NIBittensor
This page covers how to use the BittensorLLM inference runtime within LangChain. It is broken into two parts: installation and setup, and then examples of NIBittensorLLM usage.
Installation and Setupβ
- Install the Python package with
pip install langchain
Wrappersβ
LLMβ
There exists a NIBittensor LLM wrapper, which you can access with:
from langchain.llms import NIBittensorLLM
It provides a unified interface for all models:
llm = NIBittensorLLM(system_prompt="Your task is to provide concise and accurate response based on user prompt")
print(llm('Write a fibonacci function in python with golder ratio'))
Multiple responses from top miners can be accessible using the top_responses
parameter:
multi_response_llm = NIBittensorLLM(top_responses=10)
multi_resp = multi_response_llm("What is Neural Network Feeding Mechanism?")
json_multi_resp = json.loads(multi_resp)
print(json_multi_resp)