bloom-1b1-zh-error-correction-dpo / trainer_state.json
p208p2002's picture
init
6732c6e
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.0625,
"eval_steps": 500,
"global_step": 250,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.99e-07,
"logits/chosen": -9.150251388549805,
"logits/rejected": -8.951294898986816,
"logps/chosen": -39.82106399536133,
"logps/rejected": -51.287376403808594,
"loss": 0.6931,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.979999999999999e-07,
"logits/chosen": -8.701302528381348,
"logits/rejected": -8.59887981414795,
"logps/chosen": -38.8251953125,
"logps/rejected": -50.281246185302734,
"loss": 0.6902,
"rewards/accuracies": 0.9000000953674316,
"rewards/chosen": 0.005459675099700689,
"rewards/margins": 0.005827327724546194,
"rewards/rejected": -0.00036765271215699613,
"step": 2
},
{
"epoch": 0.0,
"learning_rate": 4.97e-07,
"logits/chosen": -8.877432823181152,
"logits/rejected": -8.745807647705078,
"logps/chosen": -39.213958740234375,
"logps/rejected": -53.814178466796875,
"loss": 0.6868,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.009096940979361534,
"rewards/margins": 0.012745475396513939,
"rewards/rejected": -0.003648536978289485,
"step": 3
},
{
"epoch": 0.0,
"learning_rate": 4.96e-07,
"logits/chosen": -8.850730895996094,
"logits/rejected": -8.655380249023438,
"logps/chosen": -41.56839370727539,
"logps/rejected": -49.78559112548828,
"loss": 0.685,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.010362102650105953,
"rewards/margins": 0.016426045447587967,
"rewards/rejected": -0.006063942797482014,
"step": 4
},
{
"epoch": 0.0,
"learning_rate": 4.95e-07,
"logits/chosen": -9.169770240783691,
"logits/rejected": -9.068281173706055,
"logps/chosen": -39.02599334716797,
"logps/rejected": -54.13230514526367,
"loss": 0.6802,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.017099833115935326,
"rewards/margins": 0.026187023147940636,
"rewards/rejected": -0.009087189100682735,
"step": 5
},
{
"epoch": 0.0,
"learning_rate": 4.94e-07,
"logits/chosen": -8.943150520324707,
"logits/rejected": -8.760871887207031,
"logps/chosen": -40.129783630371094,
"logps/rejected": -51.28352737426758,
"loss": 0.6759,
"rewards/accuracies": 0.8799999952316284,
"rewards/chosen": 0.018004529178142548,
"rewards/margins": 0.03507697209715843,
"rewards/rejected": -0.017072444781661034,
"step": 6
},
{
"epoch": 0.0,
"learning_rate": 4.93e-07,
"logits/chosen": -8.973767280578613,
"logits/rejected": -8.846033096313477,
"logps/chosen": -38.954933166503906,
"logps/rejected": -52.8114128112793,
"loss": 0.6694,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.01600685343146324,
"rewards/margins": 0.048318009823560715,
"rewards/rejected": -0.032311152666807175,
"step": 7
},
{
"epoch": 0.0,
"learning_rate": 4.92e-07,
"logits/chosen": -9.100973129272461,
"logits/rejected": -8.930387496948242,
"logps/chosen": -36.695037841796875,
"logps/rejected": -51.42365646362305,
"loss": 0.6664,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": 0.028073444962501526,
"rewards/margins": 0.054737381637096405,
"rewards/rejected": -0.026663940399885178,
"step": 8
},
{
"epoch": 0.0,
"learning_rate": 4.909999999999999e-07,
"logits/chosen": -8.858081817626953,
"logits/rejected": -8.691803932189941,
"logps/chosen": -39.423805236816406,
"logps/rejected": -53.124473571777344,
"loss": 0.6663,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.006643497850745916,
"rewards/margins": 0.05475940182805061,
"rewards/rejected": -0.04811590164899826,
"step": 9
},
{
"epoch": 0.0,
"learning_rate": 4.9e-07,
"logits/chosen": -8.932714462280273,
"logits/rejected": -8.752630233764648,
"logps/chosen": -40.90498733520508,
"logps/rejected": -53.95857620239258,
"loss": 0.6667,
"rewards/accuracies": 0.9000000953674316,
"rewards/chosen": 0.022381644695997238,
"rewards/margins": 0.054132066667079926,
"rewards/rejected": -0.031750429421663284,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 4.89e-07,
"logits/chosen": -8.926645278930664,
"logits/rejected": -8.795331954956055,
"logps/chosen": -39.50680923461914,
"logps/rejected": -49.81561279296875,
"loss": 0.6657,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": 0.02121429704129696,
"rewards/margins": 0.05642315000295639,
"rewards/rejected": -0.035208847373723984,
"step": 11
},
{
"epoch": 0.0,
"learning_rate": 4.879999999999999e-07,
"logits/chosen": -9.18893814086914,
"logits/rejected": -9.020618438720703,
"logps/chosen": -38.57902908325195,
"logps/rejected": -54.4058952331543,
"loss": 0.6514,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.03454957157373428,
"rewards/margins": 0.0861399695277214,
"rewards/rejected": -0.05159040167927742,
"step": 12
},
{
"epoch": 0.0,
"learning_rate": 4.87e-07,
"logits/chosen": -8.865549087524414,
"logits/rejected": -8.72575569152832,
"logps/chosen": -37.333641052246094,
"logps/rejected": -50.77653884887695,
"loss": 0.6508,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.045416563749313354,
"rewards/margins": 0.08835401386022568,
"rewards/rejected": -0.04293745011091232,
"step": 13
},
{
"epoch": 0.0,
"learning_rate": 4.86e-07,
"logits/chosen": -9.011914253234863,
"logits/rejected": -8.871015548706055,
"logps/chosen": -37.863914489746094,
"logps/rejected": -51.89360427856445,
"loss": 0.6456,
"rewards/accuracies": 0.8799999952316284,
"rewards/chosen": 0.05132318660616875,
"rewards/margins": 0.09992541372776031,
"rewards/rejected": -0.048602234572172165,
"step": 14
},
{
"epoch": 0.0,
"learning_rate": 4.85e-07,
"logits/chosen": -9.177026748657227,
"logits/rejected": -9.040088653564453,
"logps/chosen": -38.882564544677734,
"logps/rejected": -49.73377227783203,
"loss": 0.647,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.021302934736013412,
"rewards/margins": 0.09643281996250153,
"rewards/rejected": -0.07512988150119781,
"step": 15
},
{
"epoch": 0.0,
"learning_rate": 4.839999999999999e-07,
"logits/chosen": -8.8854398727417,
"logits/rejected": -8.71996784210205,
"logps/chosen": -41.88127517700195,
"logps/rejected": -53.972862243652344,
"loss": 0.6432,
"rewards/accuracies": 0.8400000333786011,
"rewards/chosen": 0.01929306425154209,
"rewards/margins": 0.10526645183563232,
"rewards/rejected": -0.08597338944673538,
"step": 16
},
{
"epoch": 0.0,
"learning_rate": 4.83e-07,
"logits/chosen": -8.87810230255127,
"logits/rejected": -8.729606628417969,
"logps/chosen": -37.907958984375,
"logps/rejected": -54.153953552246094,
"loss": 0.6297,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.026595568284392357,
"rewards/margins": 0.13360002636909485,
"rewards/rejected": -0.10700444877147675,
"step": 17
},
{
"epoch": 0.0,
"learning_rate": 4.82e-07,
"logits/chosen": -9.304192543029785,
"logits/rejected": -9.190086364746094,
"logps/chosen": -38.377891540527344,
"logps/rejected": -51.592796325683594,
"loss": 0.6391,
"rewards/accuracies": 0.8399999737739563,
"rewards/chosen": 0.06439777463674545,
"rewards/margins": 0.11593715101480484,
"rewards/rejected": -0.051539380103349686,
"step": 18
},
{
"epoch": 0.0,
"learning_rate": 4.809999999999999e-07,
"logits/chosen": -8.877577781677246,
"logits/rejected": -8.728302001953125,
"logps/chosen": -36.82625198364258,
"logps/rejected": -50.772274017333984,
"loss": 0.6317,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.04051102325320244,
"rewards/margins": 0.13047367334365845,
"rewards/rejected": -0.08996264636516571,
"step": 19
},
{
"epoch": 0.01,
"learning_rate": 4.8e-07,
"logits/chosen": -8.93675422668457,
"logits/rejected": -8.837154388427734,
"logps/chosen": -38.22481918334961,
"logps/rejected": -51.28339385986328,
"loss": 0.6327,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.024285469204187393,
"rewards/margins": 0.12898896634578705,
"rewards/rejected": -0.10470350831747055,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.79e-07,
"logits/chosen": -8.971104621887207,
"logits/rejected": -8.809735298156738,
"logps/chosen": -38.27422332763672,
"logps/rejected": -51.61455154418945,
"loss": 0.6212,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": 0.04925350099802017,
"rewards/margins": 0.1548921763896942,
"rewards/rejected": -0.10563866049051285,
"step": 21
},
{
"epoch": 0.01,
"learning_rate": 4.779999999999999e-07,
"logits/chosen": -8.943137168884277,
"logits/rejected": -8.724119186401367,
"logps/chosen": -39.740211486816406,
"logps/rejected": -53.67293167114258,
"loss": 0.6185,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.031043654307723045,
"rewards/margins": 0.15891726315021515,
"rewards/rejected": -0.12787359952926636,
"step": 22
},
{
"epoch": 0.01,
"learning_rate": 4.769999999999999e-07,
"logits/chosen": -8.783563613891602,
"logits/rejected": -8.568138122558594,
"logps/chosen": -40.75530242919922,
"logps/rejected": -55.012939453125,
"loss": 0.6142,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.022519828751683235,
"rewards/margins": 0.1669926643371582,
"rewards/rejected": -0.14447283744812012,
"step": 23
},
{
"epoch": 0.01,
"learning_rate": 4.76e-07,
"logits/chosen": -8.948726654052734,
"logits/rejected": -8.741633415222168,
"logps/chosen": -38.42156982421875,
"logps/rejected": -54.19482421875,
"loss": 0.6141,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": 0.042123936116695404,
"rewards/margins": 0.16957491636276245,
"rewards/rejected": -0.12745098769664764,
"step": 24
},
{
"epoch": 0.01,
"learning_rate": 4.7499999999999995e-07,
"logits/chosen": -8.800437927246094,
"logits/rejected": -8.649993896484375,
"logps/chosen": -39.37318420410156,
"logps/rejected": -50.02665710449219,
"loss": 0.6253,
"rewards/accuracies": 0.8400000333786011,
"rewards/chosen": 0.004943800158798695,
"rewards/margins": 0.14754608273506165,
"rewards/rejected": -0.14260227978229523,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 4.7399999999999993e-07,
"logits/chosen": -9.165999412536621,
"logits/rejected": -9.073251724243164,
"logps/chosen": -37.4705696105957,
"logps/rejected": -50.39186477661133,
"loss": 0.613,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.005681462120264769,
"rewards/margins": 0.17087292671203613,
"rewards/rejected": -0.16519147157669067,
"step": 26
},
{
"epoch": 0.01,
"learning_rate": 4.7299999999999996e-07,
"logits/chosen": -8.994885444641113,
"logits/rejected": -8.7987642288208,
"logps/chosen": -37.67160415649414,
"logps/rejected": -54.15082931518555,
"loss": 0.5968,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.03374551981687546,
"rewards/margins": 0.20702290534973145,
"rewards/rejected": -0.1732773780822754,
"step": 27
},
{
"epoch": 0.01,
"learning_rate": 4.7199999999999994e-07,
"logits/chosen": -8.93022346496582,
"logits/rejected": -8.841963768005371,
"logps/chosen": -37.873985290527344,
"logps/rejected": -56.12858200073242,
"loss": 0.6008,
"rewards/accuracies": 0.8399999737739563,
"rewards/chosen": 0.00800693966448307,
"rewards/margins": 0.2039312869310379,
"rewards/rejected": -0.19592434167861938,
"step": 28
},
{
"epoch": 0.01,
"learning_rate": 4.7099999999999997e-07,
"logits/chosen": -9.13399600982666,
"logits/rejected": -8.960123062133789,
"logps/chosen": -37.81399917602539,
"logps/rejected": -53.7081298828125,
"loss": 0.5955,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.03130952641367912,
"rewards/margins": 0.21175985038280487,
"rewards/rejected": -0.24306941032409668,
"step": 29
},
{
"epoch": 0.01,
"learning_rate": 4.6999999999999995e-07,
"logits/chosen": -8.995954513549805,
"logits/rejected": -8.84068775177002,
"logps/chosen": -38.8903923034668,
"logps/rejected": -57.47145462036133,
"loss": 0.5892,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.044602252542972565,
"rewards/margins": 0.22832927107810974,
"rewards/rejected": -0.18372702598571777,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 4.689999999999999e-07,
"logits/chosen": -8.951233863830566,
"logits/rejected": -8.74404525756836,
"logps/chosen": -39.165775299072266,
"logps/rejected": -53.60399627685547,
"loss": 0.5901,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.005719953216612339,
