Spaces:
Sleeping
Sleeping
from transformers import pipeline | |
from dotenv import find_dotenv, load_dotenv | |
from langchain import PromptTemplate, LLMChain, HuggingFaceHub | |
import streamlit as st | |
import requests | |
import os | |
load_dotenv(find_dotenv()) | |
HUGGINGFACE_API = os.getenv("HUGGINGFACE_API") | |
def image2text(url): | |
image_to_text = pipeline('image-to-text', model='Salesforce/blip-image-captioning-large') | |
text = image_to_text(url)[0]['generated_text'] | |
print(text) | |
return text | |
def generate_story(scenario, length): | |
template = """ | |
You are story teller, generate a short story in {length} words\n | |
CONTEXT:{scenario}\n | |
STORY: | |
""" | |
prompt = PromptTemplate(template=template, input_variables=["scenario","length"]) | |
llm = LLMChain(llm=HuggingFaceHub(huggingfacehub_api_token=HUGGINGFACE_API, repo_id="mistralai/Mixtral-8x7B-Instruct-v0.1"), prompt=prompt, verbose=True) | |
story = llm.predict(scenario=scenario, length=length) | |
print(story) | |
return story | |
# def text2speech(message): | |
# API_URL = "https://api-inference.huggingface.co/models/microsoft/speecht5_tts" | |
# headers = {"Authorization": f"Bearer {HUGGINGFACE_API}"} | |
# payloads = { | |
# "inputs": message | |
# } | |
# response = requests.post(API_URL,headers=headers,json=payloads) | |
# with open('audio.wav', 'wb') as file: | |
# file.write(response.content) | |
def main(): | |
st.set_page_config(page_title="Image Storyteller") | |
st.header("Image to Audio") | |
uploaded_file = st.file_uploader("Choose an Image", type="jpg") | |
length = st.number_input("Length") | |
scenario = "" | |
successful_processing = False | |
if uploaded_file is not None: | |
print(uploaded_file) | |
bytes_data = uploaded_file.getvalue() | |
with open(uploaded_file.name, "wb") as file: | |
file.write(bytes_data) | |
st.image(uploaded_file.name, caption="Uploaded Image", use_column_width=True) | |
try: | |
scenario = image2text(uploaded_file.name) | |
successful_processing = True | |
except Exception as e: | |
st.error(f"Error processing the image: {e}") | |
if successful_processing: | |
story = generate_story(scenario, length) | |
# text2speech(story) | |
with st.expander("scenario"): | |
st.write(scenario) | |
with st.expander("generated story"): | |
st.write(story) | |
# st.audio('audio.wav') | |
if __name__ == '__main__': | |
main() | |