Neural Networks in Python
Putting Neural Networks together should be easier than it usually is. The code in this repository presents simple Python snippets designed to make prototyping neural network architectures quick.
Note:
- Requires Python.
- Requires NumPy.
- Requires MatplotLib.
- Requires PyOpenCL.
- Requires DisPy.
-
net: this package contains all the modules you will need to create neural networks quickly. Because that is what you wanted.
-
bench: this package contains tools used for benchmarking new algorithms. Because you can do it better.
-
data: this package contains scripts providing access to common datasets. Because no one wants to have to go looking for them anyway.
-
tutor: this folder contains tutorials written to help you get started. Because the first time is always the hardest.
-
doc: this folder contains the documentation of the all modules we have implemented. These are the grizzly mathematical details your professor warned you about.
-
test: this folder contains tests we use to make sure everything is working the way it should. Because things do not always work the way they should.