How can I change the inference_address?

pytorch / serve

Serve, optimize and scale PyTorch models in production

https://pytorch.org/serve/

Apache License 2.0

4.23k stars 864 forks source link

How can I change the inference_address? #1911

Closed Hanyu-Jin closed 2 years ago

Hanyu-Jin commented 2 years ago

📚 The doc issue

I run the steps in Getting Started. When I run the torchserve --start --ncs --model-store model_store --models densenet161.mar, I find that http://127.0.0.1:8080 is already in use and I cannot kill the process that uses the port. I tried to change 8080 to 8085 in all files but it didn't work. The command torchserve still tried to bing to 127.0.0.1:8080. How can I change the inference_address?

Suggest a potential alternative/fix

No response

harshbafna commented 2 years ago

@Hanyu-Jin : The default inference port for TorchServe is 8080. Please refer the advanced configuration on how to configure the listening address and port for TorchServe