# kjaved0/mrcl architecture

> Architecture diagram of the GitHub repository kjaved0/mrcl, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-01). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/kjaved0/mrcl
- Repository: https://github.com/kjaved0/mrcl

## Overview

This repository is a research codebase for Online-aware Meta-learning (OML), with training and evaluation workflows for sequential classification and regression, plus pretraining and SR-NN comparison runs. The principal workflow is to configure an experiment, obtain datasets and sampled tasks, train a learner/meta-learner, and evaluate or visualize learned representations. The graph includes major README capabilities and indexed modules; edges are limited to README statements, visible calls, or SOURCE INDEX references, so unverified runtime details are omitted.

Main entry points: [`model/learner.py`](https://github.com/kjaved0/mrcl/blob/master/model/learner.py) (Neural learner), [`oml_omniglot.py`](https://github.com/kjaved0/mrcl/blob/master/oml_omniglot.py) (OML training), [`evaluate_omniglot.py`](https://github.com/kjaved0/mrcl/blob/master/evaluate_omniglot.py) (Omniglot evaluation), [`evaluate_imagenet.py`](https://github.com/kjaved0/mrcl/blob/master/evaluate_imagenet.py) (ImageNet evaluation)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_entry["Experiment workflows"]
  node_oml_train["OML training<br/>[oml_omniglot.py]"]
  node_oml_regression["Regression training<br/>[oml_regression.py]"]
  node_maml_baseline["MAML baseline"]
  node_pretrain_omni["Omniglot pretraining"]
  node_pretrain_image["ImageNet pretraining"]
  node_srnn_omni["Omniglot SR-NN<br/>[srnn_omniglot.py]"]
  node_srnn_image["ImageNet SR-NN<br/>[srnn_imagenet.py]"]
  node_eval_omni["Omniglot evaluation"]
  node_eval_image["ImageNet evaluation"]
  node_eval_reg["Regression evaluation"]
  node_visualize["Representation visualization"]
  node_class_config["Classification options<br/>[class_parser.py]"]
  node_eval_config["Evaluation options"]
  node_reg_config["Regression options<br/>[reg_parser.py]"]
end

subgraph group_data["Data and tasks"]
  node_dataset_factory["Dataset selection<br/>[datasetfactory.py]"]
  node_omniglot_data["Omniglot dataset<br/>[omniglot.py]"]
  node_imagenet_data["Mini-ImageNet dataset<br/>[miniimagenet.py]"]
  node_task_sampling["Task sampling<br/>[task_sampler.py]"]
end

subgraph group_models["Learning models"]
  node_model_factory["Model configuration<br/>[modelfactory.py]"]
  node_learner["Neural learner<br/>[learner.py]"]
  node_meta_learner["Meta-learning<br/>[meta_learner.py]"]
  node_meta_classification["Online classification"]
  node_oml_layers["OML layer operations<br/>[nn.py]"]
end

subgraph group_support["Experiment support"]
  node_experiment["Experiment tracking<br/>[experiment.py]"]
  node_color_logging["Colored logging<br/>[colorer.py]"]
end

node_researcher(("Researcher"))

