,
      "loss": 0.1539,
      "step": 108
    },
    {
      "epoch": 1.23,
      "logps_train/chosen": -138.66619873046875,
      "logps_train/ref_chosen": -120.0,
      "logps_train/ref_rejected": -126.5,
      "logps_train/rejected": -176.2493896484375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8791207075119019,
      "rewards_train/margins": 3.097185254096985,
      "rewards_train/rejected": -4.976305961608887,
      "step": 108
    },
    {
      "epoch": 1.23,
      "logps_train/chosen": -143.60552978515625,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -149.5123291015625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4019588232040405,
      "rewards_train/margins": 2.506013035774231,
      "rewards_train/rejected": -3.9079718589782715,
      "step": 108
    },
    {
      "epoch": 1.23,
      "logps_train/chosen": -157.93670654296875,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -159.3701171875,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.7084167003631592,
      "rewards_train/margins": 3.0381643772125244,
      "rewards_train/rejected": -4.746581077575684,
      "step": 108
    },
    {
      "epoch": 1.23,
      "logps_train/chosen": -199.24874877929688,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -186.7269287109375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0649135112762451,
      "rewards_train/margins": 3.8796541690826416,
      "rewards_train/rejected": -4.944567680358887,
      "step": 108
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.278444904138297e-05,
      "loss": 0.1546,
      "step": 109
    },
    {
      "epoch": 1.25,
      "logps_train/chosen": -143.3296661376953,
      "logps_train/ref_chosen": -119.5,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -155.20834350585938,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -2.4022021293640137,
      "rewards_train/margins": 2.016094207763672,
      "rewards_train/rejected": -4.4182963371276855,
      "step": 109
    },
    {
      "epoch": 1.25,
      "logps_train/chosen": -152.14297485351562,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -163.31515502929688,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.001040458679199,
      "rewards_train/margins": 2.6719303131103516,
      "rewards_train/rejected": -4.672970771789551,
      "step": 109
    },
    {
      "epoch": 1.25,
      "logps_train/chosen": -147.17257690429688,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -85.5,
      "logps_train/rejected": -133.1774444580078,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.0941858291625977,
      "rewards_train/margins": 2.6622066497802734,
      "rewards_train/rejected": -4.756392478942871,
      "step": 109
    },
    {
      "epoch": 1.25,
      "logps_train/chosen": -170.54586791992188,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -167.38327026367188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1571266651153564,
      "rewards_train/margins": 3.347363233566284,
      "rewards_train/rejected": -4.504489898681641,
      "step": 109
    },
    {
      "epoch": 1.26,
      "learning_rate": 3.248885530033004e-05,
      "loss": 0.2248,
      "step": 110
    },
    {
      "epoch": 1.26,
      "logps_train/chosen": -170.85304260253906,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -174.83726501464844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8820815086364746,
      "rewards_train/margins": 3.082406520843506,
      "rewards_train/rejected": -4.9644880294799805,
      "step": 110
    },
    {
      "epoch": 1.26,
      "logps_train/chosen": -141.23141479492188,
      "logps_train/ref_chosen": -121.0,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -149.21527099609375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.9987776279449463,
      "rewards_train/margins": 2.5813276767730713,
      "rewards_train/rejected": -4.580105304718018,
      "step": 110
    },
    {
      "epoch": 1.26,
      "logps_train/chosen": -150.4207000732422,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -141.09347534179688,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7420827150344849,
      "rewards_train/margins": 2.7829874753952026,
      "rewards_train/rejected": -4.5250701904296875,
      "step": 110
    },
    {
      "epoch": 1.26,
      "logps_train/chosen": -154.02862548828125,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -98.0,
      "logps_train/rejected": -146.64706420898438,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.1366517543792725,
      "rewards_train/margins": 2.7122819423675537,
      "rewards_train/rejected": -4.848933696746826,
      "step": 110
    },
    {
      "epoch": 1.27,
      "learning_rate": 3.219210686088278e-05,
      "loss": 0.1783,
      "step": 111
    },
    {
      "epoch": 1.27,
      "logps_train/chosen": -196.1944580078125,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -177.0728759765625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4467411041259766,
      "rewards_train/margins": 3.594043731689453,
      "rewards_train/rejected": -5.04078483581543,
      "step": 111
    },
    {
      "epoch": 1.27,
      "logps_train/chosen": -154.63021850585938,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -155.19606018066406,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4006195068359375,
      "rewards_train/margins": 2.3893957138061523,
      "rewards_train/rejected": -4.79001522064209,
      "step": 111
    },
    {
      "epoch": 1.27,
      "logps_train/chosen": -206.43435668945312,
      "logps_train/ref_chosen": -196.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -181.19366455078125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0297620296478271,
      "rewards_train/margins": 3.8278849124908447,
      "rewards_train/rejected": -4.857646942138672,
      "step": 111
    },
    {
      "epoch": 1.27,
      "logps_train/chosen": -200.61825561523438,
      "logps_train/ref_chosen": -185.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -197.92047119140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5702245235443115,
      "rewards_train/margins": 3.5200655460357666,
      "rewards_train/rejected": -5.090290069580078,
      "step": 111
    },
    {
      "epoch": 1.28,
      "learning_rate": 3.1894249478357965e-05,
      "loss": 0.1124,
      "step": 112
    },
    {
      "epoch": 1.28,
      "logps_train/chosen": -184.96241760253906,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -177.40328979492188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3922240734100342,
      "rewards_train/margins": 3.4400475025177,
      "rewards_train/rejected": -4.832271575927734,
      "step": 112
    },
    {
      "epoch": 1.28,
      "logps_train/chosen": -190.53353881835938,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -161.22642517089844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5588231086730957,
      "rewards_train/margins": 3.2116708755493164,
      "rewards_train/rejected": -4.770493984222412,
      "step": 112
    },
    {
      "epoch": 1.28,
      "logps_train/chosen": -162.96185302734375,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -184.81460571289062,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.5726509094238281,
      "rewards_train/margins": 3.1111040115356445,
      "rewards_train/rejected": -4.683754920959473,
      "step": 112
    },
    {
      "epoch": 1.28,
      "logps_train/chosen": -188.84292602539062,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -189.7190399169922,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4589011669158936,
      "rewards_train/margins": 3.8735506534576416,
      "rewards_train/rejected": -5.332451820373535,
      "step": 112
    },
    {
      "epoch": 1.29,
      "learning_rate": 3.15953290790591e-05,
      "loss": 0.1447,
      "step": 113
    },
    {
      "epoch": 1.29,
      "logps_train/chosen": -155.69085693359375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -144.44094848632812,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -2.01090669631958,
      "rewards_train/margins": 3.0220065116882324,
      "rewards_train/rejected": -5.0329132080078125,
      "step": 113
    },
    {
      "epoch": 1.29,
      "logps_train/chosen": -176.43182373046875,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -196.0265655517578,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.9890806674957275,
      "rewards_train/margins": 3.118849515914917,
      "rewards_train/rejected": -5.1079301834106445,
      "step": 113
    },
    {
      "epoch": 1.29,
      "logps_train/chosen": -147.64617919921875,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -158.9957733154297,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8295596837997437,
      "rewards_train/margins": 2.9519516229629517,
      "rewards_train/rejected": -4.781511306762695,
      "step": 113
    },
    {
      "epoch": 1.29,
      "logps_train/chosen": -201.11227416992188,
      "logps_train/ref_chosen": -191.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -172.78042602539062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9713826775550842,
      "rewards_train/margins": 3.347773015499115,
      "rewards_train/rejected": -4.319155693054199,
      "step": 113
    },
    {
      "epoch": 1.3,
      "learning_rate": 3.1295391753195047e-05,
      "loss": 0.1702,
      "step": 114
    },
    {
      "epoch": 1.3,
      "logps_train/chosen": -183.66317749023438,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -176.61851501464844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3601659536361694,
      "rewards_train/margins": 3.3909436464309692,
      "rewards_train/rejected": -4.751109600067139,
      "step": 114
    },
    {
      "epoch": 1.3,
      "logps_train/chosen": -190.912353515625,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -192.1277618408203,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.485377311706543,
      "rewards_train/margins": 2.946054458618164,
      "rewards_train/rejected": -5.431431770324707,
      "step": 114
    },
    {
      "epoch": 1.3,
      "logps_train/chosen": -142.88148498535156,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -151.7729949951172,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.4135019779205322,
      "rewards_train/margins": 3.1451447010040283,
      "rewards_train/rejected": -4.5586466789245605,
      "step": 114
    },
    {
      "epoch": 1.3,
      "logps_train/chosen": -195.8353271484375,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -157.0,
      "logps_train/rejected": -207.72579956054688,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4638062715530396,
      "rewards_train/margins": 3.5900248289108276,
      "rewards_train/rejected": -5.053831100463867,
      "step": 114
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.099448374777351e-05,
      "loss": 0.1436,
      "step": 115
    },
    {
      "epoch": 1.31,
      "logps_train/chosen": -202.21957397460938,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -158.0,
      "logps_train/rejected": -206.32733154296875,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -1.5319174528121948,
      "rewards_train/margins": 3.348861813545227,
      "rewards_train/rejected": -4.880779266357422,
      "step": 115
    },
    {
      "epoch": 1.31,
      "logps_train/chosen": -164.8944091796875,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -96.5,
      "logps_train/rejected": -145.25302124023438,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.778893232345581,
      "rewards_train/margins": 3.073557138442993,
      "rewards_train/rejected": -4.852450370788574,
      "step": 115
    },
    {
      "epoch": 1.31,
      "logps_train/chosen": -160.88668823242188,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -163.2848358154297,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -2.420285224914551,
      "rewards_train/margins": 2.5804386138916016,
      "rewards_train/rejected": -5.000723838806152,
      "step": 115
    },
    {
      "epoch": 1.31,
      "logps_train/chosen": -188.2561492919922,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -169.0,
      "logps_train/rejected": -224.9777069091797,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.849443793296814,
      "rewards_train/margins": 3.7523800134658813,
      "rewards_train/rejected": -5.601823806762695,
      "step": 115
    },
    {
      "epoch": 1.33,
      "learning_rate": 3.069265145947016e-05,
      "loss": 0.2013,
      "step": 116
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -159.26284790039062,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -178.18568420410156,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1178855895996094,
      "rewards_train/margins": 3.3582873344421387,
      "rewards_train/rejected": -5.476172924041748,
      "step": 116
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -158.87692260742188,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -163.0965118408203,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6447348594665527,
      "rewards_train/margins": 3.0411858558654785,
      "rewards_train/rejected": -4.685920715332031,
      "step": 116
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -182.80258178710938,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -177.33001708984375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4982264041900635,
      "rewards_train/margins": 3.7515709400177,
      "rewards_train/rejected": -5.249797344207764,
      "step": 116
    },
    {
      "epoch": 1.33,
      "logps_train/chosen": -167.61244201660156,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -161.88333129882812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.1869282722473145,
      "rewards_train/margins": 3.1240487098693848,
      "rewards_train/rejected": -5.310976982116699,
      "step": 116
    },
    {
      "epoch": 1.34,
      "learning_rate": 3.0389941427474873e-05,
      "loss": 0.1413,
      "step": 117
    },
    {
      "epoch": 1.34,
      "logps_train/chosen": -187.91346740722656,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -195.27867126464844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.616736650466919,
      "rewards_train/margins": 3.5172088146209717,
      "rewards_train/rejected": -5.133945465087891,
      "step": 117
    },
    {
      "epoch": 1.34,
      "logps_train/chosen": -151.50758361816406,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -100.5,
      "logps_train/rejected": -147.73008728027344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9189229011535645,
      "rewards_train/margins": 2.8163905143737793,
      "rewards_train/rejected": -4.735313415527344,
      "step": 117
    },
    {
      "epoch": 1.34,
      "logps_train/chosen": -169.54360961914062,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -167.17601013183594,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.2594144344329834,
      "rewards_train/margins": 3.2765467166900635,
      "rewards_train/rejected": -5.535961151123047,
      "step": 117
    },
    {
      "epoch": 1.34,
      "logps_train/chosen": -198.6015625,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -195.97879028320312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7592782974243164,
      "rewards_train/margins": 3.4153575897216797,
      "rewards_train/rejected": -5.174635887145996,
      "step": 117
    },
    {
      "epoch": 1.35,
      "learning_rate": 3.008640032631585e-05,
      "loss": 0.1433,
      "step": 118
    },
    {
      "epoch": 1.35,
      "logps_train/chosen": -169.9024200439453,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -191.21194458007812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6621170043945312,
      "rewards_train/margins": 3.6406192779541016,
      "rewards_train/rejected": -5.302736282348633,
      "step": 118
    },
    {
      "epoch": 1.35,
      "logps_train/chosen": -144.08741760253906,
      "logps_train/ref_chosen": -127.0,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -157.87893676757812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7102066278457642,
      "rewards_train/margins": 3.2556179761886597,
      "rewards_train/rejected": -4.965824604034424,
      "step": 118
    },
    {
      "epoch": 1.35,
      "logps_train/chosen": -137.56689453125,
      "logps_train/ref_chosen": -117.5,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -156.909423828125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0094242095947266,
      "rewards_train/margins": 3.6044678688049316,
      "rewards_train/rejected": -5.613892078399658,
      "step": 118
    },
    {
      "epoch": 1.35,
      "logps_train/chosen": -177.030029296875,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -172.6234130859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.687475323677063,
      "rewards_train/margins": 3.1036747694015503,
      "rewards_train/rejected": -4.791150093078613,
      "step": 118
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.978207495866292e-05,
      "loss": 0.146,
      "step": 119
    },
    {
      "epoch": 1.36,
      "logps_train/chosen": -167.2484130859375,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -151.47787475585938,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6930049657821655,
      "rewards_train/margins": 3.221190094947815,
      "rewards_train/rejected": -4.9141950607299805,
      "step": 119
    },
    {
      "epoch": 1.36,
      "logps_train/chosen": -178.5880584716797,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -176.48397827148438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.979020595550537,
      "rewards_train/margins": 3.617034435272217,
      "rewards_train/rejected": -5.596055030822754,
      "step": 119
    },
    {
      "epoch": 1.36,
      "logps_train/chosen": -158.26107788085938,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -180.61029052734375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.294564127922058,
      "rewards_train/margins": 3.3548184633255005,
      "rewards_train/rejected": -4.649382591247559,
      "step": 119
    },
    {
      "epoch": 1.36,
      "logps_train/chosen": -164.2959442138672,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -165.3199005126953,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3953654766082764,
      "rewards_train/margins": 2.9375030994415283,
      "rewards_train/rejected": -4.332868576049805,
      "step": 119
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.947701224811113e-05,
      "loss": 0.1248,
      "step": 120
    },
    {
      "epoch": 1.37,
      "logps_train/chosen": -174.03245544433594,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -171.17623901367188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3544423580169678,
      "rewards_train/margins": 3.606821298599243,
      "rewards_train/rejected": -4.961263656616211,
      "step": 120
    },
    {
      "epoch": 1.37,
      "logps_train/chosen": -147.8043670654297,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -166.4681396484375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.2177417278289795,
      "rewards_train/margins": 2.9022276401519775,
      "rewards_train/rejected": -4.119969367980957,
      "step": 120
    },
    {
      "epoch": 1.37,
      "logps_train/chosen": -178.89630126953125,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -170.27052307128906,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1310200691223145,
      "rewards_train/margins": 3.1832242012023926,
      "rewards_train/rejected": -4.314244270324707,
      "step": 120
    },
    {
      "epoch": 1.37,
      "logps_train/chosen": -165.63587951660156,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -154.56637573242188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.1438608169555664,
      "rewards_train/margins": 2.665705680847168,
      "rewards_train/rejected": -4.809566497802734,
      "step": 120
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.9171259231945598e-05,
      "loss": 0.1559,
      "step": 121
    },
    {
      "epoch": 1.38,
      "logps_train/chosen": -151.94607543945312,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -88.0,
      "logps_train/rejected": -136.6217041015625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5857216119766235,
      "rewards_train/margins": 3.2760576009750366,
      "rewards_train/rejected": -4.86177921295166,
      "step": 121
    },
    {
      "epoch": 1.38,
      "logps_train/chosen": -155.14697265625,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -162.53421020507812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3468749523162842,
      "rewards_train/margins": 3.6037633419036865,
      "rewards_train/rejected": -4.950638294219971,
      "step": 121
    },
    {
      "epoch": 1.38,
      "logps_train/chosen": -161.08724975585938,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -160.4564208984375,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.577230453491211,
      "rewards_train/margins": 2.754911422729492,
      "rewards_train/rejected": -4.332141876220703,
      "step": 121
    },
    {
      "epoch": 1.38,
      "logps_train/chosen": -160.4774932861328,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -114.5,
      "logps_train/rejected": -158.9612274169922,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.4526317119598389,
      "rewards_train/margins": 2.9784510135650635,
      "rewards_train/rejected": -4.431082725524902,
      "step": 121
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.8864863053888925e-05,
      "loss": 0.1559,
      "step": 122
    },
    {
      "epoch": 1.39,
      "logps_train/chosen": -166.92727661132812,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -182.114990234375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6018102169036865,
      "rewards_train/margins": 3.3425023555755615,
      "rewards_train/rejected": -4.944312572479248,
      "step": 122
    },
    {
      "epoch": 1.39,
      "logps_train/chosen": -161.64584350585938,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -146.01136779785156,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.470199465751648,
      "rewards_train/margins": 2.7532025575637817,
      "rewards_train/rejected": -4.22340202331543,
      "step": 122
    },
    {
      "epoch": 1.39,
      "logps_train/chosen": -179.35714721679688,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -87.0,
      "logps_train/rejected": -127.37224578857422,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3069069385528564,
      "rewards_train/margins": 2.7117631435394287,
      "rewards_train/rejected": -4.018670082092285,
      "step": 122
    },
    {
      "epoch": 1.39,
      "logps_train/chosen": -142.1234588623047,
      "logps_train/ref_chosen": -123.5,
      "logps_train/ref_rejected": -89.5,
      "logps_train/rejected": -134.3780059814453,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.8813402652740479,
      "rewards_train/margins": 2.604262590408325,
      "rewards_train/rejected": -4.485602855682373,
      "step": 122
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.8557870956832132e-05,
      "loss": 0.1891,
      "step": 123
    },
    {
      "epoch": 1.41,
      "logps_train/chosen": -215.79141235351562,
      "logps_train/ref_chosen": -207.0,
      "logps_train/ref_rejected": -158.0,
      "logps_train/rejected": -206.87472534179688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9100003242492676,
      "rewards_train/margins": 4.041144847869873,
      "rewards_train/rejected": -4.951145172119141,
      "step": 123
    },
    {
      "epoch": 1.41,
      "logps_train/chosen": -171.5395965576172,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -178.28213500976562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3091840744018555,
      "rewards_train/margins": 3.5175161361694336,
      "rewards_train/rejected": -4.826700210571289,
      "step": 123
    },
    {
      "epoch": 1.41,
      "logps_train/chosen": -197.67599487304688,
      "logps_train/ref_chosen": -191.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -182.3161163330078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6696504354476929,
      "rewards_train/margins": 3.831687092781067,
      "rewards_train/rejected": -4.50133752822876,
      "step": 123
    },
    {
      "epoch": 1.41,
      "logps_train/chosen": -173.24679565429688,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -164.2289276123047,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.745944619178772,
      "rewards_train/margins": 3.166388154029846,
      "rewards_train/rejected": -3.912332773208618,
      "step": 123
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.8250330275550336e-05,
      "loss": 0.0987,
      "step": 124
    },
    {
      "epoch": 1.42,
      "logps_train/chosen": -172.51718139648438,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -177.36227416992188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7524986267089844,
      "rewards_train/margins": 4.2184929847717285,
      "rewards_train/rejected": -4.970991611480713,
      "step": 124
    },
    {
      "epoch": 1.42,
      "logps_train/chosen": -141.9478302001953,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -145.900146484375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2405827045440674,
      "rewards_train/margins": 3.0773122310638428,
      "rewards_train/rejected": -4.31789493560791,
      "step": 124
    },
    {
      "epoch": 1.42,
      "logps_train/chosen": -180.90890502929688,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -169.93887329101562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6273165941238403,
      "rewards_train/margins": 3.1660832166671753,
      "rewards_train/rejected": -4.793399810791016,
      "step": 124
    },
    {
      "epoch": 1.42,
      "logps_train/chosen": -171.01747131347656,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -157.03610229492188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1220600605010986,
      "rewards_train/margins": 3.2815511226654053,
      "rewards_train/rejected": -4.403611183166504,
      "step": 124
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.7942288429404256e-05,
      "loss": 0.1097,
      "step": 125
    },
    {
      "epoch": 1.43,
      "logps_train/chosen": -175.7913818359375,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -169.33328247070312,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.0130254030227661,
      "rewards_train/margins": 2.717079997062683,
      "rewards_train/rejected": -3.730105400085449,
      "step": 125
    },
    {
      "epoch": 1.43,
      "logps_train/chosen": -193.09005737304688,
      "logps_train/ref_chosen": -186.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -190.29751586914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7705289125442505,
      "rewards_train/margins": 3.368988871574402,
      "rewards_train/rejected": -4.139517784118652,
      "step": 125
    },
    {
      "epoch": 1.43,
      "logps_train/chosen": -175.15960693359375,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -162.30618286132812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0979918241500854,
      "rewards_train/margins": 3.0310651063919067,
      "rewards_train/rejected": -4.129056930541992,
      "step": 125
    },
    {
      "epoch": 1.43,
      "logps_train/chosen": -192.91085815429688,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -156.0,
      "logps_train/rejected": -200.0994415283203,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.1479214429855347,
      "rewards_train/margins": 3.2763782739639282,
      "rewards_train/rejected": -4.424299716949463,
      "step": 125
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.7633792915028677e-05,
      "loss": 0.1349,
      "step": 126
    },
    {
      "epoch": 1.44,
      "logps_train/chosen": -185.21578979492188,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -176.3563232421875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.304196298122406,
      "rewards_train/margins": 3.450674593448639,
      "rewards_train/rejected": -3.754870891571045,
      "step": 126
    },
    {
      "epoch": 1.44,
      "logps_train/chosen": -171.53411865234375,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -186.81044006347656,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.3897385597229004,
      "rewards_train/margins": 3.4538064002990723,
      "rewards_train/rejected": -4.843544960021973,
      "step": 126
    },
    {
      "epoch": 1.44,
      "logps_train/chosen": -157.05662536621094,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -139.83226013183594,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.347580909729004,
      "rewards_train/margins": 2.9695076942443848,
      "rewards_train/rejected": -4.317088603973389,
      "step": 126
    },
    {
      "epoch": 1.44,
      "logps_train/chosen": -155.17971801757812,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -143.3584747314453,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.2133831977844238,
      "rewards_train/margins": 2.3373942375183105,
      "rewards_train/rejected": -3.5507774353027344,
      "step": 126
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.7324891299008985e-05,
      "loss": 0.2015,
      "step": 127
    },
    {
      "epoch": 1.45,
      "logps_train/chosen": -153.86216735839844,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -162.63079833984375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.480747938156128,
      "rewards_train/margins": 2.8033287525177,
      "rewards_train/rejected": -4.284076690673828,
      "step": 127
    },
    {
      "epoch": 1.45,
