WN

`model = san.san(sa_type=1, layers=[3, 2, 3, 5, 2], kernels=[3, 7, 7, 7, 7], num_classes=1000).cuda() model.load_state_dict(torch.load('./san15_patchwise/model/model_best.pth'))` RuntimeError: Error(s) in loading state_dict for SAN: Missing key(s) in state_dict: "conv_in.weight", "bn_in.weight", "bn_in.bias", "bn_in.running_mean",...

only test on new image

1

hi, if I only want to test this pretrained model(39999.npy) on new images, may I need to download the ms coco dataset?

train_autodeeplab.py: error: unrecognized arguments: --gpu_ids 0,1,2,3

2

hi,could i ask what's wrong with this?

word_to_id.txt corpus_word2vec.txt 能分享一下吗

Salians_syn

1

Could provide the citation of the dataset Salians_syn

您好，代码在san diego数据集上有运行错误，请问您有什么办法解决吗？ 错误使用 eigs>KSnonherm (line 1527) 输入矩阵为病态矩阵。它可能包含 Inf 或 NaN 值。 出错 eigs>KrylovSchur (line 1285) [V, d, isNotConverged, stopAlgorithm, VV] = KSnonherm(applyOP, applyM, n, k, ... 出错 eigs (line 126)...