File size: 3,652 Bytes
3c4c14d bf7e1be 3c4c14d ed65502 b6ca69f ed65502 3c4c14d ed65502 b6ca69f ed65502 3c4c14d b6ca69f 3c4c14d b6ca69f ed65502 3c4c14d b6ca69f ed65502 3c4c14d ed65502 b6ca69f 3c4c14d ed65502 3c4c14d a0c12a7 3c4c14d ed65502 3c4c14d 52b96f9 3c4c14d ed65502 3c4c14d ed65502 2c77ef8 3c4c14d ed65502 3c4c14d 2c77ef8 ed65502 3c4c14d ed65502 3c4c14d ed65502 |
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 |
import gradio as gr
import requests
from transformers import MarianMTModel, MarianTokenizer, AutoModelForCausalLM, AutoTokenizer
from PIL import Image
import torch
import io
import os
import base64
# Load Hugging Face API key from environment
HF_API_KEY = os.getenv("HF_API_KEY") # Add this in 'Variables and secrets' on HF Spaces
if not HF_API_KEY:
raise ValueError("HF_API_KEY is not set. Please add it in Hugging Face 'Variables and secrets'.")
# API endpoint for image generation
IMAGE_GEN_URL = "https://api-inference.huggingface.co/models/black-forest-labs/FLUX.1-schnell"
HEADERS = {"Authorization": f"Bearer {HF_API_KEY}"}
# Check if GPU is available
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# Load translation model: Tamil to English
translator_model = "Helsinki-NLP/opus-mt-mul-en"
translator = MarianMTModel.from_pretrained(translator_model).to(device)
translator_tokenizer = MarianTokenizer.from_pretrained(translator_model)
# Load text generation model
generator_model = "EleutherAI/gpt-neo-1.3B"
generator = AutoModelForCausalLM.from_pretrained(generator_model).to(device)
generator_tokenizer = AutoTokenizer.from_pretrained(generator_model)
if generator_tokenizer.pad_token is None:
generator_tokenizer.pad_token = generator_tokenizer.eos_token
generator.config.pad_token_id = generator_tokenizer.pad_token_id
def translate_tamil_to_english(text):
"""Translate Tamil text to English."""
inputs = translator_tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=128).to(device)
output = translator.generate(**inputs)
return translator_tokenizer.decode(output[0], skip_special_tokens=True)
def generate_text(prompt):
"""Generate creative text from English input."""
inputs = generator_tokenizer(prompt, return_tensors="pt", padding=True, truncation=True, max_length=128).to(device)
output = generator.generate(**inputs, max_length=100)
return generator_tokenizer.decode(output[0], skip_special_tokens=True)
def generate_image(prompt):
"""Generate image using Hugging Face inference API."""
response = requests.post(IMAGE_GEN_URL, headers=HEADERS, json={"inputs": prompt})
if response.status_code == 200:
# Check if raw image or base64 encoded
content_type = response.headers.get("content-type", "")
if "image" in content_type:
return Image.open(io.BytesIO(response.content))
else:
try:
result = response.json()
if "image" in result:
image_data = base64.b64decode(result["image"])
return Image.open(io.BytesIO(image_data))
except Exception as e:
print("Error parsing image:", e)
return Image.new("RGB", (300, 300), "red") # fallback placeholder
def process_input(tamil_text):
"""Pipeline: Translate → Generate Text → Generate Image"""
english_text = translate_tamil_to_english(tamil_text)
creative_text = generate_text(english_text)
image = generate_image(english_text)
return english_text, creative_text, image
# Create Gradio UI
interface = gr.Interface(
fn=process_input,
inputs=gr.Textbox(label="Enter Tamil Text"),
outputs=[
gr.Textbox(label="Translated English Text"),
gr.Textbox(label="Creative Text"),
gr.Image(label="Generated Image")
],
title="Tamil to English Translator & Image Generator",
description="Enter Tamil text. This app translates it to English, generates a creative description, and produces an image based on the translated text."
)
# Launch app
interface.launch()
|