Spaces:
Running
Running
File size: 2,160 Bytes
ed4d993 |
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 71 72 |
import os
from git import Repo
from langchain_community.document_loaders.generic import GenericLoader
from langchain_community.document_loaders.parsers import LanguageParser
from langchain_community.embeddings import GPT4AllEmbeddings
from langchain_community.llms.fireworks import Fireworks
from langchain_community.vectorstores import Chroma
from langchain_core.output_parsers import StrOutputParser
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.pydantic_v1 import BaseModel
from langchain_core.runnables import RunnableParallel, RunnablePassthrough
from langchain_text_splitters import Language, RecursiveCharacterTextSplitter
# Check API key
if os.environ.get("FIREWORKS_API_KEY", None) is None:
raise Exception("Missing `FIREWORKS_API_KEY` environment variable.")
# Load codebase
# Set local path
repo_path = "/Users/rlm/Desktop/tmp_repo"
# Use LangChain as an example
repo = Repo.clone_from("https://github.com/langchain-ai/langchain", to_path=repo_path)
loader = GenericLoader.from_filesystem(
repo_path + "/libs/langchain/langchain",
glob="**/*",
suffixes=[".py"],
parser=LanguageParser(language=Language.PYTHON, parser_threshold=500),
)
documents = loader.load()
# Split
python_splitter = RecursiveCharacterTextSplitter.from_language(
language=Language.PYTHON, chunk_size=2000, chunk_overlap=200
)
texts = python_splitter.split_documents(documents)
# Add to vectorDB
vectorstore = Chroma.from_documents(
documents=texts,
collection_name="codebase-rag",
embedding=GPT4AllEmbeddings(),
)
retriever = vectorstore.as_retriever()
# RAG prompt
template = """Answer the question based only on the following context:
{context}
Question: {question}
"""
prompt = ChatPromptTemplate.from_template(template)
# Initialize a Fireworks model
model = Fireworks(model="accounts/fireworks/models/llama-v2-34b-code-instruct")
# RAG chain
chain = (
RunnableParallel({"context": retriever, "question": RunnablePassthrough()})
| prompt
| model
| StrOutputParser()
)
# Add typing for input
class Question(BaseModel):
__root__: str
chain = chain.with_types(input_type=Question)
|