curl http://any-hosted-llama-stack.com/alpha/post-training/preference-optimize \
    -H 'Content-Type: application/json' \
    -d '{
      "algorithm_config": {
        "epsilon": 0,
        "gamma": 0,
        "reward_clip": 0,
        "reward_scale": 0
      },
      "finetuned_model": "finetuned_model",
      "hyperparam_search_config": {
        "foo": true
      },
      "job_uuid": "job_uuid",
      "logger_config": {
        "foo": true
      },
      "training_config": {
        "data_config": {
          "batch_size": 0,
          "dataset_id": "dataset_id",
          "shuffle": true
        },
        "gradient_accumulation_steps": 0,
        "max_steps_per_epoch": 0,
        "max_validation_steps": 0,
        "n_epochs": 0,
        "optimizer_config": {
          "lr": 0,
          "num_warmup_steps": 0,
          "optimizer_type": "adam",
          "weight_decay": 0
        }
      }
    }'