{
  "best_metric": 0.7562676668167114,
  "best_model_checkpoint": "./output/checkpoints/2024-05-27_09-00-27/checkpoint-100",
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 198,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025252525252525252,
      "grad_norm": 31.540218353271484,
      "learning_rate": 6e-06,
      "logits/chosen": 0.14386241137981415,
      "logits/rejected": -0.5877799391746521,
      "logps/chosen": -220.9837646484375,
      "logps/rejected": -171.7880096435547,
      "loss": 0.6932,
      "rewards/accuracies": 0.11249999701976776,
      "rewards/chosen": 0.004091186448931694,
      "rewards/margins": -7.568336877739057e-05,
      "rewards/rejected": 0.004166870377957821,
      "step": 5
    },
    {
      "epoch": 0.050505050505050504,
      "grad_norm": 23.380859375,
      "learning_rate": 1.6000000000000003e-05,
      "logits/chosen": -0.10832454264163971,
      "logits/rejected": -0.4757871627807617,
      "logps/chosen": -196.83120727539062,
      "logps/rejected": -171.42141723632812,
      "loss": 0.6684,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.33129221200942993,
      "rewards/margins": 0.08790162205696106,
      "rewards/rejected": 0.24339056015014648,
      "step": 10
    },
    {
      "epoch": 0.07575757575757576,
      "grad_norm": 23.615068435668945,
      "learning_rate": 2.6000000000000002e-05,
      "logits/chosen": -0.2714936137199402,
      "logits/rejected": -0.6976348161697388,
      "logps/chosen": -209.19827270507812,
      "logps/rejected": -173.0024871826172,
      "loss": 0.6943,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 1.1716283559799194,
      "rewards/margins": 0.2701141834259033,
      "rewards/rejected": 0.9015142321586609,
      "step": 15
    },
    {
      "epoch": 0.10101010101010101,
      "grad_norm": 23.117158889770508,
      "learning_rate": 3.6e-05,
      "logits/chosen": -0.4857279658317566,
      "logits/rejected": -0.9330118298530579,
      "logps/chosen": -202.5274200439453,
      "logps/rejected": -176.1457977294922,
      "loss": 0.6632,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 1.3160903453826904,
      "rewards/margins": 0.552257239818573,
      "rewards/rejected": 0.7638329863548279,
      "step": 20
    },
    {
      "epoch": 0.12626262626262627,
      "grad_norm": 31.966520309448242,
      "learning_rate": 3.997197144003557e-05,
      "logits/chosen": 0.038466982543468475,
      "logits/rejected": -0.503484845161438,
      "logps/chosen": -225.3045196533203,
      "logps/rejected": -201.3688201904297,
      "loss": 0.6399,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.8800640106201172,
      "rewards/margins": 0.41275835037231445,
      "rewards/rejected": 0.46730566024780273,
      "step": 25
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 33.95348358154297,
      "learning_rate": 3.980097021028909e-05,
      "logits/chosen": -0.036555200815200806,
      "logits/rejected": -0.4494614005088806,
      "logps/chosen": -230.3584747314453,
      "logps/rejected": -197.8057098388672,
      "loss": 0.7822,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.15881267189979553,
      "rewards/margins": 0.23805825412273407,
      "rewards/rejected": -0.07924561202526093,
      "step": 30
    },
    {
      "epoch": 0.17676767676767677,
      "grad_norm": 30.72279930114746,
      "learning_rate": 3.947586836927601e-05,
      "logits/chosen": -0.24093489348888397,
      "logits/rejected": -0.8071072697639465,
      "logps/chosen": -206.206787109375,
      "logps/rejected": -173.3004608154297,
      "loss": 0.6694,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.3433918356895447,
      "rewards/margins": 0.5247436761856079,
      "rewards/rejected": -0.18135181069374084,
      "step": 35
    },
    {
      "epoch": 0.20202020202020202,
      "grad_norm": 29.44301986694336,
      "learning_rate": 3.899919601485982e-05,
      "logits/chosen": -0.5006700754165649,
      "logits/rejected": -1.0141886472702026,
      "logps/chosen": -211.9564666748047,
      "logps/rejected": -185.8589630126953,
      "loss": 0.6946,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.9416486620903015,
