The next generation of zero-shot classifiers.
AI & ML interests
Text classification, relations extraction, NER, computational biology
Recent Activity
View all activity
Papers
The Million-Label NER: Breaking Scale Barriers with GLiNER bi-encoder
GLiClass: Generalist Lightweight Model for Sequence Classification Tasks
Multi-lingual zero-shot text classification models
Multi-task efficient zero-shot sequence classification models
A joint encoder-decoder GLiNER model for a scalable open-ontology entity recognition
The Multilingual Named Entity Recognition (NER) model which is capable of identifying any entity type.
-
knowledgator/gliner-x-large
Token Classification • 0.9B • Updated • 1.17k • 48 -
knowledgator/gliner-x-base
Token Classification • 0.5B • Updated • 1.16k • 9 -
knowledgator/gliner-x-small
Token Classification • 0.3B • Updated • 1.2k • 19 -
knowledgator/gliner-x-small-v0.5
Token Classification • Updated • 22 • 4
GLiClass with ModernBERT backbone
-
knowledgator/gliclass-modern-large-v2.0
0.4B • Updated • 145 • 5 -
knowledgator/gliclass-modern-base-v2.0
0.2B • Updated • 2.11k • 4 -
knowledgator/gliclass-base-v2.0-rac-init
Zero-Shot Classification • 0.2B • Updated • 42 • 12 -
knowledgator/gliclass-modern-base-v2.0-init
Zero-Shot Classification • 0.2B • Updated • 127 • 27
Bi-encoder and poly-encoder architectures of GLiNER
-
knowledgator/gliner-bi-small-v1.0
Token Classification • Updated • 82 • 10 -
knowledgator/gliner-bi-base-v1.0
Token Classification • Updated • 26 • 4 -
knowledgator/gliner-bi-large-v1.0
Token Classification • Updated • 52 • 26 -
knowledgator/gliner-poly-small-v1.0
Token Classification • Updated • 114 • 14
Generalist and Light-weighted Models for Zero-shot Text Classification
-
GLiClass SandBox
🌖15Classify text with zero-shot classification
-
knowledgator/gliclass-large-v1.0-init
Zero-Shot Classification • 0.4B • Updated • 47 • 15 -
knowledgator/gliclass-base-v1.0-init
Zero-Shot Classification • 0.2B • Updated • 37 • 2 -
knowledgator/gliclass-small-v1.0-init
Zero-Shot Classification • 0.1B • Updated • 129 • 6
Collection of the best zero-shot text classification models. Fine-tune them with few examples using LiqFit - https://github.com/Knowledgator/LiqFit.
-
knowledgator/comprehend_it-base
Zero-Shot Classification • Updated • 142 • 88 -
MoritzLaurer/deberta-v3-large-zeroshot-v1.1-all-33
Zero-Shot Classification • 0.4B • Updated • 2.13k • • 58 -
MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7
Zero-Shot Classification • 0.3B • Updated • 849k • • 410 -
MoritzLaurer/deberta-v3-base-zeroshot-v1.1-all-33
Zero-Shot Classification • 0.2B • Updated • 16.2k • • 30
Collection of auto-regressive models tuned for text classification
Collection of pre-trained encoder models trained on large molecules databases.
Efficient Multi-Task Safety Classification Models
-
knowledgator/opir-multitask-large-v1.0
Text Classification • 0.4B • Updated • 1.5k • 10 -
knowledgator/opir-multitask-multilang-v1.0
Text Classification • 0.3B • Updated • 924 • 20 -
knowledgator/opir-edge-multilang-v1.0
Text Classification • 0.1B • Updated • 87 • 6 -
knowledgator/opir-edge-v1.0
Text Classification • 32.7M • Updated • 139 • 9
Zero-shot joint NER and relation extraction models
-
knowledgator/gliner-relex-large-v0.5
Token Classification • 0.5B • Updated • 2.03k • 26 -
knowledgator/gliner-relex-large-v1.0
Token Classification • 0.5B • Updated • 758 • 11 -
knowledgator/gliner-relex-base-v1.0
Token Classification • 0.2B • Updated • 166 • 3 -
knowledgator/gliner-relex-multi-v1.0
Token Classification • 0.3B • Updated • 594 • 9
GLiNER-bi-Encoder models for entity linking with the GLiNKER framework
PII detection models developed in collaboration with Wordcab
-
knowledgator/gliner-pii-large-v1.0
Token Classification • Updated • 20.4k • 46 -
knowledgator/gliner-pii-base-v1.0
Token Classification • Updated • 6.67k • 18 -
knowledgator/gliner-pii-small-v1.0
Token Classification • Updated • 1.86k • 6 -
knowledgator/gliner-pii-edge-v1.0
Token Classification • Updated • 1.88k • 13
Models for zero-shot text classification that are up to 50 times faster than Cross-Encoders and show the same or higher accuracy.