      "logps_train/chosen": -186.6682586669922,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -192.19993591308594,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0969038009643555,
      "rewards_train/margins": 3.314544677734375,
      "rewards_train/rejected": -4.4114484786987305,
      "step": 127
    },
    {
      "epoch": 1.45,
      "logps_train/chosen": -153.8680419921875,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -171.44602966308594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9669804573059082,
      "rewards_train/margins": 2.993198871612549,
      "rewards_train/rejected": -4.960179328918457,
      "step": 127
    },
    {
      "epoch": 1.45,
      "logps_train/chosen": -176.20404052734375,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -169.44656372070312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6339795589447021,
      "rewards_train/margins": 3.147592306137085,
      "rewards_train/rejected": -3.781571865081787,
      "step": 127
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.701563121054695e-05,
      "loss": 0.1361,
      "step": 128
    },
    {
      "epoch": 1.46,
      "logps_train/chosen": -166.00640869140625,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -178.55386352539062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8213452100753784,
      "rewards_train/margins": 3.634822964668274,
      "rewards_train/rejected": -4.456168174743652,
      "step": 128
    },
    {
      "epoch": 1.46,
      "logps_train/chosen": -162.34912109375,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -152.22750854492188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0764155387878418,
      "rewards_train/margins": 3.124901294708252,
      "rewards_train/rejected": -4.201316833496094,
      "step": 128
    },
    {
      "epoch": 1.46,
      "logps_train/chosen": -156.54486083984375,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -149.5716094970703,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1559518575668335,
      "rewards_train/margins": 3.318982720375061,
      "rewards_train/rejected": -4.4749345779418945,
      "step": 128
    },
    {
      "epoch": 1.46,
      "logps_train/chosen": -190.66033935546875,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -178.93310546875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.6988459825515747,
      "rewards_train/margins": 3.3385089635849,
      "rewards_train/rejected": -4.037354946136475,
      "step": 128
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.6706060334116777e-05,
      "loss": 0.1665,
      "step": 129
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -153.0819854736328,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -140.51028442382812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.736420750617981,
      "rewards_train/margins": 3.3056238889694214,
      "rewards_train/rejected": -4.042044639587402,
      "step": 129
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -179.62692260742188,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -159.4015655517578,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3000822067260742,
      "rewards_train/margins": 3.679259777069092,
      "rewards_train/rejected": -4.979341983795166,
      "step": 129
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -162.6141815185547,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -156.1014862060547,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.0475513935089111,
      "rewards_train/margins": 3.2958009243011475,
      "rewards_train/rejected": -4.343352317810059,
      "step": 129
    },
    {
      "epoch": 1.47,
      "logps_train/chosen": -174.86026000976562,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -182.4564971923828,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3674092292785645,
      "rewards_train/margins": 3.3553266525268555,
      "rewards_train/rejected": -4.72273588180542,
      "step": 129
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.639622640211277e-05,
      "loss": 0.1287,
      "step": 130
    },
    {
      "epoch": 1.49,
      "logps_train/chosen": -142.3889923095703,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -153.8872528076172,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3541336059570312,
      "rewards_train/margins": 3.004904270172119,
      "rewards_train/rejected": -4.35903787612915,
      "step": 130
    },
    {
      "epoch": 1.49,
      "logps_train/chosen": -168.5977020263672,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -160.97059631347656,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.1273491382598877,
      "rewards_train/margins": 3.112093210220337,
      "rewards_train/rejected": -4.239442348480225,
      "step": 130
    },
    {
      "epoch": 1.49,
      "logps_train/chosen": -186.99581909179688,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -178.27426147460938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.5144252777099609,
      "rewards_train/margins": 3.7290172576904297,
      "rewards_train/rejected": -4.243442535400391,
      "step": 130
    },
    {
      "epoch": 1.49,
      "logps_train/chosen": -163.23785400390625,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -85.5,
      "logps_train/rejected": -131.27413940429688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0287654399871826,
      "rewards_train/margins": 3.533609628677368,
      "rewards_train/rejected": -4.562375068664551,
      "step": 130
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.6086177187489453e-05,
      "loss": 0.1228,
      "step": 131
    },
    {
      "epoch": 1.5,
      "logps_train/chosen": -147.3597412109375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -145.3800048828125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1206421852111816,
      "rewards_train/margins": 3.2051515579223633,
      "rewards_train/rejected": -4.325793743133545,
      "step": 131
    },
    {
      "epoch": 1.5,
      "logps_train/chosen": -165.80751037597656,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -178.4746551513672,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1619529724121094,
      "rewards_train/margins": 3.133852005004883,
      "rewards_train/rejected": -4.295804977416992,
      "step": 131
    },
    {
      "epoch": 1.5,
      "logps_train/chosen": -169.84341430664062,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -160.15591430664062,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.9361000061035156,
      "rewards_train/margins": 3.2722649574279785,
      "rewards_train/rejected": -4.208364963531494,
      "step": 131
    },
    {
      "epoch": 1.5,
      "logps_train/chosen": -163.28036499023438,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -147.83535766601562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.7479585409164429,
      "rewards_train/margins": 3.2120429277420044,
      "rewards_train/rejected": -3.9600014686584473,
      "step": 131
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.5775960496395564e-05,
      "loss": 0.1649,
      "step": 132
    },
    {
      "epoch": 1.51,
      "logps_train/chosen": -149.8038330078125,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -147.8526611328125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4508914947509766,
      "rewards_train/margins": 3.12939453125,
      "rewards_train/rejected": -4.580286026000977,
      "step": 132
    },
    {
      "epoch": 1.51,
      "logps_train/chosen": -158.60960388183594,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -155.93435668945312,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.1415255069732666,
      "rewards_train/margins": 2.831206798553467,
      "rewards_train/rejected": -3.9727323055267334,
      "step": 132
    },
    {
      "epoch": 1.51,
      "logps_train/chosen": -177.83819580078125,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -161.01573181152344,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.7680490612983704,
      "rewards_train/margins": 3.2229778170585632,
      "rewards_train/rejected": -3.9910268783569336,
      "step": 132
    },
    {
      "epoch": 1.51,
      "logps_train/chosen": -150.2593231201172,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -158.00173950195312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7834035754203796,
      "rewards_train/margins": 3.515306293964386,
      "rewards_train/rejected": -4.298709869384766,
      "step": 132
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.5465624160802847e-05,
      "loss": 0.1705,
      "step": 133
    },
    {
      "epoch": 1.52,
      "logps_train/chosen": -163.82518005371094,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -160.40072631835938,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.2626445293426514,
      "rewards_train/margins": 2.3889267444610596,
      "rewards_train/rejected": -3.651571273803711,
      "step": 133
    },
    {
      "epoch": 1.52,
      "logps_train/chosen": -138.35488891601562,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -150.45436096191406,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.8314854502677917,
      "rewards_train/margins": 3.542002499103546,
      "rewards_train/rejected": -4.373487949371338,
      "step": 133
    },
    {
      "epoch": 1.52,
      "logps_train/chosen": -164.03900146484375,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -152.27687072753906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3316338062286377,
      "rewards_train/margins": 3.427401304244995,
      "rewards_train/rejected": -4.759035110473633,
      "step": 133
    },
    {
      "epoch": 1.52,
      "logps_train/chosen": -193.63211059570312,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -177.69418334960938,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.9139914512634277,
      "rewards_train/margins": 3.4447827339172363,
      "rewards_train/rejected": -4.358774185180664,
      "step": 133
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.515521603113088e-05,
      "loss": 0.187,
      "step": 134
    },
    {
      "epoch": 1.53,
      "logps_train/chosen": -175.33932495117188,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -165.0,
      "logps_train/rejected": -213.67547607421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.5542943477630615,
      "rewards_train/margins": 4.358224153518677,
      "rewards_train/rejected": -4.912518501281738,
      "step": 134
    },
    {
      "epoch": 1.53,
      "logps_train/chosen": -158.3114013671875,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -173.7991943359375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7678590416908264,
      "rewards_train/margins": 4.60151332616806,
      "rewards_train/rejected": -5.369372367858887,
      "step": 134
    },
    {
      "epoch": 1.53,
      "logps_train/chosen": -131.25491333007812,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -91.0,
      "logps_train/rejected": -131.59129333496094,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3272488117218018,
      "rewards_train/margins": 2.7177212238311768,
      "rewards_train/rejected": -4.0449700355529785,
      "step": 134
    },
    {
      "epoch": 1.53,
      "logps_train/chosen": -139.65713500976562,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -143.0663299560547,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1557526588439941,
      "rewards_train/margins": 3.5598645210266113,
      "rewards_train/rejected": -4.7156171798706055,
      "step": 134
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.4844783968869126e-05,
      "loss": 0.1299,
      "step": 135
    },
    {
      "epoch": 1.54,
      "logps_train/chosen": -155.28524780273438,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -126.5,
      "logps_train/rejected": -163.39198303222656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8066498041152954,
      "rewards_train/margins": 2.9134079217910767,
      "rewards_train/rejected": -3.720057725906372,
      "step": 135
    },
    {
      "epoch": 1.54,
      "logps_train/chosen": -178.38946533203125,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -192.02011108398438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0768359899520874,
      "rewards_train/margins": 3.238333821296692,
      "rewards_train/rejected": -4.315169811248779,
      "step": 135
    },
    {
      "epoch": 1.54,
      "logps_train/chosen": -160.60389709472656,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -163.22579956054688,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4700582027435303,
      "rewards_train/margins": 2.681037187576294,
      "rewards_train/rejected": -4.151095390319824,
      "step": 135
    },
    {
      "epoch": 1.54,
      "logps_train/chosen": -191.0967559814453,
      "logps_train/ref_chosen": -185.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -190.99061584472656,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6215896606445312,
      "rewards_train/margins": 3.8812813758850098,
      "rewards_train/rejected": -4.502871036529541,
      "step": 135
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.4534375839197166e-05,
      "loss": 0.1323,
      "step": 136
    },
    {
      "epoch": 1.55,
      "logps_train/chosen": -216.13201904296875,
      "logps_train/ref_chosen": -210.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -161.54336547851562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6491394639015198,
      "rewards_train/margins": 3.5523642897605896,
      "rewards_train/rejected": -4.201503753662109,
      "step": 136
    },
    {
      "epoch": 1.55,
      "logps_train/chosen": -186.5906219482422,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -165.56106567382812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7216589450836182,
      "rewards_train/margins": 3.585913896560669,
      "rewards_train/rejected": -4.307572841644287,
      "step": 136
    },
    {
      "epoch": 1.55,
      "logps_train/chosen": -157.28329467773438,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -144.79754638671875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2453217506408691,
      "rewards_train/margins": 3.0565037727355957,
      "rewards_train/rejected": -4.301825523376465,
      "step": 136
    },
    {
      "epoch": 1.55,
      "logps_train/chosen": -146.12713623046875,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -144.28741455078125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4028985500335693,
      "rewards_train/margins": 2.723278284072876,
      "rewards_train/rejected": -4.126176834106445,
      "step": 136
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.4224039503604435e-05,
      "loss": 0.1395,
      "step": 137
    },
    {
      "epoch": 1.57,
      "logps_train/chosen": -193.05108642578125,
      "logps_train/ref_chosen": -186.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -180.34556579589844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7160476446151733,
      "rewards_train/margins": 3.606399178504944,
      "rewards_train/rejected": -4.322446823120117,
      "step": 137
    },
    {
      "epoch": 1.57,
      "logps_train/chosen": -177.28684997558594,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -169.56716918945312,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -0.8280016779899597,
      "rewards_train/margins": 3.3246129155158997,
      "rewards_train/rejected": -4.152614593505859,
      "step": 137
    },
    {
      "epoch": 1.57,
      "logps_train/chosen": -181.3519744873047,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -164.0,
      "logps_train/rejected": -209.20787048339844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.585979163646698,
      "rewards_train/margins": 3.8938897252082825,
      "rewards_train/rejected": -4.4798688888549805,
      "step": 137
    },
    {
      "epoch": 1.57,
      "logps_train/chosen": -159.36561584472656,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -163.44479370117188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.33968687057495117,
      "rewards_train/margins": 3.739558219909668,
      "rewards_train/rejected": -4.079245090484619,
      "step": 137
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.391382281251055e-05,
      "loss": 0.1132,
      "step": 138
    },
    {
      "epoch": 1.58,
      "logps_train/chosen": -160.30413818359375,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -160.53933715820312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.747503936290741,
      "rewards_train/margins": 3.6289380192756653,
      "rewards_train/rejected": -4.376441955566406,
      "step": 138
    },
    {
      "epoch": 1.58,
      "logps_train/chosen": -193.16165161132812,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -167.0458984375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.4382353723049164,
      "rewards_train/margins": 3.1267053186893463,
      "rewards_train/rejected": -3.5649406909942627,
      "step": 138
    },
    {
      "epoch": 1.58,
      "logps_train/chosen": -177.08596801757812,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -176.0380401611328,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2597697973251343,
      "rewards_train/margins": 3.0808502435684204,
      "rewards_train/rejected": -4.340620040893555,
      "step": 138
    },
    {
      "epoch": 1.58,
      "logps_train/chosen": -159.55792236328125,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -145.29629516601562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9486621618270874,
      "rewards_train/margins": 3.1882187128067017,
      "rewards_train/rejected": -4.136880874633789,
      "step": 138
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3603773597887237e-05,
      "loss": 0.1249,
      "step": 139
    },
    {
      "epoch": 1.59,
      "logps_train/chosen": -165.89608764648438,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -172.01133728027344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7296485900878906,
      "rewards_train/margins": 3.3035411834716797,
      "rewards_train/rejected": -4.03318977355957,
      "step": 139
    },
    {
      "epoch": 1.59,
      "logps_train/chosen": -144.5615234375,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -163.75762939453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.721398651599884,
      "rewards_train/margins": 3.903387725353241,
      "rewards_train/rejected": -4.624786376953125,
      "step": 139
    },
    {
      "epoch": 1.59,
      "logps_train/chosen": -169.52536010742188,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -170.62039184570312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9957976341247559,
      "rewards_train/margins": 2.806670904159546,
      "rewards_train/rejected": -3.8024685382843018,
      "step": 139
    },
    {
      "epoch": 1.59,
      "logps_train/chosen": -159.63851928710938,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -154.8424072265625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.4302584230899811,
      "rewards_train/margins": 3.5069116055965424,
      "rewards_train/rejected": -3.9371700286865234,
      "step": 139
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.329393966588323e-05,
      "loss": 0.1378,
      "step": 140
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -174.330078125,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -172.95440673828125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9353509545326233,
      "rewards_train/margins": 3.4886531233787537,
      "rewards_train/rejected": -4.424004077911377,
      "step": 140
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -166.16903686523438,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -187.2434844970703,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.7706148624420166,
      "rewards_train/margins": 3.949387788772583,
      "rewards_train/rejected": -4.7200026512146,
      "step": 140
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -163.92617797851562,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -188.00546264648438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8798237442970276,
      "rewards_train/margins": 3.081660211086273,
      "rewards_train/rejected": -3.961483955383301,
      "step": 140
    },
    {
      "epoch": 1.6,
      "logps_train/chosen": -154.75360107421875,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -161.3019561767578,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1850275993347168,
      "rewards_train/margins": 3.391695022583008,
      "rewards_train/rejected": -4.576722621917725,
      "step": 140
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.298436878945306e-05,
      "loss": 0.1181,
      "step": 141
    },
    {
      "epoch": 1.61,
      "logps_train/chosen": -142.72462463378906,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -179.27532958984375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9942886829376221,
      "rewards_train/margins": 3.0164482593536377,
      "rewards_train/rejected": -4.01073694229126,
      "step": 141
    },
    {
      "epoch": 1.61,
      "logps_train/chosen": -190.180419921875,
      "logps_train/ref_chosen": -181.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -180.17337036132812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9520259499549866,
      "rewards_train/margins": 3.4328898787498474,
      "rewards_train/rejected": -4.384915828704834,
      "step": 141
    },
    {
      "epoch": 1.61,
      "logps_train/chosen": -134.74435424804688,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -72.5,
      "logps_train/rejected": -111.79988098144531,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.6727750301361084,
      "rewards_train/margins": 2.240952968597412,
      "rewards_train/rejected": -3.9137279987335205,
      "step": 141
    },
    {
      "epoch": 1.61,
      "logps_train/chosen": -176.04776000976562,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -164.4101104736328,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1286051273345947,
      "rewards_train/margins": 3.44355845451355,
      "rewards_train/rejected": -4.5721635818481445,
      "step": 141
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.267510870099101e-05,
      "loss": 0.1654,
      "step": 142
    },
    {
      "epoch": 1.62,
      "logps_train/chosen": -188.28570556640625,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -180.6005859375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1609909534454346,
      "rewards_train/margins": 3.7696726322174072,
      "rewards_train/rejected": -4.930663585662842,
      "step": 142
    },
    {
      "epoch": 1.62,
      "logps_train/chosen": -157.38246154785156,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -157.08433532714844,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.2556779384613037,
      "rewards_train/margins": 2.78234601020813,
      "rewards_train/rejected": -4.038023948669434,
      "step": 142
    },
    {
      "epoch": 1.62,
      "logps_train/chosen": -139.43142700195312,
      "logps_train/ref_chosen": -127.5,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -141.7026824951172,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.159501314163208,
      "rewards_train/margins": 3.0188729763031006,
      "rewards_train/rejected": -4.178374290466309,
      "step": 142
    },
    {
      "epoch": 1.62,
      "logps_train/chosen": -125.58201599121094,
      "logps_train/ref_chosen": -104.0,
      "logps_train/ref_rejected": -88.0,
      "logps_train/rejected": -139.86517333984375,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -2.145481824874878,
      "rewards_train/margins": 3.053291082382202,
      "rewards_train/rejected": -5.19877290725708,
      "step": 142
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.2366207084971325e-05,
      "loss": 0.1394,
      "step": 143
    },
    {
      "epoch": 1.63,
      "logps_train/chosen": -119.32640838623047,
      "logps_train/ref_chosen": -104.0,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -134.8964385986328,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.5528064966201782,
      "rewards_train/margins": 2.5242398977279663,
      "rewards_train/rejected": -4.0770463943481445,
      "step": 143
    },
    {
      "epoch": 1.63,
      "logps_train/chosen": -166.6683349609375,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -158.14041137695312,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.3560905456542969,
      "rewards_train/margins": 2.8011627197265625,
      "rewards_train/rejected": -4.157253265380859,
      "step": 143
    },
    {
      "epoch": 1.63,
      "logps_train/chosen": -153.33035278320312,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -163.28121948242188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7734657526016235,
      "rewards_train/margins": 3.996648669242859,
      "rewards_train/rejected": -4.770114421844482,
      "step": 143
    },
    {
      "epoch": 1.63,
      "logps_train/chosen": -171.47213745117188,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -174.22564697265625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.0140111446380615,
      "rewards_train/margins": 2.929255723953247,
      "rewards_train/rejected": -3.9432668685913086,
      "step": 143
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2057711570595746e-05,
      "loss": 0.2175,
      "step": 144
    },
    {
      "epoch": 1.65,
      "logps_train/chosen": -173.53387451171875,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -161.1766357421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.8866889476776123,
      "rewards_train/margins": 3.222722291946411,
      "rewards_train/rejected": -4.109411239624023,
      "step": 144
    },
    {
      "epoch": 1.65,
      "logps_train/chosen": -141.76498413085938,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -159.65899658203125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.348568320274353,
      "rewards_train/margins": 3.194480061531067,
      "rewards_train/rejected": -4.54304838180542,
      "step": 144
    },
    {
      "epoch": 1.65,
      "logps_train/chosen": -197.79476928710938,
      "logps_train/ref_chosen": -192.0,
      "logps_train/ref_rejected": -166.0,
      "logps_train/rejected": -209.7325439453125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.5577000379562378,
      "rewards_train/margins": 3.7895785570144653,
      "rewards_train/rejected": -4.347278594970703,
      "step": 144
    },
    {
      "epoch": 1.65,
      "logps_train/chosen": -130.68243408203125,
      "logps_train/ref_chosen": -122.0,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -138.5338897705078,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8820131421089172,
      "rewards_train/margins": 3.5245015025138855,
      "rewards_train/rejected": -4.406514644622803,
      "step": 144
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.174966972444967e-05,
      "loss": 0.1446,
      "step": 145
    },
    {
      "epoch": 1.66,
      "logps_train/chosen": -163.034912109375,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -165.3939666748047,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1952872276306152,
      "rewards_train/margins": 3.1185240745544434,
      "rewards_train/rejected": -4.313811302185059,
      "step": 145
    },
    {
      "epoch": 1.66,
      "logps_train/chosen": -182.8961181640625,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -172.29110717773438,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.5673462152481079,
      "rewards_train/margins": 3.453562617301941,
      "rewards_train/rejected": -4.020908832550049,
      "step": 145
    },
    {
      "epoch": 1.66,
      "logps_train/chosen": -150.4124298095703,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -159.89599609375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4033517837524414,
      "rewards_train/margins": 3.661468505859375,
      "rewards_train/rejected": -5.064820289611816,
      "step": 145
    },
    {
      "epoch": 1.66,
      "logps_train/chosen": -184.98175048828125,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -164.98641967773438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.6009587049484253,
      "rewards_train/margins": 3.9541760683059692,
      "rewards_train/rejected": -4.5551347732543945,
      "step": 145
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.1442129043167874e-05,
      "loss": 0.1344,
      "step": 146
    },
    {
      "epoch": 1.67,
      "logps_train/chosen": -164.82098388671875,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -179.12875366210938,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3958685398101807,
      "rewards_train/margins": 3.42560076713562,
      "rewards_train/rejected": -4.821469306945801,
      "step": 146
    },
    {
      "epoch": 1.67,
      "logps_train/chosen": -143.52203369140625,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -173.89883422851562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.5310614109039307,
      "rewards_train/margins": 3.589242696762085,
      "rewards_train/rejected": -5.120304107666016,
      "step": 146
    },
    {
      "epoch": 1.67,
      "logps_train/chosen": -171.19552612304688,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -150.95005798339844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.160958170890808,
      "rewards_train/margins": 3.271157383918762,
      "rewards_train/rejected": -4.43211555480957,
      "step": 146
    },
    {
      "epoch": 1.67,
      "logps_train/chosen": -185.29385375976562,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -177.15206909179688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1695337295532227,
      "rewards_train/margins": 3.7931838035583496,
      "rewards_train/rejected": -4.962717533111572,
      "step": 146
