Files
ray-project--ray/doc/source/serve/doc_code/external_scaler_predictive.py
T
2026-07-13 13:17:40 +08:00

36 lines
756 B
Python

# __serve_example_begin__
import time
from ray import serve
from typing import Any
@serve.deployment(num_replicas=3)
class TextProcessor:
"""A simple text processing deployment that can be scaled externally."""
def __init__(self):
self.request_count = 0
def __call__(self, text: Any) -> dict:
# Simulate text processing work
time.sleep(0.1)
self.request_count += 1
return {
"request_count": self.request_count,
}
app = TextProcessor.bind()
# __serve_example_end__
def main():
import requests
serve.run(app)
# Test the deployment
resp = requests.post(
"http://localhost:8000/",
json="hello world"
)
print(f"Response: {resp.json()}")