File size: 3,043 Bytes
7336908 184dc24 7336908 184dc24 fffd8eb 94099fe 7336908 fe78835 0fe79e1 fe78835 7336908 fb8b406 7336908 df0be25 7336908 0fe79e1 7336908 fffd8eb 7336908 c1f8728 7336908 df0be25 7336908 318feab 20e5c62 318feab df0be25 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 |
from magentic import FunctionCall, OpenaiChatModel, prompt_chain
import requests
from BasicAgent import BasicAgent
import os
class CustomAgent(BasicAgent):
def __init__(self):
super().__init__()
self._system_prompt = """\
You are a general AI assistant. I will ask you a question. Provide an answer like so: YOUR FINAL ANSWER. DO NOT SAY ANYTHING EXCEPT YOUR FINAL ANSWER.
Only use the provided tools if your own knowledge would be insufficient. If you are unable to answer the question using the provided tools, give your best guess.
Your final answer should be a number OR as few words as possible OR a comma separated list of numbers and/or strings.
If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise.
If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise.
If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string.
Question: {question}
""" # System prompt, adapted from the GAIA team's example at https://huggingface.co/spaces/gaia-benchmark/leaderboard
self._model = OpenaiChatModel(
model="llama-3.1-8b-instant",
api_key=os.environ["GROQ_API_KEY"],
base_url="https://api.groq.com/openai/v1/",
temperature=0.7,
max_tokens=512
)
print("CustomAgent initialized.")
def __call__(self, question: str) -> str:
"""
this is where the magic happens
"""
print(f"Agent received question (first 50 chars): {question[:50]}...")
@prompt_chain(
template=self._system_prompt,
model=self._model,
functions=[self.websearch],
)
def _magic(question: str) -> FunctionCall | str:
"""
An agent, using websearch to gather information and
answer a user's query.
"""
...
result = _magic(question)
print("Agent returning actual answer:", result)
return result
@staticmethod
def websearch(query: str) -> str:
"""
Performs a web search using an external web search API. Note that better results may be obtained from varying the wording of the query.
Args:
query (str): The search query string.
Returns:
str: The response from the web search API as a string.
"""
try:
r = requests.post(
f"{os.environ['WEBSEARCH_API_URL']}/tools/websearch",
json={"query": query}
)
r.raise_for_status()
response = r.text
except Exception as e:
print("Error:", e)
response = "An error has occurred while using websearch, please inform the user of this"
print(f"Agent searched for {query}. Response: {response}")
|