sweta20 / detecting-cyberbullying-across-smps Goto Github PK

Ma'am, I am working on this very project as my minor project for engineering but I am facing problems in running the code. Can u please tell me on which python version have u created the pickle file and I am facing an error in the below mentioned line of code.
x_text.append(p.tokenize((data[i]['text']).encode('utf-8')))
the error is type error it says can't apply string pattern on bytes type object

What are the versions for the used libraries? A requirements.txt file with exact versions would be great for reproducibility.

Hey there,

Some colleagues and I are doing a reproducibility study for this paper. Sadly the link called "Sentiment Specific word embeddings (SSWE)" in the README is down.
I searched around and can't seem to find a backup anywhere. The file should be called "embedding-results.zip".
It'd be awesome if somebody could provide a version of the embeddings.

Thanks a lot in advance :)

In your project in the function train i got the following error

model = get_model(model_type, trainX.shape[1], vocab_size, embed_size, NUM_CLASSES, LEARN_RATE)
initial_weights = model.get_weights()
shuffle_weights(model,model.get_weights())

TypeError: get_weights() takes exactly 2 arguments (1 given)

what to do to solve the error

When I'm reproducing this article, I always get an index overflow problem, as shown below

IndexError: index 18831 is out of bounds for axis 0 with size 18831
Apply node that caused the error: AdvancedSubtensor1(embedding_2/embedding_3/embedding_4/embedding_10/embeddings, Elemwise{Cast{int32}}.0)
Toposort index: 77
Inputs types: [TensorType(float32, matrix), TensorType(int32, vector)]
Inputs shapes: [(18831, 25), (7936,)]
Inputs strides: [(100, 4), (4,)]
Inputs values: ['not shown', 'not shown']
Inputs type_num: [11, 5]
Outputs clients: [[Reshape{3}(AdvancedSubtensor1.0, MakeVector{dtype='int64'}.0)]]