The python code is the implementation of Bi-LSTM and Transformer in the binary classification tasks. The code is related to the paper Rapid detection of phase transitions from Monte Carlo samples before equilibrium.
The code Bi-LSTM.py require tensorflow-GPU >= 2.3 and the code Transformer.py code require tensorflow-GPU >= 2.6. The whole package list of conda we use in the project can be found in the 'package list.txt'
Each folder has a 'raw data' folder contains a few samples for the task, the 'process data.py' file shows how the raw data is processed into series and also visualize some configurations.