Theano

Theano is an open-source Python library and deep learning framework designed for fast numerical computation. It allows developers to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. By leveraging CPU and GPU acceleration, Theano is widely used in machine learning research, mathematical modeling, and prototyping of neural networks.

Features of Theano:

• Efficiently computes mathematical expressions involving multi-dimensional arrays (tensors).
• Supports both CPU and GPU acceleration for faster performance.
• Automatic differentiation for gradient computation in machine learning models.
• Integrates seamlessly with NumPy and other scientific Python libraries.
• Used for prototyping deep learning models and research in AI/ML.
• Open-source and maintained by the research community (with active forks like Theano-PyMC).

To check the installed version of Theano, run these commands in your environment:

$ sudo su 
$ source /opt/theano/bin/activate
$ python -c "
import pytensor as theano
import pytensor.tensor as tt
print('Successfully aliased PyTensor as theano!')
print('Version:', theano.__version__)
print('floatX:', theano.config.floatX)
print('BLAS:', theano.config.blas__ldflags)
"

Disclaimer: Theano was originally developed by the MILA lab at the University of Montreal. Official development has ended, but community forks (such as Theano-PyMC) continue to provide updates. Always refer to the official or fork documentation for the most accurate and up-to-date information.