0
I Use This!
Very Low Activity

Commits : Listings

Analyzed 2 days ago. based on code collected 2 days ago.
Aug 24, 2024 — Aug 24, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Merge pull request #155 from BrainJS/154-testing-times More... over 7 years ago
feedforward now working, nearly identical to NeuralNetwork More... over 7 years ago
killing the spy killing (-_-) More... over 7 years ago
comparing nets to nets, currently failing. More... over 7 years ago
Uses sinon More... over 7 years ago
# If applied, this commit will make life better for everyone More... over 7 years ago
Merge branch 'master' into develop More... over 7 years ago
adjust description More... over 7 years ago
add .adjustWeights(), and unit test More... over 7 years ago
Unit test .calculateDeltas(), .cacheWeights(), .cacheDeltas() More... over 7 years ago
further debugging with Garrett, the genius. More... over 7 years ago
debugging with Garrett More... over 7 years ago
Merge pull request #153 from BrainJS/152-NaN-error-rate More... over 7 years ago
fix #152 and bump version number More... over 7 years ago
starting on backpropagation via deltas More... over 7 years ago
bring recurrent net runInput forward propagation caching tested and online More... over 7 years ago
Updated to always throw an error. More... over 7 years ago
Turning input/output training data into textures to see if we can speed things up a little bit. More... over 7 years ago
Getting recurrent layers connecting, and testing a simple neural network. Started on GRU and LSTM. More... over 7 years ago
laying groundwork for deprecating the existing recurrent class family, and starting on the new foundation. Provide layer indexes for reuse elsewhere of FeedForward. More... over 7 years ago
Merge branch 'develop' into nn-gpu-layers More... over 7 years ago
Merge pull request #151 from BrainJS/catchup-nn-gpu More... over 7 years ago
With some of the changes pushed up recently, NeuralNetworkGPU lagged behind. This was mainly the prepending of "_" to internal-only methods, as well as moving things to a `trainOpts` object. This meant that the GPU was not able to work at all. This catches things up. More... over 7 years ago
increase iterations for more consistent test More... over 7 years ago
Merge pull request #149 from daffl/common-npmignore More... over 7 years ago
Add common .npmignore to exclude development files More... over 7 years ago
Handles User Object Validation #142 More... over 7 years ago
Merge pull request #141 from BrainJS/fix-training-option-setting More... over 7 years ago
Missed a _updateTrainingOpts in the merge, remade browser and dist More... over 7 years ago
Merge in changes on develop More... over 7 years ago