[BUG]: raise ValueError("No parameters found in moe_dp_group, please consider using HybridParallelPlugin instead") ValueError: No parameters found in moe_dp_group, please consider using HybridParallelPlugin instead

[klompn]

Is there an existing issue for this bug?

• I have searched the existing issues

The bug has not been fixed in the latest main branch

• I have checked the latest main branch

Do you feel comfortable sharing a concise (minimal) script that reproduces the error? :)

Yes, I will share a minimal reproducible script.

🐛 Describe the bug

我使用lora_finetune脚本微调r1时，发生上述错误，为什么检测不到moe的参数呢？ 训练环境：64npu(910B32g)，torch_npu==2.2.0 脚本命令：colossalai run --nproc_per_node 8 lora_finetune.py --pretrained /cache/DeepSeek-R1-671B-BF16/ --dataset lora_sft_data.jsonl --plugin moe --lr 2e-5 --max_length 256 -g --ep 4 --pp 16 --tp 1 --batch_size 24 --lora_rank 8 --lora_alpha 16 --num_epochs 2 --warmup_steps 8 --tensorboard_dir logs --save_dir DeepSeek-R1-bf16-lora --host ${VC_WORKER_HOSTS} --master_addr ${VC_WORKER_HOSTS%%,} --master_port 13123

Environment

No response

[klompn]

有没有大神帮助一下，现在无法定位到错误

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

Is there any master who can help? The error cannot be located now

[klompn]

BUG发生在plugin\moe_hybrid_parallel_plugin.py MoeHybridParallelZeroOptimizer类中：