Text Generation
Safetensors
Chinese
English
Aurora-dpo / checkpoint-3000 /trainer_state.json
wangrongsheng's picture
Upload 11 files
e412e5f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.3292994154935375,
"eval_steps": 500,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 9.999970269475589e-06,
"logits/chosen": -2.748720645904541,
"logits/rejected": -3.5176024436950684,
"logps/chosen": -165.87062072753906,
"logps/rejected": -23.556869506835938,
"loss": 1.5949,
"rewards/accuracies": 0.612500011920929,
"rewards/chosen": 11.860421180725098,
"rewards/margins": 2.147804021835327,
"rewards/rejected": 9.712616920471191,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 9.999881078255916e-06,
"logits/chosen": -3.261986255645752,
"logits/rejected": -4.041715145111084,
"logps/chosen": -153.11898803710938,
"logps/rejected": -24.845947265625,
"loss": 1.6208,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": 12.161478042602539,
"rewards/margins": 2.482882499694824,
"rewards/rejected": 9.678595542907715,
"step": 20
},
{
"epoch": 0.0,
"learning_rate": 9.99973242740166e-06,
"logits/chosen": -2.687460422515869,
"logits/rejected": -3.837062358856201,
"logps/chosen": -148.68341064453125,
"logps/rejected": -34.57600402832031,
"loss": 1.497,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 12.840588569641113,
"rewards/margins": 2.2458953857421875,
"rewards/rejected": 10.59469223022461,
"step": 30
},
{
"epoch": 0.0,
"learning_rate": 9.99954780509173e-06,
"logits/chosen": -2.9765782356262207,
"logits/rejected": -4.164349555969238,
"logps/chosen": -201.6632537841797,
"logps/rejected": -29.7652645111084,
"loss": 1.8816,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": 12.654102325439453,
"rewards/margins": 2.569594383239746,
"rewards/rejected": 10.084507942199707,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 9.999286186386431e-06,
"logits/chosen": -3.153574228286743,
"logits/rejected": -3.934051036834717,
"logps/chosen": -138.03778076171875,
"logps/rejected": -30.519153594970703,
"loss": 1.9103,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": 11.428102493286133,
"rewards/margins": 1.7032209634780884,
"rewards/rejected": 9.724882125854492,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 9.998965115121132e-06,
"logits/chosen": -3.201958417892456,
"logits/rejected": -4.274596691131592,
"logps/chosen": -166.93260192871094,
"logps/rejected": -35.59614944458008,
"loss": 2.0164,
"rewards/accuracies": 0.5625,
"rewards/chosen": 11.013410568237305,
"rewards/margins": 1.4899368286132812,
"rewards/rejected": 9.523473739624023,
"step": 60
},
{
"epoch": 0.01,
"learning_rate": 9.998584595114074e-06,
"logits/chosen": -3.6508994102478027,
"logits/rejected": -4.988117694854736,
"logps/chosen": -192.28738403320312,
"logps/rejected": -39.88438415527344,
"loss": 0.7081,
"rewards/accuracies": 0.7875000238418579,
"rewards/chosen": 12.313508987426758,
"rewards/margins": 4.361058712005615,
"rewards/rejected": 7.952448844909668,
"step": 70
},
{
"epoch": 0.01,
"learning_rate": 9.998144630890485e-06,
"logits/chosen": -3.2785956859588623,
"logits/rejected": -4.609368801116943,
"logps/chosen": -177.5666961669922,
"logps/rejected": -48.05603790283203,
"loss": 1.1081,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 12.321931838989258,
"rewards/margins": 4.14510440826416,
"rewards/rejected": 8.176828384399414,
"step": 80
},
{
"epoch": 0.01,
"learning_rate": 9.997645227682513e-06,
"logits/chosen": -3.4411368370056152,
"logits/rejected": -4.750131607055664,
"logps/chosen": -163.21450805664062,
"logps/rejected": -53.2952880859375,
"loss": 1.1417,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": 11.179890632629395,
"rewards/margins": 3.490216016769409,
"rewards/rejected": 7.689675331115723,
"step": 90
},
{
"epoch": 0.01,
"learning_rate": 9.997086391429161e-06,
"logits/chosen": -3.429443836212158,
"logits/rejected": -4.5238776206970215,
"logps/chosen": -176.43289184570312,
"logps/rejected": -65.92268371582031,
"loss": 0.9445,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 10.390297889709473,
"rewards/margins": 3.6286094188690186,
"rewards/rejected": 6.761689186096191,
"step": 100
},
{
"epoch": 0.01,
"learning_rate": 9.99646812877623e-06,
"logits/chosen": -3.4572105407714844,
"logits/rejected": -4.460841655731201,
"logps/chosen": -170.5079803466797,
"logps/rejected": -68.17359161376953,
"loss": 0.6313,
"rewards/accuracies": 0.8125,
"rewards/chosen": 10.23057746887207,
"rewards/margins": 4.633530616760254,
"rewards/rejected": 5.5970458984375,
"step": 110
},
{
"epoch": 0.01,
"learning_rate": 9.995790447076232e-06,
"logits/chosen": -3.5251846313476562,
"logits/rejected": -4.237984657287598,
"logps/chosen": -163.4308624267578,
"logps/rejected": -69.18915557861328,
"loss": 0.8116,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 10.12580394744873,
"rewards/margins": 4.935164451599121,
"rewards/rejected": 5.190640449523926,
"step": 120
},
{
"epoch": 0.01,
"learning_rate": 9.995053354388295e-06,
"logits/chosen": -3.2857518196105957,
"logits/rejected": -4.0882792472839355,
"logps/chosen": -211.8911590576172,
"logps/rejected": -76.1227035522461,
"loss": 0.1832,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 10.317492485046387,
"rewards/margins": 5.812389373779297,
"rewards/rejected": 4.505103588104248,
"step": 130
},
{
"epoch": 0.02,
"learning_rate": 9.994256859478082e-06,
"logits/chosen": -3.0444934368133545,
"logits/rejected": -3.676633358001709,
"logps/chosen": -229.4088134765625,
"logps/rejected": -82.24313354492188,
"loss": 0.3931,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": 10.056417465209961,
"rewards/margins": 5.8769307136535645,
"rewards/rejected": 4.17948579788208,
"step": 140
},
{
"epoch": 0.02,
"learning_rate": 9.993489232972927e-06,
"logits/chosen": -3.0874276161193848,
"logits/rejected": -3.6212425231933594,
"logps/chosen": -163.163330078125,
"logps/rejected": -76.16889953613281,
"loss": 0.7483,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 9.392515182495117,
"rewards/margins": 5.074815273284912,
"rewards/rejected": 4.3177008628845215,
"step": 150
},
{
"epoch": 0.02,
"learning_rate": 9.992579900519751e-06,
"logits/chosen": -2.9442169666290283,
"logits/rejected": -3.488318681716919,
"logps/chosen": -202.21347045898438,
"logps/rejected": -84.56549072265625,
"loss": 0.3173,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 9.955198287963867,
"rewards/margins": 6.719505310058594,
"rewards/rejected": 3.2356934547424316,
"step": 160
},
{
"epoch": 0.02,
"learning_rate": 9.99161119525913e-06,
"logits/chosen": -2.819887161254883,
"logits/rejected": -3.2742202281951904,
"logps/chosen": -212.2724151611328,
"logps/rejected": -90.73429870605469,
"loss": 0.3464,
"rewards/accuracies": 0.875,
"rewards/chosen": 9.708670616149902,
"rewards/margins": 5.832002639770508,
"rewards/rejected": 3.8766684532165527,
"step": 170
},
{
"epoch": 0.02,
"learning_rate": 9.990688606280938e-06,
"logits/chosen": -2.7344448566436768,
"logits/rejected": -2.890084743499756,
"logps/chosen": -184.69085693359375,
"logps/rejected": -89.32521057128906,
"loss": 0.6789,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": 9.630158424377441,
"rewards/margins": 5.920450687408447,
"rewards/rejected": 3.7097079753875732,
"step": 180
},
{
"epoch": 0.02,
"learning_rate": 9.989607125007361e-06,
"logits/chosen": -2.5483996868133545,
"logits/rejected": -2.572554588317871,
"logps/chosen": -172.837158203125,
"logps/rejected": -92.3455581665039,
"loss": 0.4471,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 8.161615371704102,
"rewards/margins": 4.883408069610596,
"rewards/rejected": 3.2782065868377686,
"step": 190
},
{
"epoch": 0.02,
"learning_rate": 9.98846630627921e-06,
"logits/chosen": -2.7279348373413086,
"logits/rejected": -2.8250045776367188,
"logps/chosen": -216.2030487060547,
"logps/rejected": -94.36392211914062,
"loss": 0.1686,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 9.2470703125,
"rewards/margins": 6.067578315734863,
"rewards/rejected": 3.1794912815093994,
"step": 200
},
{
"epoch": 0.02,
"learning_rate": 9.98726616366334e-06,
"logits/chosen": -2.476562261581421,
"logits/rejected": -2.579338550567627,
"logps/chosen": -169.55120849609375,
"logps/rejected": -101.41313171386719,
"loss": 0.3145,
"rewards/accuracies": 0.875,
"rewards/chosen": 8.830245018005371,
"rewards/margins": 5.8404154777526855,
"rewards/rejected": 2.9898295402526855,
"step": 210
},
{
"epoch": 0.02,
"learning_rate": 9.986006711432099e-06,
"logits/chosen": -2.6179986000061035,
"logits/rejected": -2.938459873199463,
"logps/chosen": -239.5026397705078,
"logps/rejected": -93.572998046875,
"loss": 0.136,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 10.406990051269531,
"rewards/margins": 7.794189453125,
"rewards/rejected": 2.612799882888794,
"step": 220
},
{
"epoch": 0.03,
"learning_rate": 9.984687964563159e-06,
"logits/chosen": -2.4739646911621094,
"logits/rejected": -2.5754473209381104,
"logps/chosen": -178.31881713867188,
"logps/rejected": -95.78547668457031,
"loss": 0.2827,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.740915298461914,
"rewards/margins": 6.064324855804443,
"rewards/rejected": 2.676591396331787,
"step": 230
},
{
"epoch": 0.03,
"learning_rate": 9.983309938739332e-06,
"logits/chosen": -2.6331839561462402,
"logits/rejected": -2.8605971336364746,
"logps/chosen": -229.9063720703125,
"logps/rejected": -90.2779541015625,
"loss": 0.0817,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 10.207468032836914,
"rewards/margins": 7.982739448547363,
"rewards/rejected": 2.2247281074523926,
"step": 240
},
{
"epoch": 0.03,
"learning_rate": 9.98187265034839e-06,
"logits/chosen": -2.3521595001220703,
"logits/rejected": -2.5626819133758545,
"logps/chosen": -190.8962860107422,
"logps/rejected": -104.4652099609375,
"loss": 0.2826,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 8.663321495056152,
"rewards/margins": 6.1150994300842285,
"rewards/rejected": 2.5482218265533447,
"step": 250
},
{
"epoch": 0.03,
"learning_rate": 9.980376116482871e-06,
"logits/chosen": -2.4073798656463623,
"logits/rejected": -2.531911849975586,
"logps/chosen": -204.7512664794922,
"logps/rejected": -108.78132629394531,
"loss": 0.2884,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 9.377767562866211,
"rewards/margins": 6.906778812408447,
"rewards/rejected": 2.470989227294922,
"step": 260
},
{
"epoch": 0.03,
"learning_rate": 9.978820354939866e-06,
"logits/chosen": -2.2742762565612793,
"logits/rejected": -2.1533379554748535,
"logps/chosen": -189.8150177001953,
"logps/rejected": -112.7627944946289,
"loss": 0.4643,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 9.204684257507324,
"rewards/margins": 6.9484758377075195,
"rewards/rejected": 2.2562096118927,
"step": 270
},
{
"epoch": 0.03,
"learning_rate": 9.97720538422082e-06,
"logits/chosen": -2.1974568367004395,
"logits/rejected": -2.233710765838623,
"logps/chosen": -198.45361328125,
"logps/rejected": -111.978515625,
"loss": 0.1247,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 10.231064796447754,
"rewards/margins": 8.133554458618164,
"rewards/rejected": 2.0975089073181152,
"step": 280
},
{
"epoch": 0.03,
"learning_rate": 9.975531223531301e-06,
"logits/chosen": -2.2490475177764893,
"logits/rejected": -2.30739164352417,
"logps/chosen": -187.04519653320312,
"logps/rejected": -109.82148742675781,
"loss": 0.1268,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.633919715881348,
"rewards/margins": 6.911574363708496,
"rewards/rejected": 1.7223446369171143,
"step": 290
},
{
"epoch": 0.03,
"learning_rate": 9.973797892780783e-06,
"logits/chosen": -2.035712718963623,
"logits/rejected": -1.961692452430725,
"logps/chosen": -179.4069366455078,
"logps/rejected": -116.109130859375,
"loss": 0.0938,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.934889793395996,
"rewards/margins": 8.173246383666992,
"rewards/rejected": 1.7616443634033203,
"step": 300
},
{
"epoch": 0.03,
"learning_rate": 9.972005412582397e-06,
"logits/chosen": -2.2436442375183105,
"logits/rejected": -1.8246514797210693,
"logps/chosen": -201.26980590820312,
"logps/rejected": -117.66944885253906,
"loss": 0.2073,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 8.626791954040527,
"rewards/margins": 7.419760227203369,
"rewards/rejected": 1.207031488418579,
"step": 310
},
{
"epoch": 0.04,
"learning_rate": 9.970153804252692e-06,
"logits/chosen": -1.8984334468841553,
"logits/rejected": -1.6126270294189453,
"logps/chosen": -185.87290954589844,
"logps/rejected": -117.4137954711914,
"loss": 0.278,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 8.371435165405273,
"rewards/margins": 7.151828765869141,
"rewards/rejected": 1.2196067571640015,
"step": 320
},
{
"epoch": 0.04,
"learning_rate": 9.968243089811384e-06,
"logits/chosen": -1.771867036819458,
"logits/rejected": -1.5103999376296997,
"logps/chosen": -192.2825469970703,
"logps/rejected": -118.695556640625,
"loss": 0.0729,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.544290542602539,
"rewards/margins": 7.602395057678223,
"rewards/rejected": 0.9418962597846985,
"step": 330
},
{
"epoch": 0.04,
"learning_rate": 9.96627329198109e-06,
"logits/chosen": -1.924703598022461,
"logits/rejected": -1.3986445665359497,
"logps/chosen": -203.59628295898438,
"logps/rejected": -128.64163208007812,
"loss": 0.2888,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 8.449407577514648,
"rewards/margins": 7.650280952453613,
"rewards/rejected": 0.7991273999214172,
"step": 340
},
{
"epoch": 0.04,
"learning_rate": 9.96424443418706e-06,
"logits/chosen": -1.8718311786651611,
"logits/rejected": -1.552980661392212,
"logps/chosen": -181.55087280273438,
"logps/rejected": -120.4705581665039,
"loss": 0.0762,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.385054588317871,
"rewards/margins": 7.822892665863037,
"rewards/rejected": 0.5621613264083862,
"step": 350
},
{
"epoch": 0.04,
"learning_rate": 9.962156540556895e-06,
"logits/chosen": -1.7886946201324463,
"logits/rejected": -1.5114587545394897,
"logps/chosen": -183.34066772460938,
"logps/rejected": -117.0723648071289,
"loss": 0.1247,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.962329864501953,
"rewards/margins": 7.366267204284668,
"rewards/rejected": 0.5960620045661926,
"step": 360
},
{
"epoch": 0.04,
"learning_rate": 9.960009635920264e-06,
"logits/chosen": -1.788691520690918,
"logits/rejected": -1.3814653158187866,
"logps/chosen": -197.45651245117188,
"logps/rejected": -127.6612319946289,
"loss": 0.076,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.019963264465332,
"rewards/margins": 8.529821395874023,
"rewards/rejected": 0.49014148116111755,
"step": 370
},
{
"epoch": 0.04,