"rewards/margins": 0.2242022305727005,
"rewards/rejected": -0.22992220520973206,
"step": 31
},
{
"epoch": 0.01,
"learning_rate": 4.68e-07,
"logits/chosen": -8.605962753295898,
"logits/rejected": -8.464166641235352,
"logps/chosen": -39.067291259765625,
"logps/rejected": -56.53247833251953,
"loss": 0.565,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.03435752913355827,
"rewards/margins": 0.28634151816368103,
"rewards/rejected": -0.2519839406013489,
"step": 32
},
{
"epoch": 0.01,
"learning_rate": 4.67e-07,
"logits/chosen": -8.780204772949219,
"logits/rejected": -8.59669017791748,
"logps/chosen": -39.499412536621094,
"logps/rejected": -55.1195068359375,
"loss": 0.5906,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.012880785390734673,
"rewards/margins": 0.22619274258613586,
"rewards/rejected": -0.21331194043159485,
"step": 33
},
{
"epoch": 0.01,
"learning_rate": 4.66e-07,
"logits/chosen": -9.019442558288574,
"logits/rejected": -8.858607292175293,
"logps/chosen": -39.38768768310547,
"logps/rejected": -56.47705841064453,
"loss": 0.5818,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.029814431443810463,
"rewards/margins": 0.24373717606067657,
"rewards/rejected": -0.21392273902893066,
"step": 34
},
{
"epoch": 0.01,
"learning_rate": 4.65e-07,
"logits/chosen": -8.972381591796875,
"logits/rejected": -8.825788497924805,
"logps/chosen": -39.450408935546875,
"logps/rejected": -54.7730827331543,
"loss": 0.5931,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": 0.01736309379339218,
"rewards/margins": 0.22288131713867188,
"rewards/rejected": -0.2055182158946991,
"step": 35
},
{
"epoch": 0.01,
"learning_rate": 4.64e-07,
"logits/chosen": -8.794334411621094,
"logits/rejected": -8.642388343811035,
"logps/chosen": -38.96186447143555,
"logps/rejected": -52.460182189941406,
"loss": 0.5874,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": 0.03622301667928696,
"rewards/margins": 0.23951515555381775,
"rewards/rejected": -0.20329216122627258,
"step": 36
},
{
"epoch": 0.01,
"learning_rate": 4.63e-07,
"logits/chosen": -8.897015571594238,
"logits/rejected": -8.766530990600586,
"logps/chosen": -37.4925537109375,
"logps/rejected": -50.289085388183594,
"loss": 0.5946,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": 0.008345548994839191,
"rewards/margins": 0.21366062760353088,
"rewards/rejected": -0.20531506836414337,
"step": 37
},
{
"epoch": 0.01,
"learning_rate": 4.62e-07,
"logits/chosen": -9.160964965820312,
"logits/rejected": -9.01344108581543,
"logps/chosen": -39.84162139892578,
"logps/rejected": -50.35344696044922,
"loss": 0.5878,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.03968983516097069,
"rewards/margins": 0.23294827342033386,
"rewards/rejected": -0.19325841963291168,
"step": 38
},
{
"epoch": 0.01,
"learning_rate": 4.61e-07,
"logits/chosen": -8.879249572753906,
"logits/rejected": -8.736249923706055,
"logps/chosen": -39.56095504760742,
"logps/rejected": -55.347694396972656,
"loss": 0.5662,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.02437666617333889,
"rewards/margins": 0.28601738810539246,
"rewards/rejected": -0.261640727519989,
"step": 39
},
{
"epoch": 0.01,
"learning_rate": 4.6e-07,
"logits/chosen": -9.063352584838867,
"logits/rejected": -8.963701248168945,
"logps/chosen": -38.563880920410156,
"logps/rejected": -51.63567352294922,
"loss": 0.5769,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.02472488023340702,
"rewards/margins": 0.2595635950565338,
"rewards/rejected": -0.2842884659767151,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 4.59e-07,
"logits/chosen": -9.278203964233398,
"logits/rejected": -9.08998966217041,
"logps/chosen": -39.864532470703125,
"logps/rejected": -54.92973709106445,
"loss": 0.5737,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.020174704492092133,
"rewards/margins": 0.26886457204818726,
"rewards/rejected": -0.24868984520435333,
"step": 41
},
{
"epoch": 0.01,
"learning_rate": 4.58e-07,
"logits/chosen": -8.939000129699707,
"logits/rejected": -8.744054794311523,
"logps/chosen": -37.10688400268555,
"logps/rejected": -54.631141662597656,
"loss": 0.5495,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.04707593470811844,
"rewards/margins": 0.3288743495941162,
"rewards/rejected": -0.28179842233657837,
"step": 42
},
{
"epoch": 0.01,
"learning_rate": 4.57e-07,
"logits/chosen": -8.78071403503418,
"logits/rejected": -8.645574569702148,
"logps/chosen": -38.05216598510742,
"logps/rejected": -54.2934684753418,
"loss": 0.5586,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.01882551982998848,
"rewards/margins": 0.3048241138458252,
"rewards/rejected": -0.2859985828399658,
"step": 43
},
{
"epoch": 0.01,
"learning_rate": 4.56e-07,
"logits/chosen": -8.890421867370605,
"logits/rejected": -8.79121208190918,
"logps/chosen": -36.806209564208984,
"logps/rejected": -54.96057891845703,
"loss": 0.5378,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.09500815719366074,
"rewards/margins": 0.3511715531349182,
"rewards/rejected": -0.2561633884906769,
"step": 44
},
{
"epoch": 0.01,
"learning_rate": 4.55e-07,
"logits/chosen": -8.712895393371582,
"logits/rejected": -8.633960723876953,
"logps/chosen": -38.36237335205078,
"logps/rejected": -53.05036163330078,
"loss": 0.5671,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.10673508793115616,
"rewards/margins": 0.28717535734176636,
"rewards/rejected": -0.1804402768611908,
"step": 45
},
{
"epoch": 0.01,
"learning_rate": 4.54e-07,
"logits/chosen": -9.085457801818848,
"logits/rejected": -8.948368072509766,
"logps/chosen": -40.31169128417969,
"logps/rejected": -56.712364196777344,
"loss": 0.5538,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.05474279075860977,
"rewards/margins": 0.32782307267189026,
"rewards/rejected": -0.2730802297592163,
"step": 46
},
{
"epoch": 0.01,
"learning_rate": 4.53e-07,
"logits/chosen": -8.925232887268066,
"logits/rejected": -8.793390274047852,
"logps/chosen": -40.80510711669922,
"logps/rejected": -53.9893798828125,
"loss": 0.5603,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.004703240934759378,
"rewards/margins": 0.2981411814689636,
"rewards/rejected": -0.30284440517425537,
"step": 47
},
{
"epoch": 0.01,
"learning_rate": 4.5199999999999997e-07,
"logits/chosen": -8.936556816101074,
"logits/rejected": -8.795175552368164,
"logps/chosen": -39.4090690612793,
"logps/rejected": -56.091827392578125,
"loss": 0.5701,
"rewards/accuracies": 0.8799999952316284,
"rewards/chosen": -0.02163396216928959,
"rewards/margins": 0.28110483288764954,
"rewards/rejected": -0.3027387261390686,
"step": 48
},
{
"epoch": 0.01,
"learning_rate": 4.51e-07,
"logits/chosen": -8.785165786743164,
"logits/rejected": -8.663257598876953,
"logps/chosen": -39.04924011230469,
"logps/rejected": -55.55643844604492,
"loss": 0.5291,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.07937277853488922,
"rewards/margins": 0.3831271529197693,
"rewards/rejected": -0.3037543296813965,
"step": 49
},
{
"epoch": 0.01,
"learning_rate": 4.5e-07,
"logits/chosen": -8.918477058410645,
"logits/rejected": -8.69984245300293,
"logps/chosen": -39.793434143066406,
"logps/rejected": -51.4716911315918,
"loss": 0.5581,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.030612414702773094,
"rewards/margins": 0.3040314316749573,
"rewards/rejected": -0.3346438705921173,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 4.49e-07,
"logits/chosen": -8.830556869506836,
"logits/rejected": -8.661957740783691,
"logps/chosen": -39.756195068359375,
"logps/rejected": -55.29694747924805,
"loss": 0.5275,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.03386972099542618,
"rewards/margins": 0.3903976380825043,
"rewards/rejected": -0.3565279543399811,
"step": 51
},
{
"epoch": 0.01,
"learning_rate": 4.48e-07,
"logits/chosen": -8.969281196594238,
"logits/rejected": -8.82741641998291,
"logps/chosen": -37.20234298706055,
"logps/rejected": -59.197296142578125,
"loss": 0.5128,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.04513595253229141,
"rewards/margins": 0.429348886013031,
"rewards/rejected": -0.3842129111289978,
"step": 52
},
{
"epoch": 0.01,
"learning_rate": 4.4699999999999997e-07,
"logits/chosen": -8.737951278686523,
"logits/rejected": -8.606011390686035,
"logps/chosen": -38.3763542175293,
"logps/rejected": -58.59581756591797,
"loss": 0.5162,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.030918415635824203,
"rewards/margins": 0.41947564482688904,
"rewards/rejected": -0.3885572552680969,
"step": 53
},
{
"epoch": 0.01,
"learning_rate": 4.46e-07,
"logits/chosen": -8.750028610229492,
"logits/rejected": -8.646293640136719,
"logps/chosen": -38.137672424316406,
"logps/rejected": -56.8802490234375,
"loss": 0.5144,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.08223019540309906,
"rewards/margins": 0.42382898926734924,
"rewards/rejected": -0.3415988087654114,
"step": 54
},
{
"epoch": 0.01,
"learning_rate": 4.45e-07,
"logits/chosen": -9.04769515991211,
"logits/rejected": -8.884363174438477,
"logps/chosen": -39.094200134277344,
"logps/rejected": -56.6423454284668,
"loss": 0.5296,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.01863374561071396,
"rewards/margins": 0.3876304030418396,
"rewards/rejected": -0.36899662017822266,
"step": 55
},
{
"epoch": 0.01,
"learning_rate": 4.44e-07,
"logits/chosen": -8.93348503112793,
"logits/rejected": -8.77459716796875,
"logps/chosen": -38.1727294921875,
"logps/rejected": -51.39598846435547,
"loss": 0.5394,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.08033865690231323,
"rewards/margins": 0.3627088963985443,
"rewards/rejected": -0.2823702096939087,
"step": 56
},
{
"epoch": 0.01,
"learning_rate": 4.43e-07,
"logits/chosen": -8.88620376586914,
"logits/rejected": -8.79748821258545,
"logps/chosen": -38.02767562866211,
"logps/rejected": -53.46161651611328,
"loss": 0.5248,
"rewards/accuracies": 0.8400000333786011,
"rewards/chosen": 0.07710711658000946,
"rewards/margins": 0.404231458902359,
"rewards/rejected": -0.32712429761886597,
"step": 57
},
{
"epoch": 0.01,
"learning_rate": 4.4199999999999996e-07,
"logits/chosen": -9.048208236694336,
"logits/rejected": -8.921117782592773,
"logps/chosen": -39.349849700927734,
"logps/rejected": -57.04978561401367,
"loss": 0.525,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.005890236236155033,
"rewards/margins": 0.3883860111236572,
"rewards/rejected": -0.38249582052230835,
"step": 58
},
{
"epoch": 0.01,
"learning_rate": 4.41e-07,
"logits/chosen": -9.050989151000977,
"logits/rejected": -8.864224433898926,
"logps/chosen": -38.37465286254883,
"logps/rejected": -54.69602584838867,
"loss": 0.5161,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": -0.00012345091090537608,
"rewards/margins": 0.41539669036865234,
"rewards/rejected": -0.4155201017856598,
"step": 59
},
{
"epoch": 0.01,
"learning_rate": 4.3999999999999997e-07,
"logits/chosen": -8.77641773223877,
"logits/rejected": -8.65733528137207,
"logps/chosen": -37.439788818359375,
"logps/rejected": -56.8023681640625,
"loss": 0.491,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.10770513862371445,
"rewards/margins": 0.4838793873786926,
"rewards/rejected": -0.37617427110671997,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 4.39e-07,
"logits/chosen": -8.84074878692627,
"logits/rejected": -8.694551467895508,
"logps/chosen": -40.056434631347656,
"logps/rejected": -60.27661895751953,
"loss": 0.5015,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.010883894748985767,
"rewards/margins": 0.4609171748161316,
"rewards/rejected": -0.4500332772731781,
"step": 61
},
{
"epoch": 0.02,
"learning_rate": 4.38e-07,
"logits/chosen": -8.841463088989258,
"logits/rejected": -8.691165924072266,
"logps/chosen": -38.9443473815918,
"logps/rejected": -56.7674446105957,
"loss": 0.5052,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.058075837790966034,
"rewards/margins": 0.45668134093284607,
"rewards/rejected": -0.39860549569129944,
"step": 62
},
{
"epoch": 0.02,
"learning_rate": 4.3699999999999996e-07,
"logits/chosen": -8.556544303894043,
