Skip to content

Instantly share code, notes, and snippets.

import numpy as np
import torch
import torch.nn as nn
import torch.nn.functional as F
from lanczos_utils import check_dist
EPS = float(np.finfo(np.float32).eps)
__all__ = ['LanczosNet']
Can not use cuDNN on context None: Disabled by dnn.enabled flag
Mapped name None to device cuda0: GeForce GTX 1080 Ti (0000:61:00.0)
11/16/2018 02:06:13 PM In order to work for big datasets fix https://github.com/Theano/Theano/pull/5721 should be applied to theano.
11/16/2018 02:06:13 PM loading the dataset from ./data/cmu/
11/16/2018 02:06:14 PM #labels: 129
11/16/2018 02:06:16 PM TfidfVectorizer(analyzer=u'word', binary=True, decode_error=u'strict',
dtype='float32', encoding='latin1', input=u'content',
lowercase=True, max_df=0.2, max_features=None, min_df=10,
ngram_range=(1, 1), norm='l2', preprocessor=None, smooth_idf=True,
stop_words='english', strip_accents=None, sublinear_tf=False,
Can not use cuDNN on context None: Disabled by dnn.enabled flag
Mapped name None to device cuda0: GeForce GTX TITAN X (0000:83:00.0)
11/14/2018 02:45:18 PM In order to work for big datasets fix https://github.com/Theano/Theano/pull/5721 should be applied to theano.
11/14/2018 02:45:18 PM loading data from dumped file...
11/14/2018 02:45:18 PM loading data finished!
11/14/2018 02:45:18 PM stacking training, dev and test features and creating indices...
11/14/2018 02:45:18 PM running mlp with graph conv...
11/14/2018 02:45:18 PM highway is True
11/14/2018 02:45:18 PM Graphconv model input size 9467, output size 32 and hidden layers [300, 300, 300] regul 0.0 dropout 0.5.
11/14/2018 02:45:18 PM 3 gconv layers
@felixgwu
felixgwu / treernn.py
Created April 6, 2017 21:51 — forked from volkancirik/treernn.py
Pytorch TreeRNN
"""
TreeLSTM[1] implementation in Pytorch
Based on dynet benchmarks :
https://github.com/neulab/dynet-benchmark/blob/master/dynet-py/treenn.py
https://github.com/neulab/dynet-benchmark/blob/master/chainer/treenn.py
Other References:
https://github.com/pytorch/examples/tree/master/word_language_model
https://github.com/pfnet/chainer/blob/29c67fe1f2140fa8637201505b4c5e8556fad809/chainer/functions/activation/slstm.py
https://github.com/stanfordnlp/treelstm
@felixgwu
felixgwu / fc_densenet.py
Created April 2, 2017 05:22
FC-DenseNet Implementation in PyTorch
import torch
from torch import nn
__all__ = ['FCDenseNet', 'fcdensenet_tiny', 'fcdensenet56_nodrop',
'fcdensenet56', 'fcdensenet67', 'fcdensenet103',
'fcdensenet103_nodrop']
class DenseBlock(nn.Module):