      "rewards/margins": 0.5160013437271118,
      "rewards/rejected": 0.4256472587585449,
      "step": 40
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 21.33418083190918,
      "learning_rate": 3.837466283906112e-05,
      "logits/chosen": -0.37220650911331177,
      "logits/rejected": -1.1134871244430542,
      "logps/chosen": -203.7546844482422,
      "logps/rejected": -160.15695190429688,
      "loss": 0.7258,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 1.9558101892471313,
      "rewards/margins": 0.6899305582046509,
      "rewards/rejected": 1.2658796310424805,
      "step": 45
    },
    {
      "epoch": 0.25252525252525254,
      "grad_norm": 29.768083572387695,
      "learning_rate": 3.760712925746183e-05,
      "logits/chosen": -0.40102189779281616,
      "logits/rejected": -0.8553081750869751,
      "logps/chosen": -213.06222534179688,
      "logps/rejected": -184.9990997314453,
      "loss": 0.7829,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 2.8542065620422363,
      "rewards/margins": 1.1129443645477295,
      "rewards/rejected": 1.7412618398666382,
      "step": 50
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 21.650400161743164,
      "learning_rate": 3.6702568583128715e-05,
      "logits/chosen": -0.5856252908706665,
      "logits/rejected": -1.2530784606933594,
      "logps/chosen": -204.8536834716797,
      "logps/rejected": -151.60122680664062,
      "loss": 0.6457,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.6728694438934326,
      "rewards/margins": 1.0119760036468506,
      "rewards/rejected": 0.6608934998512268,
      "step": 55
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 25.246517181396484,
      "learning_rate": 3.566802053943705e-05,
      "logits/chosen": -0.56941157579422,
      "logits/rejected": -1.02021062374115,
      "logps/chosen": -209.837158203125,
      "logps/rejected": -172.53732299804688,
      "loss": 0.763,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 1.4966418743133545,
      "rewards/margins": 0.739094078540802,
      "rewards/rejected": 0.7575478553771973,
      "step": 60
    },
    {
      "epoch": 0.3282828282828283,
      "grad_norm": 17.691490173339844,
      "learning_rate": 3.451153647357965e-05,
      "logits/chosen": -0.05042291432619095,
      "logits/rejected": -0.6024894118309021,
      "logps/chosen": -234.263427734375,
      "logps/rejected": -189.74607849121094,
      "loss": 0.7844,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 1.7601999044418335,
      "rewards/margins": 0.520693302154541,
      "rewards/rejected": 1.2395066022872925,
      "step": 65
    },
    {
      "epoch": 0.35353535353535354,
      "grad_norm": 26.0878849029541,
      "learning_rate": 3.3242116697136015e-05,
      "logits/chosen": -0.10399909317493439,
      "logits/rejected": -0.5000227093696594,
      "logps/chosen": -210.8389434814453,
      "logps/rejected": -188.54824829101562,
      "loss": 0.7101,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 1.1393743753433228,
      "rewards/margins": 0.3506864905357361,
      "rewards/rejected": 0.7886878848075867,
      "step": 70
    },
    {
      "epoch": 0.3787878787878788,
      "grad_norm": 24.767913818359375,
      "learning_rate": 3.186964044134713e-05,
      "logits/chosen": -0.6071327328681946,
      "logits/rejected": -0.8622013926506042,
      "logps/chosen": -188.81149291992188,
      "logps/rejected": -173.66650390625,
      "loss": 0.769,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.9471501111984253,
      "rewards/margins": 0.3631848394870758,
      "rewards/rejected": 0.5839653611183167,
      "step": 75
    },
    {
      "epoch": 0.40404040404040403,
      "grad_norm": 27.299219131469727,
      "learning_rate": 3.0404788972217645e-05,
      "logits/chosen": -0.47333812713623047,
      "logits/rejected": -0.855948805809021,
      "logps/chosen": -232.19448852539062,
      "logps/rejected": -185.18539428710938,
      "loss": 0.7493,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 1.0035685300827026,
      "rewards/margins": 0.4641413688659668,
      "rewards/rejected": 0.5394272208213806,
      "step": 80
    },
    {
      "epoch": 0.4292929292929293,