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.1135136946111078e-05,
      "loss": 0.1606,
      "step": 147
    },
    {
      "epoch": 1.68,
      "logps_train/chosen": -178.3898162841797,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -172.91342163085938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1439135074615479,
      "rewards_train/margins": 3.5383970737457275,
      "rewards_train/rejected": -4.682310581207275,
      "step": 147
    },
    {
      "epoch": 1.68,
      "logps_train/chosen": -175.71603393554688,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -164.0,
      "logps_train/rejected": -209.33184814453125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.5641824007034302,
      "rewards_train/margins": 4.025251746177673,
      "rewards_train/rejected": -4.5894341468811035,
      "step": 147
    },
    {
      "epoch": 1.68,
      "logps_train/chosen": -186.2715301513672,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -172.0,
      "logps_train/rejected": -216.0167999267578,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.4318392276763916,
      "rewards_train/margins": 4.031365156173706,
      "rewards_train/rejected": -4.463204383850098,
      "step": 147
    },
    {
      "epoch": 1.68,
      "logps_train/chosen": -159.61163330078125,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -172.23619079589844,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1693428754806519,
      "rewards_train/margins": 3.491873860359192,
      "rewards_train/rejected": -4.661216735839844,
      "step": 147
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.0828740768054405e-05,
      "loss": 0.094,
      "step": 148
    },
    {
      "epoch": 1.69,
      "logps_train/chosen": -146.35887145996094,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -93.0,
      "logps_train/rejected": -136.61520385742188,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -1.5266588926315308,
      "rewards_train/margins": 2.8000484704971313,
      "rewards_train/rejected": -4.326707363128662,
      "step": 148
    },
    {
      "epoch": 1.69,
      "logps_train/chosen": -187.1864013671875,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -154.72918701171875,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3573615550994873,
      "rewards_train/margins": 2.9866015911102295,
      "rewards_train/rejected": -4.343963146209717,
      "step": 148
    },
    {
      "epoch": 1.69,
      "logps_train/chosen": -170.3841552734375,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -188.68914794921875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0149770975112915,
      "rewards_train/margins": 3.598516821861267,
      "rewards_train/rejected": -4.613493919372559,
      "step": 148
    },
    {
      "epoch": 1.69,
      "logps_train/chosen": -184.0547332763672,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -186.50656127929688,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.43496501445770264,
      "rewards_train/margins": 4.409247994422913,
      "rewards_train/rejected": -4.844213008880615,
      "step": 148
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.0522987751888878e-05,
      "loss": 0.1432,
      "step": 149
    },
    {
      "epoch": 1.7,
      "logps_train/chosen": -172.67449951171875,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -175.04855346679688,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4930354356765747,
      "rewards_train/margins": 2.8857582807540894,
      "rewards_train/rejected": -4.378793716430664,
      "step": 149
    },
    {
      "epoch": 1.7,
      "logps_train/chosen": -178.00057983398438,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -162.48727416992188,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.4646331071853638,
      "rewards_train/margins": 2.847375750541687,
      "rewards_train/rejected": -4.312008857727051,
      "step": 149
    },
    {
      "epoch": 1.7,
      "logps_train/chosen": -177.27328491210938,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -173.37014770507812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8803309798240662,
      "rewards_train/margins": 3.7075143456459045,
      "rewards_train/rejected": -4.587845325469971,
      "step": 149
    },
    {
      "epoch": 1.7,
      "logps_train/chosen": -178.50416564941406,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -192.19558715820312,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.926978588104248,
      "rewards_train/margins": 3.855861186981201,
      "rewards_train/rejected": -4.782839775085449,
      "step": 149
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.0217925041337088e-05,
      "loss": 0.1603,
      "step": 150
    },
    {
      "epoch": 1.71,
      "logps_train/chosen": -182.9842071533203,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -156.3119354248047,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1948070526123047,
      "rewards_train/margins": 3.445127010345459,
      "rewards_train/rejected": -4.639934062957764,
      "step": 150
    },
    {
      "epoch": 1.71,
      "logps_train/chosen": -196.98382568359375,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -182.46234130859375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4740171432495117,
      "rewards_train/margins": 3.0879406929016113,
      "rewards_train/rejected": -4.561957836151123,
      "step": 150
    },
    {
      "epoch": 1.71,
      "logps_train/chosen": -188.139892578125,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -190.43099975585938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2782472372055054,
      "rewards_train/margins": 3.6082111597061157,
      "rewards_train/rejected": -4.886458396911621,
      "step": 150
    },
    {
      "epoch": 1.71,
      "logps_train/chosen": -205.94732666015625,
      "logps_train/ref_chosen": -193.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -199.26638793945312,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.2850637435913086,
      "rewards_train/margins": 3.3791723251342773,
      "rewards_train/rejected": -4.664236068725586,
      "step": 150
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.991359967368416e-05,
      "loss": 0.1191,
      "step": 151
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -164.7177276611328,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -163.2447509765625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.7888622283935547,
      "rewards_train/margins": 2.895867347717285,
      "rewards_train/rejected": -4.68472957611084,
      "step": 151
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -192.6652374267578,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -164.0,
      "logps_train/rejected": -208.13421630859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.8331253528594971,
      "rewards_train/margins": 3.594896078109741,
      "rewards_train/rejected": -4.428021430969238,
      "step": 151
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -179.51992797851562,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -167.351318359375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3398842811584473,
      "rewards_train/margins": 3.2446608543395996,
      "rewards_train/rejected": -4.584545135498047,
      "step": 151
    },
    {
      "epoch": 1.73,
      "logps_train/chosen": -142.92648315429688,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -151.46743774414062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6470386981964111,
      "rewards_train/margins": 3.038767099380493,
      "rewards_train/rejected": -4.685805797576904,
      "step": 151
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.9610058572525126e-05,
      "loss": 0.156,
      "step": 152
    },
    {
      "epoch": 1.74,
      "logps_train/chosen": -125.03524017333984,
      "logps_train/ref_chosen": -104.5,
      "logps_train/ref_rejected": -98.5,
      "logps_train/rejected": -143.35089111328125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.051748275756836,
      "rewards_train/margins": 2.4157614707946777,
      "rewards_train/rejected": -4.467509746551514,
      "step": 152
    },
    {
      "epoch": 1.74,
      "logps_train/chosen": -163.41842651367188,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -166.00680541992188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5096149444580078,
      "rewards_train/margins": 3.4148945808410645,
      "rewards_train/rejected": -4.924509525299072,
      "step": 152
    },
    {
      "epoch": 1.74,
      "logps_train/chosen": -184.01190185546875,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -182.28060913085938,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1560717821121216,
      "rewards_train/margins": 3.208804965019226,
      "rewards_train/rejected": -4.364876747131348,
      "step": 152
    },
    {
      "epoch": 1.74,
      "logps_train/chosen": -134.97909545898438,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -154.97637939453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7018145322799683,
      "rewards_train/margins": 2.6661349534988403,
      "rewards_train/rejected": -4.367949485778809,
      "step": 152
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.9307348540529842e-05,
      "loss": 0.199,
      "step": 153
    },
    {
      "epoch": 1.75,
      "logps_train/chosen": -184.7167205810547,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -184.70632934570312,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.5790934562683105,
      "rewards_train/margins": 3.391930103302002,
      "rewards_train/rejected": -4.9710235595703125,
      "step": 153
    },
    {
      "epoch": 1.75,
      "logps_train/chosen": -170.3045654296875,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -183.49859619140625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.2700073719024658,
      "rewards_train/margins": 3.2787787914276123,
      "rewards_train/rejected": -4.548786163330078,
      "step": 153
    },
    {
      "epoch": 1.75,
      "logps_train/chosen": -192.09291076660156,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -182.16331481933594,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.9595848321914673,
      "rewards_train/margins": 3.4182342290878296,
      "rewards_train/rejected": -4.377819061279297,
      "step": 153
    },
    {
      "epoch": 1.75,
      "logps_train/chosen": -185.59799194335938,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -156.0,
      "logps_train/rejected": -207.5470733642578,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3219197988510132,
      "rewards_train/margins": 3.890599846839905,
      "rewards_train/rejected": -5.212519645690918,
      "step": 153
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.90055162522265e-05,
      "loss": 0.1899,
      "step": 154
    },
    {
      "epoch": 1.76,
      "logps_train/chosen": -160.90257263183594,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -174.55712890625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.113547921180725,
      "rewards_train/margins": 3.4354270696640015,
      "rewards_train/rejected": -4.548974990844727,
      "step": 154
    },
    {
      "epoch": 1.76,
      "logps_train/chosen": -168.6278076171875,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -185.01776123046875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2065556049346924,
      "rewards_train/margins": 3.4405343532562256,
      "rewards_train/rejected": -4.647089958190918,
      "step": 154
    },
    {
      "epoch": 1.76,
      "logps_train/chosen": -179.93496704101562,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -182.0,
      "logps_train/rejected": -236.11505126953125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.095841407775879,
      "rewards_train/margins": 4.302578449249268,
      "rewards_train/rejected": -5.3984198570251465,
      "step": 154
    },
    {
      "epoch": 1.76,
      "logps_train/chosen": -141.23049926757812,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -173.46426391601562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.7132835388183594,
      "rewards_train/margins": 3.7317748069763184,
      "rewards_train/rejected": -5.445058345794678,
      "step": 154
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.8704608246804956e-05,
      "loss": 0.1316,
      "step": 155
    },
    {
      "epoch": 1.77,
      "logps_train/chosen": -169.93484497070312,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -159.2445526123047,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6998562812805176,
      "rewards_train/margins": 3.3152976036071777,
      "rewards_train/rejected": -5.015153884887695,
      "step": 155
    },
    {
      "epoch": 1.77,
      "logps_train/chosen": -170.24639892578125,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -179.92730712890625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5101873874664307,
      "rewards_train/margins": 3.3885018825531006,
      "rewards_train/rejected": -4.898689270019531,
      "step": 155
    },
    {
      "epoch": 1.77,
      "logps_train/chosen": -172.0523681640625,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -177.50497436523438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.215980052947998,
      "rewards_train/margins": 3.4622879028320312,
      "rewards_train/rejected": -4.678267955780029,
      "step": 155
    },
    {
      "epoch": 1.77,
      "logps_train/chosen": -167.3325958251953,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -181.28607177734375,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.4187091588974,
      "rewards_train/margins": 3.3278671503067017,
      "rewards_train/rejected": -4.746576309204102,
      "step": 155
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.840467092094091e-05,
      "loss": 0.1536,
      "step": 156
    },
    {
      "epoch": 1.78,
      "logps_train/chosen": -170.8963623046875,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -178.3448028564453,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.1062872409820557,
      "rewards_train/margins": 3.4995310306549072,
      "rewards_train/rejected": -4.605818271636963,
      "step": 156
    },
    {
      "epoch": 1.78,
      "logps_train/chosen": -152.4623260498047,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -152.89100646972656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9597091674804688,
      "rewards_train/margins": 2.9297332763671875,
      "rewards_train/rejected": -4.889442443847656,
      "step": 156
    },
    {
      "epoch": 1.78,
      "logps_train/chosen": -140.10516357421875,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -93.0,
      "logps_train/rejected": -136.63473510742188,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.0900087356567383,
      "rewards_train/margins": 3.295339584350586,
      "rewards_train/rejected": -4.385348320007324,
      "step": 156
    },
    {
      "epoch": 1.78,
      "logps_train/chosen": -178.86492919921875,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -188.87057495117188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0580618381500244,
      "rewards_train/margins": 3.5533130168914795,
      "rewards_train/rejected": -4.611374855041504,
      "step": 156
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.8105750521642034e-05,
      "loss": 0.1377,
      "step": 157
    },
    {
      "epoch": 1.79,
      "logps_train/chosen": -176.15272521972656,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -193.9800567626953,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5117566585540771,
      "rewards_train/margins": 3.3692567348480225,
      "rewards_train/rejected": -4.8810133934021,
      "step": 157
    },
    {
      "epoch": 1.79,
      "logps_train/chosen": -165.08853149414062,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -168.65086364746094,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.9819000959396362,
      "rewards_train/margins": 4.159559845924377,
      "rewards_train/rejected": -5.141459941864014,
      "step": 157
    },
    {
      "epoch": 1.79,
      "logps_train/chosen": -171.33531188964844,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -165.46160888671875,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.4948601722717285,
      "rewards_train/margins": 3.4002251625061035,
      "rewards_train/rejected": -4.895085334777832,
      "step": 157
    },
    {
      "epoch": 1.79,
      "logps_train/chosen": -158.28863525390625,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -168.30862426757812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3315372467041016,
      "rewards_train/margins": 3.280808448791504,
      "rewards_train/rejected": -4.6123456954956055,
      "step": 157
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.780789313911722e-05,
      "loss": 0.1212,
      "step": 158
    },
    {
      "epoch": 1.81,
      "logps_train/chosen": -164.3837890625,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -195.4906005859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.437110424041748,
      "rewards_train/margins": 3.6194682121276855,
      "rewards_train/rejected": -5.056578636169434,
      "step": 158
    },
    {
      "epoch": 1.81,
      "logps_train/chosen": -167.58389282226562,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -165.21438598632812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1453032493591309,
      "rewards_train/margins": 3.3836545944213867,
      "rewards_train/rejected": -4.528957843780518,
      "step": 158
    },
    {
      "epoch": 1.81,
      "logps_train/chosen": -167.3607177734375,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -95.0,
      "logps_train/rejected": -142.85699462890625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.172424077987671,
      "rewards_train/margins": 3.6320254802703857,
      "rewards_train/rejected": -4.804449558258057,
      "step": 158
    },
    {
      "epoch": 1.81,
      "logps_train/chosen": -138.3846893310547,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -89.0,
      "logps_train/rejected": -128.16195678710938,
      "rewards_train/accuracies": 0.84375,
      "rewards_train/chosen": -2.0708906650543213,
      "rewards_train/margins": 1.8548266887664795,
      "rewards_train/rejected": -3.925717353820801,
      "step": 158
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.7511144699669966e-05,
      "loss": 0.1585,
      "step": 159
    },
    {
      "epoch": 1.82,
      "logps_train/chosen": -184.75267028808594,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -182.9331512451172,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -0.784446656703949,
      "rewards_train/margins": 4.016046583652496,
      "rewards_train/rejected": -4.800493240356445,
      "step": 159
    },
    {
      "epoch": 1.82,
      "logps_train/chosen": -138.78643798828125,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -86.0,
      "logps_train/rejected": -133.27508544921875,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.0814690589904785,
      "rewards_train/margins": 2.621307849884033,
      "rewards_train/rejected": -4.702776908874512,
      "step": 159
    },
    {
      "epoch": 1.82,
      "logps_train/chosen": -174.599609375,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -172.98977661132812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.085846424102783,
      "rewards_train/margins": 3.226027488708496,
      "rewards_train/rejected": -5.311873912811279,
      "step": 159
    },
    {
      "epoch": 1.82,
      "logps_train/chosen": -154.55592346191406,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -148.77499389648438,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.5021743774414062,
      "rewards_train/margins": 2.780306339263916,
      "rewards_train/rejected": -4.282480716705322,
      "step": 159
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.7215550958617034e-05,
      "loss": 0.1462,
      "step": 160
    },
    {
      "epoch": 1.83,
      "logps_train/chosen": -162.04150390625,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -162.67465209960938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.419677495956421,
      "rewards_train/margins": 3.43099045753479,
      "rewards_train/rejected": -4.850667953491211,
      "step": 160
    },
    {
      "epoch": 1.83,
      "logps_train/chosen": -147.52159118652344,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -146.4416961669922,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.1266707181930542,
      "rewards_train/margins": 3.2333184480667114,
      "rewards_train/rejected": -4.359989166259766,
      "step": 160
    },
    {
      "epoch": 1.83,
      "logps_train/chosen": -168.31619262695312,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -152.9374542236328,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2656030654907227,
      "rewards_train/margins": 3.5036306381225586,
      "rewards_train/rejected": -4.769233703613281,
      "step": 160
    },
    {
      "epoch": 1.83,
      "logps_train/chosen": -205.88299560546875,
      "logps_train/ref_chosen": -196.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -174.3811492919922,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -0.9983091354370117,
      "rewards_train/margins": 3.4017200469970703,
      "rewards_train/rejected": -4.400029182434082,
      "step": 160
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.6921157493233532e-05,
      "loss": 0.1424,
      "step": 161
    },
    {
      "epoch": 1.84,
      "logps_train/chosen": -167.32449340820312,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -164.69781494140625,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4970972537994385,
      "rewards_train/margins": 3.72248911857605,
      "rewards_train/rejected": -5.219586372375488,
      "step": 161
    },
    {
      "epoch": 1.84,
      "logps_train/chosen": -172.23770141601562,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -108.0,
      "logps_train/rejected": -154.22283935546875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0136151313781738,
      "rewards_train/margins": 3.5768332481384277,
      "rewards_train/rejected": -4.590448379516602,
      "step": 161
    },
    {
      "epoch": 1.84,
      "logps_train/chosen": -151.36978149414062,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -171.3380126953125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3356115818023682,
      "rewards_train/margins": 3.544039487838745,
      "rewards_train/rejected": -4.879651069641113,
      "step": 161
    },
    {
      "epoch": 1.84,
      "logps_train/chosen": -199.06410217285156,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -194.74569702148438,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1116838455200195,
      "rewards_train/margins": 3.712299346923828,
      "rewards_train/rejected": -4.823983192443848,
      "step": 161
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.6628009695725346e-05,
      "loss": 0.129,
      "step": 162
    },
    {
      "epoch": 1.85,
      "logps_train/chosen": -131.08058166503906,
      "logps_train/ref_chosen": -121.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -166.07493591308594,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0106275081634521,
      "rewards_train/margins": 3.0171782970428467,
      "rewards_train/rejected": -4.027805805206299,
      "step": 162
    },
    {
      "epoch": 1.85,
      "logps_train/chosen": -137.0918426513672,
      "logps_train/ref_chosen": -115.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -155.005859375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.2230069637298584,
      "rewards_train/margins": 2.3856842517852783,
      "rewards_train/rejected": -4.608691215515137,
      "step": 162
    },
    {
      "epoch": 1.85,
      "logps_train/chosen": -135.45196533203125,
      "logps_train/ref_chosen": -121.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -156.66986083984375,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.433404564857483,
      "rewards_train/margins": 2.7299917936325073,
      "rewards_train/rejected": -4.16339635848999,
      "step": 162
    },
    {
      "epoch": 1.85,
      "logps_train/chosen": -134.3225860595703,
      "logps_train/ref_chosen": -114.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -150.91635131835938,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.0572102069854736,
      "rewards_train/margins": 2.305445909500122,
      "rewards_train/rejected": -4.362656116485596,
      "step": 162
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.6336152766230232e-05,
      "loss": 0.1969,
      "step": 163
    },
    {
      "epoch": 1.86,
      "logps_train/chosen": -142.69918823242188,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -166.99905395507812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.13105309009552,
      "rewards_train/margins": 3.8543986082077026,
      "rewards_train/rejected": -4.985451698303223,
      "step": 163
    },
    {
      "epoch": 1.86,
      "logps_train/chosen": -167.85174560546875,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -168.38937377929688,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.7078323364257812,
      "rewards_train/margins": 3.499269485473633,
      "rewards_train/rejected": -5.207101821899414,
      "step": 163
    },
    {
      "epoch": 1.86,
      "logps_train/chosen": -165.04327392578125,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -174.62939453125,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.4283990859985352,
      "rewards_train/margins": 3.1969175338745117,
      "rewards_train/rejected": -4.625316619873047,
      "step": 163
    },
    {
      "epoch": 1.86,
      "logps_train/chosen": -189.43804931640625,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -183.73580932617188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0531786680221558,
      "rewards_train/margins": 3.821672558784485,
      "rewards_train/rejected": -4.874851226806641,
      "step": 163
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.6045631705848404e-05,
      "loss": 0.1373,
      "step": 164
    },
    {
      "epoch": 1.87,
      "logps_train/chosen": -160.4321746826172,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -90.0,
      "logps_train/rejected": -136.58428955078125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9580607414245605,
      "rewards_train/margins": 2.7091808319091797,
      "rewards_train/rejected": -4.66724157333374,
      "step": 164
    },
    {
      "epoch": 1.87,
      "logps_train/chosen": -157.96083068847656,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -179.47161865234375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4216691255569458,
      "rewards_train/margins": 3.2209337949752808,
      "rewards_train/rejected": -4.642602920532227,
      "step": 164
    },
    {
      "epoch": 1.87,
      "logps_train/chosen": -156.0556640625,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -152.37881469726562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3762691020965576,
      "rewards_train/margins": 3.1942293643951416,
      "rewards_train/rejected": -4.570498466491699,
      "step": 164
    },
    {
      "epoch": 1.87,
      "logps_train/chosen": -154.2939453125,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -157.37115478515625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.6712883710861206,
      "rewards_train/margins": 2.8872863054275513,
      "rewards_train/rejected": -4.558574676513672,
      "step": 164
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.5756491309703875e-05,
      "loss": 0.1581,
      "step": 165
    },
    {
      "epoch": 1.89,
      "logps_train/chosen": -135.0899658203125,
      "logps_train/ref_chosen": -121.5,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -153.34164428710938,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.354859709739685,
      "rewards_train/margins": 3.076668381690979,
      "rewards_train/rejected": -4.431528091430664,
      "step": 165
    },
    {
      "epoch": 1.89,
      "logps_train/chosen": -184.4627227783203,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -203.6276397705078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.5243486166000366,
      "rewards_train/margins": 4.656775116920471,
      "rewards_train/rejected": -5.181123733520508,
      "step": 165
    },
    {
      "epoch": 1.89,
      "logps_train/chosen": -161.5297393798828,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -185.64584350585938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4633244276046753,
      "rewards_train/margins": 3.1752835512161255,
      "rewards_train/rejected": -4.638607978820801,
      "step": 165
    },
    {
      "epoch": 1.89,
      "logps_train/chosen": -185.7200164794922,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -151.0,
      "logps_train/rejected": -196.57159423828125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0503220558166504,
      "rewards_train/margins": 3.486915111541748,
      "rewards_train/rejected": -4.537237167358398,
      "step": 165
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.5468776160037556e-05,
      "loss": 0.1205,
      "step": 166
    },
    {
      "epoch": 1.9,
      "logps_train/chosen": -168.28759765625,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -169.3665313720703,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3095204830169678,
      "rewards_train/margins": 2.7026207447052,
      "rewards_train/rejected": -4.012141227722168,
      "step": 166
    },
    {
      "epoch": 1.9,
      "logps_train/chosen": -141.98912048339844,
      "logps_train/ref_chosen": -126.5,
      "logps_train/ref_rejected": -96.5,