"logits/rejected": -8.511371612548828,
"logps/chosen": -37.55242919921875,
"logps/rejected": -52.5374755859375,
"loss": 0.5261,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.05054298788309097,
"rewards/margins": 0.3981621265411377,
"rewards/rejected": -0.3476191461086273,
"step": 63
},
{
"epoch": 0.02,
"learning_rate": 4.36e-07,
"logits/chosen": -8.824804306030273,
"logits/rejected": -8.721410751342773,
"logps/chosen": -39.77140808105469,
"logps/rejected": -60.0277099609375,
"loss": 0.4751,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.01679672673344612,
"rewards/margins": 0.5294073820114136,
"rewards/rejected": -0.5462040901184082,
"step": 64
},
{
"epoch": 0.02,
"learning_rate": 4.3499999999999996e-07,
"logits/chosen": -8.601284980773926,
"logits/rejected": -8.539445877075195,
"logps/chosen": -40.41975402832031,
"logps/rejected": -55.65443801879883,
"loss": 0.5237,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.10446874797344208,
"rewards/margins": 0.39701706171035767,
"rewards/rejected": -0.5014857649803162,
"step": 65
},
{
"epoch": 0.02,
"learning_rate": 4.34e-07,
"logits/chosen": -9.016075134277344,
"logits/rejected": -8.911613464355469,
"logps/chosen": -36.6504020690918,
"logps/rejected": -52.8062629699707,
"loss": 0.4991,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.03334439545869827,
"rewards/margins": 0.46417728066444397,
"rewards/rejected": -0.4308328628540039,
"step": 66
},
{
"epoch": 0.02,
"learning_rate": 4.3299999999999997e-07,
"logits/chosen": -9.052962303161621,
"logits/rejected": -8.949380874633789,
"logps/chosen": -37.03533172607422,
"logps/rejected": -52.54149627685547,
"loss": 0.5236,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.004350895527750254,
"rewards/margins": 0.41073599457740784,
"rewards/rejected": -0.4063850939273834,
"step": 67
},
{
"epoch": 0.02,
"learning_rate": 4.3199999999999995e-07,
"logits/chosen": -9.015048027038574,
"logits/rejected": -8.912318229675293,
"logps/chosen": -38.99870300292969,
"logps/rejected": -52.52238082885742,
"loss": 0.517,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.050432223826646805,
"rewards/margins": 0.4209260046482086,
"rewards/rejected": -0.3704938292503357,
"step": 68
},
{
"epoch": 0.02,
"learning_rate": 4.31e-07,
"logits/chosen": -8.882599830627441,
"logits/rejected": -8.721563339233398,
"logps/chosen": -39.24352264404297,
"logps/rejected": -57.102867126464844,
"loss": 0.4889,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.05120646953582764,
"rewards/margins": 0.5033482909202576,
"rewards/rejected": -0.4521418511867523,
"step": 69
},
{
"epoch": 0.02,
"learning_rate": 4.2999999999999996e-07,
"logits/chosen": -8.721124649047852,
"logits/rejected": -8.638886451721191,
"logps/chosen": -40.13719940185547,
"logps/rejected": -56.29268264770508,
"loss": 0.5076,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.024597348645329475,
"rewards/margins": 0.4424295425415039,
"rewards/rejected": -0.46702688932418823,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 4.29e-07,
"logits/chosen": -8.832576751708984,
"logits/rejected": -8.6920747756958,
"logps/chosen": -39.40597152709961,
"logps/rejected": -57.85718536376953,
"loss": 0.4844,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.0307625625282526,
"rewards/margins": 0.5312424302101135,
"rewards/rejected": -0.5004798769950867,
"step": 71
},
{
"epoch": 0.02,
"learning_rate": 4.2799999999999997e-07,
"logits/chosen": -8.825450897216797,
"logits/rejected": -8.670401573181152,
"logps/chosen": -39.127418518066406,
"logps/rejected": -58.1724967956543,
"loss": 0.4951,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.04789305105805397,
"rewards/margins": 0.48660898208618164,
"rewards/rejected": -0.43871593475341797,
"step": 72
},
{
"epoch": 0.02,
"learning_rate": 4.2699999999999995e-07,
"logits/chosen": -9.031660079956055,
"logits/rejected": -8.88580322265625,
"logps/chosen": -40.26874542236328,
"logps/rejected": -59.174354553222656,
"loss": 0.4813,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.056357551366090775,
"rewards/margins": 0.5150011777877808,
"rewards/rejected": -0.4586435854434967,
"step": 73
},
{
"epoch": 0.02,
"learning_rate": 4.26e-07,
"logits/chosen": -9.249466896057129,
"logits/rejected": -9.104633331298828,
"logps/chosen": -38.35686492919922,
"logps/rejected": -58.423255920410156,
"loss": 0.4694,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.003165402915328741,
"rewards/margins": 0.5658671855926514,
"rewards/rejected": -0.5627016425132751,
"step": 74
},
{
"epoch": 0.02,
"learning_rate": 4.2499999999999995e-07,
"logits/chosen": -8.971678733825684,
"logits/rejected": -8.806716918945312,
"logps/chosen": -40.63257598876953,
"logps/rejected": -58.69512939453125,
"loss": 0.4714,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": -0.003097705077379942,
"rewards/margins": 0.5738621950149536,
"rewards/rejected": -0.5769599676132202,
"step": 75
},
{
"epoch": 0.02,
"learning_rate": 4.24e-07,
"logits/chosen": -8.751283645629883,
"logits/rejected": -8.613395690917969,
"logps/chosen": -41.808475494384766,
"logps/rejected": -57.4010009765625,
"loss": 0.5206,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": -0.018361693248152733,
"rewards/margins": 0.4181668758392334,
"rewards/rejected": -0.4365285038948059,
"step": 76
},
{
"epoch": 0.02,
"learning_rate": 4.2299999999999996e-07,
"logits/chosen": -8.919511795043945,
"logits/rejected": -8.767281532287598,
"logps/chosen": -38.50531005859375,
"logps/rejected": -56.9246826171875,
"loss": 0.5066,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08143829554319382,
"rewards/margins": 0.46956175565719604,
"rewards/rejected": -0.5510000586509705,
"step": 77
},
{
"epoch": 0.02,
"learning_rate": 4.2199999999999994e-07,
"logits/chosen": -9.05775260925293,
"logits/rejected": -8.882523536682129,
"logps/chosen": -37.908966064453125,
"logps/rejected": -57.79386520385742,
"loss": 0.4638,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.06249593570828438,
"rewards/margins": 0.5957901477813721,
"rewards/rejected": -0.5332942008972168,
"step": 78
},
{
"epoch": 0.02,
"learning_rate": 4.2099999999999997e-07,
"logits/chosen": -9.013116836547852,
"logits/rejected": -8.926069259643555,
"logps/chosen": -37.548606872558594,
"logps/rejected": -54.40301513671875,
"loss": 0.4867,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08341957628726959,
"rewards/margins": 0.5038079023361206,
"rewards/rejected": -0.4203883111476898,
"step": 79
},
{
"epoch": 0.02,
"learning_rate": 4.1999999999999995e-07,
"logits/chosen": -8.759903907775879,
"logits/rejected": -8.721986770629883,
"logps/chosen": -38.83586120605469,
"logps/rejected": -56.38386154174805,
"loss": 0.4749,
"rewards/accuracies": 0.9000000953674316,
"rewards/chosen": 0.03295533359050751,
"rewards/margins": 0.5568066239356995,
"rewards/rejected": -0.5238512754440308,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 4.19e-07,
"logits/chosen": -8.818650245666504,
"logits/rejected": -8.679798126220703,
"logps/chosen": -40.744956970214844,
"logps/rejected": -56.096473693847656,
"loss": 0.4843,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.10699313879013062,
"rewards/margins": 0.49514955282211304,
"rewards/rejected": -0.6021426916122437,
"step": 81
},
{
"epoch": 0.02,
"learning_rate": 4.1799999999999996e-07,
"logits/chosen": -8.959874153137207,
"logits/rejected": -8.873147010803223,
"logps/chosen": -38.775062561035156,
"logps/rejected": -56.94916534423828,
"loss": 0.4637,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.026896214112639427,
"rewards/margins": 0.5791338682174683,
"rewards/rejected": -0.5522376298904419,
"step": 82
},
{
"epoch": 0.02,
"learning_rate": 4.17e-07,
"logits/chosen": -9.120010375976562,
"logits/rejected": -8.946874618530273,
"logps/chosen": -41.79666519165039,
"logps/rejected": -59.282958984375,
"loss": 0.469,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.009585860185325146,
"rewards/margins": 0.5570787191390991,
"rewards/rejected": -0.5666645169258118,
"step": 83
},
{
"epoch": 0.02,
"learning_rate": 4.1599999999999997e-07,
"logits/chosen": -8.929158210754395,
"logits/rejected": -8.834013938903809,
"logps/chosen": -40.36653137207031,
"logps/rejected": -54.375205993652344,
"loss": 0.5029,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": -0.10653682053089142,
"rewards/margins": 0.4618176519870758,
"rewards/rejected": -0.5683544874191284,
"step": 84
},
{
"epoch": 0.02,
"learning_rate": 4.1499999999999994e-07,
"logits/chosen": -9.015276908874512,
"logits/rejected": -8.931076049804688,
"logps/chosen": -40.707420349121094,
"logps/rejected": -57.53357696533203,
"loss": 0.4619,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.034116435796022415,
"rewards/margins": 0.5871976613998413,
"rewards/rejected": -0.6213140487670898,
"step": 85
},
{
"epoch": 0.02,
"learning_rate": 4.14e-07,
"logits/chosen": -8.926106452941895,
"logits/rejected": -8.83165168762207,
"logps/chosen": -37.86637878417969,
"logps/rejected": -57.00706100463867,
"loss": 0.4474,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.026764903217554092,
"rewards/margins": 0.6376525163650513,
"rewards/rejected": -0.6108875870704651,
"step": 86
},
{
"epoch": 0.02,
"learning_rate": 4.1299999999999995e-07,
"logits/chosen": -8.976417541503906,
"logits/rejected": -8.816550254821777,
"logps/chosen": -40.38107681274414,
"logps/rejected": -58.815818786621094,
"loss": 0.4547,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": -0.023364685475826263,
"rewards/margins": 0.6099767088890076,
"rewards/rejected": -0.6333414316177368,
"step": 87
},
{
"epoch": 0.02,
"learning_rate": 4.12e-07,
"logits/chosen": -8.921110153198242,
"logits/rejected": -8.771123886108398,
"logps/chosen": -36.119667053222656,
"logps/rejected": -54.26972198486328,
"loss": 0.4564,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.024796072393655777,
"rewards/margins": 0.5949214100837708,
"rewards/rejected": -0.6197174787521362,
"step": 88
},
{
"epoch": 0.02,
"learning_rate": 4.1099999999999996e-07,
"logits/chosen": -8.972981452941895,
"logits/rejected": -8.839695930480957,
"logps/chosen": -37.952171325683594,
"logps/rejected": -53.723838806152344,
"loss": 0.4678,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.0343460738658905,
"rewards/margins": 0.5734801888465881,
"rewards/rejected": -0.5391340851783752,
"step": 89
},
{
"epoch": 0.02,
"learning_rate": 4.0999999999999994e-07,
"logits/chosen": -8.822635650634766,
"logits/rejected": -8.72290325164795,
"logps/chosen": -40.56664276123047,
"logps/rejected": -58.175697326660156,
"loss": 0.4662,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.11275799572467804,
"rewards/margins": 0.5608721971511841,
"rewards/rejected": -0.6736301779747009,
"step": 90
},
{
"epoch": 0.02,
"learning_rate": 4.0899999999999997e-07,
"logits/chosen": -8.966471672058105,
"logits/rejected": -8.847415924072266,
"logps/chosen": -39.2136344909668,
"logps/rejected": -55.99798583984375,
"loss": 0.4806,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.03947858139872551,
"rewards/margins": 0.5527079701423645,
"rewards/rejected": -0.5132293701171875,
"step": 91
},
{
"epoch": 0.02,
"learning_rate": 4.0799999999999995e-07,
"logits/chosen": -8.708093643188477,
"logits/rejected": -8.622647285461426,
"logps/chosen": -40.248023986816406,
"logps/rejected": -60.658241271972656,
"loss": 0.4263,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": 0.06861706078052521,
"rewards/margins": 0.7151986360549927,
"rewards/rejected": -0.6465815305709839,
"step": 92
},
{
"epoch": 0.02,
"learning_rate": 4.07e-07,
"logits/chosen": -8.927497863769531,
"logits/rejected": -8.764131546020508,
"logps/chosen": -41.07722473144531,
"logps/rejected": -57.40480422973633,
"loss": 0.4604,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.10462639480829239,
"rewards/margins": 0.577288806438446,
"rewards/rejected": -0.6819152235984802,
"step": 93
},
{
"epoch": 0.02,
"learning_rate": 4.06e-07,