node_researcher -->|"runs training"| node_oml_train
node_researcher -->|"evaluates models"| node_eval_omni
node_researcher -->|"visualizes models"| node_visualize
node_oml_train -->|"uses options"| node_class_config
node_oml_train -->|"uses tracking"| node_experiment
node_oml_train -->|"selects data"| node_dataset_factory
node_oml_train -->|"samples tasks"| node_task_sampling
node_oml_train -->|"builds model config"| node_model_factory
node_oml_train -->|"trains meta-learner"| node_meta_learner
node_oml_regression -->|"uses options"| node_reg_config
node_oml_regression -->|"samples tasks"| node_task_sampling
node_oml_regression -->|"trains meta-learner"| node_meta_learner
node_maml_baseline -->|"uses meta-learner"| node_meta_learner
node_pretrain_omni -->|"trains learner"| node_learner
node_pretrain_image -->|"trains learner"| node_learner
node_srnn_omni -->|"uses learner"| node_learner
node_srnn_image -->|"uses learner"| node_learner
node_eval_omni -->|"uses options"| node_eval_config
node_eval_omni -->|"selects data"| node_dataset_factory
node_eval_omni -->|"uses learner"| node_learner
node_eval_omni -->|"builds model config"| node_model_factory
node_eval_omni -->|"uses tracking"| node_experiment
node_eval_image -->|"selects data"| node_dataset_factory
node_eval_image -->|"uses dataset"| node_imagenet_data
node_eval_image -->|"uses learner"| node_learner
node_eval_image -->|"builds model config"| node_model_factory
node_eval_image -->|"uses tracking"| node_experiment
node_eval_reg -->|"uses options"| node_reg_config
node_eval_reg -->|"samples tasks"| node_task_sampling
node_eval_reg -->|"uses meta-learner"| node_meta_learner
node_eval_reg -->|"uses tracking"| node_experiment
node_visualize -->|"selects data"| node_dataset_factory
node_visualize -->|"uses learner"| node_learner
node_visualize -->|"builds model config"| node_model_factory
node_visualize -->|"uses tracking"| node_experiment
node_dataset_factory -->|"constructs dataset"| node_omniglot_data
node_learner -->|"uses layers"| node_oml_layers
node_meta_learner -->|"constructs learner"| node_learner