"learning_rate": 9.95780374580861e-06,
"logits/chosen": -1.5638988018035889,
"logits/rejected": -1.1953327655792236,
"logps/chosen": -189.52597045898438,
"logps/rejected": -122.5257797241211,
"loss": 0.2198,
"rewards/accuracies": 0.9375,
"rewards/chosen": 7.37133264541626,
"rewards/margins": 6.876678466796875,
"rewards/rejected": 0.4946538507938385,
"step": 380
},
{
"epoch": 0.04,
"learning_rate": 9.955538896454835e-06,
"logits/chosen": -1.5664819478988647,
"logits/rejected": -0.962997317314148,
"logps/chosen": -204.67657470703125,
"logps/rejected": -129.93968200683594,
"loss": 0.0446,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.449831485748291,
"rewards/margins": 7.785650730133057,
"rewards/rejected": -0.33581945300102234,
"step": 390
},
{
"epoch": 0.04,
"learning_rate": 9.95321511479301e-06,
"logits/chosen": -1.7048896551132202,
"logits/rejected": -1.1939642429351807,
"logps/chosen": -223.2345428466797,
"logps/rejected": -130.3861846923828,
"loss": 0.1745,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.562383651733398,
"rewards/margins": 8.441825866699219,
"rewards/rejected": 0.12055821716785431,
"step": 400
},
{
"epoch": 0.05,
"learning_rate": 9.950832428458025e-06,
"logits/chosen": -1.2762507200241089,
"logits/rejected": -0.9500716924667358,
"logps/chosen": -195.8381805419922,
"logps/rejected": -124.5715560913086,
"loss": 0.1329,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.099652290344238,
"rewards/margins": 7.511727809906006,
"rewards/rejected": -0.41207608580589294,
"step": 410
},
{
"epoch": 0.05,
"learning_rate": 9.948390865785295e-06,
"logits/chosen": -1.151587963104248,
"logits/rejected": -0.925134539604187,
"logps/chosen": -199.50967407226562,
"logps/rejected": -137.6339111328125,
"loss": 0.1983,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.476855278015137,
"rewards/margins": 7.712057590484619,
"rewards/rejected": -0.23520107567310333,
"step": 420
},
{
"epoch": 0.05,
"learning_rate": 9.94589045581039e-06,
"logits/chosen": -1.0321991443634033,
"logits/rejected": -0.8457671999931335,
"logps/chosen": -206.13467407226562,
"logps/rejected": -130.86544799804688,
"loss": 0.2554,
"rewards/accuracies": 0.9375,
"rewards/chosen": 7.223177433013916,
"rewards/margins": 7.446470737457275,
"rewards/rejected": -0.2232942134141922,
"step": 430
},
{
"epoch": 0.05,
"learning_rate": 9.943331228268711e-06,
"logits/chosen": -1.1853946447372437,
"logits/rejected": -0.7296536564826965,
"logps/chosen": -217.77304077148438,
"logps/rejected": -127.19853210449219,
"loss": 0.1477,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.660037040710449,
"rewards/margins": 8.213974952697754,
"rewards/rejected": -0.5539380311965942,
"step": 440
},
{
"epoch": 0.05,
"learning_rate": 9.94071321359513e-06,
"logits/chosen": -0.9670246243476868,
"logits/rejected": -0.314952552318573,
"logps/chosen": -225.93533325195312,
"logps/rejected": -154.71826171875,
"loss": 0.172,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.689838409423828,
"rewards/margins": 8.839930534362793,
"rewards/rejected": -1.1500914096832275,
"step": 450
},
{
"epoch": 0.05,
"learning_rate": 9.938036442923624e-06,
"logits/chosen": -0.9485570788383484,
"logits/rejected": -0.33327311277389526,
"logps/chosen": -225.99398803710938,
"logps/rejected": -142.39173889160156,
"loss": 0.0579,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.543333530426025,
"rewards/margins": 8.586603164672852,
"rewards/rejected": -1.0432696342468262,
"step": 460
},
{
"epoch": 0.05,
"learning_rate": 9.935300948086915e-06,
"logits/chosen": -0.8986530303955078,
"logits/rejected": -0.400226354598999,
"logps/chosen": -209.4599609375,
"logps/rejected": -130.23941040039062,
"loss": 0.1114,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.468968391418457,
"rewards/margins": 8.008504867553711,
"rewards/rejected": -0.5395378470420837,
"step": 470
},
{
"epoch": 0.05,
"learning_rate": 9.932506761616078e-06,
"logits/chosen": -0.7118775248527527,
"logits/rejected": -0.03559774532914162,
"logps/chosen": -193.45834350585938,
"logps/rejected": -131.99844360351562,
"loss": 0.0508,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.474764823913574,
"rewards/margins": 8.308823585510254,
"rewards/rejected": -0.8340586423873901,
"step": 480
},
{
"epoch": 0.05,
"learning_rate": 9.929653916740168e-06,
"logits/chosen": -1.1100711822509766,
"logits/rejected": -0.40355220437049866,
"logps/chosen": -242.05697631835938,
"logps/rejected": -139.188232421875,
"loss": 0.1127,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.726511001586914,
"rewards/margins": 9.716850280761719,
"rewards/rejected": -0.9903383255004883,
"step": 490
},
{
"epoch": 0.05,
"learning_rate": 9.92674244738581e-06,
"logits/chosen": -0.8486649394035339,
"logits/rejected": 0.05283903330564499,
"logps/chosen": -155.493408203125,
"logps/rejected": -141.4785919189453,
"loss": 0.0901,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.535587310791016,
"rewards/margins": 8.244181632995605,
"rewards/rejected": -0.7085943818092346,
"step": 500
},
{
"epoch": 0.06,
"learning_rate": 9.923772388176812e-06,
"logits/chosen": -0.907477080821991,
"logits/rejected": -0.171137273311615,
"logps/chosen": -236.19198608398438,
"logps/rejected": -133.06643676757812,
"loss": 0.0178,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.850728034973145,
"rewards/margins": 10.053266525268555,
"rewards/rejected": -1.202539086341858,
"step": 510
},
{
"epoch": 0.06,
"learning_rate": 9.92074377443374e-06,
"logits/chosen": -1.0009920597076416,
"logits/rejected": -0.20153608918190002,
"logps/chosen": -178.60147094726562,
"logps/rejected": -143.53695678710938,
"loss": 0.1976,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.2967023849487305,
"rewards/margins": 8.339073181152344,
"rewards/rejected": -1.0423698425292969,
"step": 520
},
{
"epoch": 0.06,
"learning_rate": 9.917656642173505e-06,
"logits/chosen": -1.0008546113967896,
"logits/rejected": -0.01453827042132616,
"logps/chosen": -206.38461303710938,
"logps/rejected": -144.1584014892578,
"loss": 0.1575,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.916428565979004,
"rewards/margins": 9.663307189941406,
"rewards/rejected": -0.7468785047531128,
"step": 530
},
{
"epoch": 0.06,
"learning_rate": 9.91451102810893e-06,
"logits/chosen": -1.326416015625,
"logits/rejected": -0.38943323493003845,
"logps/chosen": -220.2677764892578,
"logps/rejected": -130.0069122314453,
"loss": 0.054,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.959368705749512,
"rewards/margins": 9.750789642333984,
"rewards/rejected": -0.7914207577705383,
"step": 540
},
{
"epoch": 0.06,
"learning_rate": 9.911306969648318e-06,
"logits/chosen": -0.932703971862793,
"logits/rejected": -0.0587722584605217,
"logps/chosen": -205.8779296875,
"logps/rejected": -135.040283203125,
"loss": 0.1965,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 8.557300567626953,
"rewards/margins": 9.26187515258789,
"rewards/rejected": -0.7045742869377136,
"step": 550
},
{
"epoch": 0.06,
"learning_rate": 9.908044504895004e-06,
"logits/chosen": -0.9326754808425903,
"logits/rejected": -0.19441956281661987,
"logps/chosen": -195.15328979492188,
"logps/rejected": -131.31869506835938,
"loss": 0.0943,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.366579055786133,
"rewards/margins": 9.570085525512695,
"rewards/rejected": -1.203505277633667,
"step": 560
},
{
"epoch": 0.06,
"learning_rate": 9.904723672646903e-06,
"logits/chosen": -0.9648317098617554,
"logits/rejected": -0.06580810248851776,
"logps/chosen": -211.77633666992188,
"logps/rejected": -136.65756225585938,
"loss": 0.2817,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 8.721487045288086,
"rewards/margins": 9.342923164367676,
"rewards/rejected": -0.6214349865913391,
"step": 570
},
{
"epoch": 0.06,
"learning_rate": 9.901344512396049e-06,
"logits/chosen": -1.1457253694534302,
"logits/rejected": -0.2043023556470871,
"logps/chosen": -210.77450561523438,
"logps/rejected": -128.11520385742188,
"loss": 0.1145,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.162195205688477,
"rewards/margins": 9.146950721740723,
"rewards/rejected": -0.9847553968429565,
"step": 580
},
{
"epoch": 0.06,
"learning_rate": 9.897907064328125e-06,
"logits/chosen": -0.8582237362861633,
"logits/rejected": -0.14161944389343262,
"logps/chosen": -202.4379425048828,
"logps/rejected": -136.9176483154297,
"loss": 0.2544,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.603880882263184,
"rewards/margins": 9.390780448913574,
"rewards/rejected": -0.7868987321853638,
"step": 590
},
{
"epoch": 0.07,
"learning_rate": 9.894411369321984e-06,
"logits/chosen": -0.9913557767868042,
"logits/rejected": -0.0038151517510414124,
"logps/chosen": -182.00146484375,
"logps/rejected": -135.45352172851562,
"loss": 0.0794,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.4787445068359375,
"rewards/margins": 8.599211692810059,
"rewards/rejected": -1.120466947555542,
"step": 600
},
{
"epoch": 0.07,
"learning_rate": 9.890857468949164e-06,
"logits/chosen": -0.9070969820022583,
"logits/rejected": 0.03064725361764431,
"logps/chosen": -223.1108856201172,
"logps/rejected": -145.6663055419922,
"loss": 0.0421,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.449647903442383,
"rewards/margins": 10.343244552612305,
"rewards/rejected": -1.8935970067977905,
"step": 610
},
{
"epoch": 0.07,
"learning_rate": 9.887245405473393e-06,
"logits/chosen": -0.7473093867301941,
"logits/rejected": 0.2375013381242752,
"logps/chosen": -218.11428833007812,
"logps/rejected": -137.17213439941406,
"loss": 0.0778,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.949044704437256,
"rewards/margins": 8.823598861694336,
"rewards/rejected": -0.8745536804199219,
"step": 620
},
{
"epoch": 0.07,
"learning_rate": 9.88357522185009e-06,
"logits/chosen": -0.913166344165802,
"logits/rejected": -0.011747682467103004,
"logps/chosen": -220.34603881835938,
"logps/rejected": -141.53561401367188,
"loss": 0.0287,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.446084976196289,
"rewards/margins": 9.644007682800293,
"rewards/rejected": -1.1979233026504517,
"step": 630
},
{
"epoch": 0.07,
"learning_rate": 9.879846961725846e-06,
"logits/chosen": -0.8051003217697144,
"logits/rejected": 0.030248349532485008,
"logps/chosen": -199.5961456298828,
"logps/rejected": -155.56056213378906,
"loss": 0.0765,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.313097953796387,
"rewards/margins": 9.846205711364746,
"rewards/rejected": -1.5331077575683594,
"step": 640
},
{
"epoch": 0.07,
"learning_rate": 9.876060669437913e-06,
"logits/chosen": -0.616194486618042,
"logits/rejected": 0.2745420038700104,
"logps/chosen": -215.4296417236328,
"logps/rejected": -142.76492309570312,
"loss": 0.0848,
"rewards/accuracies": 0.9375,
"rewards/chosen": 7.84356164932251,
"rewards/margins": 9.544157981872559,
"rewards/rejected": -1.7005951404571533,
"step": 650
},
{
"epoch": 0.07,
"learning_rate": 9.872216390013675e-06,
"logits/chosen": -0.7901605367660522,
"logits/rejected": 0.41072148084640503,
"logps/chosen": -185.0421142578125,
"logps/rejected": -144.97398376464844,
"loss": 0.1577,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.704103946685791,
"rewards/margins": 9.240386009216309,
"rewards/rejected": -1.5362818241119385,
"step": 660
},
{
"epoch": 0.07,
"learning_rate": 9.868314169170108e-06,
"logits/chosen": -0.5289187431335449,
"logits/rejected": 0.397865355014801,
"logps/chosen": -219.9363555908203,
"logps/rejected": -142.4893798828125,
"loss": 0.1996,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 7.231544494628906,
"rewards/margins": 8.826276779174805,
"rewards/rejected": -1.5947315692901611,
"step": 670
},
{
"epoch": 0.07,
"learning_rate": 9.86435405331324e-06,
"logits/chosen": -0.3374841809272766,
"logits/rejected": 0.4616989195346832,
"logps/chosen": -197.3474578857422,
"logps/rejected": -141.45425415039062,
"loss": 0.0638,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.858680725097656,
"rewards/margins": 9.586578369140625,
"rewards/rejected": -1.7278966903686523,
"step": 680
},
{
"epoch": 0.08,
"learning_rate": 9.860336089537601e-06,
"logits/chosen": -0.4742531180381775,
"logits/rejected": 0.3263815939426422,
"logps/chosen": -197.79933166503906,
"logps/rejected": -145.15786743164062,
"loss": 0.1294,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.016458988189697,
"rewards/margins": 8.728564262390137,
"rewards/rejected": -1.7121059894561768,
"step": 690
},
{
"epoch": 0.08,
"learning_rate": 9.856260325625657e-06,
"logits/chosen": -0.6684874892234802,
"logits/rejected": 0.5160067677497864,
"logps/chosen": -215.1957550048828,
"logps/rejected": -157.3702392578125,
"loss": 0.0588,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 6.882597923278809,
"rewards/margins": 9.006999969482422,
"rewards/rejected": -2.1244006156921387,
"step": 700
},
{
"epoch": 0.08,
"learning_rate": 9.85212681004725e-06,
"logits/chosen": -0.7230426073074341,
"logits/rejected": 0.48965024948120117,
"logps/chosen": -238.30581665039062,
"logps/rejected": -151.21774291992188,
"loss": 0.0211,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.455827713012695,
"rewards/margins": 10.841313362121582,
"rewards/rejected": -2.385484457015991,
"step": 710
},
{
"epoch": 0.08,
"learning_rate": 9.847935591959011e-06,
"logits/chosen": -0.2026168555021286,
"logits/rejected": 0.6747661828994751,
"logps/chosen": -206.3802490234375,
"logps/rejected": -141.45103454589844,
"loss": 0.0742,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.532872676849365,
"rewards/margins": 9.587210655212402,
"rewards/rejected": -2.0543372631073,
"step": 720
},
{
"epoch": 0.08,
"learning_rate": 9.84368672120379e-06,
"logits/chosen": -0.3312041461467743,
"logits/rejected": 0.7338079214096069,
"logps/chosen": -200.46630859375,
"logps/rejected": -147.34877014160156,
"loss": 0.1475,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.3677077293396,
"rewards/margins": 9.639891624450684,
"rewards/rejected": -2.272183656692505,
"step": 730
},
{
"epoch": 0.08,
"learning_rate": 9.839380248310043e-06,
"logits/chosen": -0.16066689789295197,
"logits/rejected": 0.8108816146850586,
"logps/chosen": -175.85104370117188,
"logps/rejected": -157.10450744628906,
"loss": 0.0537,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.471951961517334,
"rewards/margins": 9.829510688781738,
"rewards/rejected": -2.3575589656829834,
"step": 740
},
{
"epoch": 0.08,
"learning_rate": 9.835016224491251e-06,
"logits/chosen": -0.3715950846672058,
"logits/rejected": 0.7917531728744507,
"logps/chosen": -183.4439697265625,
"logps/rejected": -141.3053436279297,
"loss": 0.0628,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.702794551849365,