"logits/chosen": -8.988286018371582,
"logits/rejected": -8.856369018554688,
"logps/chosen": -39.973121643066406,
"logps/rejected": -55.97014236450195,
"loss": 0.4614,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.0598948672413826,
"rewards/margins": 0.6045216917991638,
"rewards/rejected": -0.664416491985321,
"step": 94
},
{
"epoch": 0.02,
"learning_rate": 4.05e-07,
"logits/chosen": -8.822972297668457,
"logits/rejected": -8.72160816192627,
"logps/chosen": -40.39838409423828,
"logps/rejected": -58.671974182128906,
"loss": 0.4388,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": 0.012630686163902283,
"rewards/margins": 0.6781344413757324,
"rewards/rejected": -0.6655037999153137,
"step": 95
},
{
"epoch": 0.02,
"learning_rate": 4.04e-07,
"logits/chosen": -8.826386451721191,
"logits/rejected": -8.73621940612793,
"logps/chosen": -38.85763168334961,
"logps/rejected": -61.4788818359375,
"loss": 0.4197,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.0026855766773223877,
"rewards/margins": 0.7139507532119751,
"rewards/rejected": -0.7112652063369751,
"step": 96
},
{
"epoch": 0.02,
"learning_rate": 4.03e-07,
"logits/chosen": -9.003509521484375,
"logits/rejected": -8.840544700622559,
"logps/chosen": -39.192012786865234,
"logps/rejected": -59.52681350708008,
"loss": 0.4507,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": -0.04670552909374237,
"rewards/margins": 0.62091064453125,
"rewards/rejected": -0.6676161885261536,
"step": 97
},
{
"epoch": 0.02,
"learning_rate": 4.02e-07,
"logits/chosen": -9.014259338378906,
"logits/rejected": -8.890603065490723,
"logps/chosen": -39.755069732666016,
"logps/rejected": -62.2213134765625,
"loss": 0.4409,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.07653121650218964,
"rewards/margins": 0.6743559837341309,
"rewards/rejected": -0.7508872151374817,
"step": 98
},
{
"epoch": 0.02,
"learning_rate": 4.01e-07,
"logits/chosen": -8.92739486694336,
"logits/rejected": -8.727579116821289,
"logps/chosen": -38.63489532470703,
"logps/rejected": -56.489234924316406,
"loss": 0.4253,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.06340552121400833,
"rewards/margins": 0.6787285804748535,
"rewards/rejected": -0.7421342134475708,
"step": 99
},
{
"epoch": 0.03,
"learning_rate": 4e-07,
"logits/chosen": -8.823177337646484,
"logits/rejected": -8.733776092529297,
"logps/chosen": -39.35440444946289,
"logps/rejected": -59.88597869873047,
"loss": 0.4156,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": 0.06192043423652649,
"rewards/margins": 0.7568650245666504,
"rewards/rejected": -0.6949446201324463,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 3.99e-07,
"logits/chosen": -8.744458198547363,
"logits/rejected": -8.607200622558594,
"logps/chosen": -39.50025177001953,
"logps/rejected": -59.96971893310547,
"loss": 0.4466,
"rewards/accuracies": 0.8600000143051147,
"rewards/chosen": -0.02974173054099083,
"rewards/margins": 0.6540297269821167,
"rewards/rejected": -0.6837714910507202,
"step": 101
},
{
"epoch": 0.03,
"learning_rate": 3.98e-07,
"logits/chosen": -8.990362167358398,
"logits/rejected": -8.814451217651367,
"logps/chosen": -41.06502914428711,
"logps/rejected": -59.0604133605957,
"loss": 0.439,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.11213630437850952,
"rewards/margins": 0.681516170501709,
"rewards/rejected": -0.7936524152755737,
"step": 102
},
{
"epoch": 0.03,
"learning_rate": 3.97e-07,
"logits/chosen": -8.999361038208008,
"logits/rejected": -8.890585899353027,
"logps/chosen": -36.496822357177734,
"logps/rejected": -61.975677490234375,
"loss": 0.4139,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.07788603752851486,
"rewards/margins": 0.7719110250473022,
"rewards/rejected": -0.6940250992774963,
"step": 103
},
{
"epoch": 0.03,
"learning_rate": 3.96e-07,
"logits/chosen": -8.933026313781738,
"logits/rejected": -8.826833724975586,
"logps/chosen": -40.48444366455078,
"logps/rejected": -55.9661865234375,
"loss": 0.449,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.10673530399799347,
"rewards/margins": 0.6240180730819702,
"rewards/rejected": -0.730753481388092,
"step": 104
},
{
"epoch": 0.03,
"learning_rate": 3.95e-07,
"logits/chosen": -9.035877227783203,
"logits/rejected": -8.826123237609863,
"logps/chosen": -38.494232177734375,
"logps/rejected": -58.50568771362305,
"loss": 0.4465,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.05296442657709122,
"rewards/margins": 0.677048921585083,
"rewards/rejected": -0.7300133109092712,
"step": 105
},
{
"epoch": 0.03,
"learning_rate": 3.94e-07,
"logits/chosen": -8.989307403564453,
"logits/rejected": -8.91016674041748,
"logps/chosen": -41.43378448486328,
"logps/rejected": -57.25815963745117,
"loss": 0.4938,
"rewards/accuracies": 0.8799999952316284,
"rewards/chosen": -0.20345261693000793,
"rewards/margins": 0.5115060806274414,
"rewards/rejected": -0.7149587869644165,
"step": 106
},
{
"epoch": 0.03,
"learning_rate": 3.93e-07,
"logits/chosen": -8.774269104003906,
"logits/rejected": -8.619407653808594,
"logps/chosen": -39.817054748535156,
"logps/rejected": -54.45206832885742,
"loss": 0.4652,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.1449589878320694,
"rewards/margins": 0.5843663811683655,
"rewards/rejected": -0.7293254137039185,
"step": 107
},
{
"epoch": 0.03,
"learning_rate": 3.92e-07,
"logits/chosen": -8.860220909118652,
"logits/rejected": -8.775630950927734,
"logps/chosen": -38.79396438598633,
"logps/rejected": -58.30022430419922,
"loss": 0.4283,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.04749782010912895,
"rewards/margins": 0.7379117012023926,
"rewards/rejected": -0.6904138922691345,
"step": 108
},
{
"epoch": 0.03,
"learning_rate": 3.91e-07,
"logits/chosen": -8.848630905151367,
"logits/rejected": -8.661413192749023,
"logps/chosen": -40.61771774291992,
"logps/rejected": -60.2691764831543,
"loss": 0.4042,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.08625830709934235,
"rewards/margins": 0.773119330406189,
"rewards/rejected": -0.8593775629997253,
"step": 109
},
{
"epoch": 0.03,
"learning_rate": 3.8999999999999997e-07,
"logits/chosen": -8.671426773071289,
"logits/rejected": -8.512123107910156,
"logps/chosen": -40.18708038330078,
"logps/rejected": -61.94392776489258,
"loss": 0.426,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.004405555315315723,
"rewards/margins": 0.7397416234016418,
"rewards/rejected": -0.7353360056877136,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 3.89e-07,
"logits/chosen": -8.962957382202148,
"logits/rejected": -8.918905258178711,
"logps/chosen": -37.032073974609375,
"logps/rejected": -58.812171936035156,
"loss": 0.3985,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": 0.1496596783399582,
"rewards/margins": 0.8128288984298706,
"rewards/rejected": -0.663169264793396,
"step": 111
},
{
"epoch": 0.03,
"learning_rate": 3.88e-07,
"logits/chosen": -8.809492111206055,
"logits/rejected": -8.77208137512207,
"logps/chosen": -40.58615493774414,
"logps/rejected": -61.23163604736328,
"loss": 0.3824,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.0648830309510231,
"rewards/margins": 0.8464531898498535,
"rewards/rejected": -0.9113362431526184,
"step": 112
},
{
"epoch": 0.03,
"learning_rate": 3.87e-07,
"logits/chosen": -9.049692153930664,
"logits/rejected": -8.90321159362793,
"logps/chosen": -37.798988342285156,
"logps/rejected": -56.5068244934082,
"loss": 0.4262,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.009292450733482838,
"rewards/margins": 0.7117542624473572,
"rewards/rejected": -0.7210468053817749,
"step": 113
},
{
"epoch": 0.03,
"learning_rate": 3.86e-07,
"logits/chosen": -9.263845443725586,
"logits/rejected": -9.075590133666992,
"logps/chosen": -38.95356750488281,
"logps/rejected": -57.786277770996094,
"loss": 0.4206,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.047295115888118744,
"rewards/margins": 0.7150000333786011,
"rewards/rejected": -0.6677049398422241,
"step": 114
},
{
"epoch": 0.03,
"learning_rate": 3.8499999999999997e-07,
"logits/chosen": -8.880139350891113,
"logits/rejected": -8.713644981384277,
"logps/chosen": -37.136085510253906,
"logps/rejected": -57.32819747924805,
"loss": 0.4066,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.023683354258537292,
"rewards/margins": 0.790765643119812,
"rewards/rejected": -0.7670822143554688,
"step": 115
},
{
"epoch": 0.03,
"learning_rate": 3.84e-07,
"logits/chosen": -9.125120162963867,
"logits/rejected": -8.953531265258789,
"logps/chosen": -40.43872833251953,
"logps/rejected": -63.04737091064453,
"loss": 0.3631,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.0034645921550691128,
"rewards/margins": 0.9286357164382935,
"rewards/rejected": -0.9251710772514343,
"step": 116
},
{
"epoch": 0.03,
"learning_rate": 3.83e-07,
"logits/chosen": -9.222723960876465,
"logits/rejected": -9.075920104980469,
"logps/chosen": -37.43045425415039,
"logps/rejected": -54.45465087890625,
"loss": 0.4351,
"rewards/accuracies": 0.9000000953674316,
"rewards/chosen": 0.03504835441708565,
"rewards/margins": 0.7053635120391846,
"rewards/rejected": -0.670315146446228,
"step": 117
},
{
"epoch": 0.03,
"learning_rate": 3.82e-07,
"logits/chosen": -8.868099212646484,
"logits/rejected": -8.783113479614258,
"logps/chosen": -38.179161071777344,
"logps/rejected": -59.50090408325195,
"loss": 0.4082,
"rewards/accuracies": 0.8799999356269836,
"rewards/chosen": -0.0526953861117363,
"rewards/margins": 0.7886097431182861,
"rewards/rejected": -0.8413050770759583,
"step": 118
},
{
"epoch": 0.03,
"learning_rate": 3.81e-07,
"logits/chosen": -8.88872241973877,
"logits/rejected": -8.763318061828613,
"logps/chosen": -42.87889862060547,
"logps/rejected": -62.076698303222656,
"loss": 0.4254,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": -0.18682220578193665,
"rewards/margins": 0.7382919788360596,
"rewards/rejected": -0.9251142740249634,
"step": 119
},
{
"epoch": 0.03,
"learning_rate": 3.7999999999999996e-07,
"logits/chosen": -8.93816089630127,
"logits/rejected": -8.88119888305664,
"logps/chosen": -38.15512466430664,
"logps/rejected": -58.470611572265625,
"loss": 0.4107,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": 0.04111450910568237,
"rewards/margins": 0.7645248770713806,
"rewards/rejected": -0.7234103679656982,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 3.79e-07,
"logits/chosen": -8.973804473876953,
"logits/rejected": -8.872992515563965,
"logps/chosen": -39.51805114746094,
"logps/rejected": -55.87736129760742,
"loss": 0.4504,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1321883648633957,
"rewards/margins": 0.6513045430183411,
"rewards/rejected": -0.783492922782898,
"step": 121
},
{
"epoch": 0.03,
"learning_rate": 3.7799999999999997e-07,
"logits/chosen": -8.803823471069336,
"logits/rejected": -8.616742134094238,
"logps/chosen": -40.673221588134766,
"logps/rejected": -61.703948974609375,
"loss": 0.3922,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.07538942992687225,
"rewards/margins": 0.8220105171203613,
"rewards/rejected": -0.89739990234375,
"step": 122
},
{
"epoch": 0.03,
"learning_rate": 3.77e-07,
"logits/chosen": -9.03860855102539,
"logits/rejected": -8.868653297424316,
"logps/chosen": -37.68621063232422,
"logps/rejected": -58.667076110839844,
"loss": 0.3545,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.06506671011447906,
"rewards/margins": 0.9463762044906616,
"rewards/rejected": -0.8813096284866333,
"step": 123
},
{
"epoch": 0.03,
"learning_rate": 3.76e-07,
"logits/chosen": -8.842554092407227,
"logits/rejected": -8.731310844421387,
"logps/chosen": -40.247703552246094,
"logps/rejected": -59.82563018798828,
"loss": 0.3838,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.1266755610704422,
"rewards/margins": 0.8307477831840515,
"rewards/rejected": -0.9574233889579773,
"step": 124
},
{
"epoch": 0.03,
"learning_rate": 3.75e-07,
"logits/chosen": -9.012685775756836,
"logits/rejected": -8.93365478515625,
"logps/chosen": -39.423152923583984,
"logps/rejected": -66.47831726074219,