click node_oml_train "https://github.com/kjaved0/mrcl/blob/master/oml_omniglot.py"
click node_oml_regression "https://github.com/kjaved0/mrcl/blob/master/oml_regression.py"
click node_maml_baseline "https://github.com/kjaved0/mrcl/blob/master/maml-rep_omniglot.py"
click node_pretrain_omni "https://github.com/kjaved0/mrcl/blob/master/pretraining_omniglot.py"
click node_pretrain_image "https://github.com/kjaved0/mrcl/blob/master/pretraining_imagenet.py"
click node_srnn_omni "https://github.com/kjaved0/mrcl/blob/master/srnn_omniglot.py"
click node_srnn_image "https://github.com/kjaved0/mrcl/blob/master/srnn_imagenet.py"
click node_eval_omni "https://github.com/kjaved0/mrcl/blob/master/evaluate_omniglot.py"
click node_eval_image "https://github.com/kjaved0/mrcl/blob/master/evaluate_imagenet.py"
click node_eval_reg "https://github.com/kjaved0/mrcl/blob/master/evaluate_regression.py"
click node_visualize "https://github.com/kjaved0/mrcl/blob/master/visualize_representations.py"
click node_class_config "https://github.com/kjaved0/mrcl/blob/master/configs/classification/class_parser.py"
click node_eval_config "https://github.com/kjaved0/mrcl/blob/master/configs/classification/class_parser_eval.py"
click node_reg_config "https://github.com/kjaved0/mrcl/blob/master/configs/regression/reg_parser.py"
click node_dataset_factory "https://github.com/kjaved0/mrcl/blob/master/datasets/datasetfactory.py"
click node_omniglot_data "https://github.com/kjaved0/mrcl/blob/master/datasets/omniglot.py"
click node_imagenet_data "https://github.com/kjaved0/mrcl/blob/master/datasets/miniimagenet.py"
click node_task_sampling "https://github.com/kjaved0/mrcl/blob/master/datasets/task_sampler.py"
click node_model_factory "https://github.com/kjaved0/mrcl/blob/master/model/modelfactory.py"
click node_learner "https://github.com/kjaved0/mrcl/blob/master/model/learner.py"
click node_meta_learner "https://github.com/kjaved0/mrcl/blob/master/model/meta_learner.py"
click node_meta_classification "https://github.com/kjaved0/mrcl/blob/master/model/meta_classification.py"
click node_oml_layers "https://github.com/kjaved0/mrcl/blob/master/oml/nn.py"
click node_experiment "https://github.com/kjaved0/mrcl/blob/master/experiment/experiment.py"
click node_color_logging "https://github.com/kjaved0/mrcl/blob/master/utils/colorer.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_oml_train,node_oml_regression,node_maml_baseline,node_pretrain_omni,node_pretrain_image,node_srnn_omni,node_srnn_image,node_eval_omni,node_eval_image,node_eval_reg,node_visualize,node_class_config,node_eval_config,node_reg_config toneBlue
class node_dataset_factory,node_omniglot_data,node_imagenet_data,node_task_sampling toneAmber
class node_model_factory,node_learner,node_meta_learner,node_meta_classification,node_oml_layers toneMint
class node_experiment,node_color_logging toneRose
class node_researcher toneIndigo
```

## Components

### Experiment workflows

- [OML training](https://github.com/kjaved0/mrcl/blob/master/oml_omniglot.py): `oml_omniglot.py`
- [Regression training](https://github.com/kjaved0/mrcl/blob/master/oml_regression.py): `oml_regression.py`
- [MAML baseline](https://github.com/kjaved0/mrcl/blob/master/maml-rep_omniglot.py): `maml-rep_omniglot.py`
- [Omniglot pretraining](https://github.com/kjaved0/mrcl/blob/master/pretraining_omniglot.py): `pretraining_omniglot.py`
- [ImageNet pretraining](https://github.com/kjaved0/mrcl/blob/master/pretraining_imagenet.py): `pretraining_imagenet.py`
- [Omniglot SR-NN](https://github.com/kjaved0/mrcl/blob/master/srnn_omniglot.py): `srnn_omniglot.py`
- [ImageNet SR-NN](https://github.com/kjaved0/mrcl/blob/master/srnn_imagenet.py): `srnn_imagenet.py`
- [Omniglot evaluation](https://github.com/kjaved0/mrcl/blob/master/evaluate_omniglot.py): `evaluate_omniglot.py`
- [ImageNet evaluation](https://github.com/kjaved0/mrcl/blob/master/evaluate_imagenet.py): `evaluate_imagenet.py`
- [Regression evaluation](https://github.com/kjaved0/mrcl/blob/master/evaluate_regression.py): `evaluate_regression.py`
- [Representation visualization](https://github.com/kjaved0/mrcl/blob/master/visualize_representations.py): `visualize_representations.py`
- [Classification options](https://github.com/kjaved0/mrcl/blob/master/configs/classification/class_parser.py): `configs/classification/class_parser.py`
- [Evaluation options](https://github.com/kjaved0/mrcl/blob/master/configs/classification/class_parser_eval.py): `configs/classification/class_parser_eval.py`
- [Regression options](https://github.com/kjaved0/mrcl/blob/master/configs/regression/reg_parser.py): `configs/regression/reg_parser.py`

