| import streamlit as st | |
| from transformers import pipeline | |
| from huggingface_hub import login | |
| from PIL import Image | |
| import os | |
| login(token=os.getenv("HUGGINGFACE_TOKEN")) | |
| st.header("Character Captions (IN PROGRESS!)") | |
| st.write("Have a character caption any image you upload!") | |
| character = st.selectbox("Choose a character", ["rapper", "monkey", "shrek", "unintelligible"]) | |
| uploaded_img = st.file_uploader("Upload an image") | |
| if uploaded_img is not None: | |
| image = Image.open(uploaded_img) | |
| st.image(image) | |
| image_captioner = pipeline("image-to-text", model="Salesforce/blip-image-captioning-large") | |
| response = image_captioner(image) | |
| caption = response[0]['generated_text'] | |
| st.write("Caption:", caption) | |
| character_prompts = { | |
| "rapper": f"Describe this scene like you're a rapper: {caption}.", | |
| "monkey": f"Describe this scene like you're a monkey going bananas: {caption}.", | |
| "shrek": f"Describe this scene like you're Shrek: {caption}.", | |
| "unintelligible": f"Describe this scene in a way that makes no sense: {caption}." | |
| } | |
| prompt = character_prompts[character] | |
| st.write(prompt) | |
| personality = "rapper" | |
| prompt = character_prompts[personality] | |
| text_generator = pipeline("text-generation", model="meta-llama/Llama-2-7b-hf", framework="pt") | |
| prompt = character_prompts[character] | |
| st.write("Styled Prompt:", prompt) | |
| generated_text = text_generator(prompt, max_length=50, do_sample=True) | |
| styled_caption = generated_text[0]['generated_text'] | |
| st.write("Character-Styled Caption:", styled_caption) |