Parameter order in loss function of word2vec.py

[LopezGG]

I believe your inputs and labels are interchanged when you calculate the loss. Here is the corrected version

# Compute the loss, using a sample of the negative labels each time.
			if self.loss_type == 'sampled_softmax_loss':
				loss = tf.nn.sampled_softmax_loss(weights=self.weights, biases=self.biases, inputs=self.embed,
					labels=self.train_labels, num_sampled=self.n_neg_samples, num_classes=self.vocabulary_size)
			elif self.loss_type == 'nce_loss':
				loss= tf.nn.nce_loss(weights=self.weights, biases=self.biases, inputs=self.embed, 
					labels=self.train_labels, num_sampled=self.n_neg_samples,num_classes= self.vocabulary_size)
			self.loss = tf.reduce_mean(loss, name="loss")