      "logps_train/rejected": -143.87429809570312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.543443202972412,
      "rewards_train/margins": 3.1720151901245117,
      "rewards_train/rejected": -4.715458393096924,
      "step": 166
    },
    {
      "epoch": 1.9,
      "logps_train/chosen": -146.0329132080078,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -158.03030395507812,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.224214792251587,
      "rewards_train/margins": 2.7690494060516357,
      "rewards_train/rejected": -3.9932641983032227,
      "step": 166
    },
    {
      "epoch": 1.9,
      "logps_train/chosen": -185.71340942382812,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -182.1860809326172,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5620156526565552,
      "rewards_train/margins": 3.35896098613739,
      "rewards_train/rejected": -4.920976638793945,
      "step": 166
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.5182530619333169e-05,
      "loss": 0.1542,
      "step": 167
    },
    {
      "epoch": 1.91,
      "logps_train/chosen": -149.9785919189453,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -152.57215881347656,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3188555240631104,
      "rewards_train/margins": 3.0212714672088623,
      "rewards_train/rejected": -4.340126991271973,
      "step": 167
    },
    {
      "epoch": 1.91,
      "logps_train/chosen": -168.6658935546875,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -182.68592834472656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.816979169845581,
      "rewards_train/margins": 3.0047385692596436,
      "rewards_train/rejected": -4.821717739105225,
      "step": 167
    },
    {
      "epoch": 1.91,
      "logps_train/chosen": -164.0330810546875,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -181.12750244140625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.766735315322876,
      "rewards_train/margins": 3.4669129848480225,
      "rewards_train/rejected": -5.233648300170898,
      "step": 167
    },
    {
      "epoch": 1.91,
      "logps_train/chosen": -201.2147674560547,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -180.60372924804688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.19296133518219,
      "rewards_train/margins": 4.000518202781677,
      "rewards_train/rejected": -5.193479537963867,
      "step": 167
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.4897798823477043e-05,
      "loss": 0.1167,
      "step": 168
    },
    {
      "epoch": 1.92,
      "logps_train/chosen": -154.00045776367188,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -161.27391052246094,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.832126498222351,
      "rewards_train/margins": 3.5510259866714478,
      "rewards_train/rejected": -5.383152484893799,
      "step": 168
    },
    {
      "epoch": 1.92,
      "logps_train/chosen": -171.59112548828125,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -170.0,
      "logps_train/rejected": -216.93157958984375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3428027629852295,
      "rewards_train/margins": 3.343519449234009,
      "rewards_train/rejected": -4.686322212219238,
      "step": 168
    },
    {
      "epoch": 1.92,
      "logps_train/chosen": -174.04019165039062,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -172.20095825195312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2645671367645264,
      "rewards_train/margins": 3.8841536045074463,
      "rewards_train/rejected": -5.148720741271973,
      "step": 168
    },
    {
      "epoch": 1.92,
      "logps_train/chosen": -169.83084106445312,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -158.2201690673828,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.683249592781067,
      "rewards_train/margins": 3.1385172605514526,
      "rewards_train/rejected": -4.8217668533325195,
      "step": 168
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.4614624674952842e-05,
      "loss": 0.1329,
      "step": 169
    },
    {
      "epoch": 1.93,
      "logps_train/chosen": -177.6451416015625,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -176.8157958984375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3176395893096924,
      "rewards_train/margins": 3.5905025005340576,
      "rewards_train/rejected": -4.90814208984375,
      "step": 169
    },
    {
      "epoch": 1.93,
      "logps_train/chosen": -178.3532257080078,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -161.0,
      "logps_train/rejected": -208.62347412109375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.913349449634552,
      "rewards_train/margins": 3.8398191332817078,
      "rewards_train/rejected": -4.75316858291626,
      "step": 169
    },
    {
      "epoch": 1.93,
      "logps_train/chosen": -160.98016357421875,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -149.50071716308594,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.2862005233764648,
      "rewards_train/margins": 3.435941696166992,
      "rewards_train/rejected": -4.722142219543457,
      "step": 169
    },
    {
      "epoch": 1.93,
      "logps_train/chosen": -214.80264282226562,
      "logps_train/ref_chosen": -205.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -185.12115478515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0412020683288574,
      "rewards_train/margins": 4.142787456512451,
      "rewards_train/rejected": -5.183989524841309,
      "step": 169
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.4333051836072298e-05,
      "loss": 0.1112,
      "step": 170
    },
    {
      "epoch": 1.94,
      "logps_train/chosen": -177.8150634765625,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -182.40084838867188,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -1.512269139289856,
      "rewards_train/margins": 3.168343424797058,
      "rewards_train/rejected": -4.680612564086914,
      "step": 170
    },
    {
      "epoch": 1.94,
      "logps_train/chosen": -157.65682983398438,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -151.7862091064453,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.602914571762085,
      "rewards_train/margins": 3.1795151233673096,
      "rewards_train/rejected": -4.7824296951293945,
      "step": 170
    },
    {
      "epoch": 1.94,
      "logps_train/chosen": -176.32493591308594,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -95.0,
      "logps_train/rejected": -140.84335327148438,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.5847883224487305,
      "rewards_train/margins": 2.9966893196105957,
      "rewards_train/rejected": -4.581477642059326,
      "step": 170
    },
    {
      "epoch": 1.94,
      "logps_train/chosen": -140.81861877441406,
      "logps_train/ref_chosen": -122.5,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -162.77423095703125,
      "rewards_train/accuracies": 0.875,
      "rewards_train/chosen": -1.8218281269073486,
      "rewards_train/margins": 2.9552524089813232,
      "rewards_train/rejected": -4.777080535888672,
      "step": 170
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.405312372224294e-05,
      "loss": 0.1776,
      "step": 171
    },
    {
      "epoch": 1.95,
      "logps_train/chosen": -138.87081909179688,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -153.38121032714844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.489914894104004,
      "rewards_train/margins": 3.1869759559631348,
      "rewards_train/rejected": -4.676890850067139,
      "step": 171
    },
    {
      "epoch": 1.95,
      "logps_train/chosen": -162.38807678222656,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -174.22418212890625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.095973014831543,
      "rewards_train/margins": 3.369903087615967,
      "rewards_train/rejected": -4.46587610244751,
      "step": 171
    },
    {
      "epoch": 1.95,
      "logps_train/chosen": -162.04440307617188,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -163.14654541015625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0364702939987183,
      "rewards_train/margins": 3.1233197450637817,
      "rewards_train/rejected": -4.1597900390625,
      "step": 171
    },
    {
      "epoch": 1.95,
      "logps_train/chosen": -145.47537231445312,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -88.5,
      "logps_train/rejected": -128.49514770507812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4886016845703125,
      "rewards_train/margins": 2.50949764251709,
      "rewards_train/rejected": -3.9980993270874023,
      "step": 171
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.3774883495273985e-05,
      "loss": 0.155,
      "step": 172
    },
    {
      "epoch": 1.97,
      "logps_train/chosen": -164.489013671875,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -96.5,
      "logps_train/rejected": -144.42698669433594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8249762058258057,
      "rewards_train/margins": 2.9649884700775146,
      "rewards_train/rejected": -4.78996467590332,
      "step": 172
    },
    {
      "epoch": 1.97,
      "logps_train/chosen": -149.56802368164062,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -144.60494995117188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.797525405883789,
      "rewards_train/margins": 2.945197105407715,
      "rewards_train/rejected": -4.742722511291504,
      "step": 172
    },
    {
      "epoch": 1.97,
      "logps_train/chosen": -142.54705810546875,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -157.16668701171875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2089290618896484,
      "rewards_train/margins": 3.610962390899658,
      "rewards_train/rejected": -4.819891452789307,
      "step": 172
    },
    {
      "epoch": 1.97,
      "logps_train/chosen": -162.0865936279297,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -163.69024658203125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1669116020202637,
      "rewards_train/margins": 3.1393208503723145,
      "rewards_train/rejected": -4.306232452392578,
      "step": 172
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.3498374056721197e-05,
      "loss": 0.1313,
      "step": 173
    },
    {
      "epoch": 1.98,
      "logps_train/chosen": -166.25057983398438,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -172.05795288085938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.346738338470459,
      "rewards_train/margins": 3.3271727561950684,
      "rewards_train/rejected": -4.673911094665527,
      "step": 173
    },
    {
      "epoch": 1.98,
      "logps_train/chosen": -170.06703186035156,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -181.25033569335938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.454749345779419,
      "rewards_train/margins": 3.6730196475982666,
      "rewards_train/rejected": -5.1277689933776855,
      "step": 173
    },
    {
      "epoch": 1.98,
      "logps_train/chosen": -140.8558349609375,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -145.61558532714844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.617444396018982,
      "rewards_train/margins": 3.549827218055725,
      "rewards_train/rejected": -5.167271614074707,
      "step": 173
    },
    {
      "epoch": 1.98,
      "logps_train/chosen": -215.5902099609375,
      "logps_train/ref_chosen": -202.0,
      "logps_train/ref_rejected": -167.0,
      "logps_train/rejected": -217.3975067138672,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4094116687774658,
      "rewards_train/margins": 3.6104161739349365,
      "rewards_train/rejected": -5.019827842712402,
      "step": 173
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.3223638041271979e-05,
      "loss": 0.1236,
      "step": 174
    },
    {
      "epoch": 1.99,
      "logps_train/chosen": -196.11367797851562,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -175.70323181152344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3418858051300049,
      "rewards_train/margins": 3.1817080974578857,
      "rewards_train/rejected": -4.523593902587891,
      "step": 174
    },
    {
      "epoch": 1.99,
      "logps_train/chosen": -142.29310607910156,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -174.541015625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3043105602264404,
      "rewards_train/margins": 3.299423933029175,
      "rewards_train/rejected": -4.603734493255615,
      "step": 174
    },
    {
      "epoch": 1.99,
      "logps_train/chosen": -172.05809020996094,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -163.4266357421875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.619040846824646,
      "rewards_train/margins": 3.089199662208557,
      "rewards_train/rejected": -4.708240509033203,
      "step": 174
    },
    {
      "epoch": 1.99,
      "logps_train/chosen": -171.18453979492188,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -167.75234985351562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.761239767074585,
      "rewards_train/margins": 3.911907911300659,
      "rewards_train/rejected": -4.673147678375244,
      "step": 174
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.2950717810171558e-05,
      "loss": 0.124,
      "step": 175
    },
    {
      "epoch": 2.0,
      "logps_train/chosen": -178.75906372070312,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -108.5,
      "logps_train/rejected": -158.2779998779297,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.51145339012146,
      "rewards_train/margins": 3.475623846054077,
      "rewards_train/rejected": -4.987077236175537,
      "step": 175
    },
    {
      "epoch": 2.0,
      "logps_train/chosen": -144.22889709472656,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -159.568603515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2451064586639404,
      "rewards_train/margins": 3.9777700901031494,
      "rewards_train/rejected": -5.22287654876709,
      "step": 175
    },
    {
      "epoch": 2.0,
      "logps_train/chosen": -159.3805694580078,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -156.07077026367188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3049516677856445,
      "rewards_train/margins": 3.2286882400512695,
      "rewards_train/rejected": -4.533639907836914,
      "step": 175
    },
    {
      "epoch": 2.0,
      "logps_train/chosen": -160.80697631835938,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -182.2629852294922,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6911461353302002,
      "rewards_train/margins": 3.7947232723236084,
      "rewards_train/rejected": -5.485869407653809,
      "step": 175
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.2679655444691369e-05,
      "loss": 0.0755,
      "step": 176
    },
    {
      "epoch": 2.01,
      "logps_train/chosen": -161.154052734375,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -167.23745727539062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2068119049072266,
      "rewards_train/margins": 3.918104648590088,
      "rewards_train/rejected": -5.1249165534973145,
      "step": 176
    },
    {
      "epoch": 2.01,
      "logps_train/chosen": -145.90554809570312,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -144.0738525390625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.574514627456665,
      "rewards_train/margins": 3.397310972213745,
      "rewards_train/rejected": -4.97182559967041,
      "step": 176
    },
    {
      "epoch": 2.01,
      "logps_train/chosen": -192.4039306640625,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -172.21121215820312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.8327764272689819,
      "rewards_train/margins": 4.236098408699036,
      "rewards_train/rejected": -5.068874835968018,
      "step": 176
    },
    {
      "epoch": 2.01,
      "logps_train/chosen": -153.53134155273438,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -175.02682495117188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4470796585083008,
      "rewards_train/margins": 4.318201065063477,
      "rewards_train/rejected": -5.765280723571777,
      "step": 176
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.2410492739640592e-05,
      "loss": 0.085,
      "step": 177
    },
    {
      "epoch": 2.02,
      "logps_train/chosen": -191.07339477539062,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -188.9764404296875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.7616373896598816,
      "rewards_train/margins": 4.577509701251984,
      "rewards_train/rejected": -5.339147090911865,
      "step": 177
    },
    {
      "epoch": 2.02,
      "logps_train/chosen": -185.62136840820312,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -184.6957550048828,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6105749011039734,
      "rewards_train/margins": 4.272965252399445,
      "rewards_train/rejected": -4.883540153503418,
      "step": 177
    },
    {
      "epoch": 2.02,
      "logps_train/chosen": -191.00115966796875,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -179.1207733154297,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.8412289619445801,
      "rewards_train/margins": 3.8856916427612305,
      "rewards_train/rejected": -4.7269206047058105,
      "step": 177
    },
    {
      "epoch": 2.02,
      "logps_train/chosen": -163.87918090820312,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -188.97149658203125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.558229923248291,
      "rewards_train/margins": 3.8317909240722656,
      "rewards_train/rejected": -5.390020847320557,
      "step": 177
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.2143271196921831e-05,
      "loss": 0.0734,
      "step": 178
    },
    {
      "epoch": 2.03,
      "logps_train/chosen": -149.78662109375,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -154.28677368164062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2491698265075684,
      "rewards_train/margins": 3.2221827507019043,
      "rewards_train/rejected": -4.471352577209473,
      "step": 178
    },
    {
      "epoch": 2.03,
      "logps_train/chosen": -180.86485290527344,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -168.6180419921875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9179306030273438,
      "rewards_train/margins": 3.6434578895568848,
      "rewards_train/rejected": -4.5613884925842285,
      "step": 178
    },
    {
      "epoch": 2.03,
      "logps_train/chosen": -191.31576538085938,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -154.0,
      "logps_train/rejected": -205.51625061035156,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.9044272899627686,
      "rewards_train/margins": 4.209453344345093,
      "rewards_train/rejected": -5.113880634307861,
      "step": 178
    },
    {
      "epoch": 2.03,
      "logps_train/chosen": -157.76119995117188,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -172.5016632080078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2712382078170776,
      "rewards_train/margins": 3.9011937379837036,
      "rewards_train/rejected": -5.172431945800781,
      "step": 178
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.1878032019132016e-05,
      "loss": 0.0712,
      "step": 179
    },
    {
      "epoch": 2.05,
      "logps_train/chosen": -179.49224853515625,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -168.34971618652344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9605532884597778,
      "rewards_train/margins": 4.291605830192566,
      "rewards_train/rejected": -5.252159118652344,
      "step": 179
    },
    {
      "epoch": 2.05,
      "logps_train/chosen": -182.27786254882812,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -147.0,
      "logps_train/rejected": -195.55474853515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": 0.051900215446949005,
      "rewards_train/margins": 4.886940307915211,
      "rewards_train/rejected": -4.835040092468262,
      "step": 179
    },
    {
      "epoch": 2.05,
      "logps_train/chosen": -144.2906494140625,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -156.61428833007812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2723876237869263,
      "rewards_train/margins": 3.064677119255066,
      "rewards_train/rejected": -4.337064743041992,
      "step": 179
    },
    {
      "epoch": 2.05,
      "logps_train/chosen": -184.02662658691406,
      "logps_train/ref_chosen": -177.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -187.46713256835938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.6745372414588928,
      "rewards_train/margins": 4.093368470668793,
      "rewards_train/rejected": -4.7679057121276855,
      "step": 179
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.1614816103209363e-05,
      "loss": 0.0749,
      "step": 180
    },
    {
      "epoch": 2.06,
      "logps_train/chosen": -152.71478271484375,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -169.14288330078125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3001878261566162,
      "rewards_train/margins": 3.4307992458343506,
      "rewards_train/rejected": -4.730987071990967,
      "step": 180
    },
    {
      "epoch": 2.06,
      "logps_train/chosen": -175.60665893554688,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -172.89309692382812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.662753701210022,
      "rewards_train/margins": 3.5386651754379272,
      "rewards_train/rejected": -5.201418876647949,
      "step": 180
    },
    {
      "epoch": 2.06,
      "logps_train/chosen": -145.56417846679688,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -156.00181579589844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1640660762786865,
      "rewards_train/margins": 3.8189284801483154,
      "rewards_train/rejected": -4.982994556427002,
      "step": 180
    },
    {
      "epoch": 2.06,
      "logps_train/chosen": -172.25607299804688,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -152.76123046875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -0.6297088265419006,
      "rewards_train/margins": 4.331765592098236,
      "rewards_train/rejected": -4.961474418640137,
      "step": 180
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.1353664034127583e-05,
      "loss": 0.0793,
      "step": 181
    },
    {
      "epoch": 2.07,
      "logps_train/chosen": -149.04946899414062,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -154.0411376953125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2665691375732422,
      "rewards_train/margins": 3.5102977752685547,
      "rewards_train/rejected": -4.776866912841797,
      "step": 181
    },
    {
      "epoch": 2.07,
      "logps_train/chosen": -158.56723022460938,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -159.50074768066406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.218638300895691,
      "rewards_train/margins": 3.6543132066726685,
      "rewards_train/rejected": -4.872951507568359,
      "step": 181
    },
    {
      "epoch": 2.07,
      "logps_train/chosen": -147.66925048828125,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -168.28912353515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.185003638267517,
      "rewards_train/margins": 4.234533905982971,
      "rewards_train/rejected": -5.419537544250488,
      "step": 181
    },
    {
      "epoch": 2.07,
      "logps_train/chosen": -179.8079833984375,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -197.74644470214844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2968143224716187,
      "rewards_train/margins": 4.06601345539093,
      "rewards_train/rejected": -5.362827777862549,
      "step": 181
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.1094616078638123e-05,
      "loss": 0.0717,
      "step": 182
    },
    {
      "epoch": 2.08,
      "logps_train/chosen": -190.28167724609375,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -181.79185485839844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6578567028045654,
      "rewards_train/margins": 4.172305345535278,
      "rewards_train/rejected": -4.830162048339844,
      "step": 182
    },
    {
      "epoch": 2.08,
      "logps_train/chosen": -170.9873046875,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -172.44073486328125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.5471186637878418,
      "rewards_train/margins": 4.172564506530762,
      "rewards_train/rejected": -4.7196831703186035,
      "step": 182
    },
    {
      "epoch": 2.08,
      "logps_train/chosen": -170.7721405029297,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -159.09173583984375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.270280122756958,
      "rewards_train/margins": 3.9390900135040283,
      "rewards_train/rejected": -5.209370136260986,
      "step": 182
    },
    {
      "epoch": 2.08,
      "logps_train/chosen": -175.2460174560547,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -195.99169921875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.9697037935256958,
      "rewards_train/margins": 3.803244471549988,
      "rewards_train/rejected": -4.772948265075684,
      "step": 182
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.083771217906143e-05,
      "loss": 0.0599,
      "step": 183
    },
    {
      "epoch": 2.09,
      "logps_train/chosen": -190.64321899414062,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -189.67926025390625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.39599961042404175,
      "rewards_train/margins": 4.678371250629425,
      "rewards_train/rejected": -5.074370861053467,
      "step": 183
    },
    {
      "epoch": 2.09,
      "logps_train/chosen": -138.62399291992188,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -144.83941650390625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.0373992919921875,
      "rewards_train/margins": 3.527449131011963,
      "rewards_train/rejected": -4.56484842300415,
      "step": 183
    },
    {
      "epoch": 2.09,
      "logps_train/chosen": -136.85711669921875,
      "logps_train/ref_chosen": -118.5,
      "logps_train/ref_rejected": -84.0,
      "logps_train/rejected": -131.87420654296875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8283874988555908,
      "rewards_train/margins": 2.936718702316284,
      "rewards_train/rejected": -4.765106201171875,
      "step": 183
    },
    {
      "epoch": 2.09,
      "logps_train/chosen": -148.16285705566406,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -186.79367065429688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.6797617077827454,
      "rewards_train/margins": 4.7876922488212585,
      "rewards_train/rejected": -5.467453956604004,
      "step": 183
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.0582991947128324e-05,
      "loss": 0.0863,
      "step": 184
    },
    {
      "epoch": 2.1,
      "logps_train/chosen": -168.52122497558594,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -180.79949951171875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.8917717337608337,
      "rewards_train/margins": 3.991675913333893,
      "rewards_train/rejected": -4.883447647094727,
      "step": 184
    },
    {
      "epoch": 2.1,
      "logps_train/chosen": -143.18820190429688,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -94.0,
      "logps_train/rejected": -146.11032104492188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5001935958862305,
      "rewards_train/margins": 3.6904263496398926,
      "rewards_train/rejected": -5.190619945526123,
      "step": 184
    },
    {
      "epoch": 2.1,
      "logps_train/chosen": -147.32476806640625,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -184.9254150390625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.9073777198791504,
      "rewards_train/margins": 3.7832117080688477,
      "rewards_train/rejected": -5.690589427947998,
      "step": 184
    },
    {
      "epoch": 2.1,
      "logps_train/chosen": -147.30819702148438,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -155.67486572265625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4561140537261963,
      "rewards_train/margins": 3.524458646774292,
      "rewards_train/rejected": -4.980572700500488,
      "step": 184
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.0330494657872312e-05,
      "loss": 0.0853,
      "step": 185
    },
    {
      "epoch": 2.11,
      "logps_train/chosen": -162.71340942382812,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -179.08816528320312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3596220016479492,
      "rewards_train/margins": 3.8323984146118164,
      "rewards_train/rejected": -5.192020416259766,
      "step": 185
    },
    {
      "epoch": 2.11,
      "logps_train/chosen": -162.77186584472656,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -192.93026733398438,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.177870273590088,
      "rewards_train/margins": 4.452266216278076,
      "rewards_train/rejected": -5.630136489868164,
      "step": 185
    },
    {
      "epoch": 2.11,
      "logps_train/chosen": -152.37396240234375,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -195.00845336914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4057559967041016,
      "rewards_train/margins": 4.478293418884277,
      "rewards_train/rejected": -5.884049415588379,
      "step": 185
    },
    {
      "epoch": 2.11,
      "logps_train/chosen": -151.218505859375,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -170.10499572753906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1394281387329102,
      "rewards_train/margins": 3.907008647918701,
      "rewards_train/rejected": -5.046436786651611,
      "step": 185