"rewards/margins": 9.382607460021973,
"rewards/rejected": -1.6798127889633179,
"step": 750
},
{
"epoch": 0.08,
"learning_rate": 9.830594701645304e-06,
"logits/chosen": -0.3017290234565735,
"logits/rejected": 0.6022787094116211,
"logps/chosen": -198.8234100341797,
"logps/rejected": -144.2804412841797,
"loss": 0.0508,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.64525842666626,
"rewards/margins": 9.416332244873047,
"rewards/rejected": -1.7710736989974976,
"step": 760
},
{
"epoch": 0.08,
"learning_rate": 9.826115732353877e-06,
"logits/chosen": -0.36430418491363525,
"logits/rejected": 0.5193418860435486,
"logps/chosen": -256.91583251953125,
"logps/rejected": -153.57138061523438,
"loss": 0.1406,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.545983791351318,
"rewards/margins": 10.080999374389648,
"rewards/rejected": -2.535015821456909,
"step": 770
},
{
"epoch": 0.09,
"learning_rate": 9.82157936988181e-06,
"logits/chosen": -0.4545150697231293,
"logits/rejected": 0.928754985332489,
"logps/chosen": -220.9812469482422,
"logps/rejected": -150.5502166748047,
"loss": 0.0332,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.531121730804443,
"rewards/margins": 9.915712356567383,
"rewards/rejected": -2.3845906257629395,
"step": 780
},
{
"epoch": 0.09,
"learning_rate": 9.816985668176482e-06,
"logits/chosen": -0.7030404806137085,
"logits/rejected": 0.5952990055084229,
"logps/chosen": -231.7360382080078,
"logps/rejected": -148.20033264160156,
"loss": 0.152,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.060033798217773,
"rewards/margins": 9.900753021240234,
"rewards/rejected": -1.8407185077667236,
"step": 790
},
{
"epoch": 0.09,
"learning_rate": 9.812334681867152e-06,
"logits/chosen": -0.7694862484931946,
"logits/rejected": 0.30579477548599243,
"logps/chosen": -182.74615478515625,
"logps/rejected": -146.8044891357422,
"loss": 0.106,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.797210693359375,
"rewards/margins": 9.230527877807617,
"rewards/rejected": -1.4333173036575317,
"step": 800
},
{
"epoch": 0.09,
"learning_rate": 9.807626466264328e-06,
"logits/chosen": -0.6776883006095886,
"logits/rejected": 0.4075050354003906,
"logps/chosen": -188.23257446289062,
"logps/rejected": -147.1197052001953,
"loss": 0.0302,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.79939079284668,
"rewards/margins": 10.872949600219727,
"rewards/rejected": -2.073558807373047,
"step": 810
},
{
"epoch": 0.09,
"learning_rate": 9.802861077359098e-06,
"logits/chosen": -0.6511849164962769,
"logits/rejected": 0.4911740720272064,
"logps/chosen": -197.52383422851562,
"logps/rejected": -153.37123107910156,
"loss": 0.1162,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.9149274826049805,
"rewards/margins": 10.323246955871582,
"rewards/rejected": -2.4083173274993896,
"step": 820
},
{
"epoch": 0.09,
"learning_rate": 9.798038571822463e-06,
"logits/chosen": -0.8261006474494934,
"logits/rejected": 0.3168261647224426,
"logps/chosen": -200.90821838378906,
"logps/rejected": -147.97134399414062,
"loss": 0.0839,
"rewards/accuracies": 0.9375,
"rewards/chosen": 7.613385200500488,
"rewards/margins": 9.651713371276855,
"rewards/rejected": -2.038329601287842,
"step": 830
},
{
"epoch": 0.09,
"learning_rate": 9.793159007004675e-06,
"logits/chosen": -0.9683536291122437,
"logits/rejected": -0.2409909963607788,
"logps/chosen": -231.24169921875,
"logps/rejected": -154.0245361328125,
"loss": 0.0287,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.987038612365723,
"rewards/margins": 10.66617202758789,
"rewards/rejected": -1.6791340112686157,
"step": 840
},
{
"epoch": 0.09,
"learning_rate": 9.78822244093454e-06,
"logits/chosen": -0.7381896376609802,
"logits/rejected": 0.2965128421783447,
"logps/chosen": -174.1750946044922,
"logps/rejected": -151.99537658691406,
"loss": 0.0919,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.649115085601807,
"rewards/margins": 9.841694831848145,
"rewards/rejected": -2.192580223083496,
"step": 850
},
{
"epoch": 0.09,
"learning_rate": 9.783228932318736e-06,
"logits/chosen": -0.6059221625328064,
"logits/rejected": 0.08314159512519836,
"logps/chosen": -210.4627227783203,
"logps/rejected": -152.2406768798828,
"loss": 0.0535,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.937192440032959,
"rewards/margins": 9.917118072509766,
"rewards/rejected": -1.979925513267517,
"step": 860
},
{
"epoch": 0.1,
"learning_rate": 9.778178540541116e-06,
"logits/chosen": -0.730563759803772,
"logits/rejected": 0.34863799810409546,
"logps/chosen": -200.68209838867188,
"logps/rejected": -151.97323608398438,
"loss": 0.0397,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.986664772033691,
"rewards/margins": 10.529966354370117,
"rewards/rejected": -2.543302059173584,
"step": 870
},
{
"epoch": 0.1,
"learning_rate": 9.773071325662e-06,
"logits/chosen": -0.8053274154663086,
"logits/rejected": 0.12549547851085663,
"logps/chosen": -192.92202758789062,
"logps/rejected": -146.79693603515625,
"loss": 0.1526,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.616433143615723,
"rewards/margins": 10.316951751708984,
"rewards/rejected": -1.7005198001861572,
"step": 880
},
{
"epoch": 0.1,
"learning_rate": 9.767907348417458e-06,
"logits/chosen": -0.6939282417297363,
"logits/rejected": 0.5801184773445129,
"logps/chosen": -237.7701873779297,
"logps/rejected": -150.30313110351562,
"loss": 0.2363,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.603517532348633,
"rewards/margins": 10.700518608093262,
"rewards/rejected": -2.0970005989074707,
"step": 890
},
{
"epoch": 0.1,
"learning_rate": 9.762686670218588e-06,
"logits/chosen": -0.5126532912254333,
"logits/rejected": 0.5091619491577148,
"logps/chosen": -213.9130401611328,
"logps/rejected": -155.66085815429688,
"loss": 0.0143,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.648975372314453,
"rewards/margins": 11.388581275939941,
"rewards/rejected": -2.739605665206909,
"step": 900
},
{
"epoch": 0.1,
"learning_rate": 9.757409353150793e-06,
"logits/chosen": -0.4479140341281891,
"logits/rejected": 0.5997704267501831,
"logps/chosen": -206.91928100585938,
"logps/rejected": -153.65902709960938,
"loss": 0.0504,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 6.941058158874512,
"rewards/margins": 9.638825416564941,
"rewards/rejected": -2.6977667808532715,
"step": 910
},
{
"epoch": 0.1,
"learning_rate": 9.752075459973034e-06,
"logits/chosen": -0.48702773451805115,
"logits/rejected": 0.5097901225090027,
"logps/chosen": -253.9662322998047,
"logps/rejected": -161.4595947265625,
"loss": 0.0841,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.159878730773926,
"rewards/margins": 10.821348190307617,
"rewards/rejected": -2.661470890045166,
"step": 920
},
{
"epoch": 0.1,
"learning_rate": 9.746685054117086e-06,
"logits/chosen": -0.4953214228153229,
"logits/rejected": 0.5581221580505371,
"logps/chosen": -248.16482543945312,
"logps/rejected": -150.33505249023438,
"loss": 0.0869,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.8457465171813965,
"rewards/margins": 10.702252388000488,
"rewards/rejected": -2.856506109237671,
"step": 930
},
{
"epoch": 0.1,
"learning_rate": 9.741238199686789e-06,
"logits/chosen": -0.5145618915557861,
"logits/rejected": 0.7150206565856934,
"logps/chosen": -236.22683715820312,
"logps/rejected": -149.36941528320312,
"loss": 0.0362,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.276123046875,
"rewards/margins": 10.051583290100098,
"rewards/rejected": -2.775459051132202,
"step": 940
},
{
"epoch": 0.1,
"learning_rate": 9.735734961457277e-06,
"logits/chosen": -0.4139340817928314,
"logits/rejected": 0.7916821241378784,
"logps/chosen": -179.39382934570312,
"logps/rejected": -153.31869506835938,
"loss": 0.0393,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.780765533447266,
"rewards/margins": 10.50986099243164,
"rewards/rejected": -2.729094982147217,
"step": 950
},
{
"epoch": 0.11,
"learning_rate": 9.730175404874212e-06,
"logits/chosen": -0.3699759542942047,
"logits/rejected": 0.7061504125595093,
"logps/chosen": -173.6160888671875,
"logps/rejected": -157.6312713623047,
"loss": 0.15,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.363532066345215,
"rewards/margins": 10.816950798034668,
"rewards/rejected": -2.4534192085266113,
"step": 960
},
{
"epoch": 0.11,
"learning_rate": 9.724559596053009e-06,
"logits/chosen": -0.6346268057823181,
"logits/rejected": 0.48736244440078735,
"logps/chosen": -197.9464111328125,
"logps/rejected": -163.78814697265625,
"loss": 0.0861,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.08003044128418,
"rewards/margins": 10.763525009155273,
"rewards/rejected": -2.683495044708252,
"step": 970
},
{
"epoch": 0.11,
"learning_rate": 9.718887601778044e-06,
"logits/chosen": -0.4699893891811371,
"logits/rejected": 0.5934313535690308,
"logps/chosen": -172.5706787109375,
"logps/rejected": -147.60012817382812,
"loss": 0.0598,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.218860626220703,
"rewards/margins": 9.611445426940918,
"rewards/rejected": -2.3925843238830566,
"step": 980
},
{
"epoch": 0.11,
"learning_rate": 9.713159489501863e-06,
"logits/chosen": -0.7212098836898804,
"logits/rejected": 0.6830633282661438,
"logps/chosen": -195.26705932617188,
"logps/rejected": -144.74806213378906,
"loss": 0.0301,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.268791198730469,
"rewards/margins": 9.785104751586914,
"rewards/rejected": -2.5163135528564453,
"step": 990
},
{
"epoch": 0.11,
"learning_rate": 9.707956263849591e-06,
"logits/chosen": -0.653510570526123,
"logits/rejected": 0.6825904250144958,
"logps/chosen": -214.06729125976562,
"logps/rejected": -151.3599395751953,
"loss": 0.1073,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.91938591003418,
"rewards/margins": 11.606144905090332,
"rewards/rejected": -2.6867592334747314,
"step": 1000
},
{
"epoch": 0.11,
"learning_rate": 9.702121715592386e-06,
"logits/chosen": -0.6122604608535767,
"logits/rejected": 0.5413939356803894,
"logps/chosen": -245.49929809570312,
"logps/rejected": -154.56021118164062,
"loss": 0.098,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.354562759399414,
"rewards/margins": 10.709449768066406,
"rewards/rejected": -2.3548874855041504,
"step": 1010
},
{
"epoch": 0.11,
"learning_rate": 9.696231248717397e-06,
"logits/chosen": -0.3631553053855896,
"logits/rejected": 0.6392610669136047,
"logps/chosen": -209.0885467529297,
"logps/rejected": -144.05416870117188,
"loss": 0.053,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.57081413269043,
"rewards/margins": 10.535473823547363,
"rewards/rejected": -1.9646602869033813,
"step": 1020
},
{
"epoch": 0.11,
"learning_rate": 9.690284933275296e-06,
"logits/chosen": -0.7420902252197266,
"logits/rejected": 0.7464100122451782,
"logps/chosen": -223.1457061767578,
"logps/rejected": -161.8677215576172,
"loss": 0.125,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.317934036254883,
"rewards/margins": 10.725369453430176,
"rewards/rejected": -2.407435894012451,
"step": 1030
},
{
"epoch": 0.11,
"learning_rate": 9.684282839980912e-06,
"logits/chosen": -0.4254356324672699,
"logits/rejected": 0.7181139588356018,
"logps/chosen": -219.45327758789062,
"logps/rejected": -147.68211364746094,
"loss": 0.0078,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.572770118713379,
"rewards/margins": 11.732831954956055,
"rewards/rejected": -3.1600615978240967,
"step": 1040
},
{
"epoch": 0.12,
"learning_rate": 9.678225040212398e-06,
"logits/chosen": -0.22140467166900635,
"logits/rejected": 1.044932246208191,
"logps/chosen": -178.0739288330078,
"logps/rejected": -160.81057739257812,
"loss": 0.148,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.5312628746032715,
"rewards/margins": 10.080840110778809,
"rewards/rejected": -2.5495781898498535,
"step": 1050
},
{
"epoch": 0.12,
"learning_rate": 9.67211160601038e-06,
"logits/chosen": -0.3561934232711792,
"logits/rejected": 0.9342392683029175,
"logps/chosen": -224.42599487304688,
"logps/rejected": -165.5540771484375,
"loss": 0.007,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.821218490600586,
"rewards/margins": 12.073785781860352,
"rewards/rejected": -3.252568483352661,
"step": 1060
},
{
"epoch": 0.12,
"learning_rate": 9.665942610077096e-06,
"logits/chosen": -0.2816230058670044,
"logits/rejected": 0.7103188037872314,
"logps/chosen": -226.01358032226562,
"logps/rejected": -147.710205078125,
"loss": 0.1092,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.991147994995117,
"rewards/margins": 11.593085289001465,
"rewards/rejected": -2.601935863494873,
"step": 1070
},
{
"epoch": 0.12,
"learning_rate": 9.659718125775547e-06,
"logits/chosen": -0.5628015995025635,
"logits/rejected": 0.7087093591690063,
"logps/chosen": -212.25149536132812,
"logps/rejected": -164.18304443359375,
"loss": 0.2843,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.73010540008545,
"rewards/margins": 11.143708229064941,
"rewards/rejected": -2.413602352142334,
"step": 1080
},
{
"epoch": 0.12,
"learning_rate": 9.6534382271286e-06,
"logits/chosen": -0.7271457314491272,
"logits/rejected": 0.3624531626701355,
"logps/chosen": -244.6057891845703,
"logps/rejected": -150.74534606933594,
"loss": 0.0431,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.928186416625977,
"rewards/margins": 11.571906089782715,
"rewards/rejected": -2.64371919631958,
"step": 1090
},
{
"epoch": 0.12,
"learning_rate": 9.64710298881813e-06,
"logits/chosen": -0.42074766755104065,
"logits/rejected": 0.531056821346283,
"logps/chosen": -178.32199096679688,
"logps/rejected": -145.59683227539062,
"loss": 0.0996,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.120710372924805,
"rewards/margins": 10.442843437194824,
"rewards/rejected": -2.3221330642700195,
"step": 1100
},
{
"epoch": 0.12,
"learning_rate": 9.640712486184117e-06,
"logits/chosen": -0.7548767924308777,
"logits/rejected": 0.36750084161758423,
"logps/chosen": -216.7880859375,
"logps/rejected": -160.77999877929688,
"loss": 0.1199,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.382489204406738,
"rewards/margins": 11.168853759765625,
"rewards/rejected": -2.7863640785217285,
"step": 1110
},
{
"epoch": 0.12,
"learning_rate": 9.634266795223761e-06,
"logits/chosen": -0.36768269538879395,
"logits/rejected": 0.8007356524467468,
"logps/chosen": -182.25375366210938,
"logps/rejected": -149.307373046875,
"loss": 0.0616,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 6.98935079574585,
"rewards/margins": 9.277505874633789,
"rewards/rejected": -2.288156032562256,
"step": 1120
},
{
"epoch": 0.12,
"learning_rate": 9.627765992590572e-06,
"logits/chosen": -0.679458737373352,
"logits/rejected": 0.520967423915863,
"logps/chosen": -194.57183837890625,
