用diffusers实现"Tencent-Hunyuan/HunyuanDiT-Diffusers"模型，只能简单prompt，否则会出图错误

[pljj315]

简单prompt "a girl"能正常出图， 换成下面图片中的较长一点的prompt就有问题？ 代码见下图

[gnobitab]

Running this code I got two correct images,

from diffusers import HunyuanDiTPipeline
import torch
pipe = HunyuanDiTPipeline.from_pretrained("Tencent-Hunyuan/HunyuanDiT-Diffusers", torch_dtype=torch.float16)
pipe.to('cuda')
pipe.enable_model_cpu_offload()

### NOTE: HunyuanDiT supports both Chinese and English inputs
prompt = ["A male artist standing in front of an easel", "A male artist standing in front of an easel, holding a paintbrush"]
generator=torch.Generator(device="cuda").manual_seed(0)
images = pipe(height=1024, width=1024, prompt=prompt, generator=generator, num_inference_steps=25, guidance_scale=5.0).images

for i in range(2):
    images[i].save("figs/img_%d.png"%i)

could you try this code snippet with different random seeds?

[pljj315]

Running this code I got two correct images,

from diffusers import HunyuanDiTPipeline
import torch
pipe = HunyuanDiTPipeline.from_pretrained("Tencent-Hunyuan/HunyuanDiT-Diffusers", torch_dtype=torch.float16)
pipe.to('cuda')
pipe.enable_model_cpu_offload()

### NOTE: HunyuanDiT supports both Chinese and English inputs
prompt = ["A male artist standing in front of an easel", "A male artist standing in front of an easel, holding a paintbrush"]
generator=torch.Generator(device="cuda").manual_seed(0)
images = pipe(height=1024, width=1024, prompt=prompt, generator=generator, num_inference_steps=25, guidance_scale=5.0).images

for i in range(2):
    images[i].save("figs/img_%d.png"%i)

could you try this code snippet with different random seeds?

换了seed，这个情况依然存在，没有得到改善

[gnobitab]

我的这段code你跑一下看看work吗？