rabner

After updating to 2.24 LLM hangs after first response

1

**LocalAI version:** localai/localai:latest-aio-gpu-nvidia-cuda-12 (v2.24.0) **Environment, CPU architecture, OS, and Version:** Proxmox 8.3.1; Docker in unprivileged LXC; RTX 3090 driver: 565.57.01; Xeon E5v4; /models directory is persistent **Describe the bug** After...