157 lines
5.3 KiB
Python
157 lines
5.3 KiB
Python
#
|
|
# Module: gemini_model
|
|
#
|
|
# This module implements the GeminiModel class for interacting with the Gemini API.
|
|
#
|
|
# Title: Gemini Model
|
|
# Summary: Gemini model implementation.
|
|
# Authors:
|
|
# - @TheBlockRhino
|
|
# Created: 2024-12-31
|
|
# Last edited by: @TheBlockRhino
|
|
# Last edited date: 2025-01-04
|
|
# URLs:
|
|
# - https://arai-ai.io
|
|
# - https://github.com/ARAI-DevHub/arai-ai-agents
|
|
# - https://x.com/TheBlockRhino
|
|
import os
|
|
import google.generativeai as genai
|
|
from .base_model import ModelInterface
|
|
from dotenv import load_dotenv
|
|
import yaml
|
|
import sys
|
|
|
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
|
|
|
load_dotenv()
|
|
|
|
class GeminiModel(ModelInterface):
|
|
"""Gemini model implementation.
|
|
|
|
Attributes:
|
|
model (str): The name of the Gemini model to use.
|
|
"""
|
|
|
|
def __init__(self, my_api_key=None, model_name="gemini-exp-1206"):
|
|
"""Initialize the Gemini model.
|
|
|
|
Args:
|
|
api_key (str): The API key to use for the Gemini model.
|
|
model_name (str): The name of the Gemini model to use.
|
|
|
|
Example:
|
|
>>> gemini_model = GeminiModel()
|
|
"""
|
|
if my_api_key:
|
|
genai.configure(api_key=my_api_key)
|
|
else:
|
|
genai.configure(api_key=os.environ.get('GOOGLE_GEMINI_API_KEY'))
|
|
self.model = genai.GenerativeModel(model_name)
|
|
|
|
# -------------------------------------------------------------------
|
|
# Helper to generate a response to a given prompt using the Gemini API.
|
|
# -------------------------------------------------------------------
|
|
def generate_response(self, prompt, **kwargs):
|
|
"""Generate a response to a given prompt using the Gemini API.
|
|
|
|
Args:
|
|
prompt (str): The prompt to generate a response to.
|
|
**kwargs: Additional keyword arguments.
|
|
|
|
Returns:
|
|
str: The generated response.
|
|
|
|
Example:
|
|
>>> gemini_model = GeminiModel()
|
|
>>> response = gemini_model.generate_response("What is the weather in Tokyo?")
|
|
"""
|
|
if isinstance(prompt, str):
|
|
return self.generate_response_from_string(prompt, **kwargs)
|
|
elif isinstance(prompt, list[dict]):
|
|
return self.generate_response_dictionary(prompt)
|
|
|
|
# -------------------------------------------------------------------
|
|
# Helper to generate a response to a given prompt using a list of dictionaries
|
|
# -------------------------------------------------------------------
|
|
def generate_response_dictionary(self, prompt: list[dict]) -> str:
|
|
"""Generate a response to a given prompt using a list of dictionaries.
|
|
|
|
Args:
|
|
prompt (list[dict]): The prompt to generate a response to.
|
|
|
|
Returns:
|
|
str: The generated response.
|
|
|
|
Example:
|
|
>>> gemini_model = GeminiModel()
|
|
>>> response = gemini_model.generate_response_dictionary([{"role": "user", "parts": "What is the weather in Tokyo?"}])
|
|
"""
|
|
try:
|
|
response = self.model.generate_content(prompt)
|
|
return response.text.strip()
|
|
except Exception as e:
|
|
return f"Error generating response: {str(e)}"
|
|
|
|
# -------------------------------------------------------------------
|
|
# Helper to generate a response to a given prompt using a string
|
|
# -------------------------------------------------------------------
|
|
def generate_response_from_string(self, prompt, **kwargs):
|
|
"""
|
|
Description:
|
|
Generate a response to a given prompt using a string.
|
|
|
|
Args:
|
|
prompt (str): The prompt to generate a response to.
|
|
**kwargs: Additional keyword arguments.
|
|
|
|
Returns:
|
|
str: The generated response.
|
|
|
|
Example:
|
|
>>> gemini_model = GeminiModel()
|
|
>>> response = gemini_model.generate_response_from_string("What is the weather in Tokyo?")
|
|
"""
|
|
# Extract personality and style from kwargs, or use defaults from agent_template
|
|
if kwargs:
|
|
if "personality" in kwargs:
|
|
personality = kwargs.get("personality")
|
|
if "communication_style" in kwargs:
|
|
communication_style = kwargs.get("communication_style")
|
|
else:
|
|
personality = ""
|
|
communication_style = ""
|
|
|
|
try:
|
|
# instructions being sent to the ai model
|
|
messages = []
|
|
|
|
# add personality and style to the instructions
|
|
if personality or communication_style:
|
|
persona_prompt = f"{personality} {communication_style}"
|
|
messages.append({
|
|
"role": "user",
|
|
"parts": [persona_prompt]
|
|
})
|
|
|
|
# user message
|
|
messages.append({
|
|
"role": "user",
|
|
"parts": [prompt]
|
|
})
|
|
|
|
# Make sure that what is being sent to the model is correct
|
|
# print(messages)
|
|
|
|
# generate the response
|
|
response = self.model.generate_content(messages)
|
|
return response.text.strip()
|
|
|
|
except Exception as e:
|
|
return f"Error generating response: {str(e)}"
|
|
|
|
if __name__ == "__main__":
|
|
gemini_model = GeminiModel()
|
|
response = gemini_model.generate_response("Tell me 10 one liners about crypto. put them as a json object")
|
|
print(response)
|
|
|