Shengwei An

It is a bit like the issue #53 However it's not fixed completely by issue #54 Maybe we call the Case 3: ### Case 3 #### Follow these steps: 1....

I used [PyTorch resize function](https://pytorch.org/vision/stable/generated/torchvision.transforms.functional.resize.html#torchvision.transforms.functional.resize) to resize them to the same size.

I tried using [mseitzer/pytorch-fid](https://github.com/mseitzer/pytorch-fid) and got only 11.379320395008278. I sampled 50k images using ``` python scripts/image_sample.py --model_path cifar10_uncond_vlb_50M_500K.pt --image_size 32 --num_channels 128 --num_res_blocks 3 --learn_sigma True --dropout 0.3 --diffusion_steps 4000...

I got 47.87 using the following settings: ```--image_size 32 --num_channels 128 --num_res_blocks 3 --diffusion_steps 1000 --noise_schedule linear --lr 1e-4 --batch_size 64``` I trained the model on 4 GPUs for 500K...

I am also interested in how to get the prediction probability (ie, prediction confidence). From the current [`eval_prob_adaptive` function](https://github.com/diffusion-classifier/diffusion-classifier/blob/master/eval_prob_adaptive.py#L44), it seems impossible to get the probability. I tried to compute...

This is because [line 59](https://github.com/machocam/rummikub/blob/master/rummikub.py#L59) didn't check the color for every item in the sub_results: `(item0[1] != item1[1] or item1[1] == "jocker")` only compares the color of the first item...

You're welcome! Something like `item1[1] == 'jocker' or item1[1] not in [_t[1] for _t in sub_results]` should work. I was thinking about issues a pull request. But I found the...