modelscope / modelscope/DiffSynth-Studio

lora qwen-image, error: The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).

Open
#1,531 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13.1k
Forks
1.3k
Avg merge
13h 12m
Merged PRs (30d)
45

Description

nohup: ignoring input
The following values were not passed to accelerate launch and had defaults used instead:
--num_processes was set to a value of 2
More than one GPU was found, enabling multi-GPU training.
If this was unintended please pass in --num_processes=1.
--num_machines was set to a value of 1
--mixed_precision was set to a value of 'no'
--dynamo_backend was set to a value of 'no'
To avoid this warning pass in values for each of the problematic parameters or run accelerate config.
/home/zhangkailin/.conda/envs/diff/lib/python3.12/site-packages/accelerate/utils/launch.py:236: UserWarning: Port 29500 is already in use. Accelerate will attempt to launch in a standalone-like mode by finding an open port automatically for this session. If this current attempt fails, or for more control in future runs, please specify a different port (e.g., --main_process_port <your_chosen_port>) or use --main_process_port 0 for automatic selection in your launch command or Accelerate config file.
warnings.warn(
[W718 20:29:59.662875029 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:00.332290007 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:01.086728483 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:01.663832546 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:02.003531155 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:05.547585849 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:10.156080575 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:18.263436069 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).
[W718 20:30:28.254263791 socket.cpp:755] [c10d] The IPv6 network addresses of (I, 38455) cannot be retrieved (gai error: -2 - Name or service not known).

训练脚本

DIFFSYNTH_SKIP_DOWNLOAD=True \
CUDA_VISIBLE_DEVICES=0,1 \
NPROC_PER_NODE=2 \
accelerate launch examples/qwen_image/model_training/train.py \
  --dataset_base_path /appdata/zhangkailin/agent/wam/models_dd/guofeng \
  --dataset_metadata_path /appdata/zhangkailin/agent/wam/models_dd/guofeng/metadata.csv \
  --max_pixels 1572516 \
  --dataset_repeat 50 \
  --model_id_with_origin_paths "Qwen/Qwen-Image:transformer/diffusion_pytorch_model*.safetensors,Qwen/Qwen-Image:text_encoder/model*.safetensors,Qwen/Qwen-Image:vae/diffusion_pytorch_model.safetensors" \
  --learning_rate 1e-4 \
  --num_epochs 6 \
  --remove_prefix_in_ckpt "pipe.dit." \
  --save_steps 5 \
  --output_path "./models/train/Qwen-Image-2512_lora_0518" \
  --lora_base_model "dit" \
  --lora_target_modules "to_q,to_k,to_v,add_q_proj,add_k_proj,add_v_proj,to_out.0,to_add_out,img_mlp.net.2,img_mod.1,txt_mlp.net.2,txt_mod.1" \
  --lora_rank 16 \
  --find_unused_parameters \
  --dataset_num_workers 8 \
  --use_gradient_checkpointing \
  --gradient_accumulation_steps 4 \
  --find_unused_parameters \
  --wandb_project "qwen-image-guofeng" \
  --wandb_name "qwenimage-lora-codex200-0518"

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with examples/qwen_image/model_training/train.py and the reported accelerate launch command, checking how the two-process launch selects its rendezvous address and port. Reproduce the command while examining the existing port-29500 warning and IPv6 name-resolution errors; done is a confirmed cause and a documented or implemented way to launch training without the reported failure.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems, machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.