Releases: marcpinet/neuralnetlib
Releases · marcpinet/neuralnetlib
neuralnetlib 2.4.0
- fix: indentation
- ci: bump version to 2.4.0
- Merge pull request #34 from marcpinet/refactor-typo
- fix: none exception on iterable
- Merge pull request #33 from marcpinet/fix-callback-baseline-stop-training
- feat: add EarlyStopping and reformat code
- feat: some improvements
- refactor: more robust approach
- Merge pull request #32 from marcpinet/feat-earlystopping
- refactor: switch to dict pattern instead of if-else garbage block
- Merge pull request #31 from marcpinet/refactor-better-compat-handling
- docs: update readme
- Merge pull request #29 from marcpinet/docs-update-layers
- feat: add AveragePooling2D
- feat: add AveragePooling1D
- Merge pull request #30 from marcpinet/feat-averagepooling-layers
neuralnetlib 2.3.0
- feat: add Conv1D, MaxPooling1D, Embedding
- refactor: rename train to fit and add layers compat checks
- ci: remove useless .bat
- refactor: fixes and improvements
- refactor: update code for latest changes
- feat: add BatchNormalization
- ci: bump version to 2.3.0
- Merge pull request #28 from marcpinet/feat-more-layers
neuralnetlib 2.2.1
- refactor: add seed to dropout too
- ci: bump version to 2.2.1
- Merge pull request #27 from marcpinet/refactor-dropout-seed
neuralnetlib 2.2.0
neuralnetlib 2.1.0
- docs: improve cli showcase from png to gif
- Merge pull request #22 from marcpinet/docs-improve-readme-resources
- docs(fix): blue background
- Merge pull request #23 from marcpinet/docs-improve-readme-resources
- feat: add dropout layer (should I've done it earlier)
- ci: bump version to 2.1.0
- Merge pull request #24 from marcpinet/feat-dropout-layer
neuralnetlib 2.0.0
- docs: add new examples and better folder sorting
- Merge pull request #18 from marcpinet/docs-better-code-examples
- refactor: tkinter now uses real time refresh
- Merge pull request #19 from marcpinet/refactor-realtime-update-tkinter
- docs: update readme
- Merge pull request #20 from marcpinet/fix-readme
- feat: add progress bar, val_accuracy and time metrics
- feat: add Conv2D, MaxPooling2D, Flatten layers
- feat: add cnn example
- refactor: reformat and optimize code to standards
- feat: remove the need of input_shape in hidden layers + auto detection
- test(fix): update tests to match correct paramters
- feat: add str methods
- refactor: simplified example for faster runtime
- feat: update maxpooling autoguess for pool size and strides
- docs: update side note
- fix: from_config for layers
- docs: update README and resources
- docs: update gif
- refactor: changed shape for cnn impl and add comment for simple nn
- docs: update readme details
- ci: bump version to 2.0.0
- Merge pull request #21 from marcpinet/feat-cnn-layers
neuralnetlib 1.0.1
neuralnetlib 1.0.0
- ci: improve efficiency and add auto unit testing
- Merge pull request #1 from marcpinet/ci-improve-efficiency
- refactor: add shuffle to utils and better function typing
- refactor: use of shuffle inside train
- feat: better choice of hyperparameters
- fix: accuracy_score not computing properly for binary classification
- ci: new project version
- Merge pull request #2 from marcpinet/fix-accuracy-score-for-binary-classif
- ci: fix workflows edgecases and bump version to 0.4.0
- Merge pull request #8 from marcpinet/ci-fix-workflow-and-bump-version
- feat: add confusion matrix metric
- test: add test for confusion matrix metric
- Merge pull request #8 from marcpinet/ci-fix-workflow-and-bump-version
- ci: bump version to 0.4.1
- Merge pull request #9 from marcpinet/feat-confusion-matrix
- ci: fix release workflow and bump version to 1.0.0
- Merge pull request #15 from marcpinet/ci-fix-workflow-and-bump-version-to-1.0.0
neuralnetlib v0.4.0
Merge pull request #8 from marcpinet/ci-fix-workflow-and-bump-version ci: fix workflows edgecases and bump version to 0.4.0
neuralnetlib v0.3.0
Initial commit