crutcher / bimm / #227

Builds Branch Commit Type Ran Committer Via Coverage
#227 main refactor: streamline prefab lookup methods and update references - Renamed `lookup_prefab`, `try_lookup_prefab`, and `expect_lookup_prefab` to `lookup`, `try_lookup`, and `expect_lookup` for improved consistency and brevity. - Refactored ResNet e... push crutcher
69.71
#226 main refactor: rename prefab lookup methods and enhance weight descriptor with metadata - Renamed `lookup_by_name`, `try_lookup_by_name`, and `expect_lookup_by_name` to `lookup_prefab`, `try_lookup_prefab`, and `expect_lookup_prefab` for improved sema... push crutcher
69.71
#225 main refactor: remove outdated `training.rs` and centralize ResNet training logic - Deleted `training.rs` file, consolidating training functionality directly into `main.rs` for better modularity and clarity. - Updated ResNet examples to utilize centra... push crutcher
69.69
#224 main feat: add flexible prefab lookup methods and integrate structured weight handling in ResNet examples - Introduced `lookup_by_name`, `try_lookup_by_name`, and `expect_lookup_by_name` methods in prefab caching for flexible and clear prefab lookups.... push crutcher
69.94
#223 main feat: add structured prefab and weights map support - Introduced `StaticPreFabMap` and `PreFabMap` for organized prefab definitions. - Added `StaticPretrainedWeightsMap` and `PretrainedWeightsMap` for structured weight mapping. - Updated ResNet e... push crutcher
70.19
#222 main feat: update ResNet examples with configurable activation and training parameters - Added `PReluConfig` activation to ResNet initialization in examples. - Adjusted default training parameters, including epochs, workers, and drop probabilities. - ... push crutcher
72.37
#221 main refactor: rename configuration structs in ResNet framework for clarity - Renamed `Abstract`- and `Config`-based constructs to `ContractConfig` and `StructureConfig` respectively, ensuring consistent and descriptive naming. - Updated all reference... push crutcher
72.37
#220 main feat: add normalization and activation configuration to ResNet layers - Integrated `NormalizationConfig` and `ActivationConfig` into ResNet components, including `ResidualBlock`, `LayerBlock`, and `ResNet`. - Updated ResNet initialization to supp... push crutcher
72.37
#219 main refactor: replace `ConvDownsample` with `ResNetDownsample` and add kernel size configuration - Migrated from `ConvDownsample` to `ResNetDownsample` across applicable modules, updating configurations and methods accordingly. - Introduced `down_ker... push crutcher
72.44
#218 main feat: add `AvgPool2dSame` pooling layer with TensorFlow-like SAME padding - Implemented `AvgPool2dSame` structure with support for specifying kernel size, stride, and padding. - Added utility functions: `get_same_padding` and `pad_same` for dynam... push crutcher
72.82
#217 main refactor: update ResNet finetuning example parameters and dependencies - Removed unused `pretrained_classes` argument for cleanup. - Adjusted default parameters: batch size, number of workers, epochs, drop probabilities, learning rate, and artifa... push crutcher
73.05
#216 main refactor: enhance ResNet modules with utilities and Downsample implementation - Introduced `scalar_to_array` utility to improve configuration consistency across ResNet components. - Implemented `Downsample` and `DownsampleConfig` for modular down... push crutcher
73.05
#215 main Switch resnet impl to new CNA2d based impl. push crutcher
73.67
#214 main refactor: replace `ResNet` with `CNAResNet` across examples and introduce weight copy utilities - Migrated examples to use `CNAResNet` in place of `ResNet` for consistency with the updated core model. - Added utility methods for copying weights b... push crutcher
68.54
#213 main Merge remote-tracking branch 'origin/main' push crutcher
69.12