20NG Training Set Reported Results Question

[jharrang]

For your 20NG results reported in your original paper (88.6± 0.1 ), was the model trained on the full public 20NG train set, of size 11314, or were the reported results generated using the training code currently in this repo here and here, which appears to exclude the validation set from the data used to train the model? (The latter would use a training set of size 10183).

Thanks!

[allenhaozhu]

Yes, we exclude the validation set. This code is released a few days ago because some guys sent me a mail asking the detail of text classification. And after I simply checked the results, I release codes for text classification. If there is any issue please let me know.

[vencentDebug]

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument target in method wrapper_nll_loss_forward)

use parser.add_argument('--no-cuda', action='store_true', default=True, help='Disables CUDA training.') args.device = 'cpu'

[vencentDebug]

when I use only "CPU", it notices the problem. Then I use "GPU", it also happens another problem. train_feats = torch.spmm(adj, train_feats) RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and CPU! (when checking argument for argument mat2 in method wrapper__mm)