0
I Use This!
Very Low Activity

Commits : Listings

Analyzed about 3 hours ago. based on code collected about 3 hours ago.
Aug 18, 2024 — Aug 18, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
fix: Remove duplicated declaration More... almost 6 years ago
fix: #411 add `RNNTimeStep.runObjects`, and unit test fix: bump version and build More... almost 6 years ago
feat: improve performance on `NeuralNetwork` by early lookups saw nearly 10% improvement in speed in debugging #434 More... almost 6 years ago
fix: remove unneeded `size` property More... almost 6 years ago
fix: #418 missing decayRate, inputRange, learningRate on IRNNDefaultOptions More... almost 6 years ago
fix: bump version to alpha 7, and build More... almost 6 years ago
fix: `RNN.toFunction()` logic for finding previous state matrix fix: Limit iterations for LSTM and GRU, no need to run so long More... almost 6 years ago
fix: Add call to `dataFormatter.addUnrecognized()` when using tokens fix: Remove unneeded .sort() call fix: Unit test `RNN.setupData()` fix: Cleanup `RNN` unit tests More... almost 6 years ago
fix: Audit package-lock.json More... almost 6 years ago
fix: Recurrent end to end test and bundle More... almost 6 years ago
fix: Upgrade to hiddenLayerSets and Float32Array More... almost 6 years ago
fix: Remove unneeded `this.model.input` declaration More... almost 6 years ago
fix: Use null when not defining a dimension, otherwise behaviour would be to use an actual 3d array fix: `// TODO: Fix this test` fix: Throw when target layer is used with depth greater than 0 More... almost 6 years ago
fix: Add TODO, reference issue More... almost 6 years ago
fix: #435, parenthesis or, to prevent statement leak More... almost 6 years ago
fix: Measuring NeuralNetwork against NeuralNetworkGPU fix: sinon to jest upgrades fix: Skip toFunction, it may never be implemented in that class fix: Ensure `activation` is included with NeuralNetworkGPU, for JSON More... almost 6 years ago
fix: Ensure sizes are set for toSVG and json More... almost 6 years ago
feat: Get toSVG to work with all network types fix: Test toSVG to %100 for lines fix: Cleanup rendering-svg.html for more consistent fix: Rework recurrent inputs so that input, hidden, and output can still be recovered, for toSVG, and yet they can be iterated over in sets More... almost 6 years ago
fix: Correct `RNN.getModel` and `RNN.getEquation`, static and tested More... almost 6 years ago
fix: Handle partial objects as inputs in `NeuralNetwork.toFunction` output More... almost 6 years ago
Merge branch 'master' of github.com:BrainJS/brain.js More... almost 6 years ago
fix: Bump version number and build dist feat: Add parcel-plugin-bundle-visualiser to visualize dist fix: Remove lingering docs folder More... almost 6 years ago
fix: Add unit test for `brain.recurrent.RNN()` when using json More... almost 6 years ago
fix: RNNTimeStep config for training with XOR, so it learns, and extra tests fix: add `TestUtils.expectFunction` for testing with functions that have had code coverage js injected into them fix: TrainStream test to use `toBeLessThan` rather than `toBeTruthy` More... almost 6 years ago
fix: 100% coverage on unit tests of ArthurDeviationBiases fix: 100% coverage on unit tests of ArthurDeviationWeights More... almost 6 years ago
fix: `LeakyRelu` layer to have 100% code coverage Also, come up with strategy for 100% code coverage for other layers that doesn't explode when the scripting used to test many files is injected into them. More... almost 6 years ago
fix: `FeedForward._adjustWeights` to `FeedForward.adjustWeights` tests More... almost 6 years ago
fix: `Recurrent._adjustWeights` to `Recurrent.adjustWeights` More... almost 6 years ago
fix: `Recurrent._adjustWeights` to `Recurrent.adjustWeights` tests More... almost 6 years ago
fix: `Recurrent._adjustWeights` to `Recurrent.adjustWeights` More... almost 6 years ago