ganeshkharad/gk-hinglish-sentiment

The data contains only hinglish codemixed text it and was very limited may be I will Update this model if I can get good amount of data . The training data was not much as expected. The data was very much limited . The data is being used in India mostly . The model was created using BertTokenizer, BertForSequenceClassification . It is based on a small amount of reviews dataset with 3 labels . The code was created with a pre-training model card with a description of the pre-trained model card . For more information on the model, visit http://www.huggingface.co/rohanrajpal/berber-base,