{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.968,
  "eval_steps": 100,
  "global_step": 248,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000002e-07,
      "logits/chosen": 0.02732202410697937,
      "logits/rejected": 0.16736462712287903,
      "logps/chosen": -204.44515991210938,
      "logps/rejected": -186.30474853515625,
      "loss": 0.0014,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.0000000000000003e-06,
      "logits/chosen": 0.11495557427406311,
      "logits/rejected": 0.14849303662776947,
      "logps/chosen": -174.2774658203125,
      "logps/rejected": -139.304443359375,
      "loss": 0.0014,
      "rewards/accuracies": 0.3958333432674408,
      "rewards/chosen": 0.0010303932940587401,
      "rewards/margins": 0.0013937298208475113,
      "rewards/rejected": -0.0003633367014117539,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.000000000000001e-06,
      "logits/chosen": 0.19859905540943146,
      "logits/rejected": 0.2755558490753174,
      "logps/chosen": -186.06753540039062,
      "logps/rejected": -150.23538208007812,
      "loss": 0.0014,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": -0.0006829313351772726,
      "rewards/margins": 0.0010883348295465112,
      "rewards/rejected": -0.0017712658736854792,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.993800445762451e-06,
      "logits/chosen": 0.10206829011440277,
      "logits/rejected": 0.09731761366128922,
      "logps/chosen": -189.70846557617188,
      "logps/rejected": -176.63827514648438,
      "loss": 0.0014,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.0010909180855378509,
      "rewards/margins": 0.0010104707907885313,
      "rewards/rejected": 8.044719288591295e-05,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.944388344834205e-06,
      "logits/chosen": 0.21991512179374695,
      "logits/rejected": 0.13409800827503204,
      "logps/chosen": -178.78292846679688,
      "logps/rejected": -151.7918243408203,
      "loss": 0.0016,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": -0.0007118875510059297,
      "rewards/margins": -0.000794673920609057,
      "rewards/rejected": 8.278638415504247e-05,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.8465431931347904e-06,
      "logits/chosen": 0.10882525146007538,
      "logits/rejected": 0.16875343024730682,
      "logps/chosen": -185.3433074951172,
      "logps/rejected": -174.74209594726562,
      "loss": 0.0015,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": 0.0009307868895120919,
      "rewards/margins": 0.000729514576960355,
      "rewards/rejected": 0.00020127242896705866,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.702203692102539e-06,
      "logits/chosen": 0.1939707249403,
      "logits/rejected": 0.18511822819709778,
      "logps/chosen": -194.64193725585938,
      "logps/rejected": -159.30172729492188,
      "loss": 0.0015,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.0016695528756827116,
      "rewards/margins": 8.004475239431486e-05,
      "rewards/rejected": 0.0015895080287009478,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.514229781074239e-06,
      "logits/chosen": 0.1889864206314087,
      "logits/rejected": 0.1525781750679016,
      "logps/chosen": -196.69769287109375,
      "logps/rejected": -172.0983428955078,
      "loss": 0.0015,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.0013059942284598947,
      "rewards/margins": -0.0003628497361205518,
      "rewards/rejected": 0.0016688440227881074,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.286345970517195e-06,
      "logits/chosen": 0.1526193767786026,
      "logits/rejected": 0.13564926385879517,
      "logps/chosen": -176.2852020263672,
      "logps/rejected": -149.50759887695312,
      "loss": 0.0014,
      "rewards/accuracies": 0.4437499940395355,
      "rewards/chosen": 0.0020406683906912804,
      "rewards/margins": 0.0011222332250326872,
      "rewards/rejected": 0.0009184351074509323,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.023067544670082e-06,
      "logits/chosen": 0.19515976309776306,
      "logits/rejected": 0.16219770908355713,
      "logps/chosen": -183.90406799316406,
      "logps/rejected": -157.0895538330078,
      "loss": 0.0017,
      "rewards/accuracies": 0.375,
      "rewards/chosen": -0.001846942352131009,
