#!/bin/bash # set path and first lightx2v_path= model_path= export CUDA_VISIBLE_DEVICES=0 # set environment variables source ${lightx2v_path}/scripts/base/base.sh # Start API server with distributed inference service python -m lightx2v.server \ --model_cls hunyuan_video_1.5_distill \ --task t2v \ --model_path $model_path \ --config_json ${lightx2v_path}/configs/hunyuan_video_15/hunyuan_video_t2v_480p_distill.json \ --port 8000 echo "Service stopped"