Update app.py

app.py

@@ -1,75 +1,90 @@
-"""from transformers import AutoTokenizer, AutoModelForCausalLM
+import openai
 import gradio as gr
-import torch
 
-model_name = "tiiuae/falcon-7b-instruct"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    device_map="auto",
-    torch_dtype=torch.float16
-)
+# Set your OpenAI API key
+openai.api_key = "your_openai_api_key"
 
-def convert_to_spoken_hindi(formal_hindi_text):
-    prompt = f"Convert the following formal Hindi text into conversational spoken Hindi:\n\nFormal Hindi: {formal_hindi_text}\n\nSpoken Hindi:"
-    inputs = tokenizer(prompt, return_tensors="pt", padding=True, truncation=True)
-    outputs = model.generate(
-        inputs["input_ids"], 
-        attention_mask=inputs["attention_mask"], 
-        max_length=150, 
-        num_beams=5, 
-        temperature=0.7
-    )
-    spoken_hindi = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return spoken_hindi.split("Spoken Hindi:")[-1].strip()
+def translate_dutch_to_hindi(dutch_text, model="gpt-4-mini"):
+    """
+    Translates Dutch text to both conversational and formal Hindi using OpenAI.
+    
+    Args:
+    - dutch_text (str): The text in Dutch to be translated.
+    - model (str): The model to use for translation (e.g., gpt-3.5-turbo, gpt-4-mini).
+    
+    Returns:
+    - tuple: (Conversational Hindi Translation, Formal Hindi Translation)
+    """
+    try:
+        # Prompt for Conversational Hindi
+        conversational_prompt = (
+            f"Translate the following Dutch text to Hindi in a conversational style:\n\n"
+            f"Dutch: {dutch_text}\n"
+            f"Hindi:"
+        )
 
-iface = gr.Interface(
-    fn=convert_to_spoken_hindi,
-    inputs="text",
-    outputs="text",
-    title="Hindi Text Converter"
-)
+        # Prompt for Formal Hindi
+        formal_prompt = (
+            f"Translate the following Dutch text to Hindi in a formal style:\n\n"
+            f"Dutch: {dutch_text}\n"
+            f"Hindi:"
+        )
 
-iface.launch()
-from transformers import pipeline, MBartTokenizer
-import gradio as gr
+        # OpenAI API call for Conversational Hindi
+        conversational_response = openai.ChatCompletion.create(
+            model=model,
+            messages=[
+                {"role": "system", "content": "You are a professional translator."},
+                {"role": "user", "content": conversational_prompt}
+            ],
+            temperature=0.5
+        )
+        conversational_hindi = conversational_response['choices'][0]['message']['content'].strip()
+
+        # OpenAI API call for Formal Hindi
+        formal_response = openai.ChatCompletion.create(
+            model=model,
+            messages=[
+                {"role": "system", "content": "You are a professional translator."},
+                {"role": "user", "content": formal_prompt}
+            ],
+            temperature=0.5
+        )
+        formal_hindi = formal_response['choices'][0]['message']['content'].strip()
 
-# Load the pre-trained mBART model and tokenizer
-model_name = "facebook/mbart-large-50-many-to-one-mmt"
-tokenizer = MBartTokenizer.from_pretrained(model_name)
-model = pipeline("text2text-generation", model=model_name, tokenizer=tokenizer)
+        return conversational_hindi, formal_hindi
 
-def convert_to_casual_hindi(text):
-    # Use the model to transform the text dynamically
-    transformed_text = model(text)
-    return transformed_text[0]['generated_text']
+    except openai.error.OpenAIError as e:
+        error_message = f"An error occurred: {e}"
+        return error_message, error_message
 
-# Gradio interface for deployment in Hugging Face Spaces
-iface = gr.Interface(fn=convert_to_casual_hindi, inputs="text", outputs="text", title="Formal to Casual Hindi Converter")
-iface.launch()
-"""
-from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+# Define the Gradio interface
+def gradio_interface(dutch_text):
+    conversational_hindi, formal_hindi = translate_dutch_to_hindi(dutch_text)
+    return conversational_hindi, formal_hindi
 
-# Use the IndicTrans model
-model_name = "ai4bharat/indictrans-hin-eng"  # IndicTrans for Hindi
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
+# Create the Gradio app
+with gr.Blocks() as app:
+    gr.Markdown("## Dutch to Hindi Translator")
+    gr.Markdown("Enter Dutch text below and click 'Translate' to see both Conversational and Formal Hindi translations.")
 
-def formal_to_casual_hindi(input_text):
-    # Prepare input for IndicTrans
-    prompt = f"Convert formal Hindi to casual Hindi: {input_text}"
-    input_ids = tokenizer.encode(prompt, return_tensors="pt")
-    outputs = model.generate(input_ids, max_length=128, num_beams=5, early_stopping=True)
-    casual_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return casual_text
+    # Input text box for Dutch text
+    input_text = gr.Textbox(label="Enter Dutch Text", placeholder="Type Dutch text here...")
 
-# Gradio interface for the deployment
-iface = gr.Interface(
-    fn=formal_to_casual_hindi,
-    inputs="text",
-    outputs="text",
-    title="Formal to Casual Hindi Converter",
-    description="Convert formal Hindi text into conversational Hindi using AI."
-)
+    # Output text boxes for Conversational and Formal Hindi
+    output_conversational = gr.Textbox(label="Conversational Hindi", placeholder="Conversational Hindi translation will appear here...")
+    output_formal = gr.Textbox(label="Formal Hindi", placeholder="Formal Hindi translation will appear here...")
+
+    # Translate button
+    translate_button = gr.Button("Translate")
+
+    # Button click event
+    translate_button.click(
+        fn=gradio_interface, 
+        inputs=[input_text], 
+        outputs=[output_conversational, output_formal]
+    )
 
-iface.launch()
+# Launch the app
+if __name__ == "__main__":
+    app.launch()