-
GLiClass: Generalist Lightweight Model for Sequence Classification Tasks
Paper • 2508.07662 • Published • 11 -
knowledgator/gliclass-edge-v3.0
Text Classification • 32.7M • Updated • 2.18k • 18 -
knowledgator/gliclass-modern-base-v3.0
Text Classification • 0.2B • Updated • 3.13k • 7 -
knowledgator/gliclass-modern-large-v3.0
Text Classification • 0.4B • Updated • 651 • 14
Collection of high-quality GLiNER models tuned for working with biomedical data
-
GLiNER-biomed: A Suite of Efficient Models for Open Biomedical Named Entity Recognition
Paper • 2504.00676 • Published • 5 -
Ihor/gliner-biomed-large-v1.0
Token Classification • Updated • 1.52k • 19 -
Ihor/gliner-biomed-base-v1.0
Token Classification • Updated • 1.6k • 5 -
Ihor/gliner-biomed-small-v1.0
Token Classification • Updated • 919 • 3
GLiNER models based on modern encoder architectures
Collection of initial models and models that use converted decoders to encoders as backbones
-
knowledgator/Qwen-encoder-0.5B
Question Answering • 0.5B • Updated • 52 • 10 -
knowledgator/Llama-encoder-1.0B
Question Answering • 1B • Updated • 264 • 3 -
knowledgator/Sheared-LLaMA-encoder-1.3B
Question Answering • 1B • Updated • 17 • 2 -
knowledgator/Qwen-encoder-1.5B
Question Answering • 2B • Updated • 39 • 2
Collection of universal token classification (UTC) models capable in prompt-tuned manner to solve many information extraction tasks.
Knowledgator GLiNER models for information extraction
-
knowledgator/gliner-multitask-v1.0
Token Classification • Updated • 154 • 40 -
knowledgator/gliner-multitask-large-v0.5
Token Classification • Updated • 1.36k • 142 -
GLiNER HandyLab
⚡90Perform multiple NLP tasks like NER, QA, and summarization
-
GLiNER multi-task: Generalist Lightweight Model for Various Information Extraction Tasks
Paper • 2406.12925 • Published • 25
Collection of models for converting chemical formats between each other.
-
knowledgator/SMILES2IUPAC-canonical-small
Text Generation • 5.97M • Updated • 553 • 7 -
knowledgator/IUPAC2SMILES-canonical-base
Text Generation • Updated • 271 • 6 -
knowledgator/IUPAC2SMILES-canonical-small
Text Generation • 5.79M • Updated • 98 • 5 -
knowledgator/SMILES2IUPAC-canonical-base
Text Generation • Updated • 1.89k • 10
Collection of datasest for various information extraction tasks.
The next generation of zero-shot classifiers.
Efficient Multi-Task Safety Classification Models
-
knowledgator/opir-multitask-large-v1.0
Text Classification • 0.4B • Updated • 1.5k • 10 -
knowledgator/opir-multitask-multilang-v1.0
Text Classification • 0.3B • Updated • 924 • 20 -
knowledgator/opir-edge-multilang-v1.0
Text Classification • 0.1B • Updated • 87 • 6 -
knowledgator/opir-edge-v1.0
Text Classification • 32.7M • Updated • 139 • 9
Multi-lingual zero-shot text classification models
Zero-shot joint NER and relation extraction models
-
knowledgator/gliner-relex-large-v0.5
Token Classification • 0.5B • Updated • 2.03k • 26 -
knowledgator/gliner-relex-large-v1.0
Token Classification • 0.5B • Updated • 758 • 11 -
knowledgator/gliner-relex-base-v1.0
Token Classification • 0.2B • Updated • 166 • 3 -
knowledgator/gliner-relex-multi-v1.0
Token Classification • 0.3B • Updated • 594 • 9
Multi-task efficient zero-shot sequence classification models
GLiNER-bi-Encoder models for entity linking with the GLiNKER framework
PII detection models developed in collaboration with Wordcab
-
knowledgator/gliner-pii-large-v1.0
Token Classification • Updated • 20.4k • 46 -
knowledgator/gliner-pii-base-v1.0
Token Classification • Updated • 6.67k • 18 -
knowledgator/gliner-pii-small-v1.0
Token Classification • Updated • 1.86k • 6 -
knowledgator/gliner-pii-edge-v1.0
Token Classification • Updated • 1.88k • 13
A joint encoder-decoder GLiNER model for a scalable open-ontology entity recognition
Models for zero-shot text classification that are up to 50 times faster than Cross-Encoders and show the same or higher accuracy.
-
GLiClass: Generalist Lightweight Model for Sequence Classification Tasks
Paper • 2508.07662 • Published • 11 -
knowledgator/gliclass-edge-v3.0
Text Classification • 32.7M • Updated • 2.18k • 18 -
knowledgator/gliclass-modern-base-v3.0
Text Classification • 0.2B • Updated • 3.13k • 7 -
knowledgator/gliclass-modern-large-v3.0
Text Classification • 0.4B • Updated • 651 • 14
The Multilingual Named Entity Recognition (NER) model which is capable of identifying any entity type.
