train_network.py: error: unrecognized arguments: --deepspeed --zero_stage=2 --offload_optimizer_device=cpu

[lianshuailong0715]

1.i have install deepspeed with pip 2.this is my train script: accelerate launch --num_cpu_threads_per_process=2 train_network.py
--pretrained_model_name_or_path="xxx"
--train_data_dir="xxx"
--caption_extension='.txt'
--resolution="840,600"
--tokenizer_cache_dir="xxx"
--logging_dir="xxx"
--output_dir="xxx"
--save_model_as="safetensors"
--output_name="xxx"
--network_module=networks.lora
--network_alpha=64
--network_dim=128
--lr_scheduler="cosine"
--learning_rate=1e-5
--lr_warmup_steps=1000
--lr_scheduler_num_cycles=20
--train_batch_size=1
--max_train_epochs=20
--save_precision='fp16'
--save_every_n_epochs=1
--seed=1024
--optimizer_type="AdamW8bit"
--deepspeed
--zero_stage=2
--offload_optimizer_device="cpu" \

[BootsofLagrangian]

Did you update this repo? try script after git pull.