Update app.py

app.py

@@ -1,79 +1,43 @@
-
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 import nltk
-import torch
-from textblob import TextBlob
-
-nltk.download("punkt")
-
-model = AutoModelForSeq2SeqLM.from_pretrained("ramsrigouthamg/t5_paraphraser")
-tokenizer = AutoTokenizer.from_pretrained("ramsrigouthamg/t5_paraphraser", use_fast=False)
-
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-model = model.to(device)
-
-def split_into_sentences(text):
-    return nltk.tokenize.sent_tokenize(text)
+nltk.download('punkt')  # Download necessary data
 
-def paraphrase_sentence(sentence, creativity):
-    text = "paraphrase: " + sentence + " </s>"
-    encoding = tokenizer.encode_plus(text, padding="max_length", return_tensors="pt", max_length=256, truncation=True)
-    input_ids, attention_mask = encoding["input_ids"].to(device), encoding["attention_mask"].to(device)
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 
-    output = model.generate(
+# Load model and tokenizer
+model_name = "Vamsi/T5_Paraphrase_Paws"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
+
+def paraphrase(text):
+    if not text.strip():
+        return "Please enter some text to paraphrase."
+    input_text = f"paraphrase: {text} </s>"
+    encoding = tokenizer.encode_plus(
+        input_text,
+        max_length=256,
+        padding="max_length",
+        return_tensors="pt",
+        truncation=True
+    )
+    input_ids, attention_mask = encoding["input_ids"], encoding["attention_mask"]
+    outputs = model.generate(
         input_ids=input_ids,
         attention_mask=attention_mask,
         max_length=256,
-        do_sample=True,
-        top_k=120,
-        top_p=creativity,
-        early_stopping=True,
-        num_return_sequences=1
-    )
-    paraphrased = tokenizer.decode(output[0], skip_special_tokens=True, clean_up_tokenization_spaces=True)
-    return paraphrased
-
-def correct_grammar(text):
-    blob = TextBlob(text)
-    return str(blob.correct())
-
-def paraphrase_text(text, creativity=0.9, tone="neutral", improve_grammar=True, batch_size=5):
-    sentences = split_into_sentences(text)
-    results = []
-
-    for sentence in sentences:
-        para = paraphrase_sentence(sentence, creativity)
-        if improve_grammar:
-            para = correct_grammar(para)
-        results.append(para)
-
-    return " ".join(results)
-
-with gr.Blocks() as demo:
-    gr.Markdown("<h1>Paraphrasing Tool</h1><p>AI-powered rewriting with grammar improvement and creativity controls.</p>")
-
-    with gr.Row():
-        input_text = gr.Textbox(lines=8, label="Enter Text")
-        output_text = gr.Textbox(lines=8, label="Paraphrased Output")
-
-    creativity = gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Creativity (top_p)")
-    improve_grammar = gr.Checkbox(value=True, label="Improve Grammar")
-    tone = gr.Radio(["neutral", "formal", "casual"], label="Tone (placeholder)", value="neutral")
-    batch_size = gr.Slider(1, 10, value=5, step=1, label="Batch Size")
-
-    run_button = gr.Button("Paraphrase")
-
-    run_button.click(
-        paraphrase_text,
-        inputs=[input_text, creativity, tone, improve_grammar, batch_size],
-        outputs=output_text
-    )
-
-if __name__ == "__main__":
-    demo.launch(
-        server_name="0.0.0.0",
-        server_port=7860,
-        show_api=True,
-        favicon_path="favicon.ico"
+        num_beams=5,
+        num_return_sequences=1,
+        temperature=1.5
     )
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+# Gradio Interface
+demo = gr.Interface(
+    fn=paraphrase,
+    inputs=gr.Textbox(label="Enter your text to paraphrase"),
+    outputs=gr.Textbox(label="Paraphrased text"),
+    title="AI Paraphrasing Tool",
+    description="Enter your sentence or paragraph, and the model will return a paraphrased version."
+)
+
+demo.launch()