"logps/rejected": -157.29466247558594,
"loss": 0.0976,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.470559120178223,
"rewards/margins": 11.556069374084473,
"rewards/rejected": -3.0855116844177246,
"step": 1130
},
{
"epoch": 0.13,
"learning_rate": 9.621210155593457e-06,
"logits/chosen": -0.3645058274269104,
"logits/rejected": 0.6819556355476379,
"logps/chosen": -183.23855590820312,
"logps/rejected": -151.49508666992188,
"loss": 0.1729,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 6.996212005615234,
"rewards/margins": 9.404515266418457,
"rewards/rejected": -2.408303737640381,
"step": 1140
},
{
"epoch": 0.13,
"learning_rate": 9.614599362195803e-06,
"logits/chosen": -0.2955774664878845,
"logits/rejected": 0.5863500833511353,
"logps/chosen": -206.7914276123047,
"logps/rejected": -146.72183227539062,
"loss": 0.0833,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.1628999710083,
"rewards/margins": 10.706645965576172,
"rewards/rejected": -2.5437464714050293,
"step": 1150
},
{
"epoch": 0.13,
"learning_rate": 9.607933691014556e-06,
"logits/chosen": -0.5553183555603027,
"logits/rejected": 0.3617022633552551,
"logps/chosen": -198.2234649658203,
"logps/rejected": -154.90408325195312,
"loss": 0.0427,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.177728652954102,
"rewards/margins": 11.280416488647461,
"rewards/rejected": -3.1026864051818848,
"step": 1160
},
{
"epoch": 0.13,
"learning_rate": 9.601213221319271e-06,
"logits/chosen": -0.6628633141517639,
"logits/rejected": 0.6921408176422119,
"logps/chosen": -192.33914184570312,
"logps/rejected": -155.47396850585938,
"loss": 0.0349,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.054289817810059,
"rewards/margins": 11.028264045715332,
"rewards/rejected": -2.973972797393799,
"step": 1170
},
{
"epoch": 0.13,
"learning_rate": 9.594438033031188e-06,
"logits/chosen": -0.510148286819458,
"logits/rejected": 0.6729466319084167,
"logps/chosen": -183.96536254882812,
"logps/rejected": -155.4961395263672,
"loss": 0.0129,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.211958885192871,
"rewards/margins": 11.422294616699219,
"rewards/rejected": -3.210334300994873,
"step": 1180
},
{
"epoch": 0.13,
"learning_rate": 9.587608206722265e-06,
"logits/chosen": -0.6839829087257385,
"logits/rejected": 0.4797239303588867,
"logps/chosen": -209.3610076904297,
"logps/rejected": -152.60250854492188,
"loss": 0.0246,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.890170097351074,
"rewards/margins": 10.441574096679688,
"rewards/rejected": -2.5514039993286133,
"step": 1190
},
{
"epoch": 0.13,
"learning_rate": 9.580723823614229e-06,
"logits/chosen": -0.32581615447998047,
"logits/rejected": 0.7886501550674438,
"logps/chosen": -187.45545959472656,
"logps/rejected": -161.00302124023438,
"loss": 0.0698,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.390974521636963,
"rewards/margins": 10.586621284484863,
"rewards/rejected": -3.1956467628479004,
"step": 1200
},
{
"epoch": 0.13,
"learning_rate": 9.573784965577608e-06,
"logits/chosen": -0.4175148606300354,
"logits/rejected": 0.8941513299942017,
"logps/chosen": -184.9542999267578,
"logps/rejected": -170.08277893066406,
"loss": 0.0517,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 6.477563381195068,
"rewards/margins": 9.664083480834961,
"rewards/rejected": -3.186519145965576,
"step": 1210
},
{
"epoch": 0.13,
"learning_rate": 9.56679171513076e-06,
"logits/chosen": -0.4166285991668701,
"logits/rejected": 0.7068973779678345,
"logps/chosen": -211.63351440429688,
"logps/rejected": -157.5101776123047,
"loss": 0.0588,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.07889461517334,
"rewards/margins": 10.903955459594727,
"rewards/rejected": -2.825061321258545,
"step": 1220
},
{
"epoch": 0.14,
"learning_rate": 9.55974415543888e-06,
"logits/chosen": -0.5035102963447571,
"logits/rejected": 0.7554497718811035,
"logps/chosen": -246.5172576904297,
"logps/rejected": -167.3823699951172,
"loss": 0.0217,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.890531539916992,
"rewards/margins": 12.393576622009277,
"rewards/rejected": -3.5030460357666016,
"step": 1230
},
{
"epoch": 0.14,
"learning_rate": 9.552642370313032e-06,
"logits/chosen": -0.4481489658355713,
"logits/rejected": 0.6294432282447815,
"logps/chosen": -213.8181915283203,
"logps/rejected": -153.36618041992188,
"loss": 0.0668,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.723450660705566,
"rewards/margins": 12.148078918457031,
"rewards/rejected": -3.4246277809143066,
"step": 1240
},
{
"epoch": 0.14,
"learning_rate": 9.545486444209133e-06,
"logits/chosen": -0.3546793758869171,
"logits/rejected": 0.7414132356643677,
"logps/chosen": -146.98406982421875,
"logps/rejected": -143.73516845703125,
"loss": 0.134,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.439810752868652,
"rewards/margins": 9.926045417785645,
"rewards/rejected": -2.4862332344055176,
"step": 1250
},
{
"epoch": 0.14,
"learning_rate": 9.538276462226956e-06,
"logits/chosen": -0.34465163946151733,
"logits/rejected": 0.9261838793754578,
"logps/chosen": -185.11166381835938,
"logps/rejected": -161.7854766845703,
"loss": 0.0201,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.154252529144287,
"rewards/margins": 10.484378814697266,
"rewards/rejected": -3.3301258087158203,
"step": 1260
},
{
"epoch": 0.14,
"learning_rate": 9.531012510109121e-06,
"logits/chosen": -0.5041292905807495,
"logits/rejected": 0.8341606855392456,
"logps/chosen": -223.11752319335938,
"logps/rejected": -168.34884643554688,
"loss": 0.0152,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.197157382965088,
"rewards/margins": 10.75901985168457,
"rewards/rejected": -3.5618622303009033,
"step": 1270
},
{
"epoch": 0.14,
"learning_rate": 9.52369467424007e-06,
"logits/chosen": -0.5082358121871948,
"logits/rejected": 0.90031898021698,
"logps/chosen": -217.27194213867188,
"logps/rejected": -157.5386199951172,
"loss": 0.1151,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.356767177581787,
"rewards/margins": 10.65047550201416,
"rewards/rejected": -3.293708086013794,
"step": 1280
},
{
"epoch": 0.14,
"learning_rate": 9.516323041645041e-06,
"logits/chosen": -0.4632319509983063,
"logits/rejected": 0.6919183135032654,
"logps/chosen": -251.9365997314453,
"logps/rejected": -159.3051300048828,
"loss": 0.0228,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.674399375915527,
"rewards/margins": 10.893823623657227,
"rewards/rejected": -3.219423294067383,
"step": 1290
},
{
"epoch": 0.14,
"learning_rate": 9.508897699989034e-06,
"logits/chosen": -0.4946981370449066,
"logits/rejected": 0.5167168378829956,
"logps/chosen": -239.37088012695312,
"logps/rejected": -163.68618774414062,
"loss": 0.0437,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.406832695007324,
"rewards/margins": 10.362913131713867,
"rewards/rejected": -2.9560797214508057,
"step": 1300
},
{
"epoch": 0.14,
"learning_rate": 9.501418737575771e-06,
"logits/chosen": -0.3266671895980835,
"logits/rejected": 0.8748236894607544,
"logps/chosen": -215.59164428710938,
"logps/rejected": -151.0523681640625,
"loss": 0.1203,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.814599514007568,
"rewards/margins": 11.216487884521484,
"rewards/rejected": -3.401888608932495,
"step": 1310
},
{
"epoch": 0.14,
"learning_rate": 9.493886243346644e-06,
"logits/chosen": -0.4157380163669586,
"logits/rejected": 1.0467342138290405,
"logps/chosen": -251.1937255859375,
"logps/rejected": -169.79441833496094,
"loss": 0.129,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.097917556762695,
"rewards/margins": 11.96033763885498,
"rewards/rejected": -3.8624203205108643,
"step": 1320
},
{
"epoch": 0.15,
"learning_rate": 9.486300306879652e-06,
"logits/chosen": -0.46027225255966187,
"logits/rejected": 0.9197125434875488,
"logps/chosen": -213.0179901123047,
"logps/rejected": -185.57815551757812,
"loss": 0.1064,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.102902889251709,
"rewards/margins": 10.794425964355469,
"rewards/rejected": -3.6915230751037598,
"step": 1330
},
{
"epoch": 0.15,
"learning_rate": 9.478661018388343e-06,
"logits/chosen": -0.27332133054733276,
"logits/rejected": 1.0553624629974365,
"logps/chosen": -195.7577667236328,
"logps/rejected": -162.10386657714844,
"loss": 0.028,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.459615230560303,
"rewards/margins": 11.042313575744629,
"rewards/rejected": -3.5826964378356934,
"step": 1340
},
{
"epoch": 0.15,
"learning_rate": 9.470968468720737e-06,
"logits/chosen": -0.4711340069770813,
"logits/rejected": 0.6478385925292969,
"logps/chosen": -218.00247192382812,
"logps/rejected": -156.74826049804688,
"loss": 0.036,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.571695327758789,
"rewards/margins": 11.931859016418457,
"rewards/rejected": -3.360163450241089,
"step": 1350
},
{
"epoch": 0.15,
"learning_rate": 9.46322274935825e-06,
"logits/chosen": -0.8137804269790649,
"logits/rejected": 0.6631590127944946,
"logps/chosen": -238.30917358398438,
"logps/rejected": -160.2780303955078,
"loss": 0.0763,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.956851959228516,
"rewards/margins": 12.471067428588867,
"rewards/rejected": -3.514216184616089,
"step": 1360
},
{
"epoch": 0.15,
"learning_rate": 9.455423952414602e-06,
"logits/chosen": -0.45048221945762634,
"logits/rejected": 0.9729928970336914,
"logps/chosen": -211.9479217529297,
"logps/rejected": -163.4690704345703,
"loss": 0.006,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.597060203552246,
"rewards/margins": 12.271041870117188,
"rewards/rejected": -3.673980236053467,
"step": 1370
},
{
"epoch": 0.15,
"learning_rate": 9.44757217063472e-06,
"logits/chosen": -0.6646102666854858,
"logits/rejected": 0.6734617352485657,
"logps/chosen": -186.35360717773438,
"logps/rejected": -152.72317504882812,
"loss": 0.0149,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.054737091064453,
"rewards/margins": 11.063182830810547,
"rewards/rejected": -3.0084469318389893,
"step": 1380
},
{
"epoch": 0.15,
"learning_rate": 9.439667497393644e-06,
"logits/chosen": -0.509512722492218,
"logits/rejected": 0.4323197901248932,
"logps/chosen": -206.06008911132812,
"logps/rejected": -156.47021484375,
"loss": 0.0029,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.382911682128906,
"rewards/margins": 11.36556339263916,
"rewards/rejected": -2.9826526641845703,
"step": 1390
},
{
"epoch": 0.15,
"learning_rate": 9.4317100266954e-06,
"logits/chosen": -0.2360249012708664,
"logits/rejected": 0.7698749303817749,
"logps/chosen": -193.63430786132812,
"logps/rejected": -154.2914276123047,
"loss": 0.3086,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 7.249920845031738,
"rewards/margins": 9.66682243347168,
"rewards/rejected": -2.4169018268585205,
"step": 1400
},
{
"epoch": 0.15,
"learning_rate": 9.423699853171903e-06,
"logits/chosen": -0.4919816553592682,
"logits/rejected": 0.6998618841171265,
"logps/chosen": -191.9207305908203,
"logps/rejected": -161.76254272460938,
"loss": 0.0641,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.954064846038818,
"rewards/margins": 11.05593490600586,
"rewards/rejected": -3.10187029838562,
"step": 1410
},
{
"epoch": 0.16,
"learning_rate": 9.415637072081817e-06,
"logits/chosen": -0.22287793457508087,
"logits/rejected": 0.6352042555809021,
"logps/chosen": -193.3926239013672,
"logps/rejected": -156.7662811279297,
"loss": 0.0086,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.585391998291016,
"rewards/margins": 11.008899688720703,
"rewards/rejected": -3.4235072135925293,
"step": 1420
},
{
"epoch": 0.16,
"learning_rate": 9.407521779309426e-06,
"logits/chosen": -0.4121321737766266,
"logits/rejected": 0.8394039869308472,
"logps/chosen": -213.10006713867188,
"logps/rejected": -166.130615234375,
"loss": 0.011,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.29853343963623,
"rewards/margins": 11.414801597595215,
"rewards/rejected": -3.116267442703247,
"step": 1430
},
{
"epoch": 0.16,
"learning_rate": 9.39935407136349e-06,
"logits/chosen": -0.41288232803344727,
"logits/rejected": 0.7763572335243225,
"logps/chosen": -232.4348907470703,
"logps/rejected": -164.8203887939453,
"loss": 0.0195,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 9.438920974731445,
"rewards/margins": 12.723814010620117,
"rewards/rejected": -3.28489351272583,
"step": 1440
},
{
"epoch": 0.16,
"learning_rate": 9.39113404537611e-06,
"logits/chosen": -0.16888564825057983,
"logits/rejected": 1.0841331481933594,
"logps/chosen": -151.8446502685547,
"logps/rejected": -165.15887451171875,
"loss": 0.1485,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.191922187805176,
"rewards/margins": 10.471902847290039,
"rewards/rejected": -3.279982089996338,
"step": 1450
},
{
"epoch": 0.16,
"learning_rate": 9.382861799101556e-06,
"logits/chosen": -0.3828301727771759,
"logits/rejected": 0.7569293975830078,
"logps/chosen": -209.0618896484375,
"logps/rejected": -169.80148315429688,
"loss": 0.074,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.271611213684082,
"rewards/margins": 11.68904972076416,
"rewards/rejected": -3.4174396991729736,
"step": 1460
},
{
"epoch": 0.16,
"learning_rate": 9.37453743091512e-06,
"logits/chosen": -0.4373001158237457,
"logits/rejected": 0.7532221674919128,
"logps/chosen": -214.9142303466797,
"logps/rejected": -157.76080322265625,
"loss": 0.1568,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.783075332641602,
"rewards/margins": 11.846728324890137,
"rewards/rejected": -3.063652276992798,
"step": 1470
},
{
"epoch": 0.16,
"learning_rate": 9.36616103981193e-06,
"logits/chosen": -0.7200452089309692,
"logits/rejected": 0.5077813267707825,
"logps/chosen": -231.0227813720703,
"logps/rejected": -163.6404571533203,
"loss": 0.0111,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.481043815612793,
"rewards/margins": 11.727174758911133,
"rewards/rejected": -3.246131420135498,
"step": 1480
},
{
"epoch": 0.16,
"learning_rate": 9.357732725405788e-06,
"logits/chosen": -0.5195794701576233,
"logits/rejected": 0.5995023846626282,
"logps/chosen": -223.84408569335938,
"logps/rejected": -167.75845336914062,
"loss": 0.072,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.739862442016602,
"rewards/margins": 12.258071899414062,
"rewards/rejected": -3.5182101726531982,
"step": 1490
},
{
"epoch": 0.16,
"learning_rate": 9.349252587927977e-06,
"logits/chosen": -0.5706058740615845,
"logits/rejected": 0.7487145662307739,
"logps/chosen": -206.3643798828125,
"logps/rejected": -160.89007568359375,
"loss": 0.0292,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.057950973510742,
"rewards/margins": 12.710989952087402,
"rewards/rejected": -3.653038501739502,
"step": 1500
},
{
"epoch": 0.17,
