Social_Media/Stance-Detection-in-Web-and-Social-Media
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Updated README.md

Browse Source
This commit is contained in:
siddharth9820 2019-06-18 22:00:07 +05:30 committed by GitHub
parent 25a8781ee1
commit 2f43a25fcd
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 15 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
TAN/README.md
View File

@ -8,10 +8,24 @@ __Args__
2. embedding_dim - int - dimension of word_embedding
3. hidden_dim - int - dimension of LSTM hidden state
4. vocab_size - int - number of words in the vocabulary
5. n_targets - int - number of target classes
5. n_targets - int - number of dataset classes
6. embedding_matrix - numpy array dtype=float - word embedding matrix
6. dropout - The dropout to be applied before on the final hidden state(lstm)/attention-weighted hidden state (tan-,tan)
__Returns__
A torch.nn.module object for the specified version
```
def forward(self, sentence, target,verbose=False)
```
__Args__
1. sentence - a numpy array of shape [1xN] and dtype int, where N is the length of the input sentence and each entry is the corresponding index of the word in the `embedding_matrix`
2. target - a numpy array of shape [1xM] and dtype int, where M is the length of the target and each entry is the corresponding index of the word in the `embedding_matrix
__Returns__
1. target_scores - a torch float Tensor of shape [1xn_targets], where N is the number of dataset classes. This is the log likelihood probabilities of all the classes
### Running the code
To run the code you have to run *get_training_plots.py* or *early_stopping_training.py*<br>
To change models, call the classes with specific versions as mentioned above<br>