Update app.py

app.py

@@ -1,30 +1,37 @@
-from transformers import MBartForConditionalGeneration, MBart50Tokenizer, pipeline
+import openai
+from transformers import MBartForConditionalGeneration, MBart50Tokenizer
 import gradio as gr
 import requests
 import io
 from PIL import Image
 import os
-import torch
+
+# Set up your OpenAI API key (make sure it's stored as an environment variable)
+openai.api_key = os.getenv("OPENAI_API_KEY")
 
 # Load the translation model and tokenizer
 model_name = "facebook/mbart-large-50-many-to-one-mmt"
 tokenizer = MBart50Tokenizer.from_pretrained(model_name)
 model = MBartForConditionalGeneration.from_pretrained(model_name)
 
-# Use a more powerful text generation model, e.g., GPT-J-6B
-text_gen_model = "EleutherAI/gpt-j-6B"  # Or use 'EleutherAI/gpt-neox-20b' for better results
-pipe = pipeline(
-    "text-generation",
-    model=text_gen_model,
-    torch_dtype=torch.float32,
-    device_map="auto"
-)
-
 # Use the Hugging Face API key from environment variables for text-to-image model
 API_URL = "https://api-inference.huggingface.co/models/ZB-Tech/Text-to-Image"
 headers = {"Authorization": f"Bearer {os.getenv('full_token')}"}
 
-# Define the translation, text generation, and image generation function
+# Define the OpenAI GPT-3 text generation function
+def generate_with_gpt3(prompt, max_tokens=150, temperature=0.7):
+    response = openai.Completion.create(
+        engine="text-davinci-003",  # You can also use "text-davinci-002" or "curie"
+        prompt=prompt,
+        max_tokens=max_tokens,
+        temperature=temperature,
+        top_p=0.9,
+        frequency_penalty=0.0,
+        presence_penalty=0.0
+    )
+    return response.choices[0].text.strip()
+
+# Define the translation, GPT-3 text generation, and image generation function
 def translate_and_generate_image(tamil_text):
     # Step 1: Translate Tamil text to English using mbart-large-50
     tokenizer.src_lang = "ta_IN"
@@ -32,10 +39,10 @@ def translate_and_generate_image(tamil_text):
     translated_tokens = model.generate(**inputs, forced_bos_token_id=tokenizer.lang_code_to_id["en_XX"])
     translated_text = tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
 
-    # Step 2: Generate high-quality English text using GPT-J
-    prompt = f"Create a detailed description based on the following text: {translated_text}"
-    generated_text = pipe(prompt, max_length=150, temperature=0.7, top_p=0.9, top_k=50, truncation=True)[0]['generated_text']
-    
+    # Step 2: Generate high-quality descriptive text using OpenAI's GPT-3
+    prompt = f"Create a detailed and creative description based on the following text: {translated_text}"
+    generated_text = generate_with_gpt3(prompt, max_tokens=150, temperature=0.7)
+
     # Step 3: Use the generated English text to create an image
     def query(payload):
         response = requests.post(API_URL, headers=headers, json=payload)
@@ -54,9 +61,9 @@ iface = gr.Interface(
     outputs=[gr.Textbox(label="Translated English Text"), 
              gr.Textbox(label="Generated Descriptive Text"),
              gr.Image(label="Generated Image")],
-    title="Tamil to English Translation, Text Generation, and Image Creation",
-    description="Translate Tamil text to English using Facebook's mbart-large-50 model, generate high-quality text using GPT-J, and create an image using the generated text.",
+    title="Tamil to English Translation, GPT-3 Text Generation, and Image Creation",
+    description="Translate Tamil text to English using Facebook's mbart-large-50 model, generate high-quality text using GPT-3, and create an image using the generated text.",
 )
 
-# Launch Gradio app
+# Launch Gradio app without `share=True`
 iface.launch()