Create app.py

app.py

@@ -0,0 +1,60 @@
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+import fitz  # PyMuPDF
+from docx import Document
+
+# Load model and tokenizer
+model_name = "microsoft/phi-2"
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True, torch_dtype=torch.float16)
+
+def extract_text_from_pdf(file):
+    doc = fitz.open(stream=file.read(), filetype="pdf")
+    text = ""
+    for page in doc:
+        text += page.get_text()
+    return text
+
+def extract_text_from_docx(file):
+    doc = Document(file)
+    return "\n".join([paragraph.text for paragraph in doc.paragraphs])
+
+def convert_to_story(file):
+    if file is None:
+        return "Please upload a file."
+    
+    file_extension = file.name.split('.')[-1].lower()
+    
+    if file_extension == 'pdf':
+        text = extract_text_from_pdf(file)
+    elif file_extension == 'docx':
+        text = extract_text_from_docx(file)
+    else:
+        return "Unsupported file format. Please upload a PDF or DOCX file."
+
+    prompt = f"Convert the following news article into a short children's story (maximum 200 words):\n\n{text}\n\nChildren's story:"
+    
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=1024)
+    
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_new_tokens=200,
+            temperature=0.7,
+            top_p=0.95,
+            do_sample=True
+        )
+    
+    story = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return story.split("Children's story:")[-1].strip()
+
+iface = gr.Interface(
+    fn=convert_to_story,
+    inputs=gr.File(label="Upload PDF or DOCX file"),
+    outputs="text",
+    title="News to Children's Story Converter",
+    description="Upload a news article in PDF or DOCX format to convert it into a short children's story."
+)
+
+iface.launch()