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.0080259243573789e-05,
      "loss": 0.0513,
      "step": 186
    },
    {
      "epoch": 2.13,
      "logps_train/chosen": -166.97433471679688,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -182.7489471435547,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.403684139251709,
      "rewards_train/margins": 4.301093101501465,
      "rewards_train/rejected": -5.704777240753174,
      "step": 186
    },
    {
      "epoch": 2.13,
      "logps_train/chosen": -131.2030792236328,
      "logps_train/ref_chosen": -119.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -171.1971435546875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.2050743103027344,
      "rewards_train/margins": 3.998037815093994,
      "rewards_train/rejected": -5.2031121253967285,
      "step": 186
    },
    {
      "epoch": 2.13,
      "logps_train/chosen": -153.02066040039062,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -172.2113494873047,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.3235623836517334,
      "rewards_train/margins": 3.6091701984405518,
      "rewards_train/rejected": -4.932732582092285,
      "step": 186
    },
    {
      "epoch": 2.13,
      "logps_train/chosen": -162.50509643554688,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -127.5,
      "logps_train/rejected": -177.67140197753906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.7423069477081299,
      "rewards_train/margins": 4.284990072250366,
      "rewards_train/rejected": -5.027297019958496,
      "step": 186
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.832324287757158e-06,
      "loss": 0.0607,
      "step": 187
    },
    {
      "epoch": 2.14,
      "logps_train/chosen": -145.08290100097656,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -177.64239501953125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1745011806488037,
      "rewards_train/margins": 4.237786531448364,
      "rewards_train/rejected": -5.412287712097168,
      "step": 187
    },
    {
      "epoch": 2.14,
      "logps_train/chosen": -172.32980346679688,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -179.23846435546875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4877278804779053,
      "rewards_train/margins": 3.9437358379364014,
      "rewards_train/rejected": -5.431463718414307,
      "step": 187
    },
    {
      "epoch": 2.14,
      "logps_train/chosen": -153.09422302246094,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -191.51918029785156,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5256340503692627,
      "rewards_train/margins": 4.172573804855347,
      "rewards_train/rejected": -5.698207855224609,
      "step": 187
    },
    {
      "epoch": 2.14,
      "logps_train/chosen": -176.45843505859375,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -108.5,
      "logps_train/rejected": -167.15196228027344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1481870412826538,
      "rewards_train/margins": 4.72482168674469,
      "rewards_train/rejected": -5.873008728027344,
      "step": 187
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.586728019241623e-06,
      "loss": 0.0399,
      "step": 188
    },
    {
      "epoch": 2.15,
      "logps_train/chosen": -154.5926513671875,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -88.5,
      "logps_train/rejected": -142.7076873779297,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.872741937637329,
      "rewards_train/margins": 3.5749804973602295,
      "rewards_train/rejected": -5.447722434997559,
      "step": 188
    },
    {
      "epoch": 2.15,
      "logps_train/chosen": -207.6587371826172,
      "logps_train/ref_chosen": -192.0,
      "logps_train/ref_rejected": -169.0,
      "logps_train/rejected": -234.085693359375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6053264141082764,
      "rewards_train/margins": 4.921602487564087,
      "rewards_train/rejected": -6.526928901672363,
      "step": 188
    },
    {
      "epoch": 2.15,
      "logps_train/chosen": -149.80093383789062,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -179.54971313476562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6450353860855103,
      "rewards_train/margins": 3.782885432243347,
      "rewards_train/rejected": -5.427920818328857,
      "step": 188
    },
    {
      "epoch": 2.15,
      "logps_train/chosen": -163.64334106445312,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -165.87039184570312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0383586883544922,
      "rewards_train/margins": 4.294675827026367,
      "rewards_train/rejected": -5.333034515380859,
      "step": 188
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.343508306246771e-06,
      "loss": 0.0743,
      "step": 189
    },
    {
      "epoch": 2.16,
      "logps_train/chosen": -162.09934997558594,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -95.0,
      "logps_train/rejected": -154.43948364257812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7151105403900146,
      "rewards_train/margins": 4.205791234970093,
      "rewards_train/rejected": -5.920901775360107,
      "step": 189
    },
    {
      "epoch": 2.16,
      "logps_train/chosen": -187.39578247070312,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -198.8822479248047,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.358717918395996,
      "rewards_train/margins": 4.348062515258789,
      "rewards_train/rejected": -5.706780433654785,
      "step": 189
    },
    {
      "epoch": 2.16,
      "logps_train/chosen": -142.24322509765625,
      "logps_train/ref_chosen": -126.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -162.78729248046875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6396063566207886,
      "rewards_train/margins": 3.697179913520813,
      "rewards_train/rejected": -5.336786270141602,
      "step": 189
    },
    {
      "epoch": 2.16,
      "logps_train/chosen": -169.55081176757812,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -203.53314208984375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2864294052124023,
      "rewards_train/margins": 5.046573162078857,
      "rewards_train/rejected": -6.33300256729126,
      "step": 189
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.102702650553671e-06,
      "loss": 0.0536,
      "step": 190
    },
    {
      "epoch": 2.17,
      "logps_train/chosen": -187.0035400390625,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -205.14060974121094,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.8618786334991455,
      "rewards_train/margins": 4.305112600326538,
      "rewards_train/rejected": -5.166991233825684,
      "step": 190
    },
    {
      "epoch": 2.17,
      "logps_train/chosen": -148.8421630859375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -156.6807861328125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.293787956237793,
      "rewards_train/margins": 3.910130023956299,
      "rewards_train/rejected": -5.203917980194092,
      "step": 190
    },
    {
      "epoch": 2.17,
      "logps_train/chosen": -162.02560424804688,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -167.29222106933594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2023648023605347,
      "rewards_train/margins": 4.4929715394973755,
      "rewards_train/rejected": -5.69533634185791,
      "step": 190
    },
    {
      "epoch": 2.17,
      "logps_train/chosen": -149.6796417236328,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -193.026611328125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.1812937259674072,
      "rewards_train/margins": 4.417267084121704,
      "rewards_train/rejected": -5.598560810089111,
      "step": 190
    },
    {
      "epoch": 2.18,
      "learning_rate": 8.864348181722559e-06,
      "loss": 0.0645,
      "step": 191
    },
    {
      "epoch": 2.18,
      "logps_train/chosen": -165.83389282226562,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -104.0,
      "logps_train/rejected": -161.61724853515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1698150634765625,
      "rewards_train/margins": 3.5971102714538574,
      "rewards_train/rejected": -5.76692533493042,
      "step": 191
    },
    {
      "epoch": 2.18,
      "logps_train/chosen": -139.57894897460938,
      "logps_train/ref_chosen": -124.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -169.34829711914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.577218770980835,
      "rewards_train/margins": 4.069085359573364,
      "rewards_train/rejected": -5.646304130554199,
      "step": 191
    },
    {
      "epoch": 2.18,
      "logps_train/chosen": -188.30484008789062,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -201.06988525390625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2504067420959473,
      "rewards_train/margins": 4.388809680938721,
      "rewards_train/rejected": -5.639216423034668,
      "step": 191
    },
    {
      "epoch": 2.18,
      "logps_train/chosen": -193.16360473632812,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -201.50115966796875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3569847345352173,
      "rewards_train/margins": 4.465201735496521,
      "rewards_train/rejected": -5.822186470031738,
      "step": 191
    },
    {
      "epoch": 2.19,
      "learning_rate": 8.628481651367876e-06,
      "loss": 0.0616,
      "step": 192
    },
    {
      "epoch": 2.19,
      "logps_train/chosen": -127.2575912475586,
      "logps_train/ref_chosen": -104.5,
      "logps_train/ref_rejected": -89.0,
      "logps_train/rejected": -145.93881225585938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.2746853828430176,
      "rewards_train/margins": 3.400190830230713,
      "rewards_train/rejected": -5.6748762130737305,
      "step": 192
    },
    {
      "epoch": 2.19,
      "logps_train/chosen": -160.23959350585938,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -163.14401245117188,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8308442831039429,
      "rewards_train/margins": 4.124375939369202,
      "rewards_train/rejected": -5.9552202224731445,
      "step": 192
    },
    {
      "epoch": 2.19,
      "logps_train/chosen": -203.55633544921875,
      "logps_train/ref_chosen": -195.0,
      "logps_train/ref_rejected": -159.0,
      "logps_train/rejected": -215.8226776123047,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.8862971067428589,
      "rewards_train/margins": 4.777416348457336,
      "rewards_train/rejected": -5.663713455200195,
      "step": 192
    },
    {
      "epoch": 2.19,
      "logps_train/chosen": -154.07431030273438,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -111.5,
      "logps_train/rejected": -165.9130859375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1087489128112793,
      "rewards_train/margins": 4.343350887298584,
      "rewards_train/rejected": -5.452099800109863,
      "step": 192
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.395139427491517e-06,
      "loss": 0.0737,
      "step": 193
    },
    {
      "epoch": 2.21,
      "logps_train/chosen": -161.2333526611328,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -172.9558868408203,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.696040391921997,
      "rewards_train/margins": 3.8845107555389404,
      "rewards_train/rejected": -5.5805511474609375,
      "step": 193
    },
    {
      "epoch": 2.21,
      "logps_train/chosen": -179.02362060546875,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -207.66558837890625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.911834478378296,
      "rewards_train/margins": 4.097449064254761,
      "rewards_train/rejected": -6.009283542633057,
      "step": 193
    },
    {
      "epoch": 2.21,
      "logps_train/chosen": -180.44769287109375,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -193.8340301513672,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.302825689315796,
      "rewards_train/margins": 4.537803888320923,
      "rewards_train/rejected": -5.840629577636719,
      "step": 193
    },
    {
      "epoch": 2.21,
      "logps_train/chosen": -169.3468017578125,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -172.2258758544922,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8215930461883545,
      "rewards_train/margins": 4.061346769332886,
      "rewards_train/rejected": -5.88293981552124,
      "step": 193
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.164357488875348e-06,
      "loss": 0.0566,
      "step": 194
    },
    {
      "epoch": 2.22,
      "logps_train/chosen": -200.07351684570312,
      "logps_train/ref_chosen": -184.0,
      "logps_train/ref_rejected": -151.0,
      "logps_train/rejected": -209.89173889160156,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5717073678970337,
      "rewards_train/margins": 4.342369437217712,
      "rewards_train/rejected": -5.914076805114746,
      "step": 194
    },
    {
      "epoch": 2.22,
      "logps_train/chosen": -171.31915283203125,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -182.50357055664062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.122978925704956,
      "rewards_train/margins": 4.234654664993286,
      "rewards_train/rejected": -5.357633590698242,
      "step": 194
    },
    {
      "epoch": 2.22,
      "logps_train/chosen": -115.46754455566406,
      "logps_train/ref_chosen": -97.5,
      "logps_train/ref_rejected": -80.0,
      "logps_train/rejected": -135.1890106201172,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8070234060287476,
      "rewards_train/margins": 3.7197171449661255,
      "rewards_train/rejected": -5.526740550994873,
      "step": 194
    },
    {
      "epoch": 2.22,
      "logps_train/chosen": -158.11651611328125,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -179.56265258789062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8428046703338623,
      "rewards_train/margins": 4.047834157943726,
      "rewards_train/rejected": -5.890638828277588,
      "step": 194
    },
    {
      "epoch": 2.23,
      "learning_rate": 7.936171419533653e-06,
      "loss": 0.0772,
      "step": 195
    },
    {
      "epoch": 2.23,
      "logps_train/chosen": -175.13031005859375,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -89.5,
      "logps_train/rejected": -145.96676635742188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.54506254196167,
      "rewards_train/margins": 4.095754146575928,
      "rewards_train/rejected": -5.640816688537598,
      "step": 195
    },
    {
      "epoch": 2.23,
      "logps_train/chosen": -173.42425537109375,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -184.82421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6176198720932007,
      "rewards_train/margins": 4.269880890846252,
      "rewards_train/rejected": -5.887500762939453,
      "step": 195
    },
    {
      "epoch": 2.23,
      "logps_train/chosen": -196.7584686279297,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -205.75381469726562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.33219575881958,
      "rewards_train/margins": 4.877925395965576,
      "rewards_train/rejected": -6.210121154785156,
      "step": 195
    },
    {
      "epoch": 2.23,
      "logps_train/chosen": -184.88011169433594,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -196.70294189453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7086167335510254,
      "rewards_train/margins": 4.16851282119751,
      "rewards_train/rejected": -5.877129554748535,
      "step": 195
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.710616403226459e-06,
      "loss": 0.0644,
      "step": 196
    },
    {
      "epoch": 2.24,
      "logps_train/chosen": -154.9520263671875,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -183.3641815185547,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9811406135559082,
      "rewards_train/margins": 4.288285732269287,
      "rewards_train/rejected": -6.269426345825195,
      "step": 196
    },
    {
      "epoch": 2.24,
      "logps_train/chosen": -180.44895935058594,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -169.0,
      "logps_train/rejected": -230.28929138183594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -0.85615074634552,
      "rewards_train/margins": 5.313965678215027,
      "rewards_train/rejected": -6.170116424560547,
      "step": 196
    },
    {
      "epoch": 2.24,
      "logps_train/chosen": -158.23910522460938,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -179.76210021972656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5342793464660645,
      "rewards_train/margins": 3.97666597366333,
      "rewards_train/rejected": -5.5109453201293945,
      "step": 196
    },
    {
      "epoch": 2.24,
      "logps_train/chosen": -148.6269073486328,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -161.82305908203125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.707613229751587,
      "rewards_train/margins": 4.167392015457153,
      "rewards_train/rejected": -5.87500524520874,
      "step": 196
    },
    {
      "epoch": 2.25,
      "learning_rate": 7.487727218034646e-06,
      "loss": 0.0777,
      "step": 197
    },
    {
      "epoch": 2.25,
      "logps_train/chosen": -184.1099853515625,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -201.27886962890625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.521935224533081,
      "rewards_train/margins": 4.687398195266724,
      "rewards_train/rejected": -6.209333419799805,
      "step": 197
    },
    {
      "epoch": 2.25,
      "logps_train/chosen": -162.78753662109375,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -185.3925323486328,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8591243028640747,
      "rewards_train/margins": 4.596633791923523,
      "rewards_train/rejected": -6.455758094787598,
      "step": 197
    },
    {
      "epoch": 2.25,
      "logps_train/chosen": -155.07752990722656,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -191.83187866210938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9492814540863037,
      "rewards_train/margins": 4.413836717605591,
      "rewards_train/rejected": -6.3631181716918945,
      "step": 197
    },
    {
      "epoch": 2.25,
      "logps_train/chosen": -158.76815795898438,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -183.657958984375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2400364875793457,
      "rewards_train/margins": 3.898550510406494,
      "rewards_train/rejected": -6.13858699798584,
      "step": 197
    },
    {
      "epoch": 2.26,
      "learning_rate": 7.267538230997487e-06,
      "loss": 0.065,
      "step": 198
    },
    {
      "epoch": 2.26,
      "logps_train/chosen": -154.61886596679688,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -165.445556640625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5989946126937866,
      "rewards_train/margins": 3.7381139993667603,
      "rewards_train/rejected": -5.337108612060547,
      "step": 198
    },
    {
      "epoch": 2.26,
      "logps_train/chosen": -174.43276977539062,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -194.56985473632812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.813735842704773,
      "rewards_train/margins": 3.821837544441223,
      "rewards_train/rejected": -5.635573387145996,
      "step": 198
    },
    {
      "epoch": 2.26,
      "logps_train/chosen": -171.7181396484375,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -204.68441772460938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9548206329345703,
      "rewards_train/margins": 4.630368232727051,
      "rewards_train/rejected": -6.585188865661621,
      "step": 198
    },
    {
      "epoch": 2.26,
      "logps_train/chosen": -156.4467010498047,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -171.12014770507812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.553849220275879,
      "rewards_train/margins": 3.99556827545166,
      "rewards_train/rejected": -5.549417495727539,
      "step": 198
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.05008339281365e-06,
      "loss": 0.079,
      "step": 199
    },
    {
      "epoch": 2.27,
      "logps_train/chosen": -201.37559509277344,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -159.0,
      "logps_train/rejected": -222.2662811279297,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2338494062423706,
      "rewards_train/margins": 5.101566910743713,
      "rewards_train/rejected": -6.335416316986084,
      "step": 199
    },
    {
      "epoch": 2.27,
      "logps_train/chosen": -155.3743438720703,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -185.0620880126953,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0335280895233154,
      "rewards_train/margins": 4.135473966598511,
      "rewards_train/rejected": -6.169002056121826,
      "step": 199
    },
    {
      "epoch": 2.27,
      "logps_train/chosen": -214.00473022460938,
      "logps_train/ref_chosen": -203.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -188.80557250976562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.0920145511627197,
      "rewards_train/margins": 4.752117395401001,
      "rewards_train/rejected": -5.844131946563721,
      "step": 199
    },
    {
      "epoch": 2.27,
      "logps_train/chosen": -157.0347900390625,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -181.71475219726562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7769157886505127,
      "rewards_train/margins": 4.140556573867798,
      "rewards_train/rejected": -5.9174723625183105,
      "step": 199
    },
    {
      "epoch": 2.29,
      "learning_rate": 6.835396232606414e-06,
      "loss": 0.0586,
      "step": 200
    },
    {
      "epoch": 2.29,
      "logps_train/chosen": -161.72032165527344,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -177.61468505859375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.022033214569092,
      "rewards_train/margins": 4.182990550994873,
      "rewards_train/rejected": -6.205023765563965,
      "step": 200
    },
    {
      "epoch": 2.29,
      "logps_train/chosen": -134.15977478027344,
      "logps_train/ref_chosen": -110.5,
      "logps_train/ref_rejected": -97.5,
      "logps_train/rejected": -164.42269897460938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.3491806983947754,
      "rewards_train/margins": 4.333714962005615,
      "rewards_train/rejected": -6.682895660400391,
      "step": 200
    },
    {
      "epoch": 2.29,
      "logps_train/chosen": -177.84542846679688,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -192.08859252929688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0337610244750977,
      "rewards_train/margins": 4.461132049560547,
      "rewards_train/rejected": -6.4948930740356445,
      "step": 200
    },
    {
      "epoch": 2.29,
      "logps_train/chosen": -196.11642456054688,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -202.897705078125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6030492782592773,
      "rewards_train/margins": 4.695511817932129,
      "rewards_train/rejected": -6.298561096191406,
      "step": 200
    },
    {
      "epoch": 2.3,
      "learning_rate": 6.623509852753798e-06,
      "loss": 0.0547,
      "step": 201
    },
    {
      "epoch": 2.3,
      "logps_train/chosen": -207.26229858398438,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -145.0,
      "logps_train/rejected": -211.3671875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9156838655471802,
      "rewards_train/margins": 4.669862151145935,
      "rewards_train/rejected": -6.585546016693115,
      "step": 201
    },
    {
      "epoch": 2.3,
      "logps_train/chosen": -154.91036987304688,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -166.4049072265625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6626181602478027,
      "rewards_train/margins": 4.285686016082764,
      "rewards_train/rejected": -5.948304176330566,
      "step": 201
    },
    {
      "epoch": 2.3,
      "logps_train/chosen": -140.320556640625,
      "logps_train/ref_chosen": -120.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -179.07057189941406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0460212230682373,
      "rewards_train/margins": 4.754591226577759,
      "rewards_train/rejected": -6.800612449645996,
      "step": 201
    },
    {
      "epoch": 2.3,
      "logps_train/chosen": -167.5859832763672,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -170.61170959472656,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4132864475250244,
      "rewards_train/margins": 4.2639000415802,
      "rewards_train/rejected": -5.677186489105225,
      "step": 201
    },
    {
      "epoch": 2.31,
      "learning_rate": 6.414456923784593e-06,
      "loss": 0.0495,
      "step": 202
    },
    {
      "epoch": 2.31,
      "logps_train/chosen": -174.45191955566406,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -183.5615234375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.61784827709198,
      "rewards_train/margins": 4.5230690240859985,
      "rewards_train/rejected": -6.1409173011779785,
      "step": 202
    },
    {
      "epoch": 2.31,
      "logps_train/chosen": -164.99447631835938,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -214.52035522460938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3355815410614014,
      "rewards_train/margins": 4.774393320083618,
      "rewards_train/rejected": -6.1099748611450195,
      "step": 202
    },
    {
      "epoch": 2.31,
      "logps_train/chosen": -186.1416778564453,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -173.00531005859375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.7241289615631104,
      "rewards_train/margins": 3.9765965938568115,
      "rewards_train/rejected": -5.700725555419922,
      "step": 202
    },
    {
      "epoch": 2.31,
      "logps_train/chosen": -183.17779541015625,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -193.62240600585938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.05098295211792,
      "rewards_train/margins": 4.535184383392334,
      "rewards_train/rejected": -6.586167335510254,
      "step": 202
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.208269679340886e-06,
      "loss": 0.0549,
      "step": 203
    },
    {
      "epoch": 2.32,
      "logps_train/chosen": -164.665283203125,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -113.5,
      "logps_train/rejected": -175.95152282714844,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9498772621154785,
      "rewards_train/margins": 4.297813415527344,
      "rewards_train/rejected": -6.247690677642822,
      "step": 203
    },
    {
      "epoch": 2.32,
      "logps_train/chosen": -179.72842407226562,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -199.34719848632812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5472551584243774,
      "rewards_train/margins": 4.814515471458435,
      "rewards_train/rejected": -6.3617706298828125,
      "step": 203
    },
    {
      "epoch": 2.32,
      "logps_train/chosen": -139.86717224121094,
      "logps_train/ref_chosen": -115.5,
      "logps_train/ref_rejected": -98.5,
      "logps_train/rejected": -162.47723388671875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4136698246002197,
      "rewards_train/margins": 3.998114824295044,
      "rewards_train/rejected": -6.411784648895264,
      "step": 203
    },
    {
      "epoch": 2.32,
      "logps_train/chosen": -152.8355712890625,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -174.1539306640625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8427374362945557,
      "rewards_train/margins": 4.584960222244263,
      "rewards_train/rejected": -6.427697658538818,
      "step": 203
    },
    {
      "epoch": 2.33,
      "learning_rate": 6.004979911208006e-06,
      "loss": 0.0592,
      "step": 204
    },
    {
      "epoch": 2.33,
      "logps_train/chosen": -150.49539184570312,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -201.37484741210938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8733186721801758,
      "rewards_train/margins": 4.557817459106445,
      "rewards_train/rejected": -6.431136131286621,
      "step": 204
    },
    {
      "epoch": 2.33,
      "logps_train/chosen": -145.48204040527344,
      "logps_train/ref_chosen": -119.5,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -165.45367431640625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.6023051738739014,
      "rewards_train/margins": 3.6914026737213135,
      "rewards_train/rejected": -6.293707847595215,
      "step": 204
    },
    {
      "epoch": 2.33,
      "logps_train/chosen": -163.88088989257812,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -191.28944396972656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.48164439201355,
      "rewards_train/margins": 4.2371437549591064,
      "rewards_train/rejected": -6.718788146972656,
      "step": 204
    },
    {
      "epoch": 2.33,
      "logps_train/chosen": -130.41600036621094,
      "logps_train/ref_chosen": -109.0,
      "logps_train/ref_rejected": -98.5,
      "logps_train/rejected": -160.140869140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.111424207687378,
      "rewards_train/margins": 4.071851015090942,
      "rewards_train/rejected": -6.18327522277832,
      "step": 204
    },
    {
      "epoch": 2.34,
      "learning_rate": 5.804618964412586e-06,
      "loss": 0.0761,
      "step": 205
    },
    {
      "epoch": 2.34,
      "logps_train/chosen": -164.7631072998047,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -161.48898315429688,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.1265063285827637,
      "rewards_train/margins": 3.8559865951538086,
