theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-18 22:08:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

Docs:Supplement NPU training script samples and documentation instruction

Browse Source
This commit is contained in:
feng0w0
2025-12-31 10:01:21 +08:00
parent 507e7e5d36
commit 07b1f5702f
3 changed files with 5 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
docs/en/Pipeline_Usage/GPU_support.md
View File

@@ -59,7 +59,7 @@ save_video(video, "video.mp4", fps=15, quality=5)
```
### Training
NPU startup script samples have been added for each type of model,the scripts are stored in the `examples/xxx/special/npu_scripts`, for examples `examples/wanvideo/model_training/special/npu_scripts/Wan2.2-T2V-A14B-NPU.sh`.
NPU startup script samples have been added for each type of model,the scripts are stored in the `examples/xxx/special/npu_scripts`, for example `examples/wanvideo/model_training/special/npu_scripts/Wan2.2-T2V-A14B-NPU.sh`.
In the NPU training scripts, NPU specific environment variables that can optimize performance have been added, and relevant parameters have been enabled for specific models.
@@ -79,6 +79,6 @@ Set 0 or not set: indicates not enabling the binding function
2: Indicates enabling fine-grained kernel binding
#### Parameters for specific models
| Model | Parameter | Note |
|-----------|---------------------------|-------------------|
| Wan 14B系列 | --initialize_model_on_cpu | The 14B model needs to be initialized on the CPU |
| Model | Parameter | Note |
|----------------|---------------------------|-------------------|
| Wan 14B series | --initialize_model_on_cpu | The 14B model needs to be initialized on the CPU |

2
examples/qwen_image/model_training/special/npu_scripts/Qwen-Image-Edit-2509-LoRA-NPU.sh
View File

@@ -21,7 +21,7 @@ accelerate launch examples/qwen_image/model_training/train.py \
--task "sft:data_process"
accelerate launch examples/qwen_image/model_training/train.py \
--dataset_base_path "./models/train/Qwen-Image-LoRA-splited-cache" \
--dataset_base_path "./models/train/Qwen-Image-Edit-2509-LoRA-splited-cache" \
--max_pixels 1048576 \
--dataset_repeat 50 \
--model_id_with_origin_paths "Qwen/Qwen-Image-Edit-2509:transformer/diffusion_pytorch_model*.safetensors" \

1
examples/z_image/model_training/special/npu_scripts/Z-Image-Turbo-NPU.sh
View File

@@ -1,4 +1,3 @@
# This example is tested on 8*A100
export PYTORCH_NPU_ALLOC_CONF=expandable_segments:True
export CPU_AFFINITY_CONF=1