asd

app.py

@@ -1,6 +1,5 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
-import spaces
 from sentence_splitter import SentenceSplitter
 
 device = "cuda"
@@ -8,46 +7,55 @@ device = "cuda"
 tokenizer = AutoTokenizer.from_pretrained("NoaiGPT/777")
 model = AutoModelForSeq2SeqLM.from_pretrained("NoaiGPT/777").to(device)
 
-# Initialize the sentence splitter
 splitter = SentenceSplitter(language='en')
 
-@spaces.GPU
-def generate_title(text):
-    input_ids = tokenizer(f'paraphraser: {text}', return_tensors="pt", padding="longest", truncation=True, max_length=64).input_ids.to(device)
-    outputs = model.generate(
-        input_ids,
-        num_beams=8,
-        num_beam_groups=4,
-        num_return_sequences=6,
-        repetition_penalty=12.0,
-        diversity_penalty=4.0,
-        no_repeat_ngram_size=3,
-        temperature=1.1,
-        top_k=50,
-        top_p=0.95,
-        max_length=64
-    )
-    return tokenizer.batch_decode(outputs, skip_special_tokens=True)
-
-def process_text(text):
+def process_and_generate(text):
+    def generate_title(sentence):
+        input_ids = tokenizer(f'paraphraser: {sentence}', return_tensors="pt", padding="longest", truncation=True, max_length=64).input_ids.to(device)
+        outputs = model.generate(
+            input_ids,
+            num_beams=8,
+            num_beam_groups=4,
+            num_return_sequences=6,
+            repetition_penalty=12.0,
+            diversity_penalty=4.0,
+            no_repeat_ngram_size=3,
+            temperature=1.1,
+            top_k=50,
+            top_p=0.95,
+            max_length=64
+        )
+        return tokenizer.batch_decode(outputs, skip_special_tokens=True)
+
     paragraphs = text.split('\n\n')
     results = []
+    final_paragraphs = []
+
     for paragraph in paragraphs:
         sentences = splitter.split(paragraph)
         paragraph_results = []
+        final_sentences = []
+
         for sentence in sentences:
             titles = generate_title(sentence)
             paragraph_results.append(f"Original: {sentence}\nParaphrases:\n" + "\n".join(titles))
+            final_sentences.append(titles[0])  # Use the first paraphrase for the final paragraph
+
         results.append("\n\n".join(paragraph_results))
-    return "\n\n---\n\n".join(results)
+        final_paragraphs.append(" ".join(final_sentences))
+
+    detailed_output = "\n\n---\n\n".join(results)
+    final_text = "\n\n".join(final_paragraphs)
 
-def gradio_generate_title(text):
-    return process_text(text)
+    return detailed_output, final_text
 
 iface = gr.Interface(
-    fn=gradio_generate_title,
+    fn=process_and_generate,
     inputs=gr.Textbox(lines=10, label="Input Text"),
-    outputs=gr.Textbox(lines=20, label="Generated Paraphrases"),
+    outputs=[
+        gr.Textbox(lines=20, label="Detailed Paraphrases"),
+        gr.Textbox(lines=10, label="Final Paraphrased Text")
+    ],
     title="Diverse Paraphrase Generator",
     description="Generate multiple diverse paraphrases for each sentence in the input text using NoaiGPT/777 model."
 )