      "rewards_train/rejected": -5.982492923736572,
      "step": 205
    },
    {
      "epoch": 2.34,
      "logps_train/chosen": -169.73573303222656,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -160.00714111328125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.167128086090088,
      "rewards_train/margins": 3.958268642425537,
      "rewards_train/rejected": -6.125396728515625,
      "step": 205
    },
    {
      "epoch": 2.34,
      "logps_train/chosen": -210.82345581054688,
      "logps_train/ref_chosen": -191.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -185.9019012451172,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9597394466400146,
      "rewards_train/margins": 4.313043832778931,
      "rewards_train/rejected": -6.272783279418945,
      "step": 205
    },
    {
      "epoch": 2.34,
      "logps_train/chosen": -204.55963134765625,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -194.1874237060547,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.501666784286499,
      "rewards_train/margins": 4.02088475227356,
      "rewards_train/rejected": -6.522551536560059,
      "step": 205
    },
    {
      "epoch": 2.35,
      "learning_rate": 5.607217732389503e-06,
      "loss": 0.0711,
      "step": 206
    },
    {
      "epoch": 2.35,
      "logps_train/chosen": -168.4705352783203,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -167.60079956054688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6452956199645996,
      "rewards_train/margins": 4.4018449783325195,
      "rewards_train/rejected": -6.047140598297119,
      "step": 206
    },
    {
      "epoch": 2.35,
      "logps_train/chosen": -189.11264038085938,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -184.2760772705078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1286468505859375,
      "rewards_train/margins": 4.595640182495117,
      "rewards_train/rejected": -6.724287033081055,
      "step": 206
    },
    {
      "epoch": 2.35,
      "logps_train/chosen": -167.8518829345703,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -179.5463104248047,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.855109453201294,
      "rewards_train/margins": 4.18018651008606,
      "rewards_train/rejected": -6.0352959632873535,
      "step": 206
    },
    {
      "epoch": 2.35,
      "logps_train/chosen": -165.7491912841797,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -152.43096923828125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.695818305015564,
      "rewards_train/margins": 4.109387040138245,
      "rewards_train/rejected": -5.805205345153809,
      "step": 206
    },
    {
      "epoch": 2.37,
      "learning_rate": 5.412806652218469e-06,
      "loss": 0.0656,
      "step": 207
    },
    {
      "epoch": 2.37,
      "logps_train/chosen": -169.99899291992188,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -170.46067810058594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8358477354049683,
      "rewards_train/margins": 4.719790577888489,
      "rewards_train/rejected": -6.555638313293457,
      "step": 207
    },
    {
      "epoch": 2.37,
      "logps_train/chosen": -193.06130981445312,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -196.26321411132812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2590603828430176,
      "rewards_train/margins": 4.279368877410889,
      "rewards_train/rejected": -6.538429260253906,
      "step": 207
    },
    {
      "epoch": 2.37,
      "logps_train/chosen": -193.26812744140625,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -202.84500122070312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.311529517173767,
      "rewards_train/margins": 5.111006379127502,
      "rewards_train/rejected": -6.4225358963012695,
      "step": 207
    },
    {
      "epoch": 2.37,
      "logps_train/chosen": -148.03170776367188,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -186.12869262695312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.667038917541504,
      "rewards_train/margins": 4.782499313354492,
      "rewards_train/rejected": -6.449538230895996,
      "step": 207
    },
    {
      "epoch": 2.38,
      "learning_rate": 5.221415699930951e-06,
      "loss": 0.0451,
      "step": 208
    },
    {
      "epoch": 2.38,
      "logps_train/chosen": -194.32412719726562,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -189.37872314453125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.066347599029541,
      "rewards_train/margins": 4.420987129211426,
      "rewards_train/rejected": -6.487334728240967,
      "step": 208
    },
    {
      "epoch": 2.38,
      "logps_train/chosen": -185.17953491210938,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -201.2667236328125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.35320782661438,
      "rewards_train/margins": 4.4406516551971436,
      "rewards_train/rejected": -6.793859481811523,
      "step": 208
    },
    {
      "epoch": 2.38,
      "logps_train/chosen": -205.48251342773438,
      "logps_train/ref_chosen": -195.0,
      "logps_train/ref_rejected": -160.0,
      "logps_train/rejected": -219.9037628173828,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.028524398803711,
      "rewards_train/margins": 5.0177106857299805,
      "rewards_train/rejected": -6.046235084533691,
      "step": 208
    },
    {
      "epoch": 2.38,
      "logps_train/chosen": -159.63363647460938,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -179.73924255371094,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.698287844657898,
      "rewards_train/margins": 4.654713749885559,
      "rewards_train/rejected": -6.353001594543457,
      "step": 208
    },
    {
      "epoch": 2.39,
      "learning_rate": 5.033074385888189e-06,
      "loss": 0.0496,
      "step": 209
    },
    {
      "epoch": 2.39,
      "logps_train/chosen": -171.902099609375,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -179.21429443359375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.372437000274658,
      "rewards_train/margins": 4.280241966247559,
      "rewards_train/rejected": -6.652678966522217,
      "step": 209
    },
    {
      "epoch": 2.39,
      "logps_train/chosen": -148.18333435058594,
      "logps_train/ref_chosen": -132.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -195.27471923828125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6556390523910522,
      "rewards_train/margins": 4.759748101234436,
      "rewards_train/rejected": -6.415387153625488,
      "step": 209
    },
    {
      "epoch": 2.39,
      "logps_train/chosen": -197.36880493164062,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -206.9953155517578,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.864400863647461,
      "rewards_train/margins": 5.523899555206299,
      "rewards_train/rejected": -7.38830041885376,
      "step": 209
    },
    {
      "epoch": 2.39,
      "logps_train/chosen": -198.30517578125,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -149.0,
      "logps_train/rejected": -218.72198486328125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.7881360054016113,
      "rewards_train/margins": 4.146368980407715,
      "rewards_train/rejected": -6.934504985809326,
      "step": 209
    },
    {
      "epoch": 2.4,
      "learning_rate": 4.847811750231057e-06,
      "loss": 0.0851,
      "step": 210
    },
    {
      "epoch": 2.4,
      "logps_train/chosen": -196.61697387695312,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -168.0,
      "logps_train/rejected": -235.59994506835938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0059356689453125,
      "rewards_train/margins": 4.734723091125488,
      "rewards_train/rejected": -6.740658760070801,
      "step": 210
    },
    {
      "epoch": 2.4,
      "logps_train/chosen": -177.05670166015625,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -194.96536254882812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9225655794143677,
      "rewards_train/margins": 4.452094912528992,
      "rewards_train/rejected": -6.374660491943359,
      "step": 210
    },
    {
      "epoch": 2.4,
      "logps_train/chosen": -209.01028442382812,
      "logps_train/ref_chosen": -187.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -218.34353637695312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1807167530059814,
      "rewards_train/margins": 4.460277795791626,
      "rewards_train/rejected": -6.640994548797607,
      "step": 210
    },
    {
      "epoch": 2.4,
      "logps_train/chosen": -218.32777404785156,
      "logps_train/ref_chosen": -199.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -211.5435333251953,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9003549814224243,
      "rewards_train/margins": 4.402435660362244,
      "rewards_train/rejected": -6.302790641784668,
      "step": 210
    },
    {
      "epoch": 2.41,
      "learning_rate": 4.665656358402395e-06,
      "loss": 0.0516,
      "step": 211
    },
    {
      "epoch": 2.41,
      "logps_train/chosen": -173.509521484375,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -179.2303009033203,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8173097372055054,
      "rewards_train/margins": 4.784577965736389,
      "rewards_train/rejected": -6.6018877029418945,
      "step": 211
    },
    {
      "epoch": 2.41,
      "logps_train/chosen": -172.63043212890625,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -187.04351806640625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2126522064208984,
      "rewards_train/margins": 4.117969512939453,
      "rewards_train/rejected": -6.330621719360352,
      "step": 211
    },
    {
      "epoch": 2.41,
      "logps_train/chosen": -221.95367431640625,
      "logps_train/ref_chosen": -201.0,
      "logps_train/ref_rejected": -170.0,
      "logps_train/rejected": -236.30203247070312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0931222438812256,
      "rewards_train/margins": 4.478243589401245,
      "rewards_train/rejected": -6.571365833282471,
      "step": 211
    },
    {
      "epoch": 2.41,
      "logps_train/chosen": -185.05625915527344,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -126.5,
      "logps_train/rejected": -186.24945068359375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3360211849212646,
      "rewards_train/margins": 4.64351487159729,
      "rewards_train/rejected": -5.979536056518555,
      "step": 211
    },
    {
      "epoch": 2.42,
      "learning_rate": 4.486636296742506e-06,
      "loss": 0.0509,
      "step": 212
    },
    {
      "epoch": 2.42,
      "logps_train/chosen": -193.8851318359375,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -187.48406982421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2984743118286133,
      "rewards_train/margins": 4.142901420593262,
      "rewards_train/rejected": -6.441375732421875,
      "step": 212
    },
    {
      "epoch": 2.42,
      "logps_train/chosen": -180.93228149414062,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -183.302978515625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5795819759368896,
      "rewards_train/margins": 4.18977952003479,
      "rewards_train/rejected": -5.76936149597168,
      "step": 212
    },
    {
      "epoch": 2.42,
      "logps_train/chosen": -163.54148864746094,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -170.50917053222656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.1980700492858887,
      "rewards_train/margins": 3.664858818054199,
      "rewards_train/rejected": -5.862928867340088,
      "step": 212
    },
    {
      "epoch": 2.42,
      "logps_train/chosen": -172.4043731689453,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -184.43423461914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.632063388824463,
      "rewards_train/margins": 4.868025302886963,
      "rewards_train/rejected": -6.500088691711426,
      "step": 212
    },
    {
      "epoch": 2.43,
      "learning_rate": 4.3107791681585655e-06,
      "loss": 0.0701,
      "step": 213
    },
    {
      "epoch": 2.43,
      "logps_train/chosen": -182.57791137695312,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -188.65792846679688,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9670684337615967,
      "rewards_train/margins": 4.3333799839019775,
      "rewards_train/rejected": -6.300448417663574,
      "step": 213
    },
    {
      "epoch": 2.43,
      "logps_train/chosen": -178.8064422607422,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -188.4781494140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4776663780212402,
      "rewards_train/margins": 4.212725639343262,
      "rewards_train/rejected": -6.690392017364502,
      "step": 213
    },
    {
      "epoch": 2.43,
      "logps_train/chosen": -180.3985595703125,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -190.71658325195312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.1126112937927246,
      "rewards_train/margins": 4.2702765464782715,
      "rewards_train/rejected": -6.382887840270996,
      "step": 213
    },
    {
      "epoch": 2.43,
      "logps_train/chosen": -214.46804809570312,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -214.39895629882812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.7165300846099854,
      "rewards_train/margins": 4.324148416519165,
      "rewards_train/rejected": -7.04067850112915,
      "step": 213
    },
    {
      "epoch": 2.45,
      "learning_rate": 4.138112087868576e-06,
      "loss": 0.0602,
      "step": 214
    },
    {
      "epoch": 2.45,
      "logps_train/chosen": -184.88540649414062,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -208.5177001953125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1092917919158936,
      "rewards_train/margins": 5.026073694229126,
      "rewards_train/rejected": -7.1353654861450195,
      "step": 214
    },
    {
      "epoch": 2.45,
      "logps_train/chosen": -168.20083618164062,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -93.0,
      "logps_train/rejected": -153.44461059570312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.237173557281494,
      "rewards_train/margins": 3.7916626930236816,
      "rewards_train/rejected": -6.028836250305176,
      "step": 214
    },
    {
      "epoch": 2.45,
      "logps_train/chosen": -139.26658630371094,
      "logps_train/ref_chosen": -117.5,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -174.18048095703125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.160618305206299,
      "rewards_train/margins": 4.620589733123779,
      "rewards_train/rejected": -6.781208038330078,
      "step": 214
    },
    {
      "epoch": 2.45,
      "logps_train/chosen": -167.5859375,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -176.97976684570312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.373340606689453,
      "rewards_train/margins": 4.333230018615723,
      "rewards_train/rejected": -6.706570625305176,
      "step": 214
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.968661679220468e-06,
      "loss": 0.0675,
      "step": 215
    },
    {
      "epoch": 2.46,
      "logps_train/chosen": -178.07106018066406,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -178.66525268554688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.273902416229248,
      "rewards_train/margins": 4.306295394897461,
      "rewards_train/rejected": -6.580197811126709,
      "step": 215
    },
    {
      "epoch": 2.46,
      "logps_train/chosen": -186.64599609375,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -190.8783721923828,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4751794338226318,
      "rewards_train/margins": 4.996055841445923,
      "rewards_train/rejected": -6.471235275268555,
      "step": 215
    },
    {
      "epoch": 2.46,
      "logps_train/chosen": -155.9014129638672,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -111.5,
      "logps_train/rejected": -170.11361694335938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.4900078773498535,
      "rewards_train/margins": 4.389421463012695,
      "rewards_train/rejected": -5.879429340362549,
      "step": 215
    },
    {
      "epoch": 2.46,
      "logps_train/chosen": -206.13455200195312,
      "logps_train/ref_chosen": -191.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -218.8090057373047,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4799585342407227,
      "rewards_train/margins": 5.572231292724609,
      "rewards_train/rejected": -7.052189826965332,
      "step": 215
    },
    {
      "epoch": 2.47,
      "learning_rate": 3.8024540695871274e-06,
      "loss": 0.0435,
      "step": 216
    },
    {
      "epoch": 2.47,
      "logps_train/chosen": -211.99249267578125,
      "logps_train/ref_chosen": -196.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -193.9930419921875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6373107433319092,
      "rewards_train/margins": 4.699286699295044,
      "rewards_train/rejected": -6.336597442626953,
      "step": 216
    },
    {
      "epoch": 2.47,
      "logps_train/chosen": -162.2291259765625,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -182.76242065429688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9232053756713867,
      "rewards_train/margins": 4.727743148803711,
      "rewards_train/rejected": -6.650948524475098,
      "step": 216
    },
    {
      "epoch": 2.47,
      "logps_train/chosen": -186.01992797851562,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -187.78819274902344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7814736366271973,
      "rewards_train/margins": 4.773713111877441,
      "rewards_train/rejected": -6.555186748504639,
      "step": 216
    },
    {
      "epoch": 2.47,
      "logps_train/chosen": -179.65704345703125,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -193.5622100830078,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4718570709228516,
      "rewards_train/margins": 3.95780086517334,
      "rewards_train/rejected": -6.429657936096191,
      "step": 216
    },
    {
      "epoch": 2.48,
      "learning_rate": 3.6395148863377858e-06,
      "loss": 0.0425,
      "step": 217
    },
    {
      "epoch": 2.48,
      "logps_train/chosen": -185.7981719970703,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -194.738037109375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9814281463623047,
      "rewards_train/margins": 4.51200532913208,
      "rewards_train/rejected": -6.493433475494385,
      "step": 217
    },
    {
      "epoch": 2.48,
      "logps_train/chosen": -174.4874267578125,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -197.3990020751953,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0167109966278076,
      "rewards_train/margins": 4.54799485206604,
      "rewards_train/rejected": -6.564705848693848,
      "step": 217
    },
    {
      "epoch": 2.48,
      "logps_train/chosen": -186.6697998046875,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -150.0,
      "logps_train/rejected": -213.67283630371094,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.491588830947876,
      "rewards_train/margins": 4.921689748764038,
      "rewards_train/rejected": -6.413278579711914,
      "step": 217
    },
    {
      "epoch": 2.48,
      "logps_train/chosen": -185.66867065429688,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -215.50732421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.834641695022583,
      "rewards_train/margins": 4.791409254074097,
      "rewards_train/rejected": -7.62605094909668,
      "step": 217
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.4798692528866057e-06,
      "loss": 0.0431,
      "step": 218
    },
    {
      "epoch": 2.49,
      "logps_train/chosen": -158.78350830078125,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -97.0,
      "logps_train/rejected": -162.4704132080078,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4692697525024414,
      "rewards_train/margins": 4.1036505699157715,
      "rewards_train/rejected": -6.572920322418213,
      "step": 218
    },
    {
      "epoch": 2.49,
      "logps_train/chosen": -181.54193115234375,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -218.23255920410156,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7103451490402222,
      "rewards_train/margins": 5.347406983375549,
      "rewards_train/rejected": -7.0577521324157715,
      "step": 218
    },
    {
      "epoch": 2.49,
      "logps_train/chosen": -175.21749877929688,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -192.23492431640625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.2478251457214355,
      "rewards_train/margins": 4.507697582244873,
      "rewards_train/rejected": -6.755522727966309,
      "step": 218
    },
    {
      "epoch": 2.49,
      "logps_train/chosen": -182.02206420898438,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -197.85394287109375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8292454481124878,
      "rewards_train/margins": 4.512984395027161,
      "rewards_train/rejected": -6.342229843139648,
      "step": 218
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.3235417848188983e-06,
      "loss": 0.0539,
      "step": 219
    },
    {
      "epoch": 2.5,
      "logps_train/chosen": -184.61679077148438,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -177.779296875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.90328049659729,
      "rewards_train/margins": 4.3316810131073,
      "rewards_train/rejected": -6.23496150970459,
      "step": 219
    },
    {
      "epoch": 2.5,
      "logps_train/chosen": -141.23770141601562,
      "logps_train/ref_chosen": -116.0,
      "logps_train/ref_rejected": -88.5,
      "logps_train/rejected": -154.41409301757812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4947173595428467,
      "rewards_train/margins": 4.093469858169556,
      "rewards_train/rejected": -6.588187217712402,
      "step": 219
    },
    {
      "epoch": 2.5,
      "logps_train/chosen": -149.37660217285156,
      "logps_train/ref_chosen": -124.5,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -173.98623657226562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.497622013092041,
      "rewards_train/margins": 4.532507419586182,
      "rewards_train/rejected": -7.030129432678223,
      "step": 219
    },
    {
      "epoch": 2.5,
      "logps_train/chosen": -192.66302490234375,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -148.0,
      "logps_train/rejected": -212.57070922851562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0915956497192383,
      "rewards_train/margins": 4.3946027755737305,
      "rewards_train/rejected": -6.486198425292969,
      "step": 219
    },
    {
      "epoch": 2.51,
      "learning_rate": 3.170556586095699e-06,
      "loss": 0.0649,
      "step": 220
    },
    {
      "epoch": 2.51,
      "logps_train/chosen": -189.02066040039062,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -177.7373046875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.7520666122436523,
      "rewards_train/margins": 4.081625461578369,
      "rewards_train/rejected": -6.8336920738220215,
      "step": 220
    },
    {
      "epoch": 2.51,
      "logps_train/chosen": -176.2943878173828,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -187.67111206054688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.6185994148254395,
      "rewards_train/margins": 4.177418231964111,
      "rewards_train/rejected": -6.796017646789551,
      "step": 220
    },
    {
      "epoch": 2.51,
      "logps_train/chosen": -151.6334686279297,
      "logps_train/ref_chosen": -131.0,
      "logps_train/ref_rejected": -95.0,
      "logps_train/rejected": -156.61749267578125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.107999324798584,
      "rewards_train/margins": 4.088809013366699,
      "rewards_train/rejected": -6.196808338165283,
      "step": 220
    },
    {
      "epoch": 2.51,
      "logps_train/chosen": -145.2259521484375,
      "logps_train/ref_chosen": -127.0,
      "logps_train/ref_rejected": -102.0,
      "logps_train/rejected": -161.01107788085938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.803601622581482,
      "rewards_train/margins": 4.087031006813049,
      "rewards_train/rejected": -5.890632629394531,
      "step": 220
    },
    {
      "epoch": 2.53,
      "learning_rate": 3.0209372453372077e-06,
      "loss": 0.0763,
      "step": 221
    },
    {
      "epoch": 2.53,
      "logps_train/chosen": -177.82476806640625,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -166.97608947753906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1691970825195312,
      "rewards_train/margins": 4.20849084854126,
      "rewards_train/rejected": -6.377687931060791,
      "step": 221
    },
    {
      "epoch": 2.53,
      "logps_train/chosen": -163.2164306640625,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -205.69760131835938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.090000867843628,
      "rewards_train/margins": 4.568700075149536,
      "rewards_train/rejected": -6.658700942993164,
      "step": 221
    },
    {
      "epoch": 2.53,
      "logps_train/chosen": -165.90538024902344,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -179.6652069091797,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1149516105651855,
      "rewards_train/margins": 4.325690269470215,
      "rewards_train/rejected": -6.4406418800354,
      "step": 221
    },
    {
      "epoch": 2.53,
      "logps_train/chosen": -191.64077758789062,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -89.0,
      "logps_train/rejected": -155.56088256835938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.883876919746399,
      "rewards_train/margins": 4.743695616722107,
      "rewards_train/rejected": -6.627572536468506,
      "step": 221
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.8747068321856556e-06,
      "loss": 0.0509,
      "step": 222
    },
    {
      "epoch": 2.54,
      "logps_train/chosen": -175.29031372070312,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -152.0,
      "logps_train/rejected": -223.1751708984375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.350759983062744,
      "rewards_train/margins": 4.803475379943848,
      "rewards_train/rejected": -7.154235363006592,
      "step": 222
    },
    {
      "epoch": 2.54,
      "logps_train/chosen": -191.7686767578125,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -201.64723205566406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.020545721054077,
      "rewards_train/margins": 4.550428628921509,
      "rewards_train/rejected": -6.570974349975586,
      "step": 222
    },
    {
      "epoch": 2.54,
      "logps_train/chosen": -204.0155029296875,
      "logps_train/ref_chosen": -178.0,
      "logps_train/ref_rejected": -144.0,
      "logps_train/rejected": -215.03477478027344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.5512094497680664,
      "rewards_train/margins": 4.560812950134277,
      "rewards_train/rejected": -7.112022399902344,
      "step": 222
    },
    {
      "epoch": 2.54,
      "logps_train/chosen": -166.68942260742188,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -186.78805541992188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4038538932800293,
      "rewards_train/margins": 4.059913635253906,
      "rewards_train/rejected": -6.4637675285339355,
      "step": 222
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.731887893748242e-06,
      "loss": 0.0351,
      "step": 223
    },
    {
      "epoch": 2.55,
      "logps_train/chosen": -181.3331298828125,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -190.67776489257812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4753060340881348,
      "rewards_train/margins": 3.9983296394348145,
      "rewards_train/rejected": -6.473635673522949,
      "step": 223
    },
    {
      "epoch": 2.55,
      "logps_train/chosen": -164.18081665039062,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -196.8584747314453,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.398452043533325,
      "rewards_train/margins": 4.512004613876343,
      "rewards_train/rejected": -6.910456657409668,
      "step": 223
    },
    {
      "epoch": 2.55,
      "logps_train/chosen": -180.45425415039062,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -187.70980834960938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.8418128490448,
      "rewards_train/margins": 4.628679037094116,
      "rewards_train/rejected": -7.470491886138916,
      "step": 223
    },
    {
      "epoch": 2.55,
      "logps_train/chosen": -182.337890625,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -192.41957092285156,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7076172828674316,
      "rewards_train/margins": 4.944886684417725,
      "rewards_train/rejected": -6.652503967285156,
      "step": 223
    },
    {
      "epoch": 2.56,
      "learning_rate": 2.5925024511206207e-06,
      "loss": 0.0606,
      "step": 224
    },
    {
      "epoch": 2.56,
      "logps_train/chosen": -193.09463500976562,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -207.1976776123047,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4825100898742676,
      "rewards_train/margins": 4.484133243560791,
      "rewards_train/rejected": -6.966643333435059,
      "step": 224
    },
    {
      "epoch": 2.56,
      "logps_train/chosen": -160.484619140625,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -165.8876190185547,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7858647108078003,
      "rewards_train/margins": 4.05990469455719,
      "rewards_train/rejected": -5.84576940536499,
      "step": 224
    },
    {
      "epoch": 2.56,
