--deepspeed_multinode_launcher: 未找到命令

按照官方教程进行训练，启动训练脚本时提示--deepspeed_multinode_launcher: 未找到命令
以下时启动脚本配置：
num_machines=4
num_processes=$((num_machines * 8))
machine_rank=0

accelerate launch
--config_file ./configs/sft.yaml
--num_processes $num_processes
--num_machines $num_machines
--machine_rank $machine_rank \
--deepspeed_multinode_launcher standard finetune_moss.py
--model_name_or_path /root/liuliu/moss/moss-moon-003-sft-plugin
--data_dir ./sft_data
--output_dir ./ckpts/moss-moon-003-sft
--log_dir ./train_logs/moss-moon-003-sft
--n_epochs 2
--train_bsz_per_gpu 4
--eval_bsz_per_gpu 4
--learning_rate 0.000015
--eval_step 10
--save_step 10

这样的问题应该怎么解决，已经使用pip 安装了deepspeed

Start with the shown accelerate launch command, configs/sft.yaml, and finetune_moss.py, then reproduce the --deepspeed_multinode_launcher error in the documented four-machine setup. Compare the installed accelerate and deepspeed command-line options with the official tutorial; done means the launcher argument is accepted and the training script starts across the configured machines.