"learning_rate": 9.340720728226071e-06,
"logits/chosen": -0.6348392963409424,
"logits/rejected": 0.7459853887557983,
"logps/chosen": -222.1605682373047,
"logps/rejected": -167.15003967285156,
"loss": 0.0187,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.992013931274414,
"rewards/margins": 12.756234169006348,
"rewards/rejected": -3.764221668243408,
"step": 1510
},
{
"epoch": 0.17,
"learning_rate": 9.332137247762735e-06,
"logits/chosen": -0.5036366581916809,
"logits/rejected": 0.43747663497924805,
"logps/chosen": -206.8778533935547,
"logps/rejected": -163.91195678710938,
"loss": 0.0239,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.371187686920166,
"rewards/margins": 10.75023078918457,
"rewards/rejected": -3.379042863845825,
"step": 1520
},
{
"epoch": 0.17,
"learning_rate": 9.32350224861452e-06,
"logits/chosen": -0.7491061687469482,
"logits/rejected": 0.40472355484962463,
"logps/chosen": -196.0803985595703,
"logps/rejected": -162.4312744140625,
"loss": 0.005,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.126627922058105,
"rewards/margins": 12.170022964477539,
"rewards/rejected": -4.043394565582275,
"step": 1530
},
{
"epoch": 0.17,
"learning_rate": 9.314815833470646e-06,
"logits/chosen": -0.8289221525192261,
"logits/rejected": -0.022221926599740982,
"logps/chosen": -184.82785034179688,
"logps/rejected": -164.69155883789062,
"loss": 0.0055,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.109321594238281,
"rewards/margins": 12.10090446472168,
"rewards/rejected": -3.9915835857391357,
"step": 1540
},
{
"epoch": 0.17,
"learning_rate": 9.306078105631783e-06,
"logits/chosen": -0.5130811333656311,
"logits/rejected": 0.783342719078064,
"logps/chosen": -191.61032104492188,
"logps/rejected": -168.2999725341797,
"loss": 0.0233,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.567174911499023,
"rewards/margins": 12.22905445098877,
"rewards/rejected": -3.661879301071167,
"step": 1550
},
{
"epoch": 0.17,
"learning_rate": 9.297289169008826e-06,
"logits/chosen": -0.6227730512619019,
"logits/rejected": 0.2883726954460144,
"logps/chosen": -199.9819793701172,
"logps/rejected": -153.50430297851562,
"loss": 0.0051,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.932660102844238,
"rewards/margins": 11.378183364868164,
"rewards/rejected": -3.445523500442505,
"step": 1560
},
{
"epoch": 0.17,
"learning_rate": 9.288449128121651e-06,
"logits/chosen": -0.26492536067962646,
"logits/rejected": 0.8799971342086792,
"logps/chosen": -183.4518280029297,
"logps/rejected": -151.05831909179688,
"loss": 0.0694,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.4086761474609375,
"rewards/margins": 10.281698226928711,
"rewards/rejected": -2.8730225563049316,
"step": 1570
},
{
"epoch": 0.17,
"learning_rate": 9.27955808809788e-06,
"logits/chosen": -0.551371693611145,
"logits/rejected": 0.7983097434043884,
"logps/chosen": -233.4415283203125,
"logps/rejected": -164.3033905029297,
"loss": 0.106,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.844173431396484,
"rewards/margins": 12.37480354309082,
"rewards/rejected": -3.5306293964385986,
"step": 1580
},
{
"epoch": 0.17,
"learning_rate": 9.270616154671624e-06,
"logits/chosen": -0.7698153257369995,
"logits/rejected": 0.640004575252533,
"logps/chosen": -186.4699249267578,
"logps/rejected": -168.69979858398438,
"loss": 0.0583,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.0685834884643555,
"rewards/margins": 10.25090217590332,
"rewards/rejected": -3.1823184490203857,
"step": 1590
},
{
"epoch": 0.18,
"learning_rate": 9.261623434182232e-06,
"logits/chosen": -0.46172839403152466,
"logits/rejected": 0.6960395574569702,
"logps/chosen": -186.02001953125,
"logps/rejected": -162.45809936523438,
"loss": 0.0538,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.1736321449279785,
"rewards/margins": 10.610064506530762,
"rewards/rejected": -3.436432361602783,
"step": 1600
},
{
"epoch": 0.18,
"learning_rate": 9.252580033573024e-06,
"logits/chosen": -0.18526312708854675,
"logits/rejected": 0.9217432141304016,
"logps/chosen": -166.52264404296875,
"logps/rejected": -156.47720336914062,
"loss": 0.0342,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 6.552803039550781,
"rewards/margins": 10.394461631774902,
"rewards/rejected": -3.841658115386963,
"step": 1610
},
{
"epoch": 0.18,
"learning_rate": 9.243486060390014e-06,
"logits/chosen": -0.7024975419044495,
"logits/rejected": 0.3586685061454773,
"logps/chosen": -261.14373779296875,
"logps/rejected": -156.59616088867188,
"loss": 0.0398,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.696449279785156,
"rewards/margins": 12.929067611694336,
"rewards/rejected": -3.232619047164917,
"step": 1620
},
{
"epoch": 0.18,
"learning_rate": 9.23434162278064e-06,
"logits/chosen": -0.5696882009506226,
"logits/rejected": 0.5719606280326843,
"logps/chosen": -206.2845916748047,
"logps/rejected": -161.68795776367188,
"loss": 0.0043,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.744453430175781,
"rewards/margins": 12.412347793579102,
"rewards/rejected": -3.6678924560546875,
"step": 1630
},
{
"epoch": 0.18,
"learning_rate": 9.225146829492472e-06,
"logits/chosen": -0.4481044411659241,
"logits/rejected": 0.750464677810669,
"logps/chosen": -179.038330078125,
"logps/rejected": -166.00767517089844,
"loss": 0.1427,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.574297904968262,
"rewards/margins": 11.261307716369629,
"rewards/rejected": -3.6870102882385254,
"step": 1640
},
{
"epoch": 0.18,
"learning_rate": 9.215901789871924e-06,
"logits/chosen": -0.9091712236404419,
"logits/rejected": 0.3831535875797272,
"logps/chosen": -206.7153778076172,
"logps/rejected": -156.24366760253906,
"loss": 0.0878,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.617593765258789,
"rewards/margins": 11.938032150268555,
"rewards/rejected": -3.3204398155212402,
"step": 1650
},
{
"epoch": 0.18,
"learning_rate": 9.206606613862941e-06,
"logits/chosen": -0.6661981344223022,
"logits/rejected": 0.7159786224365234,
"logps/chosen": -199.4313507080078,
"logps/rejected": -173.4095458984375,
"loss": 0.0661,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.909134864807129,
"rewards/margins": 12.851089477539062,
"rewards/rejected": -3.941955089569092,
"step": 1660
},
{
"epoch": 0.18,
"learning_rate": 9.197261412005708e-06,
"logits/chosen": -0.48630446195602417,
"logits/rejected": 0.5496179461479187,
"logps/chosen": -222.904541015625,
"logps/rejected": -164.3258514404297,
"loss": 0.0076,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.863286018371582,
"rewards/margins": 12.797664642333984,
"rewards/rejected": -3.9343795776367188,
"step": 1670
},
{
"epoch": 0.18,
"learning_rate": 9.187866295435328e-06,
"logits/chosen": -0.7716933488845825,
"logits/rejected": 0.49292677640914917,
"logps/chosen": -224.15087890625,
"logps/rejected": -160.72998046875,
"loss": 0.0166,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.712728500366211,
"rewards/margins": 12.726627349853516,
"rewards/rejected": -4.013897895812988,
"step": 1680
},
{
"epoch": 0.19,
"learning_rate": 9.178421375880495e-06,
"logits/chosen": -0.985678493976593,
"logits/rejected": -0.16567879915237427,
"logps/chosen": -190.69140625,
"logps/rejected": -148.5782928466797,
"loss": 0.0516,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.473517417907715,
"rewards/margins": 11.43741226196289,
"rewards/rejected": -2.963894844055176,
"step": 1690
},
{
"epoch": 0.19,
"learning_rate": 9.168926765662176e-06,
"logits/chosen": -0.8118952512741089,
"logits/rejected": 0.2844352722167969,
"logps/chosen": -200.7667999267578,
"logps/rejected": -161.06378173828125,
"loss": 0.0476,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.1055908203125,
"rewards/margins": 12.5818510055542,
"rewards/rejected": -3.476259231567383,
"step": 1700
},
{
"epoch": 0.19,
"learning_rate": 9.159382577692264e-06,
"logits/chosen": -0.827900767326355,
"logits/rejected": 0.0768413171172142,
"logps/chosen": -229.66104125976562,
"logps/rejected": -162.81088256835938,
"loss": 0.0584,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.187432289123535,
"rewards/margins": 12.042821884155273,
"rewards/rejected": -3.855389356613159,
"step": 1710
},
{
"epoch": 0.19,
"learning_rate": 9.149788925472248e-06,
"logits/chosen": -0.6998910903930664,
"logits/rejected": 0.26318255066871643,
"logps/chosen": -178.77804565429688,
"logps/rejected": -171.1815185546875,
"loss": 0.0809,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.187336921691895,
"rewards/margins": 11.490485191345215,
"rewards/rejected": -3.303149461746216,
"step": 1720
},
{
"epoch": 0.19,
"learning_rate": 9.14014592309185e-06,
"logits/chosen": -0.831656277179718,
"logits/rejected": -0.06519290804862976,
"logps/chosen": -185.37979125976562,
"logps/rejected": -156.30807495117188,
"loss": 0.0415,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.396572113037109,
"rewards/margins": 10.937728881835938,
"rewards/rejected": -3.5411572456359863,
"step": 1730
},
{
"epoch": 0.19,
"learning_rate": 9.130453685227679e-06,
"logits/chosen": -0.8854426145553589,
"logits/rejected": 0.19448325037956238,
"logps/chosen": -204.03329467773438,
"logps/rejected": -172.7010040283203,
"loss": 0.3383,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.299615859985352,
"rewards/margins": 11.909491539001465,
"rewards/rejected": -3.609875202178955,
"step": 1740
},
{
"epoch": 0.19,
"learning_rate": 9.120712327141858e-06,
"logits/chosen": -0.7734172344207764,
"logits/rejected": 0.3606404662132263,
"logps/chosen": -226.90707397460938,
"logps/rejected": -171.28457641601562,
"loss": 0.0259,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.864892959594727,
"rewards/margins": 13.222407341003418,
"rewards/rejected": -4.35751485824585,
"step": 1750
},
{
"epoch": 0.19,
"learning_rate": 9.110921964680665e-06,
"logits/chosen": -0.752662181854248,
"logits/rejected": 0.22373339533805847,
"logps/chosen": -205.64590454101562,
"logps/rejected": -160.78933715820312,
"loss": 0.1175,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.075421333312988,
"rewards/margins": 11.642030715942383,
"rewards/rejected": -3.566610336303711,
"step": 1760
},
{
"epoch": 0.19,
"learning_rate": 9.10108271427314e-06,
"logits/chosen": -0.9537338018417358,
"logits/rejected": 0.396420419216156,
"logps/chosen": -221.5945587158203,
"logps/rejected": -171.73233032226562,
"loss": 0.039,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.998331069946289,
"rewards/margins": 12.500982284545898,
"rewards/rejected": -3.502650737762451,
"step": 1770
},
{
"epoch": 0.2,
"learning_rate": 9.091194692929716e-06,
"logits/chosen": -0.48165398836135864,
"logits/rejected": 0.6215115785598755,
"logps/chosen": -179.63136291503906,
"logps/rejected": -170.05300903320312,
"loss": 0.0076,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.977053642272949,
"rewards/margins": 11.927708625793457,
"rewards/rejected": -3.9506537914276123,
"step": 1780
},
{
"epoch": 0.2,
"learning_rate": 9.081258018240815e-06,
"logits/chosen": -0.5007386207580566,
"logits/rejected": 0.5413516759872437,
"logps/chosen": -210.7262725830078,
"logps/rejected": -165.52427673339844,
"loss": 0.0723,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.764327049255371,
"rewards/margins": 12.426679611206055,
"rewards/rejected": -3.662353038787842,
"step": 1790
},
{
"epoch": 0.2,
"learning_rate": 9.071272808375457e-06,
"logits/chosen": -0.5356274843215942,
"logits/rejected": 0.5369559526443481,
"logps/chosen": -232.5734405517578,
"logps/rejected": -174.54151916503906,
"loss": 0.0791,
"rewards/accuracies": 0.9375,
"rewards/chosen": 7.314566612243652,
"rewards/margins": 11.709010124206543,
"rewards/rejected": -4.394443511962891,
"step": 1800
},
{
"epoch": 0.2,
"learning_rate": 9.061239182079851e-06,
"logits/chosen": -0.4780815541744232,
"logits/rejected": 0.8617140054702759,
"logps/chosen": -262.71435546875,
"logps/rejected": -168.20004272460938,
"loss": 0.0126,
"rewards/accuracies": 1.0,
"rewards/chosen": 9.593928337097168,
"rewards/margins": 13.690221786499023,
"rewards/rejected": -4.096295356750488,
"step": 1810
},
{
"epoch": 0.2,
"learning_rate": 9.05115725867599e-06,
"logits/chosen": -0.3635478913784027,
"logits/rejected": 0.9611604809761047,
"logps/chosen": -232.6603546142578,
"logps/rejected": -163.62411499023438,
"loss": 0.1665,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.188954830169678,
"rewards/margins": 11.151288986206055,
"rewards/rejected": -3.962334156036377,
"step": 1820
},
{
"epoch": 0.2,
"learning_rate": 9.041027158060216e-06,
"logits/chosen": -0.3142985701560974,
"logits/rejected": 0.846194863319397,
"logps/chosen": -200.39620971679688,
"logps/rejected": -165.14682006835938,
"loss": 0.0169,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.522047519683838,
"rewards/margins": 11.614678382873535,
"rewards/rejected": -4.092629432678223,
"step": 1830
},
{
"epoch": 0.2,
"learning_rate": 9.030849000701815e-06,
"logits/chosen": -0.30892837047576904,
"logits/rejected": 0.7143265604972839,
"logps/chosen": -210.937255859375,
"logps/rejected": -162.75869750976562,
"loss": 0.0715,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.18255090713501,
"rewards/margins": 11.248716354370117,
"rewards/rejected": -4.066165447235107,
"step": 1840
},
{
"epoch": 0.2,
"learning_rate": 9.020622907641568e-06,
"logits/chosen": -0.8121980428695679,
"logits/rejected": 0.7805534601211548,
"logps/chosen": -219.88967895507812,
"logps/rejected": -177.97035217285156,
"loss": 0.0041,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.750587463378906,
"rewards/margins": 12.908777236938477,
"rewards/rejected": -4.15818977355957,
"step": 1850
},
{
"epoch": 0.2,
"learning_rate": 9.01034900049032e-06,
"logits/chosen": -0.7348997592926025,
"logits/rejected": 0.8176345825195312,
"logps/chosen": -229.6995391845703,
"logps/rejected": -180.89663696289062,
"loss": 0.069,
"rewards/accuracies": 0.9375,
"rewards/chosen": 8.388083457946777,
"rewards/margins": 12.502962112426758,
"rewards/rejected": -4.1148786544799805,
"step": 1860
},
{
"epoch": 0.21,
"learning_rate": 9.000027401427524e-06,
"logits/chosen": -0.28413471579551697,
"logits/rejected": 1.1145570278167725,
"logps/chosen": -216.53085327148438,
"logps/rejected": -160.30909729003906,
"loss": 0.0177,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.6873016357421875,
"rewards/margins": 12.045165061950684,
"rewards/rejected": -4.357863426208496,
"step": 1870
},
{
"epoch": 0.21,
"learning_rate": 8.98965823319981e-06,
"logits/chosen": -0.5494770407676697,
"logits/rejected": 0.7957988977432251,
"logps/chosen": -195.81954956054688,
"logps/rejected": -172.00186157226562,
"loss": 0.0415,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.2244873046875,