      "rewards/margins": -0.0017863952089101076,
      "rewards/rejected": -6.0547237808350474e-05,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.7296110958116845e-06,
      "logits/chosen": 0.16685011982917786,
      "logits/rejected": 0.05337408185005188,
      "logps/chosen": -177.84762573242188,
      "logps/rejected": -152.65464782714844,
      "loss": 0.0015,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": -0.0007909245905466378,
      "rewards/margins": 0.0006749060703441501,
      "rewards/rejected": -0.001465830602683127,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": 0.0015224728267639875,
      "eval_logits/rejected": 0.09820695966482162,
      "eval_logps/chosen": -306.2193603515625,
      "eval_logps/rejected": -278.5382080078125,
      "eval_loss": 0.001858623931184411,
      "eval_rewards/accuracies": 0.5149999856948853,
      "eval_rewards/chosen": 0.0015068423235788941,
      "eval_rewards/margins": 0.0008032902260310948,
      "eval_rewards/rejected": 0.0007035521557554603,
      "eval_runtime": 420.6562,
      "eval_samples_per_second": 4.754,
      "eval_steps_per_second": 1.189,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.4117911628292944e-06,
      "logits/chosen": 0.22267238795757294,
      "logits/rejected": 0.19200441241264343,
      "logps/chosen": -205.3648223876953,
      "logps/rejected": -172.17208862304688,
      "loss": 0.0015,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": -0.002072205301374197,
      "rewards/margins": -0.0004376435244921595,
      "rewards/rejected": -0.0016345620388165116,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.075905022087675e-06,
      "logits/chosen": 0.1819857358932495,
      "logits/rejected": 0.23743709921836853,
      "logps/chosen": -184.61007690429688,
      "logps/rejected": -160.8583221435547,
      "loss": 0.0013,
      "rewards/accuracies": 0.45625001192092896,
      "rewards/chosen": 0.0009784279391169548,
      "rewards/margins": 0.0018532350659370422,
      "rewards/rejected": -0.000874807417858392,
      "step": 120
    },
    {
      "epoch": 2.08,
      "learning_rate": 2.728607913349464e-06,
      "logits/chosen": 0.09743748605251312,
      "logits/rejected": 0.11662141233682632,
      "logps/chosen": -176.29867553710938,
      "logps/rejected": -140.44961547851562,
      "loss": 0.0014,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0014981284039095044,
      "rewards/margins": 0.0015437586698681116,
      "rewards/rejected": -4.563046604744159e-05,
      "step": 130
    },
    {
      "epoch": 2.24,
      "learning_rate": 2.376781173017589e-06,
      "logits/chosen": 0.03450363129377365,
      "logits/rejected": 0.09166844189167023,
      "logps/chosen": -191.46543884277344,
      "logps/rejected": -154.39138793945312,
      "loss": 0.0014,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.001418759347870946,
      "rewards/margins": 0.0005404851399362087,
      "rewards/rejected": 0.0008782741497270763,
      "step": 140
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.0273958875043877e-06,
      "logits/chosen": 0.15574321150779724,
      "logits/rejected": 0.15428531169891357,
      "logps/chosen": -179.232421875,
      "logps/rejected": -154.8232421875,
      "loss": 0.0014,
      "rewards/accuracies": 0.48124998807907104,
      "rewards/chosen": 0.0011932613560929894,
      "rewards/margins": 0.0013866318622604012,
      "rewards/rejected": -0.00019337031699251384,
      "step": 150
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.6873747682962393e-06,
      "logits/chosen": 0.2517469525337219,
      "logits/rejected": 0.1588711440563202,
      "logps/chosen": -193.44821166992188,
      "logps/rejected": -170.86611938476562,
      "loss": 0.0015,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.00030686514219269156,
      "rewards/margins": 0.0002994650858454406,
      "rewards/rejected": -0.0006063304608687758,
      "step": 160
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.363454985517803e-06,
      "logits/chosen": 0.20987768471240997,
      "logits/rejected": 0.08343996852636337,
      "logps/chosen": -187.78610229492188,
      "logps/rejected": -167.81800842285156,
      "loss": 0.0014,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": 0.002795418258756399,
      "rewards/margins": 0.0018930940423160791,
