goru001 / nlp-for-marathi Goto Github PK

I am getting an error when I'm trying to deploy the model on flask.
I didn't get an error on google colab.

I was trying to get vectors for marathi language using: vectors = get_embedding_vectors('आहेस', 'mr') & encountered with the following error

The Marathi News Dataset contains only the Train and Validation Splits. This made it hard to report accuracy and compare to the baseline (which reports the validation accuracy). Hence, there's a need to have separate Train-Validation-Test sets.
Also, the headlines contain non-Devanagari characters (mostly English) and Arabic numerals. These need to be mapped to Devanagari characters and numerals. The original dataset contains ~50 duplicates which need to be cleaned.

Following the above steps, I have created new splits of the dataset (link here) which can be used for properly reporting scores on the Classification dataset.

Can it do lemmatization and stop words classification?