### Data and tasks

- [Dataset selection](https://github.com/kjaved0/mrcl/blob/master/datasets/datasetfactory.py): `datasets/datasetfactory.py`
- [Omniglot dataset](https://github.com/kjaved0/mrcl/blob/master/datasets/omniglot.py): `datasets/omniglot.py`
- [Mini-ImageNet dataset](https://github.com/kjaved0/mrcl/blob/master/datasets/miniimagenet.py): `datasets/miniimagenet.py`
- [Task sampling](https://github.com/kjaved0/mrcl/blob/master/datasets/task_sampler.py): `datasets/task_sampler.py`

### Learning models

- [Model configuration](https://github.com/kjaved0/mrcl/blob/master/model/modelfactory.py): `model/modelfactory.py`
- [Neural learner](https://github.com/kjaved0/mrcl/blob/master/model/learner.py): `model/learner.py`
- [Meta-learning](https://github.com/kjaved0/mrcl/blob/master/model/meta_learner.py): `model/meta_learner.py`
- [Online classification](https://github.com/kjaved0/mrcl/blob/master/model/meta_classification.py): `model/meta_classification.py`
- [OML layer operations](https://github.com/kjaved0/mrcl/blob/master/oml/nn.py): `oml/nn.py`

### Experiment support

- [Experiment tracking](https://github.com/kjaved0/mrcl/blob/master/experiment/experiment.py): `experiment/experiment.py`
- [Colored logging](https://github.com/kjaved0/mrcl/blob/master/utils/colorer.py): `utils/colorer.py`

### Other components

- **Researcher**

## Connections

- Researcher → OML training: runs training
- Researcher → Omniglot evaluation: evaluates models
- Researcher → Representation visualization: visualizes models
- OML training → Classification options: uses options (evidence: `oml_omniglot.py`)
- OML training → Experiment tracking: uses tracking (evidence: `oml_omniglot.py`)
- OML training → Dataset selection: selects data (evidence: `oml_omniglot.py`)
- OML training → Task sampling: samples tasks (evidence: `oml_omniglot.py`)
- OML training → Model configuration: builds model config (evidence: `oml_omniglot.py`)
- OML training → Meta-learning: trains meta-learner (evidence: `oml_omniglot.py`)
- Regression training → Regression options: uses options (evidence: `oml_regression.py`)
- Regression training → Task sampling: samples tasks (evidence: `oml_regression.py`)
- Regression training → Meta-learning: trains meta-learner (evidence: `oml_regression.py`)
- MAML baseline → Meta-learning: uses meta-learner (evidence: `maml-rep_omniglot.py`)
- Omniglot pretraining → Neural learner: trains learner
- ImageNet pretraining → Neural learner: trains learner (evidence: `pretraining_imagenet.py`)
- Omniglot SR-NN → Neural learner: uses learner (evidence: `srnn_omniglot.py`)
- ImageNet SR-NN → Neural learner: uses learner (evidence: `srnn_imagenet.py`)
- Omniglot evaluation → Evaluation options: uses options (evidence: `evaluate_omniglot.py`)
- Omniglot evaluation → Dataset selection: selects data (evidence: `evaluate_omniglot.py`)
- Omniglot evaluation → Neural learner: uses learner (evidence: `evaluate_omniglot.py`)
- Omniglot evaluation → Model configuration: builds model config (evidence: `evaluate_omniglot.py`)
- Omniglot evaluation → Experiment tracking: uses tracking (evidence: `evaluate_omniglot.py`)
- ImageNet evaluation → Dataset selection: selects data (evidence: `evaluate_imagenet.py`)
- ImageNet evaluation → Mini-ImageNet dataset: uses dataset (evidence: `evaluate_imagenet.py`)
- ImageNet evaluation → Neural learner: uses learner (evidence: `evaluate_imagenet.py`)
- ImageNet evaluation → Model configuration: builds model config (evidence: `evaluate_imagenet.py`)
- ImageNet evaluation → Experiment tracking: uses tracking (evidence: `evaluate_imagenet.py`)
- Regression evaluation → Regression options: uses options (evidence: `evaluate_regression.py`)
- Regression evaluation → Task sampling: samples tasks (evidence: `evaluate_regression.py`)
- Regression evaluation → Meta-learning: uses meta-learner (evidence: `evaluate_regression.py`)
- Regression evaluation → Experiment tracking: uses tracking (evidence: `evaluate_regression.py`)
- Representation visualization → Dataset selection: selects data (evidence: `visualize_representations.py`)
- Representation visualization → Neural learner: uses learner (evidence: `visualize_representations.py`)
- Representation visualization → Model configuration: builds model config (evidence: `visualize_representations.py`)
- Representation visualization → Experiment tracking: uses tracking (evidence: `visualize_representations.py`)
- Dataset selection → Omniglot dataset: constructs dataset (evidence: `datasets/datasetfactory.py`)
- Neural learner → OML layer operations: uses layers (evidence: `model/learner.py`)
- Meta-learning → Neural learner: constructs learner (evidence: `model/meta_learner.py`)
