how to setup llamafile as a remote server?

[andyyeh75]

While I run llamafile on a remote server. It can be deployed normally. However, I cannot access it from a remote client in local network to http::8080. Any one can share the proper parameter to the executable for remote access? Thanks.

llama server listening at http://127.0.0.1:8080

opening browser tab... (pass --nobrowser to disable)
In the sandboxing block!
{"function":"server_cli","hostname":"127.0.0.1","level":"INFO","line":3213,"msg":"HTTP server listening","port":"8080","tid":"10733792","timestamp":1726048021}
{"function":"update_slots","level":"INFO","line":1659,"msg":"all slots are idle and system prompt is empty, clear the KV cache","tid":"10733792","timestamp":1726048021}

[Djip007]

may be try to add --host 0.0.0.0 and it may need to configure the firewall

[andyyeh75]

Thanks it works! BTW, I noticed that only CPU is utilized. How to ensure iGPU is used for acceleration? (I am using Intel Core Ultra)