HuHaibo

The question about different size images to compress

1

The 'elic' model could deal with different size of images, because the arch is not sensitive to different size inputs. But diffusion model has a definite input/output size based on...

the result that imagenet model generated is not good enough, I think the diffusion config sets more steps?

11

**the label picture is**  **the per_recon picture is**  It looks like the diffusion model has not iterated enough steps, (one thousand)

I use the code below and i dont understand how does "quality" control compress ratio. ```python from compressai.zoo import bmshj2018_factorized self.codec = bmshj2018_factorized(quality=4, metric='mse', pretrained=True, progress=True) ``` ```python @register_model("bmshj2018-factorized") class...