◻️ Use pre-trained word vectors with glove-6B
◻️ Use this Dateset for train model
pip insatll -r requirements.txt
python class_emoji_text_classification.py --vector_shape dimention --features_path features_path --infrence sentence
◻️Result with Dropout
Features Vector Dimensions | Train Loss | Train Accuracy | Test Loss | Test Accuracy | Inference Time |
---|---|---|---|---|---|
50d | 0.7244 | 77.27% | 0.7332 | 75.57% | 0.068s |
100d | 0.6523 | 78.79% | 0.6593 | 79.59% | 0.0705s |
200d | 0.3144 | 94.7% | 0.5209 | 83.67% | 0.076s |
300d | 0.2055 | 97.73% | 0.4601 | 89.8% | 0.0861s |
◻️Result without Dropout
Features Vector Dimensions | Train Loss | Train Accuracy | Test Loss | Test Accuracy | Inference Time |
---|---|---|---|---|---|
50d | 0.6304 | 83.33% | 0.7163 | 79.59% | 0.0607s |
100d | 0.4839 | 90.91% | 0.6053 | 85.71% | 0.0633s |
200d | 0.2704 | 95.45% | 0.5055 | 85.71% | 0.0717s |
300d | 0.1807 | 99.24% | 0.4456 | 87.76% | 0.082s |