      "grad_norm": 11.157441139221191,
      "learning_rate": 2.8858962463800163e-05,
      "logits/chosen": -0.3463514745235443,
      "logits/rejected": -1.0284178256988525,
      "logps/chosen": -227.76937866210938,
      "logps/rejected": -173.3926239013672,
      "loss": 0.7012,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 1.2513014078140259,
      "rewards/margins": 0.724143922328949,
      "rewards/rejected": 0.5271574854850769,
      "step": 85
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 21.36394500732422,
      "learning_rate": 2.7244191276593653e-05,
      "logits/chosen": -0.396036297082901,
      "logits/rejected": -0.6990352272987366,
      "logps/chosen": -198.6284942626953,
      "logps/rejected": -180.7591094970703,
      "loss": 0.7398,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 1.1550531387329102,
      "rewards/margins": 0.42906150221824646,
      "rewards/rejected": 0.7259916067123413,
      "step": 90
    },
    {
      "epoch": 0.4797979797979798,
      "grad_norm": 21.15500831604004,
      "learning_rate": 2.5573042331529846e-05,
      "logits/chosen": -0.5320285558700562,
      "logits/rejected": -0.9789053797721863,
      "logps/chosen": -216.59359741210938,
      "logps/rejected": -175.47976684570312,
      "loss": 0.7284,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 1.1451869010925293,
      "rewards/margins": 0.508368968963623,
      "rewards/rejected": 0.6368179321289062,
      "step": 95
    },
    {
      "epoch": 0.5050505050505051,
      "grad_norm": 25.441944122314453,
      "learning_rate": 2.385852130818994e-05,
      "logits/chosen": -0.38838592171669006,
      "logits/rejected": -0.7203149795532227,
      "logps/chosen": -212.9621124267578,
      "logps/rejected": -187.50942993164062,
      "loss": 0.7677,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 1.5237963199615479,
      "rewards/margins": 0.5711702108383179,
      "rewards/rejected": 0.95262610912323,
      "step": 100
    },
    {
      "epoch": 0.5050505050505051,
      "eval_logits/chosen": 1.4398491382598877,
      "eval_logits/rejected": 1.2089641094207764,
      "eval_logps/chosen": -209.8059844970703,
      "eval_logps/rejected": -178.4928741455078,
      "eval_loss": 0.7562676668167114,
      "eval_rewards/accuracies": 0.59375,
      "eval_rewards/chosen": 1.1455074548721313,
      "eval_rewards/margins": 0.41021832823753357,
      "eval_rewards/rejected": 0.7352891564369202,
      "eval_runtime": 274.7063,
      "eval_samples_per_second": 2.33,
      "eval_steps_per_second": 0.146,
      "step": 100
    },
    {
      "epoch": 0.5303030303030303,
      "grad_norm": 18.52869415283203,
      "learning_rate": 2.2113971428391862e-05,
      "logits/chosen": -0.5178086757659912,
      "logits/rejected": -1.0226086378097534,
      "logps/chosen": -208.1871795654297,
      "logps/rejected": -171.2388458251953,
      "loss": 0.7824,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.9819921255111694,
      "rewards/margins": 0.44732803106307983,
      "rewards/rejected": 0.5346641540527344,
      "step": 105
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 18.219402313232422,
      "learning_rate": 2.0352969612862576e-05,
      "logits/chosen": -0.5200484395027161,
      "logits/rejected": -0.8673607707023621,
      "logps/chosen": -207.2660369873047,
      "logps/rejected": -179.528564453125,
      "loss": 0.6481,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.8134071230888367,
      "rewards/margins": 0.5864471197128296,
      "rewards/rejected": 0.2269599884748459,
      "step": 110
    },
    {
      "epoch": 0.5808080808080808,
      "grad_norm": 22.682283401489258,
      "learning_rate": 1.858922081915378e-05,
      "logits/chosen": -0.3367950916290283,
      "logits/rejected": -0.7870457768440247,
      "logps/chosen": -187.28225708007812,
      "logps/rejected": -167.66563415527344,
      "loss": 0.6859,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.0122456550598145,
      "rewards/margins": 0.42765671014785767,
      "rewards/rejected": 0.5845889449119568,
      "step": 115
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 21.8991756439209,
