File size: 3,120 Bytes
673d163 0df3389 63b5e64 0df3389 673d163 0df3389 673d163 0df3389 673d163 0df3389 8c9f460 0df3389 8c9f460 0df3389 8c9f460 0df3389 673d163 3d30833 673d163 0df3389 673d163 |
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 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
from langchain import PromptTemplate, LLMChain
from langchain.llms import CTransformers
import os
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.vectorstores import Chroma
from langchain.chains import RetrievalQA
from langchain.embeddings import HuggingFaceBgeEmbeddings
from io import BytesIO
from langchain.document_loaders import PyPDFLoader
import gradio as gr
local_llm = "final_model_maybe_gguf-unsloth.Q5_K_M.gguf"
config = {
'max_new_tokens': 2048,
'repetition_penalty': 1.1,
'temperature': 0.6,
'top_k': 50,
'top_p': 0.9,
'stream': True,
'threads': int(os.cpu_count() / 2)
}
llm = CTransformers(
model=local_llm,
model_type="mistral",
lib="avx2", #for CPU use
**config
)
print("LLM Initialized...")
prompt_template = """Below is an instruction that describes a task. Write a response that appropriately completes the request.
μ μλ μ§λ¬Έμ λν΄μ context λ΄μ©μΌλ‘ λ΅λ³ν΄μ€.
### Context :
{context}
### Instruction:
{question}
### Response:
"""
model_name = "jhgan/ko-sroberta-multitask"
model_kwargs = {'device': 'cpu'}
encode_kwargs = {'normalize_embeddings': False}
embeddings = HuggingFaceBgeEmbeddings(
model_name=model_name,
model_kwargs=model_kwargs,
encode_kwargs=encode_kwargs
)
prompt = PromptTemplate(template=prompt_template, input_variables=['context', 'question'])
load_vector_store = Chroma(persist_directory="stores/pet_cosine", embedding_function=embeddings)
retriever = load_vector_store.as_retriever(search_kwargs={"k":1})
# query = "what is the fastest speed for a greyhound dog?"
# semantic_search = retriever.get_relevant_documents(query)
# print(semantic_search)
print("######################################################################")
chain_type_kwargs = {"prompt": prompt}
# qa = RetrievalQA.from_chain_type(
# llm=llm,
# chain_type="stuff",
# retriever=retriever,
# return_source_documents = True,
# chain_type_kwargs= chain_type_kwargs,
# verbose=True
# )
# response = qa(query)
# print(response)
sample_prompts = ["what is the fastest speed for a greyhound dog?", "Why should we not feed chocolates to the dogs?", "Name two factors which might contribute to why some dogs might get scared?"]
def get_response(input):
query = input
chain_type_kwargs = {"prompt": prompt}
qa = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=retriever, return_source_documents=True, chain_type_kwargs=chain_type_kwargs, verbose=True)
response = qa(query)
return response
input = gr.Text(
label="Prompt",
show_label=False,
max_lines=1,
placeholder="Enter your prompt",
container=False,
)
iface = gr.Interface(fn=get_response,
inputs=input,
outputs="text",
title="My Dog PetCare Bot",
description="This is a RAG implementation based on Zephyr 7B Beta LLM.",
examples=sample_prompts,
allow_screenshot=False,
allow_flagging=False
)
iface.launch() |