Update Gradio configuration to expose API endpoint

app.py

@@ -18,19 +18,33 @@ class GenerationRequest(BaseModel):
     early_stopping: bool = True
     no_repeat_ngram_size: int = 3
 
-@app.post("/generate")
-async def generate_text(request: GenerationRequest):
-    inputs = tokenizer(request.prompt, return_tensors="pt").to(device)
+def generate(prompt: str, max_length: int = 2048, num_beams: int = 3, early_stopping: bool = True, no_repeat_ngram_size: int = 3):
+    inputs = tokenizer(prompt, return_tensors="pt").to(device)
     outputs = model.generate(
         **inputs,
-        max_length=request.max_length,
-        num_beams=request.num_beams,
-        early_stopping=request.early_stopping,
-        no_repeat_ngram_size=request.no_repeat_ngram_size,
+        max_length=max_length,
+        num_beams=num_beams,
+        early_stopping=early_stopping,
+        no_repeat_ngram_size=no_repeat_ngram_size,
         eos_token_id=tokenizer.eos_token_id,
         pad_token_id=tokenizer.pad_token_id,
     )
     output_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return output_text
+
+iface = gr.Interface(
+    fn=generate,
+    inputs=gr.Textbox(lines=10, label="Input Prompt"),
+    outputs=gr.Textbox(label="Generated Output"),
+    title="LLaMA 7B Server",
+    description="A web interface for interacting with the LLaMA 7B model.",
+    allow_flagging="never",
+    api_open=True
+)
+
+@app.post("/generate")
+async def generate_text(request: GenerationRequest):
+    return {"generated_text": generate(**request.dict())}
     return {"generated_text": output_text}
 
 if __name__ == "__main__":