gek

About the score range

2

Hello, in some cases, I found some indicators are negative. Are they normal?

1. 用的是训练集，split=val，并且没有把image作为输入。应该是dialogs+image -> image? 2. 和vist的使用方式有挺大区别，vist用的是dialogs+image -> image

Could you provide a demo for generating image from a text? example ``` prompt = "xxx" image = model.generate(prompt) image.save("xxx.png") ```

different clipScore

8

**Describe the bug** During the evaluation of **Text-guided Image Generation**, the ClipScore generated is not consistent with the results documented in the paper. I have compared 197 images; **they are...