      "learning_rate": 1.6836451383113923e-05,
      "logits/chosen": -0.6206024885177612,
      "logits/rejected": -1.1339770555496216,
      "logps/chosen": -216.4201202392578,
      "logps/rejected": -181.8843536376953,
      "loss": 0.7218,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.1707611083984375,
      "rewards/margins": 0.4338921010494232,
      "rewards/rejected": 0.7368690371513367,
      "step": 120
    },
    {
      "epoch": 0.6313131313131313,
      "grad_norm": 26.37303352355957,
      "learning_rate": 1.5108302193984004e-05,
      "logits/chosen": -0.3522421717643738,
      "logits/rejected": -1.013496994972229,
      "logps/chosen": -231.60025024414062,
      "logps/rejected": -174.62521362304688,
      "loss": 0.631,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 1.5266729593276978,
      "rewards/margins": 0.6238378286361694,
      "rewards/rejected": 0.9028350710868835,
      "step": 125
    },
    {
      "epoch": 0.6565656565656566,
      "grad_norm": 19.788497924804688,
      "learning_rate": 1.3752556459724117e-05,
      "logits/chosen": -0.44988712668418884,
      "logits/rejected": -0.899565577507019,
      "logps/chosen": -206.1852264404297,
      "logps/rejected": -188.783447265625,
      "loss": 0.685,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.4883193969726562,
      "rewards/margins": 0.6105667352676392,
      "rewards/rejected": 0.8777526021003723,
      "step": 130
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 17.837421417236328,
      "learning_rate": 1.2102420272588178e-05,
      "logits/chosen": -0.269491583108902,
      "logits/rejected": -0.7996856570243835,
      "logps/chosen": -219.24240112304688,
      "logps/rejected": -178.4714813232422,
      "loss": 0.6356,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.4494140148162842,
      "rewards/margins": 0.8258872032165527,
      "rewards/rejected": 0.6235266923904419,
      "step": 135
    },
    {
      "epoch": 0.7070707070707071,
      "grad_norm": 30.42537498474121,
      "learning_rate": 1.0513746824428951e-05,
      "logits/chosen": -0.6660299897193909,
      "logits/rejected": -1.161827564239502,
      "logps/chosen": -225.38668823242188,
      "logps/rejected": -188.2053985595703,
      "loss": 0.7771,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.9760153889656067,
      "rewards/margins": 0.37082913517951965,
      "rewards/rejected": 0.6051862239837646,
      "step": 140
    },
    {
      "epoch": 0.7323232323232324,
      "grad_norm": 21.641803741455078,
      "learning_rate": 8.998899931103173e-06,
      "logits/chosen": -0.6154942512512207,
      "logits/rejected": -1.150761365890503,
      "logps/chosen": -205.3412628173828,
      "logps/rejected": -176.29917907714844,
      "loss": 0.6959,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.9418373107910156,
      "rewards/margins": 0.5389925241470337,
      "rewards/rejected": 0.40284472703933716,
      "step": 145
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 17.89181137084961,
      "learning_rate": 7.569668854942815e-06,
      "logits/chosen": -0.2999069094657898,
      "logits/rejected": -0.8618285059928894,
      "logps/chosen": -208.3065948486328,
      "logps/rejected": -163.25619506835938,
      "loss": 0.6386,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.9588847160339355,
      "rewards/margins": 0.6155614256858826,
      "rewards/rejected": 0.343323290348053,
      "step": 150
    },
    {
      "epoch": 0.7828282828282829,
      "grad_norm": 20.726341247558594,
      "learning_rate": 6.237176555082625e-06,
      "logits/chosen": -0.09209189563989639,
      "logits/rejected": -0.40321072936058044,
      "logps/chosen": -221.6671905517578,
      "logps/rejected": -206.04989624023438,
      "loss": 0.7131,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.885414719581604,
      "rewards/margins": 0.5608201622962952,
      "rewards/rejected": 0.32459449768066406,
      "step": 155
    },
    {
      "epoch": 0.8080808080808081,
      "grad_norm": 16.55711555480957,
      "learning_rate": 5.0117931232775e-06,
      "logits/chosen": -0.19318969547748566,
      "logits/rejected": -0.7406075596809387,
