{"id":28304768,"url":"https://github.com/freya135/optimizing-catastrophic-forgetting","last_synced_at":"2025-10-18T10:37:58.756Z","repository":{"id":291618548,"uuid":"978212937","full_name":"Freya135/Optimizing-Catastrophic-Forgetting","owner":"Freya135","description":"This project investigates various continual learning methods to mitigate catastrophic forgetting","archived":false,"fork":false,"pushed_at":"2025-05-05T16:34:27.000Z","size":1018,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-05-31T10:50:42.177Z","etag":null,"topics":["adam-optimizer","catastrophic-forgetting","dataset","elastic-weight-consolidation","f1score","learning-without-forgetting","mnist","replay-based-approach","split-cifar100","synaptic-intelligence","visualization"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Freya135.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-05-05T16:34:10.000Z","updated_at":"2025-05-05T16:38:54.000Z","dependencies_parsed_at":"2025-05-05T18:06:33.869Z","dependency_job_id":null,"html_url":"https://github.com/Freya135/Optimizing-Catastrophic-Forgetting","commit_stats":null,"previous_names":["freya135/optimizing-catastrophic-forgetting"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Freya135/Optimizing-Catastrophic-Forgetting","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Freya135%2FOptimizing-Catastrophic-Forgetting","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Freya135%2FOptimizing-Catastrophic-Forgetting/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Freya135%2FOptimizing-Catastrophic-Forgetting/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Freya135%2FOptimizing-Catastrophic-Forgetting/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Freya135","download_url":"https://codeload.github.com/Freya135/Optimizing-Catastrophic-Forgetting/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Freya135%2FOptimizing-Catastrophic-Forgetting/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":260232767,"owners_count":22978611,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["adam-optimizer","catastrophic-forgetting","dataset","elastic-weight-consolidation","f1score","learning-without-forgetting","mnist","replay-based-approach","split-cifar100","synaptic-intelligence","visualization"],"created_at":"2025-05-24T01:11:40.033Z","updated_at":"2025-10-18T10:37:58.705Z","avatar_url":"https://github.com/Freya135.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Continual Learning on Split CIFAR-100\n\nThis project investigates various continual learning methods to mitigate catastrophic forgetting using the Split CIFAR-100 benchmark. A custom multi-head CNN model is used, with each head handling a 10-class task (total of 10 tasks). Methods include:\n\n- ADAM optimizer with frozen heads\n- Replay-based approach\n- Elastic Weight Consolidation (EWC)\n- Learning Without Forgetting (LwF)\n- Synaptic Intelligence (SI)\n\n---\n\n##  File Descriptions\n\n| File                              | Description                                                                                      |\n|-----------------------------------|--------------------------------------------------------------------------------------------------|\n| `adam_optimizer+freezing.py`      | Baseline training using ADAM optimizer with frozen output heads for each task.                   |\n| `catastrophic_forgetting_demo.py` | Demonstration of catastrophic forgetting using naive sequential training without any mitigation. |\n| `elastic_weight_consolidation.py` | Implementation of EWC for continual learning to preserve important weights.                      |\n| `learning_without_forgetting.py`  | Implements LwF using distillation losses to retain knowledge of past tasks.                      |\n| `replay_based_approach.py`        | Implements replay by storing and mixing a memory buffer of past samples during training.         |\n| `si.py`                           | Synaptic Intelligence implementation to regularize changes in important parameters.              | \n| `vis.py`                          | Visualizes a sample image from the CIFAR-100 dataset with its label.                             |\n| `F1score/`                        | Stores F1 score plots for each task and approach.                                                |\n| `Graphs/`                         | Stores plots for accuracy, AUC, and confusion matrices.                                          |\n| `data/`                           | Directory automatically created by torchvision for CIFAR-100 dataset.                            |\n\n---\n\n## Setup Instructions\n\n1. **Install Requirements** (Python 3.7+ recommended):\n    ```bash\n    pip install torch torchvision scikit-learn matplotlib seaborn\n    ```\n\n2. **Run Training Scripts**\n\nEach script runs training and evaluation on Split CIFAR-100:\n\n```bash\npython adam_optimizer+freezing.py\npython replay_based_approach.py\npython elastic_weight_consolidation.py\npython learning_without_forgetting.py\npython si.py\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffreya135%2Foptimizing-catastrophic-forgetting","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ffreya135%2Foptimizing-catastrophic-forgetting","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffreya135%2Foptimizing-catastrophic-forgetting/lists"}