"loss": 0.3284,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.07251543551683426,
"rewards/margins": 1.0880799293518066,
"rewards/rejected": -1.0155645608901978,
"step": 125
},
{
"epoch": 0.03,
"learning_rate": 3.74e-07,
"logits/chosen": -8.971677780151367,
"logits/rejected": -8.89514446258545,
"logps/chosen": -39.17797088623047,
"logps/rejected": -61.886260986328125,
"loss": 0.3578,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.00424486119300127,
"rewards/margins": 0.9471753239631653,
"rewards/rejected": -0.9514201879501343,
"step": 126
},
{
"epoch": 0.03,
"learning_rate": 3.7299999999999997e-07,
"logits/chosen": -8.827725410461426,
"logits/rejected": -8.715534210205078,
"logps/chosen": -42.28189468383789,
"logps/rejected": -63.66273880004883,
"loss": 0.3951,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.21217112243175507,
"rewards/margins": 0.8158702850341797,
"rewards/rejected": -1.0280416011810303,
"step": 127
},
{
"epoch": 0.03,
"learning_rate": 3.72e-07,
"logits/chosen": -8.811687469482422,
"logits/rejected": -8.690886497497559,
"logps/chosen": -39.99266815185547,
"logps/rejected": -60.4193229675293,
"loss": 0.4187,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.16783051192760468,
"rewards/margins": 0.767844557762146,
"rewards/rejected": -0.9356750249862671,
"step": 128
},
{
"epoch": 0.03,
"learning_rate": 3.71e-07,
"logits/chosen": -8.954986572265625,
"logits/rejected": -8.789628982543945,
"logps/chosen": -38.853965759277344,
"logps/rejected": -64.25423431396484,
"loss": 0.3561,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.12899711728096008,
"rewards/margins": 1.0408105850219727,
"rewards/rejected": -0.9118132591247559,
"step": 129
},
{
"epoch": 0.03,
"learning_rate": 3.7e-07,
"logits/chosen": -8.90104866027832,
"logits/rejected": -8.75663948059082,
"logps/chosen": -38.79196548461914,
"logps/rejected": -59.23478317260742,
"loss": 0.3979,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.045634448528289795,
"rewards/margins": 0.8061555027961731,
"rewards/rejected": -0.7605210542678833,
"step": 130
},
{
"epoch": 0.03,
"learning_rate": 3.69e-07,
"logits/chosen": -9.070549011230469,
"logits/rejected": -8.883820533752441,
"logps/chosen": -40.20603942871094,
"logps/rejected": -57.377357482910156,
"loss": 0.4072,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.19748033583164215,
"rewards/margins": 0.7769988775253296,
"rewards/rejected": -0.9744793772697449,
"step": 131
},
{
"epoch": 0.03,
"learning_rate": 3.6799999999999996e-07,
"logits/chosen": -8.954313278198242,
"logits/rejected": -8.846368789672852,
"logps/chosen": -39.513771057128906,
"logps/rejected": -60.5933723449707,
"loss": 0.3613,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.09566650539636612,
"rewards/margins": 0.9306944012641907,
"rewards/rejected": -1.0263609886169434,
"step": 132
},
{
"epoch": 0.03,
"learning_rate": 3.67e-07,
"logits/chosen": -8.816751480102539,
"logits/rejected": -8.633028030395508,
"logps/chosen": -39.54310989379883,
"logps/rejected": -58.71696090698242,
"loss": 0.3859,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.03110024333000183,
"rewards/margins": 0.9095786809921265,
"rewards/rejected": -0.9406788945198059,
"step": 133
},
{
"epoch": 0.03,
"learning_rate": 3.6599999999999997e-07,
"logits/chosen": -8.940773963928223,
"logits/rejected": -8.746129035949707,
"logps/chosen": -40.17658233642578,
"logps/rejected": -62.3738899230957,
"loss": 0.3541,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.16021467745304108,
"rewards/margins": 0.984930694103241,
"rewards/rejected": -1.145145297050476,
"step": 134
},
{
"epoch": 0.03,
"learning_rate": 3.65e-07,
"logits/chosen": -8.982457160949707,
"logits/rejected": -8.930691719055176,
"logps/chosen": -41.17652893066406,
"logps/rejected": -63.67726516723633,
"loss": 0.3454,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.1305641233921051,
"rewards/margins": 0.996552586555481,
"rewards/rejected": -1.1271167993545532,
"step": 135
},
{
"epoch": 0.03,
"learning_rate": 3.64e-07,
"logits/chosen": -9.113368034362793,
"logits/rejected": -9.021788597106934,
"logps/chosen": -40.26515579223633,
"logps/rejected": -61.51102828979492,
"loss": 0.3919,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.07528214156627655,
"rewards/margins": 0.8362929224967957,
"rewards/rejected": -0.9115751385688782,
"step": 136
},
{
"epoch": 0.03,
"learning_rate": 3.6299999999999995e-07,
"logits/chosen": -9.021696090698242,
"logits/rejected": -8.881583213806152,
"logps/chosen": -39.594642639160156,
"logps/rejected": -62.238922119140625,
"loss": 0.3868,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.08284278213977814,
"rewards/margins": 0.8705474138259888,
"rewards/rejected": -0.9533903002738953,
"step": 137
},
{
"epoch": 0.03,
"learning_rate": 3.62e-07,
"logits/chosen": -9.039748191833496,
"logits/rejected": -8.940320014953613,
"logps/chosen": -40.2382926940918,
"logps/rejected": -62.49633026123047,
"loss": 0.3742,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.19811776280403137,
"rewards/margins": 0.9347583651542664,
"rewards/rejected": -1.1328761577606201,
"step": 138
},
{
"epoch": 0.03,
"learning_rate": 3.6099999999999996e-07,
"logits/chosen": -8.731492042541504,
"logits/rejected": -8.617894172668457,
"logps/chosen": -40.858238220214844,
"logps/rejected": -62.641571044921875,
"loss": 0.3566,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.1721496880054474,
"rewards/margins": 0.9204368591308594,
"rewards/rejected": -1.0925863981246948,
"step": 139
},
{
"epoch": 0.04,
"learning_rate": 3.6e-07,
"logits/chosen": -9.012569427490234,
"logits/rejected": -8.815531730651855,
"logps/chosen": -42.314231872558594,
"logps/rejected": -64.81603240966797,
"loss": 0.35,
"rewards/accuracies": 0.8799999952316284,
"rewards/chosen": -0.15320369601249695,
"rewards/margins": 1.030665397644043,
"rewards/rejected": -1.1838690042495728,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 3.5899999999999997e-07,
"logits/chosen": -8.839929580688477,
"logits/rejected": -8.719411849975586,
"logps/chosen": -40.46647262573242,
"logps/rejected": -61.498626708984375,
"loss": 0.3717,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.006344547960907221,
"rewards/margins": 0.8955361247062683,
"rewards/rejected": -0.8891915082931519,
"step": 141
},
{
"epoch": 0.04,
"learning_rate": 3.5799999999999995e-07,
"logits/chosen": -8.899998664855957,
"logits/rejected": -8.761857986450195,
"logps/chosen": -39.248046875,
"logps/rejected": -64.58299255371094,
"loss": 0.3387,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.021908091381192207,
"rewards/margins": 1.0876431465148926,
"rewards/rejected": -1.109551191329956,
"step": 142
},
{
"epoch": 0.04,
"learning_rate": 3.57e-07,
"logits/chosen": -8.721105575561523,
"logits/rejected": -8.539088249206543,
"logps/chosen": -38.27935028076172,
"logps/rejected": -60.71025848388672,
"loss": 0.3689,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.12697605788707733,
"rewards/margins": 0.9274777173995972,
"rewards/rejected": -1.054453730583191,
"step": 143
},
{
"epoch": 0.04,
"learning_rate": 3.5599999999999996e-07,
"logits/chosen": -8.694182395935059,
"logits/rejected": -8.604917526245117,
"logps/chosen": -41.218135833740234,
"logps/rejected": -63.59687042236328,
"loss": 0.3785,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21460440754890442,
"rewards/margins": 0.9406296610832214,
"rewards/rejected": -1.1552340984344482,
"step": 144
},
{
"epoch": 0.04,
"learning_rate": 3.55e-07,
"logits/chosen": -9.010725021362305,
"logits/rejected": -8.792009353637695,
"logps/chosen": -44.04743576049805,
"logps/rejected": -63.77995681762695,
"loss": 0.3911,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.310243159532547,
"rewards/margins": 0.9037361145019531,
"rewards/rejected": -1.2139792442321777,
"step": 145
},
{
"epoch": 0.04,
"learning_rate": 3.5399999999999997e-07,
"logits/chosen": -8.800936698913574,
"logits/rejected": -8.641168594360352,
"logps/chosen": -40.73430252075195,
"logps/rejected": -59.47686004638672,
"loss": 0.4198,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12597987055778503,
"rewards/margins": 0.8163179159164429,
"rewards/rejected": -0.9422977566719055,
"step": 146
},
{
"epoch": 0.04,
"learning_rate": 3.5299999999999994e-07,
"logits/chosen": -8.81165885925293,
"logits/rejected": -8.621085166931152,
"logps/chosen": -41.21516418457031,
"logps/rejected": -65.77284240722656,
"loss": 0.3477,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": 0.02493276819586754,
"rewards/margins": 1.0518163442611694,
"rewards/rejected": -1.026883602142334,
"step": 147
},
{
"epoch": 0.04,
"learning_rate": 3.52e-07,
"logits/chosen": -9.05104923248291,
"logits/rejected": -8.940778732299805,
"logps/chosen": -39.09772872924805,
"logps/rejected": -63.58698272705078,
"loss": 0.3292,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.06748518347740173,
"rewards/margins": 1.0447824001312256,
"rewards/rejected": -1.1122677326202393,
"step": 148
},
{
"epoch": 0.04,
"learning_rate": 3.5099999999999995e-07,
"logits/chosen": -8.909757614135742,
"logits/rejected": -8.688767433166504,
"logps/chosen": -40.91362762451172,
"logps/rejected": -60.848472595214844,
"loss": 0.3958,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.20450308918952942,
"rewards/margins": 0.8363680839538574,
"rewards/rejected": -1.0408711433410645,
"step": 149
},
{
"epoch": 0.04,
"learning_rate": 3.5e-07,
"logits/chosen": -9.037310600280762,
"logits/rejected": -8.962252616882324,
"logps/chosen": -40.28340530395508,
"logps/rejected": -63.94322967529297,
"loss": 0.3284,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.07565836608409882,
"rewards/margins": 1.125253438949585,
"rewards/rejected": -1.2009117603302002,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 3.4899999999999996e-07,
"logits/chosen": -8.9218168258667,
"logits/rejected": -8.84290599822998,
"logps/chosen": -39.55026626586914,
"logps/rejected": -61.443641662597656,
"loss": 0.3476,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.02470378205180168,
"rewards/margins": 0.9961774945259094,
"rewards/rejected": -1.020881175994873,
"step": 151
},
{
"epoch": 0.04,
"learning_rate": 3.4799999999999994e-07,
"logits/chosen": -8.936076164245605,
"logits/rejected": -8.845673561096191,
"logps/chosen": -38.409576416015625,
"logps/rejected": -61.9953727722168,
"loss": 0.3381,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.11342154443264008,
"rewards/margins": 0.9978846311569214,
"rewards/rejected": -1.111306071281433,
"step": 152
},
{
"epoch": 0.04,
"learning_rate": 3.4699999999999997e-07,
"logits/chosen": -8.993753433227539,
"logits/rejected": -8.88867473602295,
"logps/chosen": -41.41925811767578,
"logps/rejected": -64.46873474121094,
"loss": 0.3345,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.1233917623758316,
"rewards/margins": 1.042170763015747,
"rewards/rejected": -1.165562629699707,
"step": 153
},
{
"epoch": 0.04,
"learning_rate": 3.4599999999999995e-07,
"logits/chosen": -8.77087688446045,
"logits/rejected": -8.657720565795898,
"logps/chosen": -40.0390510559082,
"logps/rejected": -60.32958984375,
"loss": 0.3967,
"rewards/accuracies": 0.9200000762939453,
"rewards/chosen": -0.19076868891716003,
"rewards/margins": 0.9012584686279297,
"rewards/rejected": -1.0920270681381226,
"step": 154
},
{
"epoch": 0.04,
"learning_rate": 3.45e-07,
"logits/chosen": -8.897802352905273,
"logits/rejected": -8.744150161743164,
"logps/chosen": -40.92519760131836,
"logps/rejected": -65.74849700927734,
"loss": 0.3411,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.15756723284721375,
"rewards/margins": 1.0892850160598755,
"rewards/rejected": -1.2468522787094116,
"step": 155
},
{
"epoch": 0.04,
"learning_rate": 3.4399999999999996e-07,
"logits/chosen": -8.871163368225098,
"logits/rejected": -8.717877388000488,
"logps/chosen": -43.01222229003906,
"logps/rejected": -70.48704528808594,
"loss": 0.3188,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3022673428058624,