      "rewards/rejected": 0.0009023241582326591,
      "step": 170
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.062054677808238e-06,
      "logits/chosen": 0.23503074049949646,
      "logits/rejected": 0.2113850861787796,
      "logps/chosen": -186.79861450195312,
      "logps/rejected": -162.50196838378906,
      "loss": 0.0014,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": 0.0017884777626022696,
      "rewards/margins": 0.0021424146834760904,
      "rewards/rejected": -0.0003539369790814817,
      "step": 180
    },
    {
      "epoch": 3.04,
      "learning_rate": 7.891457834794711e-07,
      "logits/chosen": 0.138333261013031,
      "logits/rejected": 0.21794256567955017,
      "logps/chosen": -179.44418334960938,
      "logps/rejected": -158.67454528808594,
      "loss": 0.0014,
      "rewards/accuracies": 0.45625001192092896,
      "rewards/chosen": 0.0017921695252880454,
      "rewards/margins": 0.0020220079459249973,
      "rewards/rejected": -0.00022983844974078238,
      "step": 190
    },
    {
      "epoch": 3.2,
      "learning_rate": 5.501357126768117e-07,
      "logits/chosen": 0.2148284912109375,
      "logits/rejected": 0.1433248072862625,
      "logps/chosen": -186.4739532470703,
      "logps/rejected": -167.48558044433594,
      "loss": 0.0014,
      "rewards/accuracies": 0.4437499940395355,
      "rewards/chosen": 0.002506103366613388,
      "rewards/margins": 0.0016806632047519088,
      "rewards/rejected": 0.0008254402200691402,
      "step": 200
    },
    {
      "epoch": 3.2,
      "eval_logits/chosen": -0.00685009965673089,
      "eval_logits/rejected": 0.09018866717815399,
      "eval_logps/chosen": -306.28704833984375,
      "eval_logps/rejected": -278.6018981933594,
      "eval_loss": 0.0019364985637366772,
      "eval_rewards/accuracies": 0.5195000171661377,
      "eval_rewards/chosen": 0.0008296637679450214,
      "eval_rewards/margins": 0.0007632386405020952,
      "eval_rewards/rejected": 6.642500375164673e-05,
      "eval_runtime": 420.9095,
      "eval_samples_per_second": 4.752,
      "eval_steps_per_second": 1.188,
      "step": 200
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.4976020508682345e-07,
      "logits/chosen": 0.10354860126972198,
      "logits/rejected": 0.1740628331899643,
      "logps/chosen": -187.9175262451172,
      "logps/rejected": -160.59580993652344,
      "loss": 0.0014,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.0003708422009367496,
      "rewards/margins": 0.0009719420922920108,
      "rewards/rejected": -0.0006010999786667526,
      "step": 210
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.9198949610721273e-07,
      "logits/chosen": 0.14361225068569183,
      "logits/rejected": 0.09037239849567413,
      "logps/chosen": -183.87716674804688,
      "logps/rejected": -157.98300170898438,
      "loss": 0.0013,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": 0.0009205196984112263,
      "rewards/margins": 0.0023204255849123,
      "rewards/rejected": -0.0013999061193317175,
      "step": 220
    },
    {
      "epoch": 3.68,
      "learning_rate": 7.994965069994143e-08,
      "logits/chosen": 0.20537514984607697,
      "logits/rejected": 0.18879783153533936,
      "logps/chosen": -193.15943908691406,
      "logps/rejected": -162.2894744873047,
      "loss": 0.0015,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": -0.00020981582929380238,
      "rewards/margins": 0.0006767899030819535,
      "rewards/rejected": -0.0008866057032719254,
      "step": 230
    },
    {
      "epoch": 3.84,
      "learning_rate": 1.5860623616664183e-08,
      "logits/chosen": 0.13774822652339935,
      "logits/rejected": 0.09208400547504425,
      "logps/chosen": -189.96868896484375,
      "logps/rejected": -159.98782348632812,
      "loss": 0.0016,
      "rewards/accuracies": 0.34375,
      "rewards/chosen": -6.72071473672986e-06,
      "rewards/margins": -0.0008294621366076171,
      "rewards/rejected": 0.0008227415382862091,
      "step": 240
    },
    {
      "epoch": 3.97,
      "step": 248,
      "total_flos": 0.0,
      "train_loss": 0.0014411601897013643,
      "train_runtime": 2755.6406,
      "train_samples_per_second": 1.452,
      "train_steps_per_second": 0.09
    }
  ],
  "logging_steps": 10,
  "max_steps": 248,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}