Quicktext is a framework for developing LSTM and CNN based text classification models.
- It is easy to learn and use quicktext
- The classifiers can be added to sPacy pipeline
- It's built using PyTorch, hence has inbuilt quantization and onnx support
Install from source
pip install -q git+https://github.com/GDGVIT/quicktext.git
from quicktext import TextClassifier
from quicktext import Trainer
from quicktext.datasets import get_imdb
imdb = get_imdb()
classifier = TextClassifier(num_class=2, arch='bilstm')
trainer = Trainer(classifier)
trainer.fit(imdb.train, imdb.val, epochs=10, batch_size=64, gpus=1)- Bidirectional LSTM
- CNN 2D filters
- Fasttext
- RCNN
- Seq2Seq Attention
|
Ramaneswaran
|
Made with ❤️ by DSC VIT