"rewards/margins": 1.1483441591262817,
"rewards/rejected": -1.4506114721298218,
"step": 156
},
{
"epoch": 0.04,
"learning_rate": 3.43e-07,
"logits/chosen": -8.949113845825195,
"logits/rejected": -8.905904769897461,
"logps/chosen": -39.99720001220703,
"logps/rejected": -62.5020866394043,
"loss": 0.3482,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.06127683073282242,
"rewards/margins": 1.05168616771698,
"rewards/rejected": -1.1129629611968994,
"step": 157
},
{
"epoch": 0.04,
"learning_rate": 3.42e-07,
"logits/chosen": -8.867097854614258,
"logits/rejected": -8.736467361450195,
"logps/chosen": -41.989192962646484,
"logps/rejected": -61.25897216796875,
"loss": 0.3882,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.35165315866470337,
"rewards/margins": 0.877507209777832,
"rewards/rejected": -1.2291605472564697,
"step": 158
},
{
"epoch": 0.04,
"learning_rate": 3.41e-07,
"logits/chosen": -8.919662475585938,
"logits/rejected": -8.849380493164062,
"logps/chosen": -42.1957893371582,
"logps/rejected": -66.24508666992188,
"loss": 0.3071,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.30500465631484985,
"rewards/margins": 1.1364095211029053,
"rewards/rejected": -1.4414143562316895,
"step": 159
},
{
"epoch": 0.04,
"learning_rate": 3.4000000000000003e-07,
"logits/chosen": -8.94130802154541,
"logits/rejected": -8.818609237670898,
"logps/chosen": -40.54155731201172,
"logps/rejected": -59.704978942871094,
"loss": 0.3546,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.13368672132492065,
"rewards/margins": 0.9669790267944336,
"rewards/rejected": -1.100665807723999,
"step": 160
},
{
"epoch": 0.04,
"learning_rate": 3.39e-07,
"logits/chosen": -8.80152416229248,
"logits/rejected": -8.622817039489746,
"logps/chosen": -41.8613166809082,
"logps/rejected": -63.264976501464844,
"loss": 0.3418,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.29719024896621704,
"rewards/margins": 1.0231988430023193,
"rewards/rejected": -1.3203891515731812,
"step": 161
},
{
"epoch": 0.04,
"learning_rate": 3.38e-07,
"logits/chosen": -8.761874198913574,
"logits/rejected": -8.586349487304688,
"logps/chosen": -41.71952438354492,
"logps/rejected": -68.82637023925781,
"loss": 0.2946,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.21405473351478577,
"rewards/margins": 1.3066158294677734,
"rewards/rejected": -1.5206706523895264,
"step": 162
},
{
"epoch": 0.04,
"learning_rate": 3.37e-07,
"logits/chosen": -8.828725814819336,
"logits/rejected": -8.663387298583984,
"logps/chosen": -41.91630935668945,
"logps/rejected": -65.08641052246094,
"loss": 0.341,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.24001702666282654,
"rewards/margins": 1.04341721534729,
"rewards/rejected": -1.2834341526031494,
"step": 163
},
{
"epoch": 0.04,
"learning_rate": 3.36e-07,
"logits/chosen": -8.988265991210938,
"logits/rejected": -8.813672065734863,
"logps/chosen": -41.581756591796875,
"logps/rejected": -64.73905944824219,
"loss": 0.3243,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.14427520334720612,
"rewards/margins": 1.099812388420105,
"rewards/rejected": -1.2440874576568604,
"step": 164
},
{
"epoch": 0.04,
"learning_rate": 3.35e-07,
"logits/chosen": -9.024198532104492,
"logits/rejected": -8.907011985778809,
"logps/chosen": -40.02942657470703,
"logps/rejected": -65.40174865722656,
"loss": 0.3156,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.17336931824684143,
"rewards/margins": 1.1412004232406616,
"rewards/rejected": -1.3145698308944702,
"step": 165
},
{
"epoch": 0.04,
"learning_rate": 3.34e-07,
"logits/chosen": -8.859591484069824,
"logits/rejected": -8.78995132446289,
"logps/chosen": -37.986610412597656,
"logps/rejected": -62.487632751464844,
"loss": 0.3413,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.03164323419332504,
"rewards/margins": 1.038362741470337,
"rewards/rejected": -1.0700057744979858,
"step": 166
},
{
"epoch": 0.04,
"learning_rate": 3.33e-07,
"logits/chosen": -8.95915412902832,
"logits/rejected": -8.862491607666016,
"logps/chosen": -39.90531539916992,
"logps/rejected": -68.55361938476562,
"loss": 0.2611,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.09153137356042862,
"rewards/margins": 1.3365256786346436,
"rewards/rejected": -1.4280569553375244,
"step": 167
},
{
"epoch": 0.04,
"learning_rate": 3.32e-07,
"logits/chosen": -8.84840202331543,
"logits/rejected": -8.720226287841797,
"logps/chosen": -39.774906158447266,
"logps/rejected": -63.923797607421875,
"loss": 0.2924,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.15840193629264832,
"rewards/margins": 1.228715181350708,
"rewards/rejected": -1.3871170282363892,
"step": 168
},
{
"epoch": 0.04,
"learning_rate": 3.31e-07,
"logits/chosen": -8.834859848022461,
"logits/rejected": -8.646490097045898,
"logps/chosen": -40.75969696044922,
"logps/rejected": -68.73313903808594,
"loss": 0.289,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.2145496904850006,
"rewards/margins": 1.282721757888794,
"rewards/rejected": -1.4972714185714722,
"step": 169
},
{
"epoch": 0.04,
"learning_rate": 3.3e-07,
"logits/chosen": -9.14974594116211,
"logits/rejected": -9.094747543334961,
"logps/chosen": -38.71310806274414,
"logps/rejected": -64.7855453491211,
"loss": 0.3276,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.10842056572437286,
"rewards/margins": 1.1217743158340454,
"rewards/rejected": -1.2301948070526123,
"step": 170
},
{
"epoch": 0.04,
"learning_rate": 3.29e-07,
"logits/chosen": -8.805329322814941,
"logits/rejected": -8.60169506072998,
"logps/chosen": -44.541385650634766,
"logps/rejected": -67.90604400634766,
"loss": 0.3276,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.33684179186820984,
"rewards/margins": 1.0830628871917725,
"rewards/rejected": -1.4199049472808838,
"step": 171
},
{
"epoch": 0.04,
"learning_rate": 3.28e-07,
"logits/chosen": -9.00810432434082,
"logits/rejected": -8.874260902404785,
"logps/chosen": -42.70119857788086,
"logps/rejected": -66.67572784423828,
"loss": 0.3219,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.28449511528015137,
"rewards/margins": 1.1121952533721924,
"rewards/rejected": -1.3966902494430542,
"step": 172
},
{
"epoch": 0.04,
"learning_rate": 3.27e-07,
"logits/chosen": -9.011045455932617,
"logits/rejected": -8.86266803741455,
"logps/chosen": -40.55815887451172,
"logps/rejected": -65.76628875732422,
"loss": 0.3108,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.15313352644443512,
"rewards/margins": 1.2283756732940674,
"rewards/rejected": -1.3815090656280518,
"step": 173
},
{
"epoch": 0.04,
"learning_rate": 3.26e-07,
"logits/chosen": -8.745307922363281,
"logits/rejected": -8.638575553894043,
"logps/chosen": -40.17422866821289,
"logps/rejected": -61.25606155395508,
"loss": 0.3363,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.26441556215286255,
"rewards/margins": 1.0637372732162476,
"rewards/rejected": -1.3281527757644653,
"step": 174
},
{
"epoch": 0.04,
"learning_rate": 3.25e-07,
"logits/chosen": -8.917795181274414,
"logits/rejected": -8.75191593170166,
"logps/chosen": -40.356971740722656,
"logps/rejected": -65.25245666503906,
"loss": 0.3272,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.21726162731647491,
"rewards/margins": 1.1349842548370361,
"rewards/rejected": -1.3522460460662842,
"step": 175
},
{
"epoch": 0.04,
"learning_rate": 3.24e-07,
"logits/chosen": -9.057024002075195,
"logits/rejected": -9.007242202758789,
"logps/chosen": -41.4436149597168,
"logps/rejected": -59.33845138549805,
"loss": 0.3784,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.20186138153076172,
"rewards/margins": 0.8808485865592957,
"rewards/rejected": -1.082709789276123,
"step": 176
},
{
"epoch": 0.04,
"learning_rate": 3.23e-07,
"logits/chosen": -9.150751113891602,
"logits/rejected": -8.990005493164062,
"logps/chosen": -41.529197692871094,
"logps/rejected": -64.6326904296875,
"loss": 0.3474,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.1397334188222885,
"rewards/margins": 1.0998492240905762,
"rewards/rejected": -1.2395826578140259,
"step": 177
},
{
"epoch": 0.04,
"learning_rate": 3.22e-07,
"logits/chosen": -8.985628128051758,
"logits/rejected": -8.924480438232422,
"logps/chosen": -39.99455261230469,
"logps/rejected": -62.32940673828125,
"loss": 0.3442,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.21177975833415985,
"rewards/margins": 1.0999975204467773,
"rewards/rejected": -1.311777114868164,
"step": 178
},
{
"epoch": 0.04,
"learning_rate": 3.21e-07,
"logits/chosen": -9.033208847045898,
"logits/rejected": -8.933832168579102,
"logps/chosen": -42.78688049316406,
"logps/rejected": -65.41902160644531,
"loss": 0.3323,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.42959874868392944,
"rewards/margins": 1.0489178895950317,
"rewards/rejected": -1.478516697883606,
"step": 179
},
{
"epoch": 0.04,
"learning_rate": 3.2e-07,
"logits/chosen": -8.921372413635254,
"logits/rejected": -8.761728286743164,
"logps/chosen": -41.17156982421875,
"logps/rejected": -68.08998107910156,
"loss": 0.2719,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.20495982468128204,
"rewards/margins": 1.340659737586975,
"rewards/rejected": -1.5456194877624512,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 3.19e-07,
"logits/chosen": -8.87382698059082,
"logits/rejected": -8.771355628967285,
"logps/chosen": -41.562965393066406,
"logps/rejected": -64.83119201660156,
"loss": 0.2963,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.17605173587799072,
"rewards/margins": 1.220849633216858,
"rewards/rejected": -1.3969013690948486,
"step": 181
},
{
"epoch": 0.05,
"learning_rate": 3.18e-07,
"logits/chosen": -8.82695198059082,
"logits/rejected": -8.737783432006836,
"logps/chosen": -39.56995391845703,
"logps/rejected": -62.3057746887207,
"loss": 0.3402,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.20497837662696838,
"rewards/margins": 1.1743295192718506,
"rewards/rejected": -1.3793079853057861,
"step": 182
},
{
"epoch": 0.05,
"learning_rate": 3.17e-07,
"logits/chosen": -9.018879890441895,
"logits/rejected": -8.862287521362305,
"logps/chosen": -42.85093688964844,
"logps/rejected": -72.0836181640625,
"loss": 0.3212,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.2954815924167633,
"rewards/margins": 1.268796682357788,
"rewards/rejected": -1.5642781257629395,
"step": 183
},
{
"epoch": 0.05,
"learning_rate": 3.1599999999999997e-07,
"logits/chosen": -8.963022232055664,
"logits/rejected": -8.82475471496582,
"logps/chosen": -38.94888687133789,
"logps/rejected": -68.8856201171875,
"loss": 0.2643,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.0330241322517395,
"rewards/margins": 1.4156899452209473,
"rewards/rejected": -1.4487141370773315,
"step": 184
},
{
"epoch": 0.05,
"learning_rate": 3.15e-07,
"logits/chosen": -9.073816299438477,
"logits/rejected": -8.94584846496582,
"logps/chosen": -44.071678161621094,
"logps/rejected": -70.720947265625,
"loss": 0.2948,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.4320756494998932,
"rewards/margins": 1.216040849685669,
"rewards/rejected": -1.6481164693832397,
"step": 185
},
{
"epoch": 0.05,
"learning_rate": 3.14e-07,
"logits/chosen": -9.04081916809082,
"logits/rejected": -8.882933616638184,
"logps/chosen": -39.92203140258789,
"logps/rejected": -66.04690551757812,
"loss": 0.3199,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.12839896976947784,
"rewards/margins": 1.1672937870025635,
"rewards/rejected": -1.2956926822662354,
"step": 186
},
{
"epoch": 0.05,
"learning_rate": 3.13e-07,
"logits/chosen": -8.965912818908691,
"logits/rejected": -8.788244247436523,
"logps/chosen": -40.17821502685547,
"logps/rejected": -65.7179946899414,
"loss": 0.2863,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.1306789070367813,
"rewards/margins": 1.2615444660186768,
"rewards/rejected": -1.3922232389450073,
"step": 187
},
{
"epoch": 0.05,
"learning_rate": 3.12e-07,
"logits/chosen": -8.968905448913574,
"logits/rejected": -8.836429595947266,
"logps/chosen": -40.19866180419922,
"logps/rejected": -65.63804626464844,