      "logps_train/chosen": -203.03643798828125,
      "logps_train/ref_chosen": -188.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -178.2891082763672,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5149726867675781,
      "rewards_train/margins": 4.285617828369141,
      "rewards_train/rejected": -5.800590515136719,
      "step": 224
    },
    {
      "epoch": 2.56,
      "logps_train/chosen": -177.3469696044922,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -200.88308715820312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.568193197250366,
      "rewards_train/margins": 4.182811975479126,
      "rewards_train/rejected": -6.751005172729492,
      "step": 224
    },
    {
      "epoch": 2.57,
      "learning_rate": 2.45657199599148e-06,
      "loss": 0.0759,
      "step": 225
    },
    {
      "epoch": 2.57,
      "logps_train/chosen": -173.48208618164062,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -146.0,
      "logps_train/rejected": -207.01780700683594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8036763668060303,
      "rewards_train/margins": 4.3025963306427,
      "rewards_train/rejected": -6.1062726974487305,
      "step": 225
    },
    {
      "epoch": 2.57,
      "logps_train/chosen": -188.08656311035156,
      "logps_train/ref_chosen": -175.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -201.78933715820312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3506485223770142,
      "rewards_train/margins": 4.938831686973572,
      "rewards_train/rejected": -6.289480209350586,
      "step": 225
    },
    {
      "epoch": 2.57,
      "logps_train/chosen": -116.22421264648438,
      "logps_train/ref_chosen": -89.0,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -164.86378479003906,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.71127986907959,
      "rewards_train/margins": 3.7973580360412598,
      "rewards_train/rejected": -6.50863790512085,
      "step": 225
    },
    {
      "epoch": 2.57,
      "logps_train/chosen": -160.8203125,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -185.1500244140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3180701732635498,
      "rewards_train/margins": 4.57881760597229,
      "rewards_train/rejected": -5.89688777923584,
      "step": 225
    },
    {
      "epoch": 2.58,
      "learning_rate": 2.324117487328789e-06,
      "loss": 0.0696,
      "step": 226
    },
    {
      "epoch": 2.58,
      "logps_train/chosen": -172.0726776123047,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -184.76156616210938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2580490112304688,
      "rewards_train/margins": 4.4107346534729,
      "rewards_train/rejected": -6.668783664703369,
      "step": 226
    },
    {
      "epoch": 2.58,
      "logps_train/chosen": -170.86233520507812,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -213.73233032226562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.606351375579834,
      "rewards_train/margins": 5.066394329071045,
      "rewards_train/rejected": -7.672745704650879,
      "step": 226
    },
    {
      "epoch": 2.58,
      "logps_train/chosen": -155.60812377929688,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -110.5,
      "logps_train/rejected": -177.47715759277344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.057589054107666,
      "rewards_train/margins": 4.667080402374268,
      "rewards_train/rejected": -6.724669456481934,
      "step": 226
    },
    {
      "epoch": 2.58,
      "logps_train/chosen": -188.93655395507812,
      "logps_train/ref_chosen": -173.0,
      "logps_train/ref_rejected": -123.5,
      "logps_train/rejected": -175.76377868652344,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6039090156555176,
      "rewards_train/margins": 3.646785259246826,
      "rewards_train/rejected": -5.250694274902344,
      "step": 226
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.1951593481481237e-06,
      "loss": 0.0553,
      "step": 227
    },
    {
      "epoch": 2.59,
      "logps_train/chosen": -165.5153350830078,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -106.5,
      "logps_train/rejected": -174.52703857421875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7658891677856445,
      "rewards_train/margins": 5.043651580810547,
      "rewards_train/rejected": -6.809540748596191,
      "step": 227
    },
    {
      "epoch": 2.59,
      "logps_train/chosen": -160.86737060546875,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -181.62149047851562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4635443687438965,
      "rewards_train/margins": 4.161396026611328,
      "rewards_train/rejected": -6.624940395355225,
      "step": 227
    },
    {
      "epoch": 2.59,
      "logps_train/chosen": -167.77220153808594,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -99.0,
      "logps_train/rejected": -166.71109008789062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4367904663085938,
      "rewards_train/margins": 4.332364082336426,
      "rewards_train/rejected": -6.7691545486450195,
      "step": 227
    },
    {
      "epoch": 2.59,
      "logps_train/chosen": -157.61546325683594,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -115.0,
      "logps_train/rejected": -174.14193725585938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6726795434951782,
      "rewards_train/margins": 4.217974066734314,
      "rewards_train/rejected": -5.890653610229492,
      "step": 227
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.0697174623636794e-06,
      "loss": 0.0618,
      "step": 228
    },
    {
      "epoch": 2.61,
      "logps_train/chosen": -177.1771697998047,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -185.34759521484375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.171818733215332,
      "rewards_train/margins": 3.963038444519043,
      "rewards_train/rejected": -6.134857177734375,
      "step": 228
    },
    {
      "epoch": 2.61,
      "logps_train/chosen": -184.7438507080078,
      "logps_train/ref_chosen": -165.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -185.59344482421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.953291416168213,
      "rewards_train/margins": 4.560252666473389,
      "rewards_train/rejected": -6.513544082641602,
      "step": 228
    },
    {
      "epoch": 2.61,
      "logps_train/chosen": -173.85379028320312,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -178.25070190429688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.407742500305176,
      "rewards_train/margins": 4.145502090454102,
      "rewards_train/rejected": -6.553244590759277,
      "step": 228
    },
    {
      "epoch": 2.61,
      "logps_train/chosen": -158.80514526367188,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -188.6862335205078,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4136208295822144,
      "rewards_train/margins": 4.584104657173157,
      "rewards_train/rejected": -5.997725486755371,
      "step": 228
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.947811171722397e-06,
      "loss": 0.0588,
      "step": 229
    },
    {
      "epoch": 2.62,
      "logps_train/chosen": -171.05941772460938,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -162.3818817138672,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7870943546295166,
      "rewards_train/margins": 4.338154077529907,
      "rewards_train/rejected": -6.125248432159424,
      "step": 229
    },
    {
      "epoch": 2.62,
      "logps_train/chosen": -142.09471130371094,
      "logps_train/ref_chosen": -114.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -192.31854248046875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.8323230743408203,
      "rewards_train/margins": 4.283173084259033,
      "rewards_train/rejected": -7.1154961585998535,
      "step": 229
    },
    {
      "epoch": 2.62,
      "logps_train/chosen": -179.9519500732422,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -205.62069702148438,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.694608449935913,
      "rewards_train/margins": 4.932012319564819,
      "rewards_train/rejected": -6.626620769500732,
      "step": 229
    },
    {
      "epoch": 2.62,
      "logps_train/chosen": -160.87034606933594,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -112.5,
      "logps_train/rejected": -182.16702270507812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.20851993560791,
      "rewards_train/margins": 4.76931619644165,
      "rewards_train/rejected": -6.9778361320495605,
      "step": 229
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.8294592728216765e-06,
      "loss": 0.0485,
      "step": 230
    },
    {
      "epoch": 2.63,
      "logps_train/chosen": -157.534423828125,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -108.5,
      "logps_train/rejected": -174.2003173828125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9986941814422607,
      "rewards_train/margins": 4.557275056838989,
      "rewards_train/rejected": -6.55596923828125,
      "step": 230
    },
    {
      "epoch": 2.63,
      "logps_train/chosen": -168.04296875,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -107.5,
      "logps_train/rejected": -171.24734497070312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8662118911743164,
      "rewards_train/margins": 4.531960487365723,
      "rewards_train/rejected": -6.398172378540039,
      "step": 230
    },
    {
      "epoch": 2.63,
      "logps_train/chosen": -185.285400390625,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -109.0,
      "logps_train/rejected": -170.35165405273438,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.4988040924072266,
      "rewards_train/margins": 4.609604835510254,
      "rewards_train/rejected": -6.1084089279174805,
      "step": 230
    },
    {
      "epoch": 2.63,
      "logps_train/chosen": -173.4306640625,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -188.17437744140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2563962936401367,
      "rewards_train/margins": 4.628278732299805,
      "rewards_train/rejected": -6.884675025939941,
      "step": 230
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.7146800142111535e-06,
      "loss": 0.0496,
      "step": 231
    },
    {
      "epoch": 2.64,
      "logps_train/chosen": -171.4075469970703,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -121.5,
      "logps_train/rejected": -183.84478759765625,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4142889976501465,
      "rewards_train/margins": 3.8182373046875,
      "rewards_train/rejected": -6.2325263023376465,
      "step": 231
    },
    {
      "epoch": 2.64,
      "logps_train/chosen": -164.04263305664062,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -173.68544006347656,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1521155834198,
      "rewards_train/margins": 4.001340627670288,
      "rewards_train/rejected": -6.153456211090088,
      "step": 231
    },
    {
      "epoch": 2.64,
      "logps_train/chosen": -179.4583282470703,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -180.95101928710938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6806962490081787,
      "rewards_train/margins": 4.379248857498169,
      "rewards_train/rejected": -6.059945106506348,
      "step": 231
    },
    {
      "epoch": 2.64,
      "logps_train/chosen": -181.32591247558594,
      "logps_train/ref_chosen": -160.0,
      "logps_train/ref_rejected": -135.0,
      "logps_train/rejected": -201.79678344726562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1640372276306152,
      "rewards_train/margins": 4.5218915939331055,
      "rewards_train/rejected": -6.685928821563721,
      "step": 231
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.6034910935789627e-06,
      "loss": 0.0576,
      "step": 232
    },
    {
      "epoch": 2.65,
      "logps_train/chosen": -177.09873962402344,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -197.26693725585938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.002061605453491,
      "rewards_train/margins": 4.639084577560425,
      "rewards_train/rejected": -6.641146183013916,
      "step": 232
    },
    {
      "epoch": 2.65,
      "logps_train/chosen": -167.2610321044922,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -136.0,
      "logps_train/rejected": -206.99166870117188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.846024990081787,
      "rewards_train/margins": 5.2769694328308105,
      "rewards_train/rejected": -7.122994422912598,
      "step": 232
    },
    {
      "epoch": 2.65,
      "logps_train/chosen": -135.89576721191406,
      "logps_train/ref_chosen": -107.5,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -161.03927612304688,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.8300557136535645,
      "rewards_train/margins": 3.8118605613708496,
      "rewards_train/rejected": -6.641916275024414,
      "step": 232
    },
    {
      "epoch": 2.65,
      "logps_train/chosen": -162.01553344726562,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -202.14938354492188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.1653234958648682,
      "rewards_train/margins": 5.1688525676727295,
      "rewards_train/rejected": -6.334176063537598,
      "step": 232
    },
    {
      "epoch": 2.66,
      "learning_rate": 1.4959096550229645e-06,
      "loss": 0.0669,
      "step": 233
    },
    {
      "epoch": 2.66,
      "logps_train/chosen": -180.497314453125,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -137.0,
      "logps_train/rejected": -197.39334106445312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7840571403503418,
      "rewards_train/margins": 4.3093791007995605,
      "rewards_train/rejected": -6.093436241149902,
      "step": 233
    },
    {
      "epoch": 2.66,
      "logps_train/chosen": -181.78704833984375,
      "logps_train/ref_chosen": -161.0,
      "logps_train/ref_rejected": -119.5,
      "logps_train/rejected": -179.69802856445312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.05507230758667,
      "rewards_train/margins": 3.9480795860290527,
      "rewards_train/rejected": -6.003151893615723,
      "step": 233
    },
    {
      "epoch": 2.66,
      "logps_train/chosen": -177.9119110107422,
      "logps_train/ref_chosen": -152.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -173.1656951904297,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.5379676818847656,
      "rewards_train/margins": 4.04774284362793,
      "rewards_train/rejected": -6.585710525512695,
      "step": 233
    },
    {
      "epoch": 2.66,
      "logps_train/chosen": -191.3437042236328,
      "logps_train/ref_chosen": -174.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -195.78623962402344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7578561305999756,
      "rewards_train/margins": 4.763178110122681,
      "rewards_train/rejected": -6.521034240722656,
      "step": 233
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.391952286407311e-06,
      "loss": 0.0564,
      "step": 234
    },
    {
      "epoch": 2.67,
      "logps_train/chosen": -164.625244140625,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -191.79592895507812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1351804733276367,
      "rewards_train/margins": 4.474490165710449,
      "rewards_train/rejected": -6.609670639038086,
      "step": 234
    },
    {
      "epoch": 2.67,
      "logps_train/chosen": -192.8685302734375,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -180.1429443359375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.744861364364624,
      "rewards_train/margins": 4.417285203933716,
      "rewards_train/rejected": -6.16214656829834,
      "step": 234
    },
    {
      "epoch": 2.67,
      "logps_train/chosen": -172.70458984375,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -175.58883666992188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7502193450927734,
      "rewards_train/margins": 4.853781700134277,
      "rewards_train/rejected": -6.604001045227051,
      "step": 234
    },
    {
      "epoch": 2.67,
      "logps_train/chosen": -179.72857666015625,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -183.2381134033203,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.149763584136963,
      "rewards_train/margins": 4.196936130523682,
      "rewards_train/rejected": -6.3466997146606445,
      "step": 234
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.2916350168047681e-06,
      "loss": 0.0664,
      "step": 235
    },
    {
      "epoch": 2.69,
      "logps_train/chosen": -172.16232299804688,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -190.31521606445312,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.4478726387023926,
      "rewards_train/margins": 4.224077224731445,
      "rewards_train/rejected": -6.671949863433838,
      "step": 235
    },
    {
      "epoch": 2.69,
      "logps_train/chosen": -183.98605346679688,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -184.5833740234375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0531957149505615,
      "rewards_train/margins": 4.704164743423462,
      "rewards_train/rejected": -6.757360458374023,
      "step": 235
    },
    {
      "epoch": 2.69,
      "logps_train/chosen": -150.15863037109375,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -99.5,
      "logps_train/rejected": -157.22695922851562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0300607681274414,
      "rewards_train/margins": 3.7460880279541016,
      "rewards_train/rejected": -5.776148796081543,
      "step": 235
    },
    {
      "epoch": 2.69,
      "logps_train/chosen": -198.83621215820312,
      "logps_train/ref_chosen": -183.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -208.21908569335938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.5560832023620605,
      "rewards_train/margins": 5.394535064697266,
      "rewards_train/rejected": -6.950618267059326,
      "step": 235
    },
    {
      "epoch": 2.7,
      "learning_rate": 1.1949733140252466e-06,
      "loss": 0.0635,
      "step": 236
    },
    {
      "epoch": 2.7,
      "logps_train/chosen": -161.04502868652344,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -163.6453857421875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8402457237243652,
      "rewards_train/margins": 4.072827339172363,
      "rewards_train/rejected": -5.9130730628967285,
      "step": 236
    },
    {
      "epoch": 2.7,
      "logps_train/chosen": -184.62203979492188,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -216.21878051757812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.894137382507324,
      "rewards_train/margins": 4.415827751159668,
      "rewards_train/rejected": -7.309965133666992,
      "step": 236
    },
    {
      "epoch": 2.7,
      "logps_train/chosen": -194.06765747070312,
      "logps_train/ref_chosen": -182.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -192.44342041015625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.2655562162399292,
      "rewards_train/margins": 4.643270134925842,
      "rewards_train/rejected": -5.9088263511657715,
      "step": 236
    },
    {
      "epoch": 2.7,
      "logps_train/chosen": -214.92184448242188,
      "logps_train/ref_chosen": -190.0,
      "logps_train/ref_rejected": -139.0,
      "logps_train/rejected": -202.62722778320312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.510640859603882,
      "rewards_train/margins": 3.877863645553589,
      "rewards_train/rejected": -6.388504505157471,
      "step": 236
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.1019820822307985e-06,
      "loss": 0.0649,
      "step": 237
    },
    {
      "epoch": 2.71,
      "logps_train/chosen": -126.3829574584961,
      "logps_train/ref_chosen": -105.5,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -177.629638671875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.102309465408325,
      "rewards_train/margins": 4.272714376449585,
      "rewards_train/rejected": -6.37502384185791,
      "step": 237
    },
    {
      "epoch": 2.71,
      "logps_train/chosen": -138.21041870117188,
      "logps_train/ref_chosen": -118.0,
      "logps_train/ref_rejected": -110.0,
      "logps_train/rejected": -169.5449981689453,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0431365966796875,
      "rewards_train/margins": 3.923570156097412,
      "rewards_train/rejected": -5.9667067527771,
      "step": 237
    },
    {
      "epoch": 2.71,
      "logps_train/chosen": -161.8151092529297,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -105.0,
      "logps_train/rejected": -169.7493438720703,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1523118019104004,
      "rewards_train/margins": 4.343618869781494,
      "rewards_train/rejected": -6.4959306716918945,
      "step": 237
    },
    {
      "epoch": 2.71,
      "logps_train/chosen": -216.87832641601562,
      "logps_train/ref_chosen": -204.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -197.9646453857422,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.3093161582946777,
      "rewards_train/margins": 5.389882564544678,
      "rewards_train/rejected": -6.6991987228393555,
      "step": 237
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.0126756596375686e-06,
      "loss": 0.0598,
      "step": 238
    },
    {
      "epoch": 2.72,
      "logps_train/chosen": -166.60894775390625,
      "logps_train/ref_chosen": -147.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -183.59124755859375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9279100894927979,
      "rewards_train/margins": 4.252699136734009,
      "rewards_train/rejected": -6.180609226226807,
      "step": 238
    },
    {
      "epoch": 2.72,
      "logps_train/chosen": -139.39071655273438,
      "logps_train/ref_chosen": -116.5,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -172.66722106933594,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.307724952697754,
      "rewards_train/margins": 4.875207901000977,
      "rewards_train/rejected": -7.1829328536987305,
      "step": 238
    },
    {
      "epoch": 2.72,
      "logps_train/chosen": -171.77474975585938,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -163.83929443359375,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.127732276916504,
      "rewards_train/margins": 3.6769256591796875,
      "rewards_train/rejected": -5.804657936096191,
      "step": 238
    },
    {
      "epoch": 2.72,
      "logps_train/chosen": -172.7614288330078,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -157.0,
      "logps_train/rejected": -227.40777587890625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8960649967193604,
      "rewards_train/margins": 5.144322156906128,
      "rewards_train/rejected": -7.040387153625488,
      "step": 238
    },
    {
      "epoch": 2.73,
      "learning_rate": 9.270678163050217e-07,
      "loss": 0.0751,
      "step": 239
    },
    {
      "epoch": 2.73,
      "logps_train/chosen": -163.55027770996094,
      "logps_train/ref_chosen": -137.0,
      "logps_train/ref_rejected": -82.5,
      "logps_train/rejected": -145.41427612304688,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.624851703643799,
      "rewards_train/margins": 3.6589112281799316,
      "rewards_train/rejected": -6.2837629318237305,
      "step": 239
    },
    {
      "epoch": 2.73,
      "logps_train/chosen": -203.0478515625,
      "logps_train/ref_chosen": -189.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -199.60626220703125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.3887698650360107,
      "rewards_train/margins": 5.202104806900024,
      "rewards_train/rejected": -6.590874671936035,
      "step": 239
    },
    {
      "epoch": 2.73,
      "logps_train/chosen": -178.8606719970703,
      "logps_train/ref_chosen": -158.0,
      "logps_train/ref_rejected": -131.0,
      "logps_train/rejected": -203.78274536132812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.051692008972168,
      "rewards_train/margins": 5.257539749145508,
      "rewards_train/rejected": -7.309231758117676,
      "step": 239
    },
    {
      "epoch": 2.73,
      "logps_train/chosen": -162.5720977783203,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -175.08572387695312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2999839782714844,
      "rewards_train/margins": 4.601556777954102,
      "rewards_train/rejected": -6.901540756225586,
      "step": 239
    },
    {
      "epoch": 2.74,
      "learning_rate": 8.451717520127273e-07,
      "loss": 0.0755,
      "step": 240
    },
    {
      "epoch": 2.74,
      "logps_train/chosen": -180.50900268554688,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -114.0,
      "logps_train/rejected": -180.9222412109375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7348852157592773,
      "rewards_train/margins": 4.948159217834473,
      "rewards_train/rejected": -6.68304443359375,
      "step": 240
    },
    {
      "epoch": 2.74,
      "logps_train/chosen": -189.4424285888672,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -140.0,
      "logps_train/rejected": -208.51254272460938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2249059677124023,
      "rewards_train/margins": 4.5786919593811035,
      "rewards_train/rejected": -6.803597927093506,
      "step": 240
    },
    {
      "epoch": 2.74,
      "logps_train/chosen": -154.6390380859375,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -186.6710205078125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6452138423919678,
      "rewards_train/margins": 4.573157072067261,
      "rewards_train/rejected": -6.2183709144592285,
      "step": 240
    },
    {
      "epoch": 2.74,
      "logps_train/chosen": -167.7239990234375,
      "logps_train/ref_chosen": -143.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -192.18130493164062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4205942153930664,
      "rewards_train/margins": 5.258059501647949,
      "rewards_train/rejected": -7.678653717041016,
      "step": 240
    },
    {
      "epoch": 2.75,
      "learning_rate": 7.670000942251287e-07,
      "loss": 0.0483,
      "step": 241
    },
    {
      "epoch": 2.75,
      "logps_train/chosen": -177.83676147460938,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -197.16542053222656,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9257657527923584,
      "rewards_train/margins": 4.465288400650024,
      "rewards_train/rejected": -6.391054153442383,
      "step": 241
    },
    {
      "epoch": 2.75,
      "logps_train/chosen": -186.260498046875,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -185.99197387695312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.358275890350342,
      "rewards_train/margins": 4.20889139175415,
      "rewards_train/rejected": -6.567167282104492,
      "step": 241
    },
    {
      "epoch": 2.75,
      "logps_train/chosen": -159.35272216796875,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -101.5,
      "logps_train/rejected": -167.78614807128906,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.3801445960998535,
      "rewards_train/margins": 4.245735168457031,
      "rewards_train/rejected": -6.625879764556885,
      "step": 241
    },
    {
      "epoch": 2.75,
      "logps_train/chosen": -162.38101196289062,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -181.56993103027344,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7641749382019043,
      "rewards_train/margins": 4.599165439605713,
      "rewards_train/rejected": -6.363340377807617,
      "step": 241
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.92564896144493e-07,
      "loss": 0.0347,
      "step": 242
    },
    {
      "epoch": 2.77,
      "logps_train/chosen": -169.59744262695312,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -183.14720153808594,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.599441409111023,
      "rewards_train/margins": 4.840278744697571,
      "rewards_train/rejected": -6.439720153808594,
      "step": 242
    },
    {
      "epoch": 2.77,
      "logps_train/chosen": -156.12826538085938,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -109.5,
      "logps_train/rejected": -175.92604064941406,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.272787570953369,
      "rewards_train/margins": 4.405375957489014,
      "rewards_train/rejected": -6.678163528442383,
      "step": 242
    },
    {
      "epoch": 2.77,
      "logps_train/chosen": -196.60855102539062,
      "logps_train/ref_chosen": -170.0,
      "logps_train/ref_rejected": -127.0,
      "logps_train/rejected": -193.12551879882812,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.7350728511810303,
      "rewards_train/margins": 3.8505260944366455,
      "rewards_train/rejected": -6.585598945617676,
      "step": 242
    },
    {
      "epoch": 2.77,
      "logps_train/chosen": -161.61679077148438,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -195.109130859375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.331406354904175,
      "rewards_train/margins": 4.148061990737915,
      "rewards_train/rejected": -6.47946834564209,
      "step": 242
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.218776348524663e-07,
      "loss": 0.0655,
      "step": 243
    },
    {
      "epoch": 2.78,
      "logps_train/chosen": -182.87452697753906,
      "logps_train/ref_chosen": -163.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -194.0487060546875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9816908836364746,
      "rewards_train/margins": 5.13365364074707,
      "rewards_train/rejected": -7.115344524383545,
      "step": 243
    },
    {
      "epoch": 2.78,