"rewards/margins": 11.97068977355957,
"rewards/rejected": -4.746201992034912,
"step": 1880
},
{
"epoch": 0.21,
"learning_rate": 8.979241619119496e-06,
"logits/chosen": -0.652524471282959,
"logits/rejected": 0.37755849957466125,
"logps/chosen": -210.0800323486328,
"logps/rejected": -173.26296997070312,
"loss": 0.0172,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.477156639099121,
"rewards/margins": 12.708356857299805,
"rewards/rejected": -4.231201648712158,
"step": 1890
},
{
"epoch": 0.21,
"learning_rate": 8.968777683063142e-06,
"logits/chosen": -0.9727234840393066,
"logits/rejected": 0.3874947428703308,
"logps/chosen": -241.3202362060547,
"logps/rejected": -169.428955078125,
"loss": 0.0129,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.604389190673828,
"rewards/margins": 13.190521240234375,
"rewards/rejected": -4.586132526397705,
"step": 1900
},
{
"epoch": 0.21,
"learning_rate": 8.958266549470072e-06,
"logits/chosen": -0.6917303800582886,
"logits/rejected": 0.503921627998352,
"logps/chosen": -199.63735961914062,
"logps/rejected": -166.35006713867188,
"loss": 0.0064,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.385486602783203,
"rewards/margins": 12.161643981933594,
"rewards/rejected": -3.7761566638946533,
"step": 1910
},
{
"epoch": 0.21,
"learning_rate": 8.947708343340891e-06,
"logits/chosen": -0.4662698209285736,
"logits/rejected": 0.6918913125991821,
"logps/chosen": -176.79347229003906,
"logps/rejected": -166.1067352294922,
"loss": 0.0586,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.096259117126465,
"rewards/margins": 11.056158065795898,
"rewards/rejected": -2.959900140762329,
"step": 1920
},
{
"epoch": 0.21,
"learning_rate": 8.937103190236003e-06,
"logits/chosen": -0.32194486260414124,
"logits/rejected": 0.7147929072380066,
"logps/chosen": -215.06838989257812,
"logps/rejected": -168.64984130859375,
"loss": 0.0501,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.570442199707031,
"rewards/margins": 11.79222583770752,
"rewards/rejected": -4.221782684326172,
"step": 1930
},
{
"epoch": 0.21,
"learning_rate": 8.92645121627411e-06,
"logits/chosen": -0.20297428965568542,
"logits/rejected": 0.9233037233352661,
"logps/chosen": -241.4458770751953,
"logps/rejected": -169.84255981445312,
"loss": 0.0118,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.750960350036621,
"rewards/margins": 12.908734321594238,
"rewards/rejected": -4.157774925231934,
"step": 1940
},
{
"epoch": 0.21,
"learning_rate": 8.915752548130721e-06,
"logits/chosen": -0.6144320964813232,
"logits/rejected": 0.5129267573356628,
"logps/chosen": -235.00973510742188,
"logps/rejected": -155.31793212890625,
"loss": 0.0595,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.357634544372559,
"rewards/margins": 12.143450736999512,
"rewards/rejected": -3.785815477371216,
"step": 1950
},
{
"epoch": 0.22,
"learning_rate": 8.905007313036644e-06,
"logits/chosen": -0.571103572845459,
"logits/rejected": 0.44340604543685913,
"logps/chosen": -252.47384643554688,
"logps/rejected": -152.4981689453125,
"loss": 0.167,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.95770263671875,
"rewards/margins": 12.356184005737305,
"rewards/rejected": -3.3984806537628174,
"step": 1960
},
{
"epoch": 0.22,
"learning_rate": 8.894215638776468e-06,
"logits/chosen": -0.4927874505519867,
"logits/rejected": 0.810386061668396,
"logps/chosen": -181.13320922851562,
"logps/rejected": -157.927978515625,
"loss": 0.0427,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.887845516204834,
"rewards/margins": 11.767987251281738,
"rewards/rejected": -3.8801403045654297,
"step": 1970
},
{
"epoch": 0.22,
"learning_rate": 8.883377653687046e-06,
"logits/chosen": -0.3371856212615967,
"logits/rejected": 0.7843553423881531,
"logps/chosen": -237.75802612304688,
"logps/rejected": -162.60653686523438,
"loss": 0.0209,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.312698364257812,
"rewards/margins": 12.682243347167969,
"rewards/rejected": -4.369544982910156,
"step": 1980
},
{
"epoch": 0.22,
"learning_rate": 8.87249348665597e-06,
"logits/chosen": -0.3437243103981018,
"logits/rejected": 0.6998990178108215,
"logps/chosen": -202.02743530273438,
"logps/rejected": -166.65127563476562,
"loss": 0.0705,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.631216526031494,
"rewards/margins": 11.797039985656738,
"rewards/rejected": -4.165823936462402,
"step": 1990
},
{
"epoch": 0.22,
"learning_rate": 8.861563267120041e-06,
"logits/chosen": -0.4705849587917328,
"logits/rejected": 0.5384072661399841,
"logps/chosen": -192.47850036621094,
"logps/rejected": -172.2180938720703,
"loss": 0.0345,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.673575401306152,
"rewards/margins": 12.357210159301758,
"rewards/rejected": -4.6836347579956055,
"step": 2000
},
{
"epoch": 0.22,
"learning_rate": 8.850587125063717e-06,
"logits/chosen": -0.33058807253837585,
"logits/rejected": 0.971091091632843,
"logps/chosen": -214.54623413085938,
"logps/rejected": -168.44432067871094,
"loss": 0.0881,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.962454319000244,
"rewards/margins": 12.074080467224121,
"rewards/rejected": -4.111626625061035,
"step": 2010
},
{
"epoch": 0.22,
"learning_rate": 8.839565191017587e-06,
"logits/chosen": -0.2974284887313843,
"logits/rejected": 0.97636878490448,
"logps/chosen": -235.8816680908203,
"logps/rejected": -175.85360717773438,
"loss": 0.0187,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.393792152404785,
"rewards/margins": 12.942235946655273,
"rewards/rejected": -4.548444747924805,
"step": 2020
},
{
"epoch": 0.22,
"learning_rate": 8.828497596056797e-06,
"logits/chosen": -0.1643526256084442,
"logits/rejected": 0.8214351534843445,
"logps/chosen": -181.15838623046875,
"logps/rejected": -162.52178955078125,
"loss": 0.1844,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.370779514312744,
"rewards/margins": 11.121458053588867,
"rewards/rejected": -3.7506790161132812,
"step": 2030
},
{
"epoch": 0.22,
"learning_rate": 8.817384471799514e-06,
"logits/chosen": -0.3292265832424164,
"logits/rejected": 1.094087839126587,
"logps/chosen": -208.61074829101562,
"logps/rejected": -167.3018035888672,
"loss": 0.0633,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 7.627663612365723,
"rewards/margins": 12.10447883605957,
"rewards/rejected": -4.476813316345215,
"step": 2040
},
{
"epoch": 0.23,
"learning_rate": 8.806225950405338e-06,
"logits/chosen": -0.27208006381988525,
"logits/rejected": 0.9757376909255981,
"logps/chosen": -216.1806182861328,
"logps/rejected": -167.65142822265625,
"loss": 0.0108,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.379110813140869,
"rewards/margins": 11.649734497070312,
"rewards/rejected": -4.270624160766602,
"step": 2050
},
{
"epoch": 0.23,
"learning_rate": 8.79502216457375e-06,
"logits/chosen": -0.47350209951400757,
"logits/rejected": 0.7839899659156799,
"logps/chosen": -212.70156860351562,
"logps/rejected": -170.24105834960938,
"loss": 0.0054,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.414898872375488,
"rewards/margins": 12.99805736541748,
"rewards/rejected": -4.583159446716309,
"step": 2060
},
{
"epoch": 0.23,
"learning_rate": 8.783773247542518e-06,
"logits/chosen": -0.690665066242218,
"logits/rejected": 0.4333856701850891,
"logps/chosen": -233.6276397705078,
"logps/rejected": -180.33018493652344,
"loss": 0.0056,
"rewards/accuracies": 1.0,
"rewards/chosen": 9.300657272338867,
"rewards/margins": 13.976263046264648,
"rewards/rejected": -4.675606727600098,
"step": 2070
},
{
"epoch": 0.23,
"learning_rate": 8.772479333086122e-06,
"logits/chosen": -0.5696179270744324,
"logits/rejected": 0.5036565661430359,
"logps/chosen": -208.5790252685547,
"logps/rejected": -160.26483154296875,
"loss": 0.0131,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.516992568969727,
"rewards/margins": 12.505514144897461,
"rewards/rejected": -3.9885215759277344,
"step": 2080
},
{
"epoch": 0.23,
"learning_rate": 8.761140555514167e-06,
"logits/chosen": -0.8938825726509094,
"logits/rejected": 0.0002821937086991966,
"logps/chosen": -224.6837158203125,
"logps/rejected": -163.0015869140625,
"loss": 0.0345,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 9.163957595825195,
"rewards/margins": 13.306193351745605,
"rewards/rejected": -4.14223575592041,
"step": 2090
},
{
"epoch": 0.23,
"learning_rate": 8.74975704966977e-06,
"logits/chosen": -0.7335391640663147,
"logits/rejected": 0.13731873035430908,
"logps/chosen": -202.0205078125,
"logps/rejected": -184.16671752929688,
"loss": 0.0502,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.3814058303833,
"rewards/margins": 12.815424919128418,
"rewards/rejected": -4.434020042419434,
"step": 2100
},
{
"epoch": 0.23,
"learning_rate": 8.738328950927971e-06,
"logits/chosen": -0.6339744329452515,
"logits/rejected": 0.1389215737581253,
"logps/chosen": -214.9847869873047,
"logps/rejected": -174.8680877685547,
"loss": 0.0095,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.587041854858398,
"rewards/margins": 12.97746467590332,
"rewards/rejected": -4.390421390533447,
"step": 2110
},
{
"epoch": 0.23,
"learning_rate": 8.726856395194119e-06,
"logits/chosen": -0.8502337336540222,
"logits/rejected": 0.08457791805267334,
"logps/chosen": -240.1321563720703,
"logps/rejected": -180.6337432861328,
"loss": 0.1664,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.162525177001953,
"rewards/margins": 12.321939468383789,
"rewards/rejected": -4.159415245056152,
"step": 2120
},
{
"epoch": 0.23,
"learning_rate": 8.715339518902253e-06,
"logits/chosen": -0.9064286947250366,
"logits/rejected": -0.1819077879190445,
"logps/chosen": -198.82833862304688,
"logps/rejected": -167.35934448242188,
"loss": 0.093,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.25260066986084,
"rewards/margins": 12.027249336242676,
"rewards/rejected": -3.7746481895446777,
"step": 2130
},
{
"epoch": 0.23,
"learning_rate": 8.703778459013479e-06,
"logits/chosen": -0.9572340846061707,
"logits/rejected": -0.146784707903862,
"logps/chosen": -186.52105712890625,
"logps/rejected": -160.72584533691406,
"loss": 0.0249,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.873948574066162,
"rewards/margins": 12.232105255126953,
"rewards/rejected": -4.358155727386475,
"step": 2140
},
{
"epoch": 0.24,
"learning_rate": 8.692173353014351e-06,
"logits/chosen": -0.9497688412666321,
"logits/rejected": -0.1295793503522873,
"logps/chosen": -215.9934539794922,
"logps/rejected": -175.27833557128906,
"loss": 0.0175,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.687699317932129,
"rewards/margins": 13.0452299118042,
"rewards/rejected": -4.357530117034912,
"step": 2150
},
{
"epoch": 0.24,
"learning_rate": 8.68052433891522e-06,
"logits/chosen": -0.9528650045394897,
"logits/rejected": -0.11717786639928818,
"logps/chosen": -220.6671905517578,
"logps/rejected": -152.87850952148438,
"loss": 0.1651,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.587167739868164,
"rewards/margins": 11.920230865478516,
"rewards/rejected": -3.3330631256103516,
"step": 2160
},
{
"epoch": 0.24,
"learning_rate": 8.668831555248607e-06,
"logits/chosen": -0.7263249158859253,
"logits/rejected": 0.23681822419166565,
"logps/chosen": -177.57666015625,
"logps/rejected": -185.82980346679688,
"loss": 0.1393,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.666592597961426,
"rewards/margins": 11.784173011779785,
"rewards/rejected": -4.117580890655518,
"step": 2170
},
{
"epoch": 0.24,
"learning_rate": 8.657095141067546e-06,
"logits/chosen": -0.43287286162376404,
"logits/rejected": 0.7004810571670532,
"logps/chosen": -189.4801025390625,
"logps/rejected": -163.88528442382812,
"loss": 0.0881,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.4667816162109375,
"rewards/margins": 11.954621315002441,
"rewards/rejected": -4.487839698791504,
"step": 2180
},
{
"epoch": 0.24,
"learning_rate": 8.645315235943939e-06,
"logits/chosen": -0.768271267414093,
"logits/rejected": 0.12874235212802887,
"logps/chosen": -217.19259643554688,
"logps/rejected": -167.7063751220703,
"loss": 0.0842,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.23132610321045,
"rewards/margins": 12.512828826904297,
"rewards/rejected": -4.2815022468566895,
"step": 2190
},
{
"epoch": 0.24,
"learning_rate": 8.633491979966888e-06,
"logits/chosen": -0.9630998373031616,
"logits/rejected": 0.4160458445549011,
"logps/chosen": -229.47616577148438,
"logps/rejected": -186.0326690673828,
"loss": 0.0697,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.416041374206543,
"rewards/margins": 12.873327255249023,
"rewards/rejected": -4.457285404205322,
"step": 2200
},
{
"epoch": 0.24,
"learning_rate": 8.621625513741033e-06,
"logits/chosen": -0.5632399320602417,
"logits/rejected": 0.3805318772792816,
"logps/chosen": -228.39486694335938,
"logps/rejected": -169.94186401367188,
"loss": 0.1161,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.571549415588379,
"rewards/margins": 12.16765022277832,
"rewards/rejected": -4.596099853515625,
"step": 2210
},
{
"epoch": 0.24,
"learning_rate": 8.60971597838488e-06,
"logits/chosen": -0.6236270666122437,
"logits/rejected": 0.4582180976867676,
"logps/chosen": -238.4077606201172,
"logps/rejected": -174.6767578125,
"loss": 0.006,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.159147262573242,
"rewards/margins": 13.25964641571045,
"rewards/rejected": -5.100498676300049,
"step": 2220
},
{
"epoch": 0.24,
"learning_rate": 8.59776351552912e-06,
"logits/chosen": -0.395003080368042,
"logits/rejected": 0.7398403882980347,
"logps/chosen": -170.48580932617188,
"logps/rejected": -165.9557342529297,
"loss": 0.0223,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.435987949371338,
"rewards/margins": 12.090422630310059,
"rewards/rejected": -4.654435157775879,
"step": 2230
},
{
"epoch": 0.25,
"learning_rate": 8.585768267314949e-06,
"logits/chosen": -0.37745702266693115,
"logits/rejected": 0.6807891130447388,
"logps/chosen": -190.44113159179688,
"logps/rejected": -174.80258178710938,
"loss": 0.0156,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.024377346038818,
"rewards/margins": 11.530973434448242,
"rewards/rejected": -4.506596565246582,
"step": 2240
},
{
"epoch": 0.25,
"learning_rate": 8.573730376392376e-06,
"logits/chosen": -0.5341759920120239,
"logits/rejected": 0.3510711193084717,
"logps/chosen": -202.7803192138672,
"logps/rejected": -181.39561462402344,
"loss": 0.0102,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.461994171142578,
"rewards/margins": 13.099119186401367,
"rewards/rejected": -4.637126922607422,
"step": 2250
},
{
"epoch": 0.25,
"learning_rate": 8.561649985918526e-06,
"logits/chosen": -0.5021628141403198,
"logits/rejected": 0.3783276081085205,
