On this tutorial, we construct a complicated computer-use agent from scratch that may purpose, plan, and carry out digital actions utilizing a neighborhood open-weight mannequin. We create a miniature simulated desktop, equip it with a software interface, and design an clever agent that may analyze its surroundings, determine on actions like clicking or typing, and execute them step-by-step. By the top, we see how the agent interprets targets corresponding to opening emails or taking notes, demonstrating how a neighborhood language mannequin can mimic interactive reasoning and process execution. Try the FULL CODES right here.
!pip set up -q transformers speed up sentencepiece nest_asyncio
import torch, asyncio, uuid
from transformers import pipeline
import nest_asyncio
nest_asyncio.apply()
We arrange the environment by putting in important libraries corresponding to Transformers, Speed up, and Nest Asyncio, which allow us to run native fashions and asynchronous duties seamlessly in Colab. We put together the runtime in order that the upcoming elements of our agent can work effectively with out exterior dependencies. Try the FULL CODES right here.
class LocalLLM:
def __init__(self, model_name=”google/flan-t5-small”, max_new_tokens=128):
self.pipe = pipeline(“text2text-generation”, mannequin=model_name, system=0 if torch.cuda.is_available() else -1)
self.max_new_tokens = max_new_tokens
def generate(self, immediate: str) -> str:
out = self.pipe(immediate, max_new_tokens=self.max_new_tokens, temperature=0.0)[0][“generated_text”]
return out.strip()
class VirtualComputer:
def __init__(self):
self.apps = {“browser”: “https://example.com”, “notes”: “”, “mail”: [“Welcome to CUA”, “Invoice #221”, “Weekly Report”]}
self.focus = “browser”
self.display screen = “Browser open at https://example.comnSearch bar targeted.”
self.action_log = []
def screenshot(self):
return f”FOCUS:{self.focus}nSCREEN:n{self.display screen}nAPPS:{checklist(self.apps.keys())}”
def click on(self, goal:str):
if goal in self.apps:
self.focus = goal
if goal==”browser”:
self.display screen = f”Browser tab: {self.apps[‘browser’]}nAddress bar targeted.”
elif goal==”notes”:
self.display screen = f”Notes AppnCurrent notes:n{self.apps[‘notes’]}”
elif goal==”mail”:
inbox = “n”.be part of(f”- {s}” for s in self.apps[‘mail’])
self.display screen = f”Mail App Inbox:n{inbox}n(Learn-only preview)”
else:
self.display screen += f”nClicked ‘{goal}’.”
self.action_log.append({“sort”:”click on”,”goal”:goal})
def sort(self, textual content:str):
if self.focus==”browser”:
self.apps[“browser”] = textual content
self.display screen = f”Browser tab now at {textual content}nPage headline: Instance Area”
elif self.focus==”notes”:
self.apps[“notes”] += (“n”+textual content)
self.display screen = f”Notes AppnCurrent notes:n{self.apps[‘notes’]}”
else:
self.display screen += f”nTyped ‘{textual content}’ however no editable area.”
self.action_log.append({“sort”:”sort”,”textual content”:textual content})
We outline the core elements, a light-weight native mannequin, and a digital laptop. We use Flan-T5 as our reasoning engine and create a simulated desktop that may open apps, show screens, and reply to typing and clicking actions. Try the FULL CODES right here.
class ComputerTool:
def __init__(self, laptop:VirtualComputer):
self.laptop = laptop
def run(self, command:str, argument:str=””):
if command==”click on”:
self.laptop.click on(argument)
return {“standing”:”accomplished”,”outcome”:f”clicked {argument}”}
if command==”sort”:
self.laptop.sort(argument)
return {“standing”:”accomplished”,”outcome”:f”typed {argument}”}
if command==”screenshot”:
snap = self.laptop.screenshot()
return {“standing”:”accomplished”,”outcome”:snap}
return {“standing”:”error”,”outcome”:f”unknown command {command}”}
We introduce the ComputerTool interface, which acts because the communication bridge between the agent’s reasoning and the digital desktop. We outline high-level operations corresponding to click on, sort, and screenshot, enabling the agent to work together with the surroundings in a structured method. Try the FULL CODES right here.
