litellm based custom llm provider

[staru09]

Related to #16.

I had tested it with openai api using this script

import os
from core.llm_provider import LiteLLMProvider
from dotenv import load_dotenv
load_dotenv()

provider = LiteLLMProvider()
out = provider.chat_completion(
    messages=[{"role":"user","content":"Say 'litellm-ok'"}],
    model="gpt-4o-mini",
    max_tokens=16
)
print(out)

and got this output

Litellm-ok!

please test it with other providers as well.

[Aaditya-G]

@staru09 kindly checkout requirements.txt, please rebase with main if possible, everything else lgtm.

[staru09]

Done @Aaditya-G

[Aaditya-G]

LGTM