inEXASCALE/pychop
A Python package for simulating low precision arithmetic in scientific computing and machine learning
This tool helps scientists, machine learning engineers, and researchers simulate how calculations will behave when using reduced-precision numbers (like 16-bit or 8-bit instead of standard 32-bit or 64-bit). You input your existing numerical data, often in NumPy arrays or PyTorch/JAX tensors, and it converts them into various low-precision formats. This allows you to evaluate the trade-offs in speed, memory, and accuracy without needing specialized hardware.
Use this if you need to understand how reducing the precision of numbers will impact the stability, convergence, or accuracy of your scientific simulations, machine learning models, or data processing workflows.
Not ideal if you are looking for a tool to automatically optimize your code for existing high-precision hardware, or if you need to perform actual low-precision computations on specialized hardware directly.
Stars
12
Forks
6
Language
Python
License
MIT
Category
Last pushed
Mar 13, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/inEXASCALE/pychop"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
sbi-dev/sbi
sbi is a Python package for simulation-based inference, designed to meet the needs of both...
SMTorg/smt
Surrogate Modeling Toolbox
reservoirpy/reservoirpy
A simple and flexible code for Reservoir Computing architectures like Echo State Networks
GPflow/GPflow
Gaussian processes in TensorFlow
dswah/pyGAM
[CONTRIBUTORS WELCOME] Generalized Additive Models in Python