环境
系统：CentOS-7 CPU : E5-2680V4 14核28线程 内存：DDR4 2133 32G * 2 显卡：Tesla V100-32G【PG503】 (水冷) 驱动: 535 CUDA: 12.2依赖
pipinstalldiffusers -i https://mirrors.aliyun.com/pypi/simple下载模型
pipinstallmodelscope -i https://mirrors.aliyun.com/pypi/simple modelscope download --model Tongyi-MAI/Z-Image-Turbo --local_dir /models/Z-Image-Turbo代码
importtorch from diffusersimportZImagePipeline# 1. Load the pipeline# Use bfloat16 for optimal performance on supported GPUspipe=ZImagePipeline.from_pretrained("/models/Z-Image-Turbo",torch_dtype=torch.bfloat16,low_cpu_mem_usage=False,)pipe.to("cuda")# [Optional] Attention Backend# Diffusers uses SDPA by default. Switch to Flash Attention for better efficiency if supported:# pipe.transformer.set_attention_backend("flash") # Enable Flash-Attention-2# pipe.transformer.set_attention_backend("_flash_3") # Enable Flash-Attention-3# [Optional] Model Compilation# Compiling the DiT model accelerates inference, but the first run will take longer to compile.# pipe.transformer.compile()# [Optional] CPU Offloading# Enable CPU offloading for memory-constrained devices.# pipe.enable_model_cpu_offload()prompt="Young Chinese woman in red Hanfu, intricate embroidery. Impeccable makeup, red floral forehead pattern. Elaborate high bun, golden phoenix headdress, red flowers, beads. Holds round folding fan with lady, trees, bird. Neon lightning-bolt lamp (⚡️), bright yellow glow, above extended left palm. Soft-lit outdoor night background, silhouetted tiered pagoda (西安大雁塔), blurred colorful distant lights."# 2. Generate Imageimage=pipe(prompt=prompt,height=1024,width=1024,num_inference_steps=9,# This actually results in 8 DiT forwardsguidance_scale=0.0,# Guidance should be 0 for the Turbo modelsgenerator=torch.Generator("cuda").manual_seed(42),).images[0]image.save("example.png")运行
exportHF_ENDPOINT=https://hf-mirror.com python run-1.py参数
1024 * 1024
时间
(comfyui)[root@ai-server Z-Image]# python run-1.pyLoading pipeline components...:20%|█████████▏|1/5[00:00<00:01,2.15it/s]`torch_dtype`is deprecated!Use`dtype`instead!Loading checkpoint shards:100%|███████████████████████████████████████████████████|3/3[00:00<00:00,47.55it/s]Loading checkpoint shards:100%|███████████████████████████████████████████████████|3/3[00:09<00:00,3.11s/it]Loading pipeline components...:100%|██████████████████████████████████████████████|5/5[00:10<00:00,2.06s/it]100%|██████████████████████████████████████████████████████████████████████████████|9/9[00:57<00:00,6.35s/it]GPU
Tue Dec1623:15:592025+---------------------------------------------------------------------------------------+|NVIDIA-SMI535.129.03 Driver Version:535.129.03 CUDA Version:12.2||-----------------------------------------+----------------------+----------------------+|GPU Name Persistence-M|Bus-Id Disp.A|Volatile Uncorr. ECC||Fan Temp Perf Pwr:Usage/Cap|Memory-Usage|GPU-Util Compute M.||||MIG M.||=========================================+======================+======================||0Tesla PG503-216 On|00000000:04:00.0 Off|0||N/A 35C P0 232W / 250W|29740MiB / 32768MiB|100% Default||||N/A|+-----------------------------------------+----------------------+----------------------+效果
总结
速度还是可以的，1分钟。
最高占用30G,100%GPU占用。
效果后面继续测试，后续加入到ComfyUI中。