"logps/chosen": -190.25570678710938,
"logps/rejected": -164.2488250732422,
"loss": 0.0259,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.367849826812744,
"rewards/margins": 11.650557518005371,
"rewards/rejected": -4.282708168029785,
"step": 2260
},
{
"epoch": 0.25,
"learning_rate": 8.549527239555933e-06,
"logits/chosen": -0.8880068063735962,
"logits/rejected": 0.04517006874084473,
"logps/chosen": -218.488525390625,
"logps/rejected": -173.66485595703125,
"loss": 0.0597,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.142195701599121,
"rewards/margins": 12.357624053955078,
"rewards/rejected": -4.215428352355957,
"step": 2270
},
{
"epoch": 0.25,
"learning_rate": 8.537362281470842e-06,
"logits/chosen": -0.8509469032287598,
"logits/rejected": 0.16993197798728943,
"logps/chosen": -225.26181030273438,
"logps/rejected": -165.80532836914062,
"loss": 0.0196,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.537097930908203,
"rewards/margins": 12.063167572021484,
"rewards/rejected": -4.526069641113281,
"step": 2280
},
{
"epoch": 0.25,
"learning_rate": 8.525155256331485e-06,
"logits/chosen": -0.6800307631492615,
"logits/rejected": 0.25085148215293884,
"logps/chosen": -221.555419921875,
"logps/rejected": -167.46531677246094,
"loss": 0.0208,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.689160346984863,
"rewards/margins": 11.93071174621582,
"rewards/rejected": -4.24155330657959,
"step": 2290
},
{
"epoch": 0.25,
"learning_rate": 8.512906309306369e-06,
"logits/chosen": -0.4905932545661926,
"logits/rejected": 0.41362684965133667,
"logps/chosen": -190.53590393066406,
"logps/rejected": -179.59628295898438,
"loss": 0.0392,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.085442543029785,
"rewards/margins": 13.140932083129883,
"rewards/rejected": -5.055489540100098,
"step": 2300
},
{
"epoch": 0.25,
"learning_rate": 8.500615586062537e-06,
"logits/chosen": -0.7107552886009216,
"logits/rejected": 0.2486438751220703,
"logps/chosen": -223.39205932617188,
"logps/rejected": -165.79603576660156,
"loss": 0.0127,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.861292839050293,
"rewards/margins": 12.904462814331055,
"rewards/rejected": -5.043169975280762,
"step": 2310
},
{
"epoch": 0.25,
"learning_rate": 8.48828323276385e-06,
"logits/chosen": -0.2899216115474701,
"logits/rejected": 0.7743111252784729,
"logps/chosen": -195.27996826171875,
"logps/rejected": -168.7857666015625,
"loss": 0.046,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.390793800354004,
"rewards/margins": 12.747232437133789,
"rewards/rejected": -4.356438636779785,
"step": 2320
},
{
"epoch": 0.26,
"learning_rate": 8.475909396069243e-06,
"logits/chosen": -0.8413451910018921,
"logits/rejected": 0.5604926347732544,
"logps/chosen": -224.261962890625,
"logps/rejected": -183.05560302734375,
"loss": 0.0574,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.465588569641113,
"rewards/margins": 13.44543743133545,
"rewards/rejected": -4.979849815368652,
"step": 2330
},
{
"epoch": 0.26,
"learning_rate": 8.463494223130972e-06,
"logits/chosen": -0.8377776145935059,
"logits/rejected": 0.26796481013298035,
"logps/chosen": -233.9268035888672,
"logps/rejected": -167.1359405517578,
"loss": 0.0115,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.433236122131348,
"rewards/margins": 12.992330551147461,
"rewards/rejected": -4.559096336364746,
"step": 2340
},
{
"epoch": 0.26,
"learning_rate": 8.451037861592882e-06,
"logits/chosen": -0.531271755695343,
"logits/rejected": 0.5945345163345337,
"logps/chosen": -233.77487182617188,
"logps/rejected": -171.58950805664062,
"loss": 0.0076,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.894737243652344,
"rewards/margins": 12.823348999023438,
"rewards/rejected": -4.9286112785339355,
"step": 2350
},
{
"epoch": 0.26,
"learning_rate": 8.438540459588635e-06,
"logits/chosen": -0.3315597176551819,
"logits/rejected": 0.8130376935005188,
"logps/chosen": -179.4055938720703,
"logps/rejected": -166.93429565429688,
"loss": 0.1071,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 6.321532726287842,
"rewards/margins": 10.406835556030273,
"rewards/rejected": -4.085303783416748,
"step": 2360
},
{
"epoch": 0.26,
"learning_rate": 8.426002165739959e-06,
"logits/chosen": -0.46008366346359253,
"logits/rejected": 0.7065877318382263,
"logps/chosen": -208.07479858398438,
"logps/rejected": -170.52133178710938,
"loss": 0.0107,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.816873073577881,
"rewards/margins": 12.425461769104004,
"rewards/rejected": -4.608588218688965,
"step": 2370
},
{
"epoch": 0.26,
"learning_rate": 8.413423129154873e-06,
"logits/chosen": -0.6410793662071228,
"logits/rejected": 0.3276979923248291,
"logps/chosen": -234.51693725585938,
"logps/rejected": -169.89077758789062,
"loss": 0.0061,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.718478202819824,
"rewards/margins": 13.924470901489258,
"rewards/rejected": -5.205992221832275,
"step": 2380
},
{
"epoch": 0.26,
"learning_rate": 8.40080349942592e-06,
"logits/chosen": -0.5077573657035828,
"logits/rejected": 0.3929711878299713,
"logps/chosen": -192.05772399902344,
"logps/rejected": -169.24960327148438,
"loss": 0.1019,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.895243167877197,
"rewards/margins": 12.026467323303223,
"rewards/rejected": -4.131225109100342,
"step": 2390
},
{
"epoch": 0.26,
"learning_rate": 8.388143426628384e-06,
"logits/chosen": -0.6147021651268005,
"logits/rejected": 0.23026438057422638,
"logps/chosen": -228.5024871826172,
"logps/rejected": -170.653076171875,
"loss": 0.0201,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.201008796691895,
"rewards/margins": 13.116500854492188,
"rewards/rejected": -4.915493488311768,
"step": 2400
},
{
"epoch": 0.26,
"learning_rate": 8.375443061318504e-06,
"logits/chosen": -0.6069657206535339,
"logits/rejected": 0.5055216550827026,
"logps/chosen": -182.94375610351562,
"logps/rejected": -178.98904418945312,
"loss": 0.0367,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.235411167144775,
"rewards/margins": 11.979175567626953,
"rewards/rejected": -4.743765830993652,
"step": 2410
},
{
"epoch": 0.27,
"learning_rate": 8.36270255453169e-06,
"logits/chosen": -0.6311702728271484,
"logits/rejected": 0.5873549580574036,
"logps/chosen": -216.9287567138672,
"logps/rejected": -169.9967803955078,
"loss": 0.0112,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.0543851852417,
"rewards/margins": 13.188531875610352,
"rewards/rejected": -5.1341471672058105,
"step": 2420
},
{
"epoch": 0.27,
"learning_rate": 8.349922057780723e-06,
"logits/chosen": -0.5530081987380981,
"logits/rejected": 0.7694286108016968,
"logps/chosen": -223.22305297851562,
"logps/rejected": -187.7498016357422,
"loss": 0.0015,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.594778060913086,
"rewards/margins": 13.714917182922363,
"rewards/rejected": -5.120139122009277,
"step": 2430
},
{
"epoch": 0.27,
"learning_rate": 8.337101723053949e-06,
"logits/chosen": -0.4461049437522888,
"logits/rejected": 0.6591795682907104,
"logps/chosen": -230.85140991210938,
"logps/rejected": -166.4679718017578,
"loss": 0.0101,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.5826416015625,
"rewards/margins": 13.19091510772705,
"rewards/rejected": -4.608273506164551,
"step": 2440
},
{
"epoch": 0.27,
"learning_rate": 8.324241702813477e-06,
"logits/chosen": -0.4857296049594879,
"logits/rejected": 0.5430890321731567,
"logps/chosen": -187.19033813476562,
"logps/rejected": -172.521728515625,
"loss": 0.0192,
"rewards/accuracies": 1.0,
"rewards/chosen": 6.923466682434082,
"rewards/margins": 11.60053539276123,
"rewards/rejected": -4.677069187164307,
"step": 2450
},
{
"epoch": 0.27,
"learning_rate": 8.311342149993366e-06,
"logits/chosen": -0.9044029116630554,
"logits/rejected": 0.05248773843050003,
"logps/chosen": -217.3992156982422,
"logps/rejected": -167.77630615234375,
"loss": 0.0012,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.785954475402832,
"rewards/margins": 13.33287525177002,
"rewards/rejected": -4.5469207763671875,
"step": 2460
},
{
"epoch": 0.27,
"learning_rate": 8.298403217997805e-06,
"logits/chosen": -1.2175676822662354,
"logits/rejected": -0.6908454895019531,
"logps/chosen": -269.8515930175781,
"logps/rejected": -162.03921508789062,
"loss": 0.0145,
"rewards/accuracies": 1.0,
"rewards/chosen": 9.291863441467285,
"rewards/margins": 13.004359245300293,
"rewards/rejected": -3.7124953269958496,
"step": 2470
},
{
"epoch": 0.27,
"learning_rate": 8.285425060699291e-06,
"logits/chosen": -1.0893975496292114,
"logits/rejected": -0.1912747323513031,
"logps/chosen": -194.61566162109375,
"logps/rejected": -182.98367309570312,
"loss": 0.0215,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.65707778930664,
"rewards/margins": 12.879467964172363,
"rewards/rejected": -4.222390174865723,
"step": 2480
},
{
"epoch": 0.27,
"learning_rate": 8.272407832436784e-06,
"logits/chosen": -0.6341437697410583,
"logits/rejected": 0.3419879674911499,
"logps/chosen": -171.78280639648438,
"logps/rejected": -165.11734008789062,
"loss": 0.019,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.791851043701172,
"rewards/margins": 12.213276863098145,
"rewards/rejected": -4.421425819396973,
"step": 2490
},
{
"epoch": 0.27,
"learning_rate": 8.2593516880139e-06,
"logits/chosen": -0.5838027000427246,
"logits/rejected": 0.4249269962310791,
"logps/chosen": -202.65524291992188,
"logps/rejected": -171.35055541992188,
"loss": 0.1298,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.352968215942383,
"rewards/margins": 12.428132057189941,
"rewards/rejected": -4.0751633644104,
"step": 2500
},
{
"epoch": 0.28,
"learning_rate": 8.246256782697045e-06,
"logits/chosen": -0.7317477464675903,
"logits/rejected": 0.41557279229164124,
"logps/chosen": -209.4647674560547,
"logps/rejected": -174.3729248046875,
"loss": 0.0038,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.367586135864258,
"rewards/margins": 12.99022102355957,
"rewards/rejected": -4.622633934020996,
"step": 2510
},
{
"epoch": 0.28,
"learning_rate": 8.23312327221358e-06,
"logits/chosen": -0.692236065864563,
"logits/rejected": 0.46434178948402405,
"logps/chosen": -251.1358184814453,
"logps/rejected": -174.54110717773438,
"loss": 0.0319,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.932543754577637,
"rewards/margins": 13.945353507995605,
"rewards/rejected": -5.012808799743652,
"step": 2520
},
{
"epoch": 0.28,
"learning_rate": 8.219951312749967e-06,
"logits/chosen": -0.6690990328788757,
"logits/rejected": 0.569442868232727,
"logps/chosen": -217.0177001953125,
"logps/rejected": -177.09072875976562,
"loss": 0.032,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.575159072875977,
"rewards/margins": 13.195083618164062,
"rewards/rejected": -4.619925022125244,
"step": 2530
},
{
"epoch": 0.28,
"learning_rate": 8.20674106094991e-06,
"logits/chosen": -0.4990970492362976,
"logits/rejected": 0.4127965569496155,
"logps/chosen": -195.69264221191406,
"logps/rejected": -175.87966918945312,
"loss": 0.0311,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.397913455963135,
"rewards/margins": 11.768786430358887,
"rewards/rejected": -4.370872974395752,
"step": 2540
},
{
"epoch": 0.28,
"learning_rate": 8.193492673912495e-06,
"logits/chosen": -0.5780926942825317,
"logits/rejected": 0.48505130410194397,
"logps/chosen": -183.80203247070312,
"logps/rejected": -175.343994140625,
"loss": 0.1525,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.53159236907959,
"rewards/margins": 13.045809745788574,
"rewards/rejected": -4.514214515686035,
"step": 2550
},
{
"epoch": 0.28,
"learning_rate": 8.18020630919032e-06,
"logits/chosen": -0.5260237455368042,
"logits/rejected": 0.5624647736549377,
"logps/chosen": -175.13485717773438,
"logps/rejected": -170.07009887695312,
"loss": 0.008,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.5792131423950195,
"rewards/margins": 11.603866577148438,
"rewards/rejected": -4.024654388427734,
"step": 2560
},
{
"epoch": 0.28,
"learning_rate": 8.166882124787623e-06,
"logits/chosen": -0.47336331009864807,
"logits/rejected": 0.6014837026596069,
"logps/chosen": -202.11178588867188,
"logps/rejected": -182.88864135742188,
"loss": 0.0248,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.935551643371582,
"rewards/margins": 12.890029907226562,
"rewards/rejected": -4.954479217529297,
"step": 2570
},
{
"epoch": 0.28,
"learning_rate": 8.153520279158396e-06,
"logits/chosen": -0.34318774938583374,
"logits/rejected": 0.6932908892631531,
"logps/chosen": -174.56124877929688,
"logps/rejected": -174.02755737304688,
"loss": 0.0931,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 6.897540092468262,
"rewards/margins": 11.409690856933594,
"rewards/rejected": -4.512150764465332,
"step": 2580
},
{
"epoch": 0.28,
"learning_rate": 8.140120931204512e-06,
"logits/chosen": -0.6497650742530823,
"logits/rejected": 0.5149052739143372,
"logps/chosen": -243.260009765625,
"logps/rejected": -170.65261840820312,
"loss": 0.0969,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.357889175415039,
"rewards/margins": 12.79127311706543,
"rewards/rejected": -4.433384418487549,
"step": 2590
},
{
"epoch": 0.29,
"learning_rate": 8.126684240273827e-06,
"logits/chosen": -0.6540253162384033,
"logits/rejected": 0.3523900508880615,
"logps/chosen": -200.621337890625,
"logps/rejected": -171.65789794921875,
"loss": 0.0253,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.952597618103027,
"rewards/margins": 12.75285530090332,
"rewards/rejected": -4.800256729125977,
"step": 2600
},
{
"epoch": 0.29,
"learning_rate": 8.113210366158289e-06,
"logits/chosen": -0.9135942459106445,
"logits/rejected": -0.017345678061246872,
"logps/chosen": -194.4808807373047,
"logps/rejected": -160.66390991210938,
"loss": 0.1005,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.563905239105225,
"rewards/margins": 11.910043716430664,
"rewards/rejected": -4.346137523651123,
"step": 2610
},
{
"epoch": 0.29,
"learning_rate": 8.099699469092035e-06,
"logits/chosen": -1.0905654430389404,
"logits/rejected": -0.7532132863998413,
"logps/chosen": -199.40904235839844,
"logps/rejected": -162.73121643066406,
"loss": 0.0104,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.886347770690918,
"rewards/margins": 13.453600883483887,
"rewards/rejected": -4.567252159118652,
"step": 2620
},
{
"epoch": 0.29,
"learning_rate": 8.086151709749488e-06,
"logits/chosen": -0.793353796005249,
"logits/rejected": 0.19234681129455566,
"logps/chosen": -172.5845489501953,
"logps/rejected": -171.7484893798828,
"loss": 0.0845,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.292520523071289,
"rewards/margins": 12.046894073486328,
"rewards/rejected": -3.7543742656707764,
