| # Competitions |
|
|
| Create a competition checklist: |
|
|
| - [ ] Create a space using https://huggingface.co/new-space |
| - [ ] Create a private dataset using https://huggingface.co/new-dataset |
| - [ ] Create a public dataset using https://huggingface.co/new-dataset |
| - [ ] Add the following secrets to the space: |
| - `AUTOTRAIN_TOKEN`: the token of the user that will be used to create autotrain projects |
| - `AUTOTRAIN_USERNAME`: the username of the user that will be used to create autotrain projects |
| - `COMPETITION_ID`: Private dataset created previously, e.g.: `my_org/my_private_dataset` |
|
|
| - Private dataset structure: |
| |
| ```bash |
| . |
| ├── README.md |
| ├── submission_info/ |
| ├── submissions/ |
| ├── conf.json |
| ├── solution.csv |
| ├── COMPETITION_DESC.md |
| ├── SUBMISSION_DESC.md |
| |
| ```` |
| |
| - Example content for `conf.json`: |
|
|
| ``` |
| { |
| "SUBMISSION_LIMIT": 5, |
| "SELECTION_LIMIT": 2, |
| "END_DATE": "2022-11-23", |
| "EVAL_HIGHER_IS_BETTER": 1, |
| "DATASET": "abhishek/test_competition_dataset", |
| "COMPETITION_NAME": "Shoes vs Boots vs Sandals", |
| "SUBMISSION_COLUMNS": "id,target", |
| "EVAL_METRIC": "accuracy_score" |
| } |
| ``` |
| |
| - Public dataset structure: |
| |
| ```bash |
| . |
| ├── README.md |
| ├── training files (folder/zip/csv) |
| ├── test files (folder/zip/csv) |
| ├── training labels |
| ├── sample_submission.csv |
| |
| ``` |
| |