      "logps/chosen": -207.5948944091797,
      "logps/rejected": -176.14010620117188,
      "loss": 0.6493,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.8107932209968567,
      "rewards/margins": 0.5306531190872192,
      "rewards/rejected": 0.28014007210731506,
      "step": 160
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 17.550378799438477,
      "learning_rate": 3.903055078893489e-06,
      "logits/chosen": -0.05737120658159256,
      "logits/rejected": -0.6579563617706299,
      "logps/chosen": -220.5160675048828,
      "logps/rejected": -174.6505584716797,
      "loss": 0.7329,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.9277563095092773,
      "rewards/margins": 0.7453610301017761,
      "rewards/rejected": 0.18239526450634003,
      "step": 165
    },
    {
      "epoch": 0.8585858585858586,
      "grad_norm": 33.4338493347168,
      "learning_rate": 2.919591151157475e-06,
      "logits/chosen": -0.6612521409988403,
      "logits/rejected": -1.0171642303466797,
      "logps/chosen": -191.87083435058594,
      "logps/rejected": -168.1193084716797,
      "loss": 0.7216,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.65160071849823,
      "rewards/margins": 0.4179254472255707,
      "rewards/rejected": 0.2336752861738205,
      "step": 170
    },
    {
      "epoch": 0.8838383838383839,
      "grad_norm": 18.22390365600586,
      "learning_rate": 2.069055126263433e-06,
      "logits/chosen": -0.7078531384468079,
      "logits/rejected": -0.9975414276123047,
      "logps/chosen": -177.2061004638672,
      "logps/rejected": -164.01715087890625,
      "loss": 0.7113,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.7737497091293335,
      "rewards/margins": 0.5676254034042358,
      "rewards/rejected": 0.20612427592277527,
      "step": 175
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 17.45758056640625,
      "learning_rate": 1.3580662819512093e-06,
      "logits/chosen": -0.49047690629959106,
      "logits/rejected": -0.9685350656509399,
      "logps/chosen": -211.9286346435547,
      "logps/rejected": -177.24661254882812,
      "loss": 0.6725,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.8009859323501587,
      "rewards/margins": 0.5601747632026672,
      "rewards/rejected": 0.24081113934516907,
      "step": 180
    },
    {
      "epoch": 0.9343434343434344,
      "grad_norm": 24.04217529296875,
      "learning_rate": 7.92157873124415e-07,
      "logits/chosen": -0.42594170570373535,
      "logits/rejected": -1.1140912771224976,
      "logps/chosen": -225.57559204101562,
      "logps/rejected": -178.7427520751953,
      "loss": 0.645,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.0804827213287354,
      "rewards/margins": 0.6804853081703186,
      "rewards/rejected": 0.39999738335609436,
      "step": 185
    },
    {
      "epoch": 0.9595959595959596,
      "grad_norm": 30.535690307617188,
      "learning_rate": 3.757340694169109e-07,
      "logits/chosen": -0.4965648651123047,
      "logits/rejected": -0.7092806696891785,
      "logps/chosen": -183.77874755859375,
      "logps/rejected": -176.67739868164062,
      "loss": 0.7775,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.7033542394638062,
      "rewards/margins": 0.2690446078777313,
      "rewards/rejected": 0.4343096613883972,
      "step": 190
    },
    {
      "epoch": 0.9848484848484849,
      "grad_norm": 17.108673095703125,
      "learning_rate": 1.1203567984036101e-07,
      "logits/chosen": -0.38576817512512207,
      "logits/rejected": -1.0843629837036133,
      "logps/chosen": -222.65896606445312,
      "logps/rejected": -185.77786254882812,
      "loss": 0.5822,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.9926906824111938,
      "rewards/margins": 0.9567824602127075,
      "rewards/rejected": 0.03590827062726021,
      "step": 195
    },
    {
      "epoch": 1.0,
      "step": 198,
      "total_flos": 0.0,
      "train_loss": 0.7032270744593456,
      "train_runtime": 3353.7827,
      "train_samples_per_second": 0.943,
      "train_steps_per_second": 0.059
    }
  ],
  "logging_steps": 5,
  "max_steps": 198,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}