"loss": 0.2997,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.29229894280433655,
"rewards/margins": 1.2268226146697998,
"rewards/rejected": -1.5191216468811035,
"step": 188
},
{
"epoch": 0.05,
"learning_rate": 3.1099999999999997e-07,
"logits/chosen": -8.949649810791016,
"logits/rejected": -8.85619068145752,
"logps/chosen": -41.16625213623047,
"logps/rejected": -67.67311096191406,
"loss": 0.3123,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.18751493096351624,
"rewards/margins": 1.2116743326187134,
"rewards/rejected": -1.3991893529891968,
"step": 189
},
{
"epoch": 0.05,
"learning_rate": 3.1e-07,
"logits/chosen": -9.069915771484375,
"logits/rejected": -8.9552640914917,
"logps/chosen": -44.67123031616211,
"logps/rejected": -69.6094970703125,
"loss": 0.3179,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.36990422010421753,
"rewards/margins": 1.1963659524917603,
"rewards/rejected": -1.566270112991333,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 3.09e-07,
"logits/chosen": -8.953503608703613,
"logits/rejected": -8.81824779510498,
"logps/chosen": -43.4686279296875,
"logps/rejected": -65.56145477294922,
"loss": 0.2937,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.3817153573036194,
"rewards/margins": 1.2739273309707642,
"rewards/rejected": -1.6556425094604492,
"step": 191
},
{
"epoch": 0.05,
"learning_rate": 3.08e-07,
"logits/chosen": -8.856042861938477,
"logits/rejected": -8.653620719909668,
"logps/chosen": -40.48377990722656,
"logps/rejected": -65.58942413330078,
"loss": 0.3144,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.21695208549499512,
"rewards/margins": 1.223385214805603,
"rewards/rejected": -1.4403371810913086,
"step": 192
},
{
"epoch": 0.05,
"learning_rate": 3.07e-07,
"logits/chosen": -8.951238632202148,
"logits/rejected": -8.893243789672852,
"logps/chosen": -38.89543914794922,
"logps/rejected": -65.236572265625,
"loss": 0.2907,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.14968658983707428,
"rewards/margins": 1.2427290678024292,
"rewards/rejected": -1.3924156427383423,
"step": 193
},
{
"epoch": 0.05,
"learning_rate": 3.0599999999999996e-07,
"logits/chosen": -9.175559043884277,
"logits/rejected": -9.138572692871094,
"logps/chosen": -41.56648635864258,
"logps/rejected": -62.5894889831543,
"loss": 0.3235,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3439091444015503,
"rewards/margins": 1.1237269639968872,
"rewards/rejected": -1.467635989189148,
"step": 194
},
{
"epoch": 0.05,
"learning_rate": 3.05e-07,
"logits/chosen": -9.081843376159668,
"logits/rejected": -8.967177391052246,
"logps/chosen": -48.0760383605957,
"logps/rejected": -72.55915069580078,
"loss": 0.3347,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.599833607673645,
"rewards/margins": 1.1599971055984497,
"rewards/rejected": -1.7598304748535156,
"step": 195
},
{
"epoch": 0.05,
"learning_rate": 3.0399999999999997e-07,
"logits/chosen": -8.88502025604248,
"logits/rejected": -8.776427268981934,
"logps/chosen": -38.89107894897461,
"logps/rejected": -62.98077392578125,
"loss": 0.3103,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.08193059265613556,
"rewards/margins": 1.204545021057129,
"rewards/rejected": -1.2864755392074585,
"step": 196
},
{
"epoch": 0.05,
"learning_rate": 3.03e-07,
"logits/chosen": -8.958200454711914,
"logits/rejected": -8.758167266845703,
"logps/chosen": -42.52730941772461,
"logps/rejected": -68.24816131591797,
"loss": 0.2818,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.2780629098415375,
"rewards/margins": 1.3868484497070312,
"rewards/rejected": -1.6649112701416016,
"step": 197
},
{
"epoch": 0.05,
"learning_rate": 3.02e-07,
"logits/chosen": -9.064251899719238,
"logits/rejected": -9.055684089660645,
"logps/chosen": -40.26599884033203,
"logps/rejected": -68.03285217285156,
"loss": 0.2815,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.21159246563911438,
"rewards/margins": 1.3151299953460693,
"rewards/rejected": -1.5267223119735718,
"step": 198
},
{
"epoch": 0.05,
"learning_rate": 3.0099999999999996e-07,
"logits/chosen": -8.8607177734375,
"logits/rejected": -8.776520729064941,
"logps/chosen": -44.09022521972656,
"logps/rejected": -67.35237884521484,
"loss": 0.2963,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3733174204826355,
"rewards/margins": 1.302872896194458,
"rewards/rejected": -1.6761903762817383,
"step": 199
},
{
"epoch": 0.05,
"learning_rate": 3e-07,
"logits/chosen": -9.326008796691895,
"logits/rejected": -9.195959091186523,
"logps/chosen": -42.66015625,
"logps/rejected": -69.38017272949219,
"loss": 0.29,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.3082619607448578,
"rewards/margins": 1.301923394203186,
"rewards/rejected": -1.6101853847503662,
"step": 200
},
{
"epoch": 0.05,
"learning_rate": 2.9899999999999996e-07,
"logits/chosen": -8.926416397094727,
"logits/rejected": -8.793083190917969,
"logps/chosen": -42.86346435546875,
"logps/rejected": -69.28981018066406,
"loss": 0.2735,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.30893146991729736,
"rewards/margins": 1.3181147575378418,
"rewards/rejected": -1.6270462274551392,
"step": 201
},
{
"epoch": 0.05,
"learning_rate": 2.98e-07,
"logits/chosen": -9.045461654663086,
"logits/rejected": -8.975128173828125,
"logps/chosen": -41.371551513671875,
"logps/rejected": -68.18250274658203,
"loss": 0.3091,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3913578391075134,
"rewards/margins": 1.3062489032745361,
"rewards/rejected": -1.6976070404052734,
"step": 202
},
{
"epoch": 0.05,
"learning_rate": 2.9699999999999997e-07,
"logits/chosen": -8.808774948120117,
"logits/rejected": -8.637701034545898,
"logps/chosen": -41.52623748779297,
"logps/rejected": -65.2273941040039,
"loss": 0.3081,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.23487603664398193,
"rewards/margins": 1.2760727405548096,
"rewards/rejected": -1.510948896408081,
"step": 203
},
{
"epoch": 0.05,
"learning_rate": 2.9599999999999995e-07,
"logits/chosen": -8.662897109985352,
"logits/rejected": -8.514671325683594,
"logps/chosen": -40.43410873413086,
"logps/rejected": -63.4935188293457,
"loss": 0.3236,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": -0.180062934756279,
"rewards/margins": 1.2099792957305908,
"rewards/rejected": -1.3900420665740967,
"step": 204
},
{
"epoch": 0.05,
"learning_rate": 2.95e-07,
"logits/chosen": -9.048240661621094,
"logits/rejected": -8.950238227844238,
"logps/chosen": -38.42674255371094,
"logps/rejected": -70.53987121582031,
"loss": 0.2272,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.2043372094631195,
"rewards/margins": 1.5515215396881104,
"rewards/rejected": -1.7558587789535522,
"step": 205
},
{
"epoch": 0.05,
"learning_rate": 2.9399999999999996e-07,
"logits/chosen": -8.96391487121582,
"logits/rejected": -8.855020523071289,
"logps/chosen": -41.72402572631836,
"logps/rejected": -68.75076293945312,
"loss": 0.2608,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.2568601965904236,
"rewards/margins": 1.3788115978240967,
"rewards/rejected": -1.6356719732284546,
"step": 206
},
{
"epoch": 0.05,
"learning_rate": 2.93e-07,
"logits/chosen": -8.95437240600586,
"logits/rejected": -8.87646484375,
"logps/chosen": -43.06665802001953,
"logps/rejected": -68.59609985351562,
"loss": 0.2886,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3358538746833801,
"rewards/margins": 1.31618332862854,
"rewards/rejected": -1.652037262916565,
"step": 207
},
{
"epoch": 0.05,
"learning_rate": 2.9199999999999997e-07,
"logits/chosen": -8.841405868530273,
"logits/rejected": -8.739511489868164,
"logps/chosen": -42.112754821777344,
"logps/rejected": -64.5459976196289,
"loss": 0.322,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3095633387565613,
"rewards/margins": 1.1418513059616089,
"rewards/rejected": -1.4514145851135254,
"step": 208
},
{
"epoch": 0.05,
"learning_rate": 2.9099999999999995e-07,
"logits/chosen": -9.11125373840332,
"logits/rejected": -9.027490615844727,
"logps/chosen": -41.227901458740234,
"logps/rejected": -76.45286560058594,
"loss": 0.236,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.18211723864078522,
"rewards/margins": 1.6697490215301514,
"rewards/rejected": -1.8518664836883545,
"step": 209
},
{
"epoch": 0.05,
"learning_rate": 2.9e-07,
"logits/chosen": -8.93098258972168,
"logits/rejected": -8.775957107543945,
"logps/chosen": -42.168487548828125,
"logps/rejected": -69.60896301269531,
"loss": 0.2531,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.1441860944032669,
"rewards/margins": 1.5427669286727905,
"rewards/rejected": -1.6869529485702515,
"step": 210
},
{
"epoch": 0.05,
"learning_rate": 2.8899999999999995e-07,
"logits/chosen": -9.149178504943848,
"logits/rejected": -8.981005668640137,
"logps/chosen": -40.37749481201172,
"logps/rejected": -66.62581634521484,
"loss": 0.276,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.16469226777553558,
"rewards/margins": 1.4120306968688965,
"rewards/rejected": -1.5767228603363037,
"step": 211
},
{
"epoch": 0.05,
"learning_rate": 2.88e-07,
"logits/chosen": -8.969169616699219,
"logits/rejected": -8.803339004516602,
"logps/chosen": -42.27989959716797,
"logps/rejected": -64.9478759765625,
"loss": 0.2654,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.25626304745674133,
"rewards/margins": 1.4114668369293213,
"rewards/rejected": -1.6677299737930298,
"step": 212
},
{
"epoch": 0.05,
"learning_rate": 2.8699999999999996e-07,
"logits/chosen": -8.94532585144043,
"logits/rejected": -8.8807373046875,
"logps/chosen": -44.27782440185547,
"logps/rejected": -70.2257308959961,
"loss": 0.2728,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.2951999306678772,
"rewards/margins": 1.3818528652191162,
"rewards/rejected": -1.6770527362823486,
"step": 213
},
{
"epoch": 0.05,
"learning_rate": 2.8599999999999994e-07,
"logits/chosen": -9.086102485656738,
"logits/rejected": -9.030981063842773,
"logps/chosen": -41.30907440185547,
"logps/rejected": -68.33830261230469,
"loss": 0.2695,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.1512337625026703,
"rewards/margins": 1.4189387559890747,
"rewards/rejected": -1.570172667503357,
"step": 214
},
{
"epoch": 0.05,
"learning_rate": 2.8499999999999997e-07,
"logits/chosen": -8.850401878356934,
"logits/rejected": -8.755792617797852,
"logps/chosen": -43.39167022705078,
"logps/rejected": -70.92607879638672,
"loss": 0.2675,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.1908183991909027,
"rewards/margins": 1.4917595386505127,
"rewards/rejected": -1.6825778484344482,
"step": 215
},
{
"epoch": 0.05,
"learning_rate": 2.8399999999999995e-07,
"logits/chosen": -9.259855270385742,
"logits/rejected": -9.203694343566895,
"logps/chosen": -41.756351470947266,
"logps/rejected": -65.8165512084961,
"loss": 0.2729,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.2136426419019699,
"rewards/margins": 1.3828479051589966,
"rewards/rejected": -1.5964906215667725,
"step": 216
},
{
"epoch": 0.05,
"learning_rate": 2.83e-07,
"logits/chosen": -8.922069549560547,
"logits/rejected": -8.816206932067871,
"logps/chosen": -43.09095001220703,
"logps/rejected": -70.11869049072266,
"loss": 0.2481,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.40523457527160645,
"rewards/margins": 1.536644697189331,
"rewards/rejected": -1.9418792724609375,
"step": 217
},
{
"epoch": 0.05,
"learning_rate": 2.8199999999999996e-07,
"logits/chosen": -9.18004322052002,
"logits/rejected": -9.122257232666016,
"logps/chosen": -40.0262451171875,
"logps/rejected": -73.48004150390625,
"loss": 0.2586,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18847300112247467,
"rewards/margins": 1.6103322505950928,
"rewards/rejected": -1.7988052368164062,
"step": 218
},
{
"epoch": 0.05,
"learning_rate": 2.8100000000000004e-07,
"logits/chosen": -8.809640884399414,
"logits/rejected": -8.642433166503906,
"logps/chosen": -46.24394989013672,
"logps/rejected": -66.08168029785156,
"loss": 0.3238,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.5799109935760498,
"rewards/margins": 1.1493396759033203,
