Extraction
Return the text spans that match each label (person, organization, date, money, …). Jev and the pure decision models never return spans; only the GLiNER family does.
On Metatext decisions, GLiNER2 multi leads with 1.000 micro-F1.
Metatext decisions
Every model is zero-shot on this suite.
| # | Model | Micro-F1 |
|---|---|---|
| 1 | GLiNER2 multiFastino | 1.000 |
| 2 | GLiNER2 largeFastino | 0.968 |
| 3 | GLiNER 2.5 baseFastino | 0.968 |
| 4 | GLiNER2 baseFastino | 0.935 |
| 5 | GLiNER 2.5 multiFastino | 0.885 |
| 6 | GLiNER 2.5 smallFastino | 0.812 |
| 7 | GLiGuard 300MFastino | 0.000 |
Frequently asked questions
What is a Extraction question?
Return the text spans that match each label (person, organization, date, money, …). Jev and the pure decision models never return spans; only the GLiNER family does.
Which model is best at extraction?
GLiNER2 multi leads on Metatext decisions at 1.000 mean accuracy over the extraction task.
Related
A question with a fixed set of named options and a description for each; the model returns a probability per option and the answer is the argmax.
An ordered scale (none / low / medium / high, 1–5, …). The model returns a distribution over levels and the answer is the expected level. The primitive relevance classifiers cannot do.
A statement about the state, answered with a probability that it holds. Named after the Jev primitive; the threshold for "yes" is 0.5 unless the caller sets one.