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_key = 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_key, 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 Story") | |
| length = st.number_input("Length") | |
| if not length: | |
| length = 10 | |
| uploaded_file = st.file_uploader("Choose an Image", type="jpg") | |
| 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() | |