"rewards/rejected": -1.7292506694793701,
"step": 219
},
{
"epoch": 0.06,
"learning_rate": 2.8e-07,
"logits/chosen": -9.087625503540039,
"logits/rejected": -8.999017715454102,
"logps/chosen": -44.835086822509766,
"logps/rejected": -69.23603057861328,
"loss": 0.3145,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.4856896996498108,
"rewards/margins": 1.271304726600647,
"rewards/rejected": -1.7569944858551025,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 2.79e-07,
"logits/chosen": -8.82785415649414,
"logits/rejected": -8.75294017791748,
"logps/chosen": -44.45280838012695,
"logps/rejected": -73.3116226196289,
"loss": 0.237,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.446541965007782,
"rewards/margins": 1.5875802040100098,
"rewards/rejected": -2.0341219902038574,
"step": 221
},
{
"epoch": 0.06,
"learning_rate": 2.7800000000000003e-07,
"logits/chosen": -8.659734725952148,
"logits/rejected": -8.634805679321289,
"logps/chosen": -42.231929779052734,
"logps/rejected": -70.79386901855469,
"loss": 0.2399,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.25388625264167786,
"rewards/margins": 1.5821778774261475,
"rewards/rejected": -1.836064100265503,
"step": 222
},
{
"epoch": 0.06,
"learning_rate": 2.77e-07,
"logits/chosen": -9.120695114135742,
"logits/rejected": -9.018156051635742,
"logps/chosen": -41.204193115234375,
"logps/rejected": -70.23939514160156,
"loss": 0.2549,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.20924563705921173,
"rewards/margins": 1.5008659362792969,
"rewards/rejected": -1.7101116180419922,
"step": 223
},
{
"epoch": 0.06,
"learning_rate": 2.7600000000000004e-07,
"logits/chosen": -8.943544387817383,
"logits/rejected": -8.886691093444824,
"logps/chosen": -44.56146240234375,
"logps/rejected": -69.162841796875,
"loss": 0.2689,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.40339189767837524,
"rewards/margins": 1.4689480066299438,
"rewards/rejected": -1.8723399639129639,
"step": 224
},
{
"epoch": 0.06,
"learning_rate": 2.75e-07,
"logits/chosen": -8.663228988647461,
"logits/rejected": -8.65444278717041,
"logps/chosen": -38.287235260009766,
"logps/rejected": -70.95658874511719,
"loss": 0.2141,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": 0.003157383296638727,
"rewards/margins": 1.7416331768035889,
"rewards/rejected": -1.7384757995605469,
"step": 225
},
{
"epoch": 0.06,
"learning_rate": 2.74e-07,
"logits/chosen": -9.011740684509277,
"logits/rejected": -8.81816577911377,
"logps/chosen": -41.13055419921875,
"logps/rejected": -70.431884765625,
"loss": 0.253,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.17377158999443054,
"rewards/margins": 1.6204302310943604,
"rewards/rejected": -1.7942014932632446,
"step": 226
},
{
"epoch": 0.06,
"learning_rate": 2.73e-07,
"logits/chosen": -8.906305313110352,
"logits/rejected": -8.782539367675781,
"logps/chosen": -41.5731086730957,
"logps/rejected": -68.40547180175781,
"loss": 0.3045,
"rewards/accuracies": 0.9199999570846558,
"rewards/chosen": -0.27805295586586,
"rewards/margins": 1.4335918426513672,
"rewards/rejected": -1.7116448879241943,
"step": 227
},
{
"epoch": 0.06,
"learning_rate": 2.72e-07,
"logits/chosen": -8.857791900634766,
"logits/rejected": -8.7904052734375,
"logps/chosen": -44.0427131652832,
"logps/rejected": -67.93760681152344,
"loss": 0.3059,
"rewards/accuracies": 0.9399999380111694,
"rewards/chosen": -0.4859923720359802,
"rewards/margins": 1.3006880283355713,
"rewards/rejected": -1.7866804599761963,
"step": 228
},
{
"epoch": 0.06,
"learning_rate": 2.7100000000000003e-07,
"logits/chosen": -8.68824577331543,
"logits/rejected": -8.578463554382324,
"logps/chosen": -46.805423736572266,
"logps/rejected": -71.96556854248047,
"loss": 0.2496,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.6673830151557922,
"rewards/margins": 1.489000678062439,
"rewards/rejected": -2.156383752822876,
"step": 229
},
{
"epoch": 0.06,
"learning_rate": 2.7e-07,
"logits/chosen": -8.906216621398926,
"logits/rejected": -8.847137451171875,
"logps/chosen": -45.05828857421875,
"logps/rejected": -71.52420043945312,
"loss": 0.2385,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.435865581035614,
"rewards/margins": 1.5502560138702393,
"rewards/rejected": -1.9861215353012085,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 2.69e-07,
"logits/chosen": -8.947155952453613,
"logits/rejected": -8.829044342041016,
"logps/chosen": -41.90214920043945,
"logps/rejected": -73.6261978149414,
"loss": 0.2404,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3273487687110901,
"rewards/margins": 1.621421456336975,
"rewards/rejected": -1.9487701654434204,
"step": 231
},
{
"epoch": 0.06,
"learning_rate": 2.68e-07,
"logits/chosen": -8.829843521118164,
"logits/rejected": -8.652304649353027,
"logps/chosen": -41.917579650878906,
"logps/rejected": -71.42601776123047,
"loss": 0.2304,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.40560126304626465,
"rewards/margins": 1.6656455993652344,
"rewards/rejected": -2.07124662399292,
"step": 232
},
{
"epoch": 0.06,
"learning_rate": 2.67e-07,
"logits/chosen": -9.078448295593262,
"logits/rejected": -9.015963554382324,
"logps/chosen": -48.53789138793945,
"logps/rejected": -72.73210144042969,
"loss": 0.2697,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.6256522536277771,
"rewards/margins": 1.4912573099136353,
"rewards/rejected": -2.1169095039367676,
"step": 233
},
{
"epoch": 0.06,
"learning_rate": 2.66e-07,
"logits/chosen": -9.04401683807373,
"logits/rejected": -8.977219581604004,
"logps/chosen": -44.62416458129883,
"logps/rejected": -71.57743835449219,
"loss": 0.2474,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.5261238813400269,
"rewards/margins": 1.5465754270553589,
"rewards/rejected": -2.072699546813965,
"step": 234
},
{
"epoch": 0.06,
"learning_rate": 2.65e-07,
"logits/chosen": -8.7246732711792,
"logits/rejected": -8.663068771362305,
"logps/chosen": -42.98565673828125,
"logps/rejected": -69.6664047241211,
"loss": 0.2359,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.28909361362457275,
"rewards/margins": 1.6054264307022095,
"rewards/rejected": -1.8945200443267822,
"step": 235
},
{
"epoch": 0.06,
"learning_rate": 2.64e-07,
"logits/chosen": -9.035811424255371,
"logits/rejected": -8.968504905700684,
"logps/chosen": -42.082191467285156,
"logps/rejected": -69.3706283569336,
"loss": 0.2806,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3632660508155823,
"rewards/margins": 1.3896889686584473,
"rewards/rejected": -1.7529550790786743,
"step": 236
},
{
"epoch": 0.06,
"learning_rate": 2.63e-07,
"logits/chosen": -8.899124145507812,
"logits/rejected": -8.845288276672363,
"logps/chosen": -42.73535919189453,
"logps/rejected": -73.52890014648438,
"loss": 0.2707,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3918378949165344,
"rewards/margins": 1.5124340057373047,
"rewards/rejected": -1.9042720794677734,
"step": 237
},
{
"epoch": 0.06,
"learning_rate": 2.62e-07,
"logits/chosen": -9.026718139648438,
"logits/rejected": -8.953622817993164,
"logps/chosen": -40.321632385253906,
"logps/rejected": -68.15009307861328,
"loss": 0.2917,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.21047565340995789,
"rewards/margins": 1.3489338159561157,
"rewards/rejected": -1.5594093799591064,
"step": 238
},
{
"epoch": 0.06,
"learning_rate": 2.61e-07,
"logits/chosen": -9.139519691467285,
"logits/rejected": -9.046548843383789,
"logps/chosen": -39.14946365356445,
"logps/rejected": -66.32569122314453,
"loss": 0.2374,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.24171173572540283,
"rewards/margins": 1.5724042654037476,
"rewards/rejected": -1.8141158819198608,
"step": 239
},
{
"epoch": 0.06,
"learning_rate": 2.6e-07,
"logits/chosen": -8.669023513793945,
"logits/rejected": -8.563835144042969,
"logps/chosen": -40.61223602294922,
"logps/rejected": -69.82514190673828,
"loss": 0.2413,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3103567063808441,
"rewards/margins": 1.6561682224273682,
"rewards/rejected": -1.9665250778198242,
"step": 240
},
{
"epoch": 0.06,
"learning_rate": 2.59e-07,
"logits/chosen": -8.794702529907227,
"logits/rejected": -8.615911483764648,
"logps/chosen": -43.99329376220703,
"logps/rejected": -66.2555923461914,
"loss": 0.2858,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3557160496711731,
"rewards/margins": 1.3395440578460693,
"rewards/rejected": -1.6952602863311768,
"step": 241
},
{
"epoch": 0.06,
"learning_rate": 2.58e-07,
"logits/chosen": -8.988102912902832,
"logits/rejected": -8.940208435058594,
"logps/chosen": -42.496700286865234,
"logps/rejected": -69.17347717285156,
"loss": 0.272,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.460458904504776,
"rewards/margins": 1.4812017679214478,
"rewards/rejected": -1.9416606426239014,
"step": 242
},
{
"epoch": 0.06,
"learning_rate": 2.57e-07,
"logits/chosen": -9.062856674194336,
"logits/rejected": -8.94471263885498,
"logps/chosen": -42.636497497558594,
"logps/rejected": -71.41357421875,
"loss": 0.221,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.4040011465549469,
"rewards/margins": 1.7163587808609009,
"rewards/rejected": -2.1203601360321045,
"step": 243
},
{
"epoch": 0.06,
"learning_rate": 2.56e-07,
"logits/chosen": -9.013895034790039,
"logits/rejected": -9.061233520507812,
"logps/chosen": -46.01321792602539,
"logps/rejected": -71.0923843383789,
"loss": 0.3042,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.4178338646888733,
"rewards/margins": 1.3820786476135254,
"rewards/rejected": -1.799912452697754,
"step": 244
},
{
"epoch": 0.06,
"learning_rate": 2.55e-07,
"logits/chosen": -9.007227897644043,
"logits/rejected": -8.854522705078125,
"logps/chosen": -42.58584213256836,
"logps/rejected": -68.90918731689453,
"loss": 0.2832,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.3119141161441803,
"rewards/margins": 1.5015567541122437,
"rewards/rejected": -1.8134710788726807,
"step": 245
},
{
"epoch": 0.06,
"learning_rate": 2.5399999999999997e-07,
"logits/chosen": -9.030736923217773,
"logits/rejected": -8.89038372039795,
"logps/chosen": -37.5427131652832,
"logps/rejected": -67.01028442382812,
"loss": 0.2396,
"rewards/accuracies": 0.940000057220459,
"rewards/chosen": -0.014831873588263988,
"rewards/margins": 1.6802629232406616,
"rewards/rejected": -1.6950948238372803,
"step": 246
},
{
"epoch": 0.06,
"learning_rate": 2.53e-07,
"logits/chosen": -8.911919593811035,
"logits/rejected": -8.769577026367188,
"logps/chosen": -43.36934280395508,
"logps/rejected": -71.1383056640625,
"loss": 0.2516,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.3899439573287964,
"rewards/margins": 1.618249535560608,
"rewards/rejected": -2.0081934928894043,
"step": 247
},
{
"epoch": 0.06,
"learning_rate": 2.52e-07,
"logits/chosen": -8.795068740844727,
"logits/rejected": -8.727156639099121,
"logps/chosen": -43.60725402832031,
"logps/rejected": -73.10511016845703,
"loss": 0.2449,
"rewards/accuracies": 0.9800000190734863,
"rewards/chosen": -0.5213699340820312,
"rewards/margins": 1.5963902473449707,
"rewards/rejected": -2.117760181427002,
"step": 248
},
{
"epoch": 0.06,
"learning_rate": 2.51e-07,
"logits/chosen": -8.8865385055542,
"logits/rejected": -8.710765838623047,
"logps/chosen": -44.19552230834961,
"logps/rejected": -74.49358367919922,
"loss": 0.224,
"rewards/accuracies": 0.9600000381469727,
"rewards/chosen": -0.5294827222824097,
"rewards/margins": 1.6507676839828491,
"rewards/rejected": -2.180250406265259,
"step": 249
},
{
"epoch": 0.06,
"learning_rate": 2.5e-07,
"logits/chosen": -8.938467979431152,
"logits/rejected": -8.918638229370117,
"logps/chosen": -42.69424819946289,
"logps/rejected": -77.01347351074219,
"loss": 0.1974,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.47195687890052795,
"rewards/margins": 1.8267240524291992,
"rewards/rejected": -2.2986807823181152,
"step": 250
}
],
"logging_steps": 1,
"max_steps": 500,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"total_flos": 0.0,
"train_batch_size": 5,
"trial_name": null,
"trial_params": null
}