gpleiss/temperature_scaling
A simple way to calibrate your neural network.
This helps machine learning engineers and researchers fine-tune their classification models. It takes the output probabilities from a trained neural network and adjusts them so that the model's stated confidence aligns with its actual accuracy. The result is a more reliable model where the predicted probabilities are trustworthy.
1,167 stars. No commits in the last 6 months.
Use this if you need your neural network's confidence scores to accurately reflect the likelihood of correct predictions, especially for applications where trusting probabilities is crucial.
Not ideal if you are looking for a currently maintained, standalone package, as this repository is unmaintained and primarily serves as a demonstration.
Stars
1,167
Forks
169
Language
Python
License
MIT
Category
Last pushed
Jul 26, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/gpleiss/temperature_scaling"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
facebookincubator/MCGrad
MCGrad is a scalable and easy-to-use tool for multicalibration. It ensures your ML model...
dholzmueller/probmetrics
Post-hoc calibration methods and metrics for classification
yfzhang114/Generalization-Causality
关于domain generalization，domain adaptation，causality，robutness，prompt，optimization，generative...
hollance/reliability-diagrams
Reliability diagrams visualize whether a classifier model needs calibration
Affirm/splinator
Splinator: probabilistic calibration with regression splines