src/genai/utils/models_loader.py

import os
import requests
import numpy as np
from langchain_groq import ChatGroq
from langchain_openai import ChatOpenAI
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_anthropic import ChatAnthropic
from langchain_openai import OpenAIEmbeddings
from huggingface_hub import login
from dotenv import load_dotenv
load_dotenv()
os.environ['HUGGINGFACEHUB_ACCESS_TOKEN']=os.getenv('HUGGINGFACEHUB_ACCESS_TOKEN')
login(os.environ['HUGGINGFACEHUB_ACCESS_TOKEN'])
os.environ['GROQ_API_KEY']=os.getenv('GROQ_API_KEY')

embedding_model = OpenAIEmbeddings(model="text-embedding-3-small", dimensions=1536)
llm_anthropic = ChatAnthropic(model='claude-3-5-sonnet-20241022', temperature=0.7, max_tokens=500)
llm_gemini = ChatGoogleGenerativeAI(model="gemini-1.5-flash")

llm = ChatGroq(
    model="llama-3.1-8b-instant",
    temperature=0.7,
)

llm_gpt = ChatOpenAI(
    model="gpt-4o-mini",
    temperature=0.3,
)


class HFEmbeddingAPI:
    def __init__(self, api_url, token):
        self.api_url = api_url
        self.headers = {
            "Authorization": f"Bearer {os.environ.get('HUGGINGFACEHUB_ACCESS_TOKEN')}",
        }

    def encode(self, texts):
        if isinstance(texts, str):
            texts = [texts]
        response = requests.post(
            self.api_url,
            headers=self.headers,
            json={"inputs": texts}
        )
        response.raise_for_status()
        embeddings=response.json()
        return np.array(embeddings[0]) if len(embeddings) == 1 else np.array(embeddings)

# Instantiate your API-backed "SentenceTransformer"
ST = HFEmbeddingAPI(
    api_url="https://router.huggingface.co/hf-inference/models/mixedbread-ai/mxbai-embed-large-v1/pipeline/feature-extraction",
    token=os.environ.get('HUGGINGFACEHUB_ACCESS_TOKEN')
)

improver_llm = llm_anthropic
ideator_llm = llm_anthropic
critic_llm = llm_anthropic
validator_llm = llm_anthropic