{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9748953974895398,
  "eval_steps": 500,
  "global_step": 118,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016736401673640166,
      "grad_norm": 118.10906601297802,
      "learning_rate": 8.333333333333332e-09,
      "logits/chosen": -2.0019314289093018,
      "logits/rejected": -1.9766970872879028,
      "logps/chosen": -309.98992919921875,
      "logps/pi_response": -172.89300537109375,
      "logps/ref_response": -172.89300537109375,
      "logps/rejected": -664.0586547851562,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16736401673640167,
      "grad_norm": 119.81663055217236,
      "learning_rate": 8.333333333333334e-08,
      "logits/chosen": -1.9912662506103516,
      "logits/rejected": -1.9100477695465088,
      "logps/chosen": -339.4380187988281,
      "logps/pi_response": -145.0324249267578,
      "logps/ref_response": -144.99766540527344,
      "logps/rejected": -690.2503051757812,
      "loss": 0.6896,
      "rewards/accuracies": 0.5208333134651184,
      "rewards/chosen": -0.005152760073542595,
      "rewards/margins": 0.007620910182595253,
      "rewards/rejected": -0.012773669324815273,
      "step": 10
    },
    {
      "epoch": 0.33472803347280333,
      "grad_norm": 61.89513911795292,
      "learning_rate": 9.860114570402053e-08,
      "logits/chosen": -1.942497968673706,
      "logits/rejected": -1.850992202758789,
      "logps/chosen": -344.79351806640625,
      "logps/pi_response": -157.76609802246094,
      "logps/ref_response": -152.63552856445312,
      "logps/rejected": -725.7340698242188,
      "loss": 0.6134,
      "rewards/accuracies": 0.7281249761581421,
      "rewards/chosen": -0.1949012279510498,
      "rewards/margins": 0.2145640105009079,
      "rewards/rejected": -0.4094652235507965,
      "step": 20
    },
    {
      "epoch": 0.502092050209205,
      "grad_norm": 41.75290331160563,
      "learning_rate": 9.305218058836777e-08,
      "logits/chosen": -1.8073937892913818,
      "logits/rejected": -1.748516321182251,
      "logps/chosen": -373.4971923828125,
      "logps/pi_response": -154.7528839111328,
      "logps/ref_response": -136.81463623046875,
      "logps/rejected": -826.9029541015625,
      "loss": 0.5314,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.5469005703926086,
      "rewards/margins": 0.8107492327690125,
      "rewards/rejected": -1.357649803161621,
      "step": 30
    },
    {
      "epoch": 0.6694560669456067,
      "grad_norm": 60.21218903804784,
      "learning_rate": 8.374915007591053e-08,
      "logits/chosen": -1.7726547718048096,
      "logits/rejected": -1.6935310363769531,
      "logps/chosen": -415.587890625,
      "logps/pi_response": -174.06515502929688,
      "logps/ref_response": -150.7990264892578,
      "logps/rejected": -888.4119262695312,
      "loss": 0.5313,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.7993820905685425,
      "rewards/margins": 1.1050827503204346,
      "rewards/rejected": -1.9044649600982666,
      "step": 40
    },
    {
      "epoch": 0.8368200836820083,
      "grad_norm": 60.309635521609,
      "learning_rate": 7.150326011382603e-08,
      "logits/chosen": -1.7549054622650146,
      "logits/rejected": -1.6928848028182983,
      "logps/chosen": -417.91644287109375,
      "logps/pi_response": -171.1042938232422,
      "logps/ref_response": -149.61317443847656,
      "logps/rejected": -928.3387451171875,
      "loss": 0.5012,
      "rewards/accuracies": 0.7718750238418579,
      "rewards/chosen": -0.7739541530609131,
      "rewards/margins": 1.2372756004333496,
      "rewards/rejected": -2.0112297534942627,
      "step": 50
    },
    {
      "epoch": 1.00418410041841,
      "grad_norm": 43.55054669917714,
      "learning_rate": 5.738232820012406e-08,
      "logits/chosen": -1.7201545238494873,
