Sends the inputs through the `model`.

Loads an agent from a local folder. Args: folder (`str` or `Path`): The folder where the agent is saved. **kwargs: Ad

Creates a [`Tool`] from a Space given its id on the Hub. Args: space_id (`str`): The id of the Space on

Get set of questions that have already been answered Args: file_path: Path to results file Returns: Set of alre

Factory function to get the appropriate experiment processor class Args: experiment_type: Type of experiment ("reasoning" or "ag

Extracts all the libraries (not relative imports) that are imported in a code. Args: code (`str`): Code text to inspect. Return

This function generates a JSON schema for a given function, based on its docstring and type hints. This is mostly used for passing lists of t

Initialize the CodeAgent with the specified model.

Initialize the Selenium WebDriver.

Compare two methods by their source code.

Load API key from file

Load OpenAI API key from file

Load the system prompt from YAML file

Load system prompt from YAML file Args: prompt_name: Name of the prompt file (without extension) Returns: Syste

Convert non-serializable types to serializable formats.

Extracts the JSON blob from the input and returns the JSON data and the rest of the input.

Prepare output path for experiment results Args: dataset_file: Path to dataset file model_id: Model ID log_folde

Process a dataset of entries Args: entries: List of dataset entries output_file: Path to output file

Process a single entry with its own model instance

Process a QA experiment using the appropriate processor Args: dataset_file: Path to dataset file model_kwargs: Model con

Provide the final answer to the task, based on the logs of the agent's interactions. Args: task (`str`): Task to perform

Extract ChatMessage objects from agent steps with proper nesting

Register the prefix of thought, which will be used only in the first step.

Reset cost tracker state Args: cost_threshold: Maximum cost allowed before stopping (in USD)

Attempt to serialize an object to JSON. If it fails, convert to string.

Compatibility wrapper for agent experiments

Compatibility wrapper for reasoning experiments

Saves the relevant code files for your agent. This will copy the code of your agent in `output_dir` as well as autogenerate: - a `to

Save the agent's memory to a log file in the logs directory. Args: task (`str`): The task that was executed, used for the log fil

Send variables to the kernel namespace using pickle.

Instantiates the `pre_processor`, `model` and `post_processor` if necessary.

Initialize the OpenAI model for scoring

To be implemented in children classes. Should return either None if the step is not final.

Perform one step in the ReAct framework: the agent thinks, acts, and observes the result. Returns None if the step is not final.

Runs an agent with the given task and streams the messages from the agent as gradio ChatMessages.

Convert the agent to a dictionary representation. Returns: `dict`: Dictionary representation of the agent.

Returns the "raw" version of that object. In the case of an AgentImage, it is a PIL.Image.Image.

Returns the stringified version of that object. In the case of an AgentImage, it is a path to the serialized version of the image.

Returns the stringified version of that object. In the case of an AgentAudio, it is a path to the serialized version of the audio.

Adds additional prompting for the managed agent, runs it, and wraps the output. This method is called only by a managed agent.

Process the input messages and return the model's response. Parameters: messages (`List[Dict[str, str]]`): A list

Check if code is a final answer and run it accordingly

Implements the += operator