{"id":15906816,"url":"https://github.com/csinva/matrix-completion-llm","last_synced_at":"2025-04-02T23:26:26.949Z","repository":{"id":228588619,"uuid":"709854761","full_name":"csinva/matrix-completion-llm","owner":"csinva","description":"Training LLMs for matrix completion","archived":false,"fork":false,"pushed_at":"2023-12-02T01:24:36.000Z","size":3888,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-02-08T13:43:00.368Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/csinva.png","metadata":{"files":{"readme":"readme.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","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}},"created_at":"2023-10-25T14:27:13.000Z","updated_at":"2024-03-20T01:16:01.000Z","dependencies_parsed_at":null,"dependency_job_id":"fbb47f6c-7995-4775-aa31-61687a8ce7c2","html_url":"https://github.com/csinva/matrix-completion-llm","commit_stats":null,"previous_names":["csinva/matrix-completion-llm"],"tags_count":0,"template":false,"template_full_name":"csinva/cookiecutter-ml-research","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/csinva%2Fmatrix-completion-llm","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/csinva%2Fmatrix-completion-llm/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/csinva%2Fmatrix-completion-llm/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/csinva%2Fmatrix-completion-llm/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/csinva","download_url":"https://codeload.github.com/csinva/matrix-completion-llm/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246907998,"owners_count":20853172,"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":[],"created_at":"2024-10-06T13:41:50.778Z","updated_at":"2025-04-02T23:26:26.927Z","avatar_url":"https://github.com/csinva.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# matrix-completion-llms\nBased of [cookiecutter-data-science](https://github.com/drivendata/cookiecutter-data-science).\n\n# Organization\n- `mcllm`: contains main code for modeling (e.g. model architecture)\n- `experiments`: code for runnning experiments (e.g. loading data, training models, evaluating models)\n- `scripts`: scripts for hyperparameter sweeps (python scripts that launch jobs in `experiments` folder with different hyperparams)\n- `notebooks`: jupyter notebooks for analyzing results and making figures\n- `tests`: unit tests\n\n# Setup\n- clone and run `pip install -e .`, resulting in a package named `mcllm` that can be imported\n    - see `setup.py` for dependencies, not all are required\n- example run: run `python scripts/01_train_basic_models.py` (which calls `experiments/01_train_model.py` then view the results in `notebooks/01_model_results.ipynb`\n- keep tests upated and run using `pytest`\n\n# Features\n- scripts sweep over hyperparameters using easy-to-specify python code\n- experiments automatically cache runs that have already completed\n    - caching uses the (**non-default**) arguments in the argparse namespace\n- notebooks can easily evaluate results aggregated over multiple experiments using pandas\n\n# Guidelines\n- See some useful packages [here](https://csinva.io/blog/misc/ml_coding_tips)\n- Avoid notebooks whenever possible (ideally, only for analyzing results, making figures)\n- Paths should be specified relative to a file's location (e.g. `os.path.join(os.path.dirname(__file__), 'data')`)\n- Naming variables: use the main thing first followed by the modifiers (e.g. `X_train`, `acc_test`)\n    - binary arguments should start with the word \"use\" (e.g. `--use_caching`) and take values 0 or 1\n- Use logging instead of print\n- Use argparse and sweep over hyperparams using python scripts (or custom things, like [amulet](https://amulet-docs.azurewebsites.net/main/index.html))\n    - Note, arguments get passed as strings so shouldn't pass args that aren't primitives or a list of primitives (more complex structures should be handled in the experiments code)\n- Each run should save a single pickle file of its results\n- All experiments that depend on each other should run end-to-end with one script (caching things along the way)\n- Keep updated requirements in setup.py\n- Follow sklearn apis whenever possible\n- Use Huggingface whenever possible, then pytorch\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcsinva%2Fmatrix-completion-llm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcsinva%2Fmatrix-completion-llm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcsinva%2Fmatrix-completion-llm/lists"}