      "logits/rejected": -1.660457968711853,
      "logps/chosen": -363.9744873046875,
      "logps/pi_response": -158.1030731201172,
      "logps/ref_response": -138.2023468017578,
      "logps/rejected": -817.6099853515625,
      "loss": 0.4936,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.6088830232620239,
      "rewards/margins": 1.028526782989502,
      "rewards/rejected": -1.6374098062515259,
      "step": 60
    },
    {
      "epoch": 1.1715481171548117,
      "grad_norm": 39.95163711230341,
      "learning_rate": 4.2617671799875946e-08,
      "logits/chosen": -1.768969178199768,
      "logits/rejected": -1.6774520874023438,
      "logps/chosen": -409.07965087890625,
      "logps/pi_response": -178.50694274902344,
      "logps/ref_response": -157.27854919433594,
      "logps/rejected": -837.8878784179688,
      "loss": 0.4784,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.6469767689704895,
      "rewards/margins": 1.0017244815826416,
      "rewards/rejected": -1.6487010717391968,
      "step": 70
    },
    {
      "epoch": 1.3389121338912133,
      "grad_norm": 54.62109419599635,
      "learning_rate": 2.8496739886173992e-08,
      "logits/chosen": -1.7565717697143555,
      "logits/rejected": -1.6744270324707031,
      "logps/chosen": -414.45989990234375,
      "logps/pi_response": -171.05670166015625,
      "logps/ref_response": -150.3345947265625,
      "logps/rejected": -869.4351806640625,
      "loss": 0.4682,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": -0.6550418138504028,
      "rewards/margins": 1.0387578010559082,
      "rewards/rejected": -1.693799614906311,
      "step": 80
    },
    {
      "epoch": 1.506276150627615,
      "grad_norm": 47.344285151326375,
      "learning_rate": 1.6250849924089483e-08,
      "logits/chosen": -1.7577073574066162,
      "logits/rejected": -1.6519912481307983,
      "logps/chosen": -414.08984375,
      "logps/pi_response": -160.68191528320312,
      "logps/ref_response": -142.70220947265625,
      "logps/rejected": -822.3709716796875,
      "loss": 0.4783,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.6929503083229065,
      "rewards/margins": 0.93775475025177,
      "rewards/rejected": -1.6307048797607422,
      "step": 90
    },
    {
      "epoch": 1.6736401673640167,
      "grad_norm": 54.4733158152604,
      "learning_rate": 6.947819411632222e-09,
      "logits/chosen": -1.8023641109466553,
      "logits/rejected": -1.725804328918457,
      "logps/chosen": -395.9789733886719,
      "logps/pi_response": -165.1318817138672,
      "logps/ref_response": -145.55709838867188,
      "logps/rejected": -855.1980590820312,
      "loss": 0.4809,
      "rewards/accuracies": 0.815625011920929,
      "rewards/chosen": -0.6401264071464539,
      "rewards/margins": 1.0448520183563232,
      "rewards/rejected": -1.6849782466888428,
      "step": 100
    },
    {
      "epoch": 1.8410041841004183,
      "grad_norm": 38.619103871977664,
      "learning_rate": 1.3988542959794625e-09,
      "logits/chosen": -1.7393662929534912,
      "logits/rejected": -1.651567816734314,
      "logps/chosen": -383.98895263671875,
      "logps/pi_response": -170.61387634277344,
      "logps/ref_response": -149.04595947265625,
      "logps/rejected": -837.4610595703125,
      "loss": 0.4635,
      "rewards/accuracies": 0.7906249761581421,
      "rewards/chosen": -0.6302980780601501,
      "rewards/margins": 1.0058103799819946,
      "rewards/rejected": -1.6361083984375,
      "step": 110
    },
    {
      "epoch": 1.9748953974895398,
      "step": 118,
      "total_flos": 0.0,
      "train_loss": 0.5197100033194332,
      "train_runtime": 5205.1722,
      "train_samples_per_second": 5.872,
      "train_steps_per_second": 0.023
    }
  ],
  "logging_steps": 10,
  "max_steps": 118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}