class ComputerAgent:
def __init__(self, llm:LocalLLM, software:ComputerTool, max_trajectory_budget:float=5.0):
self.llm = llm
self.software = software
self.max_trajectory_budget = max_trajectory_budget
async def run(self, messages):
user_goal = messages[-1][“content”]
steps_remaining = int(self.max_trajectory_budget)
output_events = []
total_prompt_tokens = 0
total_completion_tokens = 0
whereas steps_remaining>0:
display screen = self.software.laptop.screenshot()
immediate = (
“You’re a computer-use agent.n”
f”Person objective: {user_goal}n”
f”Present display screen:n{display screen}nn”
“Suppose step-by-step.n”
“Reply with: ACTION ARG THEN .n”
)
thought = self.llm.generate(immediate)
total_prompt_tokens += len(immediate.cut up())
total_completion_tokens += len(thought.cut up())
motion=”screenshot”; arg=””; assistant_msg=”Working…”
for line in thought.splitlines():
if line.strip().startswith(“ACTION “):
after = line.cut up(“ACTION “,1)[1]
motion = after.cut up()[0].strip()
if “ARG ” in line:
half = line.cut up(“ARG “,1)[1]
if ” THEN ” partially:
arg = half.cut up(” THEN “)[0].strip()
else:
arg = half.strip()
if “THEN ” in line:
assistant_msg = line.cut up(“THEN “,1)[1].strip()
output_events.append({“abstract”:[{“text”:assistant_msg,”type”:”summary_text”}],”sort”:”reasoning”})
call_id = “call_”+uuid.uuid4().hex[:16]
tool_res = self.software.run(motion, arg)
output_events.append({“motion”:{“sort”:motion,”textual content”:arg},”call_id”:call_id,”standing”:tool_res[“status”],”sort”:”computer_call”})
snap = self.software.laptop.screenshot()
output_events.append({“sort”:”computer_call_output”,”call_id”:call_id,”output”:{“sort”:”input_image”,”image_url”:snap}})
output_events.append({“sort”:”message”,”position”:”assistant”,”content material”:[{“type”:”output_text”,”text”:assistant_msg}]})
if “accomplished” in assistant_msg.decrease() or “right here is” in assistant_msg.decrease():
break
steps_remaining -= 1
utilization = {“prompt_tokens”: total_prompt_tokens,”completion_tokens”: total_completion_tokens,”total_tokens”: total_prompt_tokens + total_completion_tokens,”response_cost”: 0.0}
yield {“output”: output_events, “utilization”: utilization}
We assemble the ComputerAgent, which serves because the system’s clever controller. We program it to purpose about targets, determine which actions to take, execute these by means of the software interface, and report every interplay as a step in its decision-making course of. Try the FULL CODES right here.
async def main_demo():
laptop = VirtualComputer()
software = ComputerTool(laptop)
llm = LocalLLM()
agent = ComputerAgent(llm, software, max_trajectory_budget=4)
messages=[{“role”:”user”,”content”:”Open mail, read inbox subjects, and summarize.”}]
async for lead to agent.run(messages):
print(“==== STREAM RESULT ====”)
for occasion in outcome[“output”]:
if occasion[“type”]==”computer_call”:
a = occasion.get(“motion”,{})
print(f”[TOOL CALL] {a.get(‘sort’)} -> {a.get(‘textual content’)} [{event.get(‘status’)}]”)
if occasion[“type”]==”computer_call_output”:
snap = occasion[“output”][“image_url”]
print(“SCREEN AFTER ACTION:n”, snap[:400],”…n”)
if occasion[“type”]==”message”:
print(“ASSISTANT:”, occasion[“content”][0][“text”], “n”)
print(“USAGE:”, outcome[“usage”])
loop = asyncio.get_event_loop()
loop.run_until_complete(main_demo())
We deliver all the pieces collectively by working the demo, the place the agent interprets a person’s request and performs duties on the digital laptop. We observe it producing reasoning, executing instructions, updating the digital display screen, and attaining its objective in a transparent, step-by-step method.
In conclusion, we applied the essence of a computer-use agent able to autonomous reasoning and interplay. We witness how native language fashions like Flan-T5 can powerfully simulate desktop-level automation inside a protected, text-based sandbox. This venture helps us perceive the structure behind clever brokers corresponding to these in computer-use brokers, bridging pure language reasoning with digital software management. It lays a powerful basis for extending these capabilities towards real-world, multimodal, and safe automation methods.
Try the FULL CODES right here. Be at liberty to take a look at our GitHub Web page for Tutorials, Codes and Notebooks. Additionally, be at liberty to observe us on Twitter and don’t neglect to hitch our 100k+ ML SubReddit and Subscribe to our Publication. Wait! are you on telegram? now you possibly can be part of us on telegram as properly.
Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is dedicated to harnessing the potential of Synthetic Intelligence for social good. His most up-to-date endeavor is the launch of an Synthetic Intelligence Media Platform, Marktechpost, which stands out for its in-depth protection of machine studying and deep studying information that’s each technically sound and simply comprehensible by a large viewers. The platform boasts of over 2 million month-to-month views, illustrating its reputation amongst audiences.
🙌 Comply with MARKTECHPOST: Add us as a most well-liked supply on Google.
