Update app.py

app.py

@@ -3,14 +3,13 @@ from transformers import AutoTokenizer, AutoModel
 from sklearn.metrics.pairwise import cosine_similarity
 import torch
 import numpy as np
-import json
-import requests
+from gradio_client import Client
 from functools import lru_cache
 
 # Cache the model and tokenizer using lru_cache
 @lru_cache(maxsize=1)
 def load_model_and_tokenizer():
-    model_name = "./all-MiniLM-L6-v2"  # Adjust if needed
+    model_name = "./all-MiniLM-L6-v2"  # Replace with your Space and model path
     tokenizer = AutoTokenizer.from_pretrained(model_name)
     model = AutoModel.from_pretrained(model_name)
     return tokenizer, model
@@ -20,82 +19,19 @@ tokenizer, model = load_model_and_tokenizer()
 
 # Precompute label embeddings
 labels = [
-    "aerospace", "anatomy", "anthropology", "art", "automotive", "blockchain", 
-    "biology", "chemistry", "cryptocurrency", "data science", "design", "e-commerce",
-    "education", "engineering", "entertainment", "environment", "fashion", "finance", 
-    "food commerce", "gaming", "healthcare", "history", "information technology", 
-    "legal", "machine learning", "marketing", "medicine", "music", "philosophy", 
-    "physics", "politics", "real estate", "retail", "robotics", "social media", 
-    "sports", "technical", "tourism", "travel"
+    "aerospace", "anatomy", "anthropology", "art", 
+    "automotive", "blockchain", "biology", "chemistry", 
+    "cryptocurrency", "data science", "design", "e-commerce",
+    "education", "engineering", "entertainment", "environment",
+    "fashion", "finance", "food commerce", "gaming",
+    "healthcare", "history", "information technology", 
+    "legal", "machine learning", "marketing", "medicine",
+    "music", "philosophy", "physics", "politics", "real estate", "retail", 
+    "robotics", "social media", "sports", "technical",
+    "tourism", "travel"
 ]
 
 tones = [
     "formal", "positive", "negative", "poetic", "polite", "subtle", "casual", "neutral",
-    "informal", "pompous", "sustained", "rude"
-]
-
-# Compute label embeddings
-def get_label_embeddings():
-    with torch.no_grad():
-        tokenized = tokenizer(labels, padding=True, truncation=True, return_tensors="pt")
-        label_embeddings = model(**tokenized).last_hidden_state[:, 0, :].numpy()
-    return label_embeddings
-
-label_embeddings = get_label_embeddings()
-
-def detect_context(text: str):
-    # Encode input text
-    tokenized = tokenizer(text, padding=True, truncation=True, return_tensors="pt")
-    with torch.no_grad():
-        text_embedding = model(**tokenized).last_hidden_state[:, 0, :].numpy()
-
-    # Compute similarity scores
-    similarities = cosine_similarity(text_embedding, label_embeddings)[0]
-    
-    # Get best matching context
-    best_index = np.argmax(similarities)
-    detected_context = labels[best_index]
-
-    return detected_context
-
-def process_and_translate(text: str):
-    detected_context = detect_context(text)
-    try:
-        print(f"Sending to space_7: {text}")  # Debugging
-
-        translation_response = requests.post(
-            "https://api.gradio.app/v2/Frenchizer/space_18/predict",
-            json={"data": [text]}  # Make sure this is correctly formatted
-        )
-
-        print("Raw response from space_7:", translation_response.text)  # Debugging
-
-        if translation_response.status_code != 200 or not translation_response.text.strip():
-            return json.dumps({
-                "error": "space_7 returned an empty response",
-                "context": detected_context
-            })
-
-        response_json = translation_response.json()
-        return json.dumps({
-            "context": detected_context,
-            "translation": response_json.get("data", [""])[0]
-        })
-
-    except Exception as e:
-        return json.dumps({
-            "error": f"Exception: {str(e)}",
-            "context": detected_context
-        })
-
-
-
-# Define Gradio interface
-with gr.Blocks() as interface:
-    input_text = gr.Textbox(label="Input Text")
-    output_json = gr.JSON(label="Context & Translation")
-    process_button = gr.Button("Process & Translate")
-    process_button.click(fn=process_and_translate, inputs=[input_text], outputs=[output_json])
-
-if __name__ == "__main__":
-    interface.launch()
+    "informal", "pompous", "sustained", "rude", "sustained",
+interface.launch()