GLiNER2.5 Small V1
GLiNER2.5 Small is a 74M parameter span free information extraction model that predicts entity boundaries directly instead of enumerating candidate spans, enabling linear scaling with document length. It performs named entity recognition, relation extraction, and constrained classification over long documents in a single pass.
- Type
- Text
- Weights
- Open
- Downloads
- 9.9k
- Likes
- 26
- Pipeline
- token classification
- License
- apache-2.0
Model hub
fastino/gliner2.5-small-v1
- Repo
- fastino/gliner2.5-small-v1
- Library
- gliner2
- Access
- Public
Technical tags and languages
Papers
1
More from Fastino
7
-
626k DL
gliner2-base-v1
License apache-2.0 · 625,874 downloads · by fastino
-
167k DL
gliner2-large-v1
License apache-2.0 · 166,891 downloads · by fastino
-
197k DL
gliner2-multi-v1
License apache-2.0 · 197,039 downloads · by fastino
-
161k DL
gliner2-privacy-filter-PII-multi
Text
token classification · License apache-2.0 · 161,418 downloads · by fastino
-
28k DL
GLiNER2.5 Base V1
Text
GLiNER2.5 Base is a 0.2B parameter span free information extraction model that predicts entity boundaries dir…
-
252k DL
GLiNER2.5 Multi V1
Text
GLiNER2.5 Multi is a 0.3B parameter multilingual span free information extraction model that predicts entity…
About Fastino
1
Task-optimized language models (TLMs) that are more accurate, faster, and safer than traditional LLMs, running without GPUs