      "logps_train/chosen": -156.08993530273438,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -178.7113037109375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.6412432193756104,
      "rewards_train/margins": 4.133403539657593,
      "rewards_train/rejected": -6.774646759033203,
      "step": 243
    },
    {
      "epoch": 2.78,
      "logps_train/chosen": -211.40530395507812,
      "logps_train/ref_chosen": -194.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -185.21580505371094,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.7184962034225464,
      "rewards_train/margins": 4.194722294807434,
      "rewards_train/rejected": -5.9132184982299805,
      "step": 243
    },
    {
      "epoch": 2.78,
      "logps_train/chosen": -179.36248779296875,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -175.9487762451172,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.673431396484375,
      "rewards_train/margins": 4.250351905822754,
      "rewards_train/rejected": -6.923783302307129,
      "step": 243
    },
    {
      "epoch": 2.79,
      "learning_rate": 5.549492095404202e-07,
      "loss": 0.0632,
      "step": 244
    },
    {
      "epoch": 2.79,
      "logps_train/chosen": -172.94845581054688,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -193.4479217529297,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8069548606872559,
      "rewards_train/margins": 5.433931350708008,
      "rewards_train/rejected": -7.240886211395264,
      "step": 244
    },
    {
      "epoch": 2.79,
      "logps_train/chosen": -161.23968505859375,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -179.7657470703125,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.9360785484313965,
      "rewards_train/margins": 3.958073139190674,
      "rewards_train/rejected": -5.89415168762207,
      "step": 244
    },
    {
      "epoch": 2.79,
      "logps_train/chosen": -160.264892578125,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -106.0,
      "logps_train/rejected": -171.85986328125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4065914154052734,
      "rewards_train/margins": 4.195216178894043,
      "rewards_train/rejected": -6.601807594299316,
      "step": 244
    },
    {
      "epoch": 2.79,
      "logps_train/chosen": -235.1175079345703,
      "logps_train/ref_chosen": -219.0,
      "logps_train/ref_rejected": -176.0,
      "logps_train/rejected": -239.72584533691406,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.5730059146881104,
      "rewards_train/margins": 4.821647882461548,
      "rewards_train/rejected": -6.394653797149658,
      "step": 244
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.917899398289377e-07,
      "loss": 0.0628,
      "step": 245
    },
    {
      "epoch": 2.8,
      "logps_train/chosen": -171.4566650390625,
      "logps_train/ref_chosen": -150.0,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -161.31747436523438,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.139416217803955,
      "rewards_train/margins": 3.9931116104125977,
      "rewards_train/rejected": -6.132527828216553,
      "step": 245
    },
    {
      "epoch": 2.8,
      "logps_train/chosen": -160.1334228515625,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -197.39840698242188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1715452671051025,
      "rewards_train/margins": 4.618248701095581,
      "rewards_train/rejected": -6.789793968200684,
      "step": 245
    },
    {
      "epoch": 2.8,
      "logps_train/chosen": -176.39114379882812,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -128.0,
      "logps_train/rejected": -195.78604125976562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1625027656555176,
      "rewards_train/margins": 4.600754737854004,
      "rewards_train/rejected": -6.7632575035095215,
      "step": 245
    },
    {
      "epoch": 2.8,
      "logps_train/chosen": -156.68441772460938,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -160.67849731445312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8903658390045166,
      "rewards_train/margins": 4.183562994003296,
      "rewards_train/rejected": -6.0739288330078125,
      "step": 245
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.324095641766168e-07,
      "loss": 0.0534,
      "step": 246
    },
    {
      "epoch": 2.81,
      "logps_train/chosen": -177.7391357421875,
      "logps_train/ref_chosen": -159.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -191.41738891601562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8221569061279297,
      "rewards_train/margins": 5.090432167053223,
      "rewards_train/rejected": -6.912589073181152,
      "step": 246
    },
    {
      "epoch": 2.81,
      "logps_train/chosen": -133.89585876464844,
      "logps_train/ref_chosen": -111.5,
      "logps_train/ref_rejected": -101.0,
      "logps_train/rejected": -163.14993286132812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2561392784118652,
      "rewards_train/margins": 3.9409842491149902,
      "rewards_train/rejected": -6.1971235275268555,
      "step": 246
    },
    {
      "epoch": 2.81,
      "logps_train/chosen": -184.8391876220703,
      "logps_train/ref_chosen": -168.0,
      "logps_train/ref_rejected": -164.0,
      "logps_train/rejected": -227.5182342529297,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6263017654418945,
      "rewards_train/margins": 4.725424289703369,
      "rewards_train/rejected": -6.351726055145264,
      "step": 246
    },
    {
      "epoch": 2.81,
      "logps_train/chosen": -177.89773559570312,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -138.0,
      "logps_train/rejected": -202.14389038085938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0483689308166504,
      "rewards_train/margins": 4.4206223487854,
      "rewards_train/rejected": -6.468991279602051,
      "step": 246
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.768172383785268e-07,
      "loss": 0.0606,
      "step": 247
    },
    {
      "epoch": 2.82,
      "logps_train/chosen": -166.12185668945312,
      "logps_train/ref_chosen": -145.0,
      "logps_train/ref_rejected": -100.0,
      "logps_train/rejected": -160.4178466796875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0884547233581543,
      "rewards_train/margins": 3.920907974243164,
      "rewards_train/rejected": -6.009362697601318,
      "step": 247
    },
    {
      "epoch": 2.82,
      "logps_train/chosen": -172.78741455078125,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -119.0,
      "logps_train/rejected": -189.86375427246094,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.8326473236083984,
      "rewards_train/margins": 4.249040126800537,
      "rewards_train/rejected": -7.0816874504089355,
      "step": 247
    },
    {
      "epoch": 2.82,
      "logps_train/chosen": -185.13803100585938,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -112.0,
      "logps_train/rejected": -178.22286987304688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.280600070953369,
      "rewards_train/margins": 4.309672832489014,
      "rewards_train/rejected": -6.590272903442383,
      "step": 247
    },
    {
      "epoch": 2.82,
      "logps_train/chosen": -153.21363830566406,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -171.15322875976562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.703980803489685,
      "rewards_train/margins": 4.131362557411194,
      "rewards_train/rejected": -5.835343360900879,
      "step": 247
    },
    {
      "epoch": 2.83,
      "learning_rate": 3.2502153415447656e-07,
      "loss": 0.0635,
      "step": 248
    },
    {
      "epoch": 2.83,
      "logps_train/chosen": -163.0394744873047,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -76.0,
      "logps_train/rejected": -140.48609924316406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.749880313873291,
      "rewards_train/margins": 3.7217764854431152,
      "rewards_train/rejected": -6.471656799316406,
      "step": 248
    },
    {
      "epoch": 2.83,
      "logps_train/chosen": -166.8461151123047,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -129.0,
      "logps_train/rejected": -198.37579345703125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.565225601196289,
      "rewards_train/margins": 4.396816253662109,
      "rewards_train/rejected": -6.962041854858398,
      "step": 248
    },
    {
      "epoch": 2.83,
      "logps_train/chosen": -207.8005828857422,
      "logps_train/ref_chosen": -190.0,
      "logps_train/ref_rejected": -153.0,
      "logps_train/rejected": -223.14645385742188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7702926397323608,
      "rewards_train/margins": 5.2681804895401,
      "rewards_train/rejected": -7.038473129272461,
      "step": 248
    },
    {
      "epoch": 2.83,
      "logps_train/chosen": -154.43231201171875,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -96.0,
      "logps_train/rejected": -154.94419860839844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8259937763214111,
      "rewards_train/margins": 4.078387022018433,
      "rewards_train/rejected": -5.904380798339844,
      "step": 248
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.770304378273553e-07,
      "loss": 0.0647,
      "step": 249
    },
    {
      "epoch": 2.85,
      "logps_train/chosen": -145.6243438720703,
      "logps_train/ref_chosen": -127.5,
      "logps_train/ref_rejected": -103.5,
      "logps_train/rejected": -160.12098693847656,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.8032550811767578,
      "rewards_train/margins": 3.8635311126708984,
      "rewards_train/rejected": -5.666786193847656,
      "step": 249
    },
    {
      "epoch": 2.85,
      "logps_train/chosen": -179.10397338867188,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -195.87139892578125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2270493507385254,
      "rewards_train/margins": 4.725861549377441,
      "rewards_train/rejected": -6.952910900115967,
      "step": 249
    },
    {
      "epoch": 2.85,
      "logps_train/chosen": -160.42538452148438,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -105.5,
      "logps_train/rejected": -172.30838012695312,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4366796016693115,
      "rewards_train/margins": 4.239910364151001,
      "rewards_train/rejected": -6.6765899658203125,
      "step": 249
    },
    {
      "epoch": 2.85,
      "logps_train/chosen": -163.18991088867188,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -120.5,
      "logps_train/rejected": -182.28509521484375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.6743369102478027,
      "rewards_train/margins": 4.484737873077393,
      "rewards_train/rejected": -6.159074783325195,
      "step": 249
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.3285134909173112e-07,
      "loss": 0.0667,
      "step": 250
    },
    {
      "epoch": 2.86,
      "logps_train/chosen": -194.73187255859375,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -118.0,
      "logps_train/rejected": -188.28573608398438,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9335401058197021,
      "rewards_train/margins": 5.130775690078735,
      "rewards_train/rejected": -7.0643157958984375,
      "step": 250
    },
    {
      "epoch": 2.86,
      "logps_train/chosen": -167.18817138671875,
      "logps_train/ref_chosen": -142.0,
      "logps_train/ref_rejected": -116.0,
      "logps_train/rejected": -181.6185302734375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.518524169921875,
      "rewards_train/margins": 4.051629066467285,
      "rewards_train/rejected": -6.57015323638916,
      "step": 250
    },
    {
      "epoch": 2.86,
      "logps_train/chosen": -194.84786987304688,
      "logps_train/ref_chosen": -176.0,
      "logps_train/ref_rejected": -162.0,
      "logps_train/rejected": -225.637451171875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.871652364730835,
      "rewards_train/margins": 4.4969751834869385,
      "rewards_train/rejected": -6.368627548217773,
      "step": 250
    },
    {
      "epoch": 2.86,
      "logps_train/chosen": -161.91610717773438,
      "logps_train/ref_chosen": -140.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -195.36077880859375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1696693897247314,
      "rewards_train/margins": 4.929694890975952,
      "rewards_train/rejected": -7.099364280700684,
      "step": 250
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.924910798728946e-07,
      "loss": 0.0365,
      "step": 251
    },
    {
      "epoch": 2.87,
      "logps_train/chosen": -176.8163299560547,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -183.77633666992188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.261936664581299,
      "rewards_train/margins": 4.28908634185791,
      "rewards_train/rejected": -6.551023006439209,
      "step": 251
    },
    {
      "epoch": 2.87,
      "logps_train/chosen": -173.4698486328125,
      "logps_train/ref_chosen": -157.0,
      "logps_train/ref_rejected": -122.0,
      "logps_train/rejected": -184.99172973632812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6786491870880127,
      "rewards_train/margins": 4.63419508934021,
      "rewards_train/rejected": -6.312844276428223,
      "step": 251
    },
    {
      "epoch": 2.87,
      "logps_train/chosen": -168.40023803710938,
      "logps_train/ref_chosen": -146.0,
      "logps_train/ref_rejected": -120.0,
      "logps_train/rejected": -183.943603515625,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.20320725440979,
      "rewards_train/margins": 4.175772428512573,
      "rewards_train/rejected": -6.378979682922363,
      "step": 251
    },
    {
      "epoch": 2.87,
      "logps_train/chosen": -184.357666015625,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -185.84925842285156,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.7965086698532104,
      "rewards_train/margins": 4.701600193977356,
      "rewards_train/rejected": -6.498108863830566,
      "step": 251
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.559558532765404e-07,
      "loss": 0.0579,
      "step": 252
    },
    {
      "epoch": 2.88,
      "logps_train/chosen": -133.98130798339844,
      "logps_train/ref_chosen": -114.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -166.95147705078125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.0248889923095703,
      "rewards_train/margins": 4.195259094238281,
      "rewards_train/rejected": -6.220148086547852,
      "step": 252
    },
    {
      "epoch": 2.88,
      "logps_train/chosen": -198.66932678222656,
      "logps_train/ref_chosen": -179.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -210.49432373046875,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9354877471923828,
      "rewards_train/margins": 4.8838677406311035,
      "rewards_train/rejected": -6.819355487823486,
      "step": 252
    },
    {
      "epoch": 2.88,
      "logps_train/chosen": -173.53561401367188,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -117.0,
      "logps_train/rejected": -179.3866729736328,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.9231053590774536,
      "rewards_train/margins": 4.311460375785828,
      "rewards_train/rejected": -6.234565734863281,
      "step": 252
    },
    {
      "epoch": 2.88,
      "logps_train/chosen": -192.16253662109375,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -188.32643127441406,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.32460355758667,
      "rewards_train/margins": 4.6487135887146,
      "rewards_train/rejected": -6.9733171463012695,
      "step": 252
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.23251302629232e-07,
      "loss": 0.0477,
      "step": 253
    },
    {
      "epoch": 2.89,
      "logps_train/chosen": -164.695068359375,
      "logps_train/ref_chosen": -149.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -183.24545288085938,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.560620665550232,
      "rewards_train/margins": 4.226998925209045,
      "rewards_train/rejected": -5.787619590759277,
      "step": 253
    },
    {
      "epoch": 2.89,
      "logps_train/chosen": -164.260986328125,
      "logps_train/ref_chosen": -139.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -173.06460571289062,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.5073485374450684,
      "rewards_train/margins": 4.087198734283447,
      "rewards_train/rejected": -6.594547271728516,
      "step": 253
    },
    {
      "epoch": 2.89,
      "logps_train/chosen": -151.60467529296875,
      "logps_train/ref_chosen": -128.0,
      "logps_train/ref_rejected": -116.5,
      "logps_train/rejected": -179.28170776367188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.3181824684143066,
      "rewards_train/margins": 3.9910435676574707,
      "rewards_train/rejected": -6.309226036071777,
      "step": 253
    },
    {
      "epoch": 2.89,
      "logps_train/chosen": -191.03347778320312,
      "logps_train/ref_chosen": -166.0,
      "logps_train/ref_rejected": -124.0,
      "logps_train/rejected": -191.88462829589844,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.450467109680176,
      "rewards_train/margins": 4.3493733406066895,
      "rewards_train/rejected": -6.799840450286865,
      "step": 253
    },
    {
      "epoch": 2.9,
      "learning_rate": 9.438247060979955e-08,
      "loss": 0.0777,
      "step": 254
    },
    {
      "epoch": 2.9,
      "logps_train/chosen": -172.36068725585938,
      "logps_train/ref_chosen": -153.0,
      "logps_train/ref_rejected": -107.0,
      "logps_train/rejected": -175.3718719482422,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.9357753992080688,
      "rewards_train/margins": 4.9021934270858765,
      "rewards_train/rejected": -6.837968826293945,
      "step": 254
    },
    {
      "epoch": 2.9,
      "logps_train/chosen": -179.45162963867188,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -117.5,
      "logps_train/rejected": -178.16188049316406,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.3817834854125977,
      "rewards_train/margins": 3.6969776153564453,
      "rewards_train/rejected": -6.078761100769043,
      "step": 254
    },
    {
      "epoch": 2.9,
      "logps_train/chosen": -167.01654052734375,
      "logps_train/ref_chosen": -141.0,
      "logps_train/ref_rejected": -111.0,
      "logps_train/rejected": -179.34791564941406,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.5837817192077637,
      "rewards_train/margins": 4.215170383453369,
      "rewards_train/rejected": -6.798952102661133,
      "step": 254
    },
    {
      "epoch": 2.9,
      "logps_train/chosen": -170.49327087402344,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -165.78018188476562,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -2.2194879055023193,
      "rewards_train/margins": 4.055551767349243,
      "rewards_train/rejected": -6.2750396728515625,
      "step": 254
    },
    {
      "epoch": 2.91,
      "learning_rate": 6.935380847182815e-08,
      "loss": 0.0963,
      "step": 255
    },
    {
      "epoch": 2.91,
      "logps_train/chosen": -184.44064331054688,
      "logps_train/ref_chosen": -162.0,
      "logps_train/ref_rejected": -143.0,
      "logps_train/rejected": -211.0648956298828,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.2682833671569824,
      "rewards_train/margins": 4.527073860168457,
      "rewards_train/rejected": -6.7953572273254395,
      "step": 255
    },
    {
      "epoch": 2.91,
      "logps_train/chosen": -171.735595703125,
      "logps_train/ref_chosen": -148.0,
      "logps_train/ref_rejected": -121.0,
      "logps_train/rejected": -186.7560272216797,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.400291919708252,
      "rewards_train/margins": 4.154388904571533,
      "rewards_train/rejected": -6.554680824279785,
      "step": 255
    },
    {
      "epoch": 2.91,
      "logps_train/chosen": -175.62689208984375,
      "logps_train/ref_chosen": -151.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -202.15704345703125,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.4636664390563965,
      "rewards_train/margins": 4.4941277503967285,
      "rewards_train/rejected": -6.957794189453125,
      "step": 255
    },
    {
      "epoch": 2.91,
      "logps_train/chosen": -186.03994750976562,
      "logps_train/ref_chosen": -169.0,
      "logps_train/ref_rejected": -130.0,
      "logps_train/rejected": -196.6202392578125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.696962833404541,
      "rewards_train/margins": 4.931369781494141,
      "rewards_train/rejected": -6.628332614898682,
      "step": 255
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.816917535731547e-08,
      "loss": 0.0631,
      "step": 256
    },
    {
      "epoch": 2.93,
      "logps_train/chosen": -153.82058715820312,
      "logps_train/ref_chosen": -135.0,
      "logps_train/ref_rejected": -124.5,
      "logps_train/rejected": -184.61001586914062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.905739426612854,
      "rewards_train/margins": 4.107263922691345,
      "rewards_train/rejected": -6.013003349304199,
      "step": 256
    },
    {
      "epoch": 2.93,
      "logps_train/chosen": -152.47775268554688,
      "logps_train/ref_chosen": -130.0,
      "logps_train/ref_rejected": -89.0,
      "logps_train/rejected": -147.11019897460938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.2787318229675293,
      "rewards_train/margins": 3.5291028022766113,
      "rewards_train/rejected": -5.807834625244141,
      "step": 256
    },
    {
      "epoch": 2.93,
      "logps_train/chosen": -163.71652221679688,
      "logps_train/ref_chosen": -138.0,
      "logps_train/ref_rejected": -134.0,
      "logps_train/rejected": -201.04147338867188,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.5193088054656982,
      "rewards_train/margins": 4.136402368545532,
      "rewards_train/rejected": -6.6557111740112305,
      "step": 256
    },
    {
      "epoch": 2.93,
      "logps_train/chosen": -191.13845825195312,
      "logps_train/ref_chosen": -172.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -182.69961547851562,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8818151950836182,
      "rewards_train/margins": 4.146349668502808,
      "rewards_train/rejected": -6.028164863586426,
      "step": 256
    },
    {
      "epoch": 2.94,
      "learning_rate": 3.083183770162812e-08,
      "loss": 0.0762,
      "step": 257
    },
    {
      "epoch": 2.94,
      "logps_train/chosen": -165.29769897460938,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -193.4224395751953,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.089618444442749,
      "rewards_train/margins": 3.9063212871551514,
      "rewards_train/rejected": -5.9959397315979,
      "step": 257
    },
    {
      "epoch": 2.94,
      "logps_train/chosen": -175.6675567626953,
      "logps_train/ref_chosen": -154.0,
      "logps_train/ref_rejected": -142.0,
      "logps_train/rejected": -212.16360473632812,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.1210525035858154,
      "rewards_train/margins": 4.94413685798645,
      "rewards_train/rejected": -7.065189361572266,
      "step": 257
    },
    {
      "epoch": 2.94,
      "logps_train/chosen": -146.52691650390625,
      "logps_train/ref_chosen": -129.0,
      "logps_train/ref_rejected": -102.5,
      "logps_train/rejected": -156.18231201171875,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.743121862411499,
      "rewards_train/margins": 3.6365344524383545,
      "rewards_train/rejected": -5.3796563148498535,
      "step": 257
    },
    {
      "epoch": 2.94,
      "logps_train/chosen": -187.83929443359375,
      "logps_train/ref_chosen": -167.0,
      "logps_train/ref_rejected": -122.5,
      "logps_train/rejected": -187.14642333984375,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0775322914123535,
      "rewards_train/margins": 4.39111328125,
      "rewards_train/rejected": -6.4686455726623535,
      "step": 257
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.73444687298685e-08,
      "loss": 0.077,
      "step": 258
    },
    {
      "epoch": 2.95,
      "logps_train/chosen": -177.5012969970703,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -104.5,
      "logps_train/rejected": -168.67471313476562,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.313605308532715,
      "rewards_train/margins": 4.108870983123779,
      "rewards_train/rejected": -6.422476291656494,
      "step": 258
    },
    {
      "epoch": 2.95,
      "logps_train/chosen": -150.55606079101562,
      "logps_train/ref_chosen": -134.0,
      "logps_train/ref_rejected": -132.0,
      "logps_train/rejected": -195.2097930908203,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.6292637586593628,
      "rewards_train/margins": 4.666714787483215,
      "rewards_train/rejected": -6.295978546142578,
      "step": 258
    },
    {
      "epoch": 2.95,
      "logps_train/chosen": -166.3050537109375,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -125.0,
      "logps_train/rejected": -192.82681274414062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.15824031829834,
      "rewards_train/margins": 4.585964679718018,
      "rewards_train/rejected": -6.744204998016357,
      "step": 258
    },
    {
      "epoch": 2.95,
      "logps_train/chosen": -156.59942626953125,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -103.0,
      "logps_train/rejected": -170.30538940429688,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.3700997829437256,
      "rewards_train/margins": 4.345986604690552,
      "rewards_train/rejected": -6.716086387634277,
      "step": 258
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.709148044679481e-09,
      "loss": 0.0518,
      "step": 259
    },
    {
      "epoch": 2.96,
      "logps_train/chosen": -173.42506408691406,
      "logps_train/ref_chosen": -155.0,
      "logps_train/ref_rejected": -125.5,
      "logps_train/rejected": -185.6907196044922,
      "rewards_train/accuracies": 0.9375,
      "rewards_train/chosen": -1.8300071954727173,
      "rewards_train/margins": 4.202877402305603,
      "rewards_train/rejected": -6.03288459777832,
      "step": 259
    },
    {
      "epoch": 2.96,
      "logps_train/chosen": -176.98544311523438,
      "logps_train/ref_chosen": -156.0,
      "logps_train/ref_rejected": -113.0,
      "logps_train/rejected": -179.13150024414062,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.050905704498291,
      "rewards_train/margins": 4.533289432525635,
      "rewards_train/rejected": -6.584195137023926,
      "step": 259
    },
    {
      "epoch": 2.96,
      "logps_train/chosen": -193.09963989257812,
      "logps_train/ref_chosen": -171.0,
      "logps_train/ref_rejected": -115.5,
      "logps_train/rejected": -172.74990844726562,
      "rewards_train/accuracies": 0.90625,
      "rewards_train/chosen": -2.1755151748657227,
      "rewards_train/margins": 3.523548126220703,
      "rewards_train/rejected": -5.699063301086426,
      "step": 259
    },
    {
      "epoch": 2.96,
      "logps_train/chosen": -161.4818115234375,
      "logps_train/ref_chosen": -136.0,
      "logps_train/ref_rejected": -126.0,
      "logps_train/rejected": -198.05450439453125,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.5415897369384766,
      "rewards_train/margins": 4.642866134643555,
      "rewards_train/rejected": -7.184455871582031,
      "step": 259
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.9273613056008944e-09,
      "loss": 0.0788,
      "step": 260
    },
    {
      "epoch": 2.97,
      "logps_train/chosen": -185.74722290039062,
      "logps_train/ref_chosen": -164.0,
      "logps_train/ref_rejected": -123.0,
      "logps_train/rejected": -188.15093994140625,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -2.189370632171631,
      "rewards_train/margins": 4.313614845275879,
      "rewards_train/rejected": -6.50298547744751,
      "step": 260
    },
    {
      "epoch": 2.97,
      "logps_train/chosen": -200.0399932861328,
      "logps_train/ref_chosen": -180.0,
      "logps_train/ref_rejected": -118.5,
      "logps_train/rejected": -180.8087615966797,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -2.0358355045318604,
      "rewards_train/margins": 4.2098352909088135,
      "rewards_train/rejected": -6.245670795440674,
      "step": 260
    },
    {
      "epoch": 2.97,
      "logps_train/chosen": -152.52565002441406,
      "logps_train/ref_chosen": -133.0,
      "logps_train/ref_rejected": -94.5,
      "logps_train/rejected": -153.40609741210938,
      "rewards_train/accuracies": 0.96875,
      "rewards_train/chosen": -1.9625990390777588,
      "rewards_train/margins": 3.9296958446502686,
      "rewards_train/rejected": -5.892294883728027,
      "step": 260
    },
    {
      "epoch": 2.97,
      "logps_train/chosen": -162.84246826171875,
      "logps_train/ref_chosen": -144.0,
      "logps_train/ref_rejected": -133.0,
      "logps_train/rejected": -196.57708740234375,
      "rewards_train/accuracies": 1.0,
      "rewards_train/chosen": -1.8697924613952637,
      "rewards_train/margins": 4.492800235748291,
      "rewards_train/rejected": -6.362592697143555,
      "step": 260
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0,
      "loss": 0.06,
      "step": 261
    },
    {
      "epoch": 2.98,
      "step": 261,
      "total_flos": 0.0,
      "train_loss": 0.26605752715662523,
      "train_runtime": 15002.1853,
      "train_samples_per_second": 2.236,
      "train_steps_per_second": 0.017
    }
  ],
  "max_steps": 261,
  "num_train_epochs": 3,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}