This project utilizes LSTM (Long Short-Term Memory) networks and GloVe (Global Vectors for Word Representation) Word Embedding Vectors for analyzing sentiment on Twitter data.
Before running the sentiment analysis, ensure you have the necessary dataset and dependencies installed.
The GloVe word embedding vectors can be downloaded and unzipped using the following command:
curl -O http://downloads.cs.stanford.edu/nlp/data/glove.6B.zip
unzip glove.6B.zip -d path/to/destination
This project is compatible with CUDA 11.2 for GPU acceleration on Native Windows environments and requires TensorFlow version 2.10.0. Ensure you have the correct TensorFlow versions installed by running:
pip install tensorflow==2.10.0 tensorflow-gpu==2.10.0