Datasets:
|
Download artifacts/ablations/benchmark.md from Alaeddinnn/OpenRelief: direct link, hf CLI and curl.
- Browser
- Download file 553 Bytes
-
https://huggingface.co/datasets/Alaeddinnn/OpenRelief/resolve/main/artifacts/ablations/benchmark.md
- Command line
-
hf download hf://datasets/Alaeddinnn/OpenRelief/artifacts/ablations/benchmark.md
-
curl -L -o benchmark.md https://huggingface.co/datasets/Alaeddinnn/OpenRelief/resolve/main/artifacts/ablations/benchmark.md
553 Bytes
| # Classical exploratory ablations | |
| Evaluation partition: validation. | |
| Logistic regression; all preprocessing fitted on training only. These are not OpenTSLM results. | |
| No hyperparameter selection used test data. Ablations measure association, not causality. | |
| | Features | Macro-F1 (present classes) | Accuracy | | |
| |---|---:|---:| | |
| | all_sources | 0.5844 | 0.7000 | | |
| | food_history_only | 0.6032 | 0.7339 | | |
| | without_shipping | 0.5927 | 0.7036 | | |
| | without_conflict | 0.5113 | 0.6304 | | |
| | without_rainfall | 0.5615 | 0.6772 | | |
| | without_prices | 0.6104 | 0.7315 | | |