All benchmarksrecall_at_1000 (higher is better)
MTEB AppsRetrieval (default), recall_at_1000
As the model cards publish them, each model’s best result. Facthory has not run them; results are self-reported unless the hub verified them.
| # | Model | Result | Source |
|---|---|---|---|
| 1 | granite-embedding-125m-english ibm-granite · test | 71.71 | Model card |
| 2 | granite-embedding-278m-multilingual ibm-granite · test | 52.72 | Model card |
| 3 | granite-embedding-107m-multilingual ibm-granite · test | 45.21 | Model card |
