Vincent

> ERROR:torch.distributed.elastic.multiprocessing.api:failed i also met the same question? Any help?

> Hi, I want to train the model on the costom dataset including non-realistic images (painting, comic). I think your code requires pretrained model (ImageNet-1k) to build the SimMIM, but...

@larsh0103 Thx your tips! Do u know how to render the image from the text-to-emotion script?

> Thx，It works！

> Hi, we think single-time lama inpainting is enough for most cases. We didn't use lama-refine for saving processing time. Hi, @geekyutao If i want to use the refine, how...

> I didn't save weights of that fine-tuned model, but you can try to fine-tune the model by your self. You can try slightly larger batch size and more iteration...

> It seems like it is caused by out-of-memory error. Could you check if your GPU memory is larger than at least 6 GB ?看起来它是由内存不足错误引起的。你能检查一下你的 GPU 内存是否大于至少 6 GB 吗？...

> some problem, while set enable_queue=False,the "Expecting value: line 1 column 1 (char 0)" not occurred, but press diag it ,the "ValueError: Need to enable queue to use generators" error...

> Hi, We have already supported customized video evaluation for vbench_i2v. You can refer to this link for detailed usage: [Evaluate Your Own Videos](https://github.com/Vchitect/VBench/tree/master/vbench2_beta_i2v#evaluate-your-own-videos).您好，我们已支持为 vbench_i2v 提供定制视频评估。您可以参考此链接获取详细使用方法：评估您的视频。 请问一下为什么customized video evaluation支持的维度比较少，而且没有text-video 对齐的指标呢？