Delete app.py

app.py

@@ -1,92 +0,0 @@
-import gradio as gr
-from gtts import gTTS
-import os
-from PIL import Image
-from transformers import pipeline, AutoModelForSeq2SeqLM, AutoTokenizer, BlipProcessor, BlipForConditionalGeneration
-
-# Define translation models per language pair
-translation_models = {
-    "French": "Helsinki-NLP/opus-mt-en-fr",
-    "Spanish": "Helsinki-NLP/opus-mt-en-es",
-    "Mandarin": "Helsinki-NLP/opus-mt-en-zh",
-    "Vietnamese": "Helsinki-NLP/opus-mt-en-vi",
-    "Tagalog": "Helsinki-NLP/opus-mt-en-tl"
-}
-
-# Directory to store local models
-local_model_dir = "./models"
-os.makedirs(local_model_dir, exist_ok=True)
-
-# Load translation models and tokenizers locally
-translation_pipelines = {}
-for lang, model_name in translation_models.items():
-    local_path = os.path.join(local_model_dir, model_name.replace('/', '_'))
-
-    if not os.path.exists(local_path):
-        model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
-        tokenizer = AutoTokenizer.from_pretrained(model_name)
-        model.save_pretrained(local_path)
-        tokenizer.save_pretrained(local_path)
-    else:
-        model = AutoModelForSeq2SeqLM.from_pretrained(local_path)
-        tokenizer = AutoTokenizer.from_pretrained(local_path)
-
-    translation_pipelines[lang] = pipeline("translation", model=model, tokenizer=tokenizer)
-
-# Load image captioning model
-caption_processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
-caption_model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
-
-def generate_caption(image):
-    inputs = caption_processor(image, return_tensors="pt")
-    output = caption_model.generate(**inputs)
-    caption = caption_processor.decode(output[0], skip_special_tokens=True)
-    return caption
-
-# Function to translate text using locally stored models
-def translate_text(sentence, target_lang):
-    if target_lang in translation_pipelines:
-        model = translation_pipelines[target_lang]
-        translation = model(sentence, max_length=100, truncation=True)
-        return translation[0]['translation_text']
-    return sentence  # Return original if no model found
-
-# Language options
-language_options = {
-    "English": "en", "French": "fr", "Spanish": "es", "Mandarin": "zh", "Vietnamese": "vi", "Tagalog": "tl"
-}
-
-# Main function for Gradio
-def process_image(image, l1, l2):
-    if not image:
-        return "Please upload an image.", None, None
-
-    caption_en = generate_caption(image)
-    caption_l2 = translate_text(caption_en, l2) if l2 != "English" else caption_en
-
-    # Generate TTS
-    tts = gTTS(text=caption_l2, lang=language_options[l2])
-    audio_path = "caption_audio.mp3"
-    tts.save(audio_path)
-
-    return caption_l2, audio_path, caption_en
-
-# Gradio Interface
-interface = gr.Interface(
-    fn=process_image,
-    inputs=[
-        gr.Image(type="pil", sources=["upload"]),  # Only allows upload and drag-and-drop
-        gr.Dropdown(list(language_options.keys()), label="Your Native Language (L1)"),
-        gr.Dropdown(list(language_options.keys()), label="Target Language (L2)")
-    ],
-    outputs=[
-        gr.Textbox(label="Generated Caption (L2)"),
-        gr.Audio(label="Listen to the Caption"),
-        gr.Textbox(label="Generated Caption (L1)")
-    ],
-    title="Image Caption Generator for Language Learning",
-    description="Upload an image, select your native language (L1) and target language (L2), and get a generated caption with text-to-speech audio."
-)
-
-# Launch Gradio
-interface.launch()