-
knowledgator/gliner-x-large
Token Classification • 0.9B • Updated • 1.17k • 48 -
knowledgator/gliner-x-base
Token Classification • 0.5B • Updated • 1.16k • 9 -
knowledgator/gliner-x-small
Token Classification • 0.3B • Updated • 1.2k • 19 -
knowledgator/gliner-x-small-v0.5
Token Classification • Updated • 22 • 4
Collection of high-quality GLiNER models tuned for working with biomedical data
-
GLiNER-biomed: A Suite of Efficient Models for Open Biomedical Named Entity Recognition
Paper • 2504.00676 • Published • 5 -
Ihor/gliner-biomed-large-v1.0
Token Classification • Updated • 1.52k • 19 -
Ihor/gliner-biomed-base-v1.0
Token Classification • Updated • 1.6k • 5 -
Ihor/gliner-biomed-small-v1.0
Token Classification • Updated • 919 • 3
GLiClass with ModernBERT backbone
-
knowledgator/gliclass-modern-large-v2.0
0.4B • Updated • 145 • 5 -
knowledgator/gliclass-modern-base-v2.0
0.2B • Updated • 2.11k • 4 -
knowledgator/gliclass-base-v2.0-rac-init
Zero-Shot Classification • 0.2B • Updated • 42 • 12 -
knowledgator/gliclass-modern-base-v2.0-init
Zero-Shot Classification • 0.2B • Updated • 127 • 27
GLiNER models based on modern encoder architectures
Bi-encoder and poly-encoder architectures of GLiNER
-
knowledgator/gliner-bi-small-v1.0
Token Classification • Updated • 82 • 10 -
knowledgator/gliner-bi-base-v1.0
Token Classification • Updated • 26 • 4 -
knowledgator/gliner-bi-large-v1.0
Token Classification • Updated • 52 • 26 -
knowledgator/gliner-poly-small-v1.0
Token Classification • Updated • 114 • 14
Collection of initial models and models that use converted decoders to encoders as backbones
-
knowledgator/Qwen-encoder-0.5B
Question Answering • 0.5B • Updated • 52 • 10 -
knowledgator/Llama-encoder-1.0B
Question Answering • 1B • Updated • 264 • 3 -
knowledgator/Sheared-LLaMA-encoder-1.3B
Question Answering • 1B • Updated • 17 • 2 -
knowledgator/Qwen-encoder-1.5B
Question Answering • 2B • Updated • 39 • 2
Generalist and Light-weighted Models for Zero-shot Text Classification
-
GLiClass SandBox
🌖15Classify text with zero-shot classification
-
knowledgator/gliclass-large-v1.0-init
Zero-Shot Classification • 0.4B • Updated • 47 • 15 -
knowledgator/gliclass-base-v1.0-init
Zero-Shot Classification • 0.2B • Updated • 37 • 2 -
knowledgator/gliclass-small-v1.0-init
Zero-Shot Classification • 0.1B • Updated • 129 • 6
Collection of universal token classification (UTC) models capable in prompt-tuned manner to solve many information extraction tasks.
Collection of the best zero-shot text classification models. Fine-tune them with few examples using LiqFit - https://github.com/Knowledgator/LiqFit.
-
knowledgator/comprehend_it-base
Zero-Shot Classification • Updated • 142 • 88 -
MoritzLaurer/deberta-v3-large-zeroshot-v1.1-all-33
Zero-Shot Classification • 0.4B • Updated • 2.13k • • 58 -
MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7
Zero-Shot Classification • 0.3B • Updated • 849k • • 410 -
MoritzLaurer/deberta-v3-base-zeroshot-v1.1-all-33
Zero-Shot Classification • 0.2B • Updated • 16.2k • • 30
Knowledgator GLiNER models for information extraction
-
knowledgator/gliner-multitask-v1.0
Token Classification • Updated • 154 • 40 -
knowledgator/gliner-multitask-large-v0.5
Token Classification • Updated • 1.36k • 142 -
GLiNER HandyLab
⚡90Perform multiple NLP tasks like NER, QA, and summarization
-
GLiNER multi-task: Generalist Lightweight Model for Various Information Extraction Tasks
Paper • 2406.12925 • Published • 25
Collection of auto-regressive models tuned for text classification
Collection of models for converting chemical formats between each other.
-
knowledgator/SMILES2IUPAC-canonical-small
Text Generation • 5.97M • Updated • 553 • 7 -
knowledgator/IUPAC2SMILES-canonical-base
Text Generation • Updated • 271 • 6 -
knowledgator/IUPAC2SMILES-canonical-small
Text Generation • 5.79M • Updated • 98 • 5 -
knowledgator/SMILES2IUPAC-canonical-base
Text Generation • Updated • 1.89k • 10
Collection of pre-trained encoder models trained on large molecules databases.
Collection of datasest for various information extraction tasks.