Update app.py

app.py

@@ -1,45 +1,3 @@
-import gradio as gr
-from transformers import AutoModelForCausalLM, AutoTokenizer
-from peft import PeftModel, PeftConfig
-
-# Base model and adapter model names
-base_model_name = "unsloth/Qwen2.5-7B-Instruct-bnb-4bit"
-adapter_model_name = "djmax13/qween7.5-arabic-story-teller-bnb-4bit"
-
-# Load base model
-base_model = AutoModelForCausalLM.from_pretrained(base_model_name)
-tokenizer = AutoTokenizer.from_pretrained(base_model_name)
-
-# Load LoRA configuration
-config = PeftConfig.from_pretrained(adapter_model_name)
-
-# Load LoRA adapter and merge it with the base model
-model = PeftModel.from_pretrained(base_model, adapter_model_name)
-model = model.merge_and_unload() # Optional: Merge adapter weights into base model for potential speedup
-
-def generate_text(prompt):
-    input_ids = tokenizer.encode(prompt, return_tensors="pt").to(model.device) # Move input to model's device
-
-    # Generate text (you might need to adjust generation parameters)
-    output = model.generate(
-        input_ids,
-        max_length=150,  # Adjust as needed
-        num_return_sequences=1,
-        no_repeat_ngram_size=2,
-        top_k=50,
-        top_p=0.95,
-        temperature=0.7,
-    )
-
-    generated_text = tokenizer.decode(output[0], skip_special_tokens=True)
-    return generated_text
-
-iface = gr.Interface(
-    fn=generate_text,
-    inputs=gr.Textbox(lines=5, placeholder="Enter your story prompt here..."),
-    outputs=gr.Textbox(),
-    title="Arabic Story Teller",
-    description="A Qwen2.5-7B model finetuned for Arabic story generation using LoRA.",
-)
-
-iface.launch()
+# Load model directly
+from transformers import AutoModel
+model = AutoModel.from_pretrained("GhadyIbra250/qwen_arabic_stories")