"step": 2630
},
{
"epoch": 0.29,
"learning_rate": 8.072567249243444e-06,
"logits/chosen": -1.1524503231048584,
"logits/rejected": -0.16285154223442078,
"logps/chosen": -229.02871704101562,
"logps/rejected": -168.34915161132812,
"loss": 0.099,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 9.448681831359863,
"rewards/margins": 13.124113082885742,
"rewards/rejected": -3.6754310131073,
"step": 2640
},
{
"epoch": 0.29,
"learning_rate": 8.058946249123155e-06,
"logits/chosen": -1.0330545902252197,
"logits/rejected": -0.21650974452495575,
"logps/chosen": -231.02783203125,
"logps/rejected": -186.7714385986328,
"loss": 0.0003,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.94650650024414,
"rewards/margins": 14.374897956848145,
"rewards/rejected": -5.428391456604004,
"step": 2650
},
{
"epoch": 0.29,
"learning_rate": 8.04528887137241e-06,
"logits/chosen": -0.8955181241035461,
"logits/rejected": -0.10996793210506439,
"logps/chosen": -197.42031860351562,
"logps/rejected": -171.0691680908203,
"loss": 0.0405,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.19475269317627,
"rewards/margins": 12.796459197998047,
"rewards/rejected": -4.601706027984619,
"step": 2660
},
{
"epoch": 0.29,
"learning_rate": 8.031595278407613e-06,
"logits/chosen": -1.051721215248108,
"logits/rejected": -0.3572627604007721,
"logps/chosen": -206.19784545898438,
"logps/rejected": -167.8656005859375,
"loss": 0.0466,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 9.205878257751465,
"rewards/margins": 13.923624992370605,
"rewards/rejected": -4.717745780944824,
"step": 2670
},
{
"epoch": 0.29,
"learning_rate": 8.017865633075846e-06,
"logits/chosen": -0.9133942723274231,
"logits/rejected": -0.07959525287151337,
"logps/chosen": -202.22283935546875,
"logps/rejected": -167.33413696289062,
"loss": 0.0702,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.584909439086914,
"rewards/margins": 12.664779663085938,
"rewards/rejected": -4.079871654510498,
"step": 2680
},
{
"epoch": 0.3,
"learning_rate": 8.004100098652926e-06,
"logits/chosen": -0.7203928232192993,
"logits/rejected": -0.05325474217534065,
"logps/chosen": -182.52957153320312,
"logps/rejected": -163.93081665039062,
"loss": 0.1392,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 8.526405334472656,
"rewards/margins": 12.884794235229492,
"rewards/rejected": -4.358389854431152,
"step": 2690
},
{
"epoch": 0.3,
"learning_rate": 7.990298838841481e-06,
"logits/chosen": -0.976810097694397,
"logits/rejected": -0.26609936356544495,
"logps/chosen": -211.3890838623047,
"logps/rejected": -176.94076538085938,
"loss": 0.0172,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.936230659484863,
"rewards/margins": 12.897746086120605,
"rewards/rejected": -3.961515426635742,
"step": 2700
},
{
"epoch": 0.3,
"learning_rate": 7.976462017768984e-06,
"logits/chosen": -0.7487978935241699,
"logits/rejected": 0.2039092481136322,
"logps/chosen": -196.99449157714844,
"logps/rejected": -177.93943786621094,
"loss": 0.0199,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.814328193664551,
"rewards/margins": 12.83709716796875,
"rewards/rejected": -5.022768497467041,
"step": 2710
},
{
"epoch": 0.3,
"learning_rate": 7.962589799985817e-06,
"logits/chosen": -0.9832037091255188,
"logits/rejected": 0.010686754249036312,
"logps/chosen": -230.860595703125,
"logps/rejected": -179.5846710205078,
"loss": 0.1128,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.199392318725586,
"rewards/margins": 12.86755657196045,
"rewards/rejected": -4.6681647300720215,
"step": 2720
},
{
"epoch": 0.3,
"learning_rate": 7.948682350463299e-06,
"logits/chosen": -0.6001917719841003,
"logits/rejected": 0.41285085678100586,
"logps/chosen": -200.84994506835938,
"logps/rejected": -172.69180297851562,
"loss": 0.0687,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.3350114822387695,
"rewards/margins": 12.243888854980469,
"rewards/rejected": -4.908877372741699,
"step": 2730
},
{
"epoch": 0.3,
"learning_rate": 7.934739834591743e-06,
"logits/chosen": -0.7725926637649536,
"logits/rejected": 0.5752295255661011,
"logps/chosen": -219.83322143554688,
"logps/rejected": -185.51937866210938,
"loss": 0.0206,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.2027106285095215,
"rewards/margins": 12.59252643585205,
"rewards/rejected": -5.389814853668213,
"step": 2740
},
{
"epoch": 0.3,
"learning_rate": 7.920762418178467e-06,
"logits/chosen": -0.7085618376731873,
"logits/rejected": 0.6622563600540161,
"logps/chosen": -198.0343017578125,
"logps/rejected": -173.04983520507812,
"loss": 0.0028,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.681931495666504,
"rewards/margins": 12.312196731567383,
"rewards/rejected": -4.630265235900879,
"step": 2750
},
{
"epoch": 0.3,
"learning_rate": 7.906750267445842e-06,
"logits/chosen": -0.5810704827308655,
"logits/rejected": 0.5701958537101746,
"logps/chosen": -191.88255310058594,
"logps/rejected": -167.3490447998047,
"loss": 0.1238,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.207613468170166,
"rewards/margins": 12.253547668457031,
"rewards/rejected": -5.045934677124023,
"step": 2760
},
{
"epoch": 0.3,
"learning_rate": 7.892703549029302e-06,
"logits/chosen": -0.618598997592926,
"logits/rejected": 0.6705722212791443,
"logps/chosen": -200.2035675048828,
"logps/rejected": -180.4983673095703,
"loss": 0.0843,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.773738861083984,
"rewards/margins": 13.073773384094238,
"rewards/rejected": -5.3000359535217285,
"step": 2770
},
{
"epoch": 0.31,
"learning_rate": 7.878622429975369e-06,
"logits/chosen": -0.4437601566314697,
"logits/rejected": 0.6392486691474915,
"logps/chosen": -186.42111206054688,
"logps/rejected": -176.2037811279297,
"loss": 0.1384,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 7.519711971282959,
"rewards/margins": 12.133156776428223,
"rewards/rejected": -4.613443374633789,
"step": 2780
},
{
"epoch": 0.31,
"learning_rate": 7.864507077739667e-06,
"logits/chosen": -0.520032525062561,
"logits/rejected": 0.4803471565246582,
"logps/chosen": -231.8350372314453,
"logps/rejected": -176.01327514648438,
"loss": 0.1153,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 9.256172180175781,
"rewards/margins": 14.020452499389648,
"rewards/rejected": -4.764278888702393,
"step": 2790
},
{
"epoch": 0.31,
"learning_rate": 7.850357660184924e-06,
"logits/chosen": -0.6796492338180542,
"logits/rejected": 0.6617184281349182,
"logps/chosen": -227.59487915039062,
"logps/rejected": -177.05235290527344,
"loss": 0.0812,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.946455478668213,
"rewards/margins": 13.443826675415039,
"rewards/rejected": -5.497370719909668,
"step": 2800
},
{
"epoch": 0.31,
"learning_rate": 7.836174345578982e-06,
"logits/chosen": -0.7396942973136902,
"logits/rejected": 0.3257351517677307,
"logps/chosen": -213.6725311279297,
"logps/rejected": -181.46981811523438,
"loss": 0.0114,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.922917366027832,
"rewards/margins": 13.603009223937988,
"rewards/rejected": -4.680091857910156,
"step": 2810
},
{
"epoch": 0.31,
"learning_rate": 7.821957302592791e-06,
"logits/chosen": -0.8451019525527954,
"logits/rejected": 0.3018733561038971,
"logps/chosen": -227.48641967773438,
"logps/rejected": -178.04055786132812,
"loss": 0.0381,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.504293441772461,
"rewards/margins": 13.973077774047852,
"rewards/rejected": -5.468785762786865,
"step": 2820
},
{
"epoch": 0.31,
"learning_rate": 7.807706700298412e-06,
"logits/chosen": -0.7736309766769409,
"logits/rejected": 0.44020017981529236,
"logps/chosen": -212.9801483154297,
"logps/rejected": -175.34637451171875,
"loss": 0.0634,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.439635276794434,
"rewards/margins": 13.418614387512207,
"rewards/rejected": -4.978979110717773,
"step": 2830
},
{
"epoch": 0.31,
"learning_rate": 7.793422708166996e-06,
"logits/chosen": -0.7209655046463013,
"logits/rejected": 0.425056517124176,
"logps/chosen": -216.6347198486328,
"logps/rejected": -179.53964233398438,
"loss": 0.1194,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.9016618728637695,
"rewards/margins": 12.855749130249023,
"rewards/rejected": -4.95408821105957,
"step": 2840
},
{
"epoch": 0.31,
"learning_rate": 7.779105496066772e-06,
"logits/chosen": -0.6034275889396667,
"logits/rejected": 0.4955766797065735,
"logps/chosen": -208.5738067626953,
"logps/rejected": -185.705322265625,
"loss": 0.0025,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.3899245262146,
"rewards/margins": 12.2137451171875,
"rewards/rejected": -4.823821067810059,
"step": 2850
},
{
"epoch": 0.31,
"learning_rate": 7.76475523426103e-06,
"logits/chosen": -0.6930392980575562,
"logits/rejected": 0.28452375531196594,
"logps/chosen": -206.74337768554688,
"logps/rejected": -181.8390350341797,
"loss": 0.03,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.6725358963012695,
"rewards/margins": 12.631759643554688,
"rewards/rejected": -4.959224700927734,
"step": 2860
},
{
"epoch": 0.32,
"learning_rate": 7.750372093406095e-06,
"logits/chosen": -0.6658349633216858,
"logits/rejected": 0.31616711616516113,
"logps/chosen": -224.6175079345703,
"logps/rejected": -184.55096435546875,
"loss": 0.0427,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.468997955322266,
"rewards/margins": 14.208582878112793,
"rewards/rejected": -5.739585876464844,
"step": 2870
},
{
"epoch": 0.32,
"learning_rate": 7.735956244549294e-06,
"logits/chosen": -0.7174617052078247,
"logits/rejected": 0.31869566440582275,
"logps/chosen": -221.75466918945312,
"logps/rejected": -173.40158081054688,
"loss": 0.0961,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.690484046936035,
"rewards/margins": 13.39197063446045,
"rewards/rejected": -4.701488494873047,
"step": 2880
},
{
"epoch": 0.32,
"learning_rate": 7.721507859126927e-06,
"logits/chosen": -0.6412464380264282,
"logits/rejected": 0.23269173502922058,
"logps/chosen": -191.8711700439453,
"logps/rejected": -181.15994262695312,
"loss": 0.0611,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.7382941246032715,
"rewards/margins": 13.182668685913086,
"rewards/rejected": -5.444373607635498,
"step": 2890
},
{
"epoch": 0.32,
"learning_rate": 7.707027108962222e-06,
"logits/chosen": -0.4695025384426117,
"logits/rejected": 0.48480135202407837,
"logps/chosen": -224.146240234375,
"logps/rejected": -182.19253540039062,
"loss": 0.0342,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 8.21843147277832,
"rewards/margins": 13.430513381958008,
"rewards/rejected": -5.212081432342529,
"step": 2900
},
{
"epoch": 0.32,
"learning_rate": 7.692514166263296e-06,
"logits/chosen": -0.731987476348877,
"logits/rejected": 0.23414798080921173,
"logps/chosen": -183.3697509765625,
"logps/rejected": -180.24000549316406,
"loss": 0.0521,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.768662452697754,
"rewards/margins": 12.571802139282227,
"rewards/rejected": -4.80313777923584,
"step": 2910
},
{
"epoch": 0.32,
"learning_rate": 7.677969203621113e-06,
"logits/chosen": -0.8885566592216492,
"logits/rejected": 0.126034677028656,
"logps/chosen": -222.8698272705078,
"logps/rejected": -186.18331909179688,
"loss": 0.0438,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.468908309936523,
"rewards/margins": 13.904070854187012,
"rewards/rejected": -5.435164451599121,
"step": 2920
},
{
"epoch": 0.32,
"learning_rate": 7.663392394007418e-06,
"logits/chosen": -1.0950114727020264,
"logits/rejected": -0.08405710756778717,
"logps/chosen": -244.9022979736328,
"logps/rejected": -170.50209045410156,
"loss": 0.0613,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": 8.325568199157715,
"rewards/margins": 13.444650650024414,
"rewards/rejected": -5.119081497192383,
"step": 2930
},
{
"epoch": 0.32,
"learning_rate": 7.648783910772687e-06,
"logits/chosen": -0.7430115938186646,
"logits/rejected": 0.2351471483707428,
"logps/chosen": -236.56448364257812,
"logps/rejected": -171.637451171875,
"loss": 0.0975,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 8.123806953430176,
"rewards/margins": 13.381278991699219,
"rewards/rejected": -5.257473945617676,
"step": 2940
},
{
"epoch": 0.32,
"learning_rate": 7.634143927644066e-06,
"logits/chosen": -0.6513711214065552,
"logits/rejected": 0.3646014332771301,
"logps/chosen": -179.13702392578125,
"logps/rejected": -180.0673065185547,
"loss": 0.0043,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.1201629638671875,
"rewards/margins": 12.562201499938965,
"rewards/rejected": -5.4420390129089355,
"step": 2950
},
{
"epoch": 0.32,
"learning_rate": 7.6194726187233094e-06,
"logits/chosen": -0.4454960823059082,
"logits/rejected": 0.47287583351135254,
"logps/chosen": -201.38282775878906,
"logps/rejected": -178.86441040039062,
"loss": 0.1987,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 7.348745822906494,
"rewards/margins": 12.077168464660645,
"rewards/rejected": -4.728421211242676,
"step": 2960
},
{
"epoch": 0.33,
"learning_rate": 7.604770158484697e-06,
"logits/chosen": -0.7678300738334656,
"logits/rejected": 0.533244252204895,
"logps/chosen": -206.36178588867188,
"logps/rejected": -189.08424377441406,
"loss": 0.001,
"rewards/accuracies": 1.0,
"rewards/chosen": 8.136065483093262,
"rewards/margins": 13.558685302734375,
"rewards/rejected": -5.4226179122924805,
"step": 2970
},
{
"epoch": 0.33,
"learning_rate": 7.590036721772971e-06,
"logits/chosen": -0.5914806127548218,
"logits/rejected": 0.5361270904541016,
"logps/chosen": -185.54403686523438,
"logps/rejected": -189.84043884277344,
"loss": 0.0353,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": 7.511700630187988,
"rewards/margins": 13.264554023742676,
"rewards/rejected": -5.752852916717529,
"step": 2980
},
{
"epoch": 0.33,
"learning_rate": 7.57527248380125e-06,
"logits/chosen": -0.5882383584976196,
"logits/rejected": 0.5471413135528564,
"logps/chosen": -196.56063842773438,
"logps/rejected": -181.27120971679688,
"loss": 0.0239,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.016335487365723,
"rewards/margins": 12.353779792785645,
"rewards/rejected": -5.3374457359313965,
"step": 2990
},
{
"epoch": 0.33,
"learning_rate": 7.560477620148953e-06,
"logits/chosen": -0.6498584151268005,
"logits/rejected": 0.411933571100235,
"logps/chosen": -203.97779846191406,
"logps/rejected": -184.84317016601562,
"loss": 0.0208,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": 7.5401105880737305,
"rewards/margins": 13.002525329589844,
"rewards/rejected": -5.462414264678955,
"step": 3000
}
],
"logging_steps": 10,
"max_steps": 9110,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 3.7568556896322847e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}