Gradio_UI.py

#!/usr/bin/env python
# coding=utf-8
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import mimetypes
import os
import re
import shutil
from typing import Optional

from smolagents.agent_types import AgentAudio, AgentImage, AgentText, handle_agent_output_types
from smolagents.agents import ActionStep, MultiStepAgent
from smolagents.memory import MemoryStep
from smolagents.utils import _is_package_available

def pull_messages_from_step(
    step_log: MemoryStep,
):
    """Extract ChatMessage objects from agent steps with proper nesting"""
    import gradio as gr

    if isinstance(step_log, ActionStep):
        step_number = f"Step {step_log.step_number}" if step_log.step_number is not None else ""
        yield gr.ChatMessage(role="assistant", content=f"**{step_number}**")

        if hasattr(step_log, "model_output") and step_log.model_output is not None:
            model_output = step_log.model_output.strip()
            yield gr.ChatMessage(role="assistant", content=model_output)

        if hasattr(step_log, "tool_calls") and step_log.tool_calls is not None:
            first_tool_call = step_log.tool_calls[0]
            parent_id = f"call_{len(step_log.tool_calls)}"
            args = first_tool_call.arguments
            content = str(args.get("answer", str(args))) if isinstance(args, dict) else str(args).strip()

            parent_message_tool = gr.ChatMessage(
                role="assistant",
                content=content,
                metadata={"title": f"🛠️ Used tool {first_tool_call.name}", "id": parent_id, "status": "pending"},
            )
            yield parent_message_tool

            if hasattr(step_log, "observations") and step_log.observations:
                log_content = step_log.observations.strip()
                yield gr.ChatMessage(
                    role="assistant",
                    content=f"{log_content}",
                    metadata={"title": "📝 Execution Logs", "parent_id": parent_id, "status": "done"},
                )
            parent_message_tool.metadata["status"] = "done"

        elif hasattr(step_log, "error") and step_log.error is not None:
            yield gr.ChatMessage(role="assistant", content=str(step_log.error), metadata={"title": "💥 Error"})

def stream_to_gradio(
    agent,
    task: str,
    reset_agent_memory: bool = False,
    additional_args: Optional[dict] = None,
):
    import gradio as gr

    for step_log in agent.run(task, stream=True, reset=reset_agent_memory, additional_args=additional_args):
        for message in pull_messages_from_step(step_log):
            yield message

    final_answer = handle_agent_output_types(step_log)
    if isinstance(final_answer, AgentText):
        yield gr.ChatMessage(role="assistant", content={"path": final_answer.to_string(), "mime_type": "image/png"})
    elif isinstance(final_answer, AgentImage):
        yield gr.ChatMessage(role="assistant", content={"path": final_answer.to_string(), "mime_type": "image/png"})
    elif isinstance(final_answer, AgentAudio):
        yield gr.ChatMessage(role="assistant", content={"path": final_answer.to_string(), "mime_type": "audio/wav"})
    else:
        yield gr.ChatMessage(role="assistant", content=f"**Final answer:** {str(final_answer)}")

class GradioUI:
    def __init__(self, agent: MultiStepAgent, file_upload_folder: str | None = None):
        if not _is_package_available("gradio"):
            raise ModuleNotFoundError("Please install 'gradio' extra to use the GradioUI: `pip install 'smolagents[gradio]'`")
        self.agent = agent
        self.file_upload_folder = file_upload_folder
        if self.file_upload_folder is not None:
            if not os.path.exists(file_upload_folder):
                os.mkdir(file_upload_folder)

    def interact_with_agent(self, prompt, messages):
        import gradio as gr

        messages.append(gr.ChatMessage(role="user", content=prompt))
        yield messages
        for msg in stream_to_gradio(self.agent, task=prompt, reset_agent_memory=False):
            messages.append(msg)
            yield messages
        yield messages

    def launch(self, **kwargs):
        import gradio as gr

        with gr.Blocks(fill_height=True) as demo:
            stored_messages = gr.State([])
            chatbot = gr.Chatbot(label="Agent", type="messages")
            text_input = gr.Textbox(lines=1, label="Chat Message", placeholder="Ask me anything! Try: 'Show me the first logo of Toyota.'")
            text_input.submit(self.interact_with_agent, [text_input, stored_messages], [chatbot])
        demo.launch(debug=True, share=True, **kwargs)

__all__ = ["stream_to_gradio", "GradioUI"]