An error raised when the interpreter cannot evaluate a Python expression, due to syntax error or unsupported operations.

Exception raised for errors in parsing type hints to generate JSON schemas

In this agent, the tool calls will be formulated by the LLM in code format, then parsed and executed. Args: tools (`list[Tool]`): [`

Base class for other agent-related exceptions

Model to use [vLLM](https://docs.vllm.ai/) server endpoint for fast LLM inference and serving. Parameters: model_id (`str`):

Exception raised for errors in parsing in the agent

This model connects to an OpenAI-compatible API server. Parameters: model_id (`str`): The model identifier to use on the serv

Audio type returned by the agent.

Exception raised for errors in generation in the agent

Exception raised for errors when executing a tool

Exception raised for errors in parsing docstrings to generate JSON schemas

Checks that a method - only uses defined names - contains no local imports (e.g. numpy is ok but local_script is not)

Model to use [vLLM](https://docs.vllm.ai/) for fast LLM inference and serving. Parameters: model_id (`str`): The Hugging Face

Exception raised for errors in execution in the agent

Image type returned by the agent. Behaves as a PIL.Image.Image.

Exception raised for errors in execution in the agent

Text type returned by the agent. Behaves as a string.

Exception raised for errors when incorrect arguments are passed to the tool

Minimal config class (simulating your argparse) Replace this with your real arguments or load them dynamically.

Singleton class for tracking costs across threads This class provides thread-safe cost tracking capabilities for experiments. It's i

Data collator used for completion tasks. It ensures that all the tokens of the labels are set to an 'ignore_index' when they do not come from

Executes Python code using Jupyter Kernel Gateway in a Docker container.

Executes Python code using E2B. Args: additional_imports (`list[str]`): Additional imports to install. logger (`Logger`): Lo

Prompt templates for the final answer. Args: pre_messages (`str`): Pre-messages prompt. post_messages (`str`): Post-messages

A class to interact with Hugging Face's Inference API for language model interaction. This model allows you to communicate with Hugging Face's mo

Model to use [LiteLLM Python SDK](https://docs.litellm.ai/docs/#litellm-python-sdk) to access hundreds of LLMs. Parameters: model_id (`st

Prompt templates for the managed agent. Args: task (`str`): Task prompt. report (`str`): Report prompt.

Prompt templates for the planning step. Args: initial_facts (`str`): Initial facts prompt. initial_plan (`str`): Initial pla

Prompt templates for the agent. Args: system_prompt (`str`): System prompt. planning ([`~agents.PlanningPromptTemplate`]): P

A class that uses Hugging Face's Transformers library for language model interaction. This model allows you to load and use Hugging Face's models

Processor for agent experiments This processor handles agent-based experiments where the model interacts with tools to solve problems.

Abstract class to be reimplemented to define types that can be returned by agents. These objects serve three purposes: - They behave as

This model connects to an Azure OpenAI deployment. Parameters: model_id (`str`): The model deployment name to use when connec

Base class for experiment processors This abstract class defines the interface and common functionality for all experiment processor

A one-line interface to launch your agent in Gradio

A class to interact with models loaded using MLX on Apple silicon. > [!TIP] > You must have `mlx-lm` installed on your machine. Please run `p

Agent class that solves the given task step by step, using the ReAct framework: While the objective is not reached, the agent will perform a

A [`Tool`] tailored towards Transformer models. On top of the class attributes of the base class [`Tool`], you will need to specify: - *

Processor for reasoning experiments This processor handles direct reasoning-based QA experiments, where the model is given a question an

A base class for the functions used by the agent. Subclass this and implement the `forward` method as well as the following class attributes:

This agent uses JSON-like tool calls, using method `model.get_tool_call` to leverage the LLM engine's tool calling capabilities. Args:

Tool collections enable loading a collection of tools in the agent's toolbox. Collections can be loaded from a collection in the Hub or from

WikipediaSearchTool searches Wikipedia and returns a summary or full text of the given topic, along with the page URL. Attributes: u