Generate a validation prompt for the language model. The thought state is unpacked to allow for additional keyword arguments

Inits the prompter.

Inits the response cache.

Initialize the ChatGPT instance with configuration, model details, and caching options. :param config_path: Path to the configuratio

Initialize the ChatGPT instance with configuration, model details, and caching options. :param config_path: Path to the configuratio

Initialize the AbstractLanguageModel instance with configuration, model details, and caching options. :param config_path: Path to th

Initialize an instance of the Llama2HF class with configuration, model details, and caching options. :param config_path: Path to the

Initialize the Controller instance with the language model, operations graph, prompter, parser, and problem parameters. :par

Initializes a new Operation instance with a unique id, and empty predecessors and successors.

Initializes a new ValidateAndImprove operation. :param num_samples: Number of samples to use for validation. Defaults to 1.

Initializes a new Generate operation. :param num_branches_prompt: Number of responses that each prompt should generate (passed to pr

Initializes a new Improve operation.

Initializes a new Aggregate operation. :param num_responses: Number of responses to use for aggregation. Defaults to 1. :typ

Initializes a new KeepBestN operation. :param n: Maximum number of thoughts to keep. :type n: int :param higher_is_b

Initializes a new KeepValid operation.

Initializes a new GroundTruth operation. :param ground_truth_evaluator: A function to evaluate if a thought solves the problem.

Initializes a new Selector operation. :param selector: A function to select thoughts from the predecessors' thoughts. :type

Initializes a new Graph of Operations instance with empty operations, roots, and leaves. The roots are the entry points in the graph

Initializes a new Thought instance with a state and various default flags. :param state: The state of the thought. Defaults to None.

Executes the scoring operation by scoring the thoughts from the predecessors. If combined scoring is used, the thoughts are scored to

Executes the ValidateAndImprove operation by validating and improving the predecessors' thoughts. If a validation function is provide

Executes the Generate operation by generating thoughts from the predecessors. The thoughts are generated by prompting the LM with the

Executes the Improve operation by improving the predecessors' thoughts. The thoughts are improved by prompting the LM with the predec

Executes the Aggregate operation by aggregating the predecessors' thoughts. The thoughts are aggregated by prompting the LM with the

Executes the KeepBestN operation by keeping the best N thoughts from the predecessors according to their score. :param lm: The langu

Executes the KeepValid operation by keeping the valid thoughts from the predecessors. Keeps unvalidated thoughts as well. :p

Executes the GroundTruth operation by evaluating the predecessors' thoughts using the ground truth evaluator function. :param lm: Th

Executes the Selector operation by selecting thoughts from the predecessors using the selector function. If the Selector has no prede

Add an operation to the graph considering its predecessors and successors. Adjust roots and leaves based on the added operation's pos

Generate an aggregation prompt for the language model. :param state_dicts: The thought states that should be aggregated. :ty

Clear the response cache.