Chunking on CoNLL 2003 (test)
92.64ScoreMAGNET
Evaluation results
| Method | Links | |
|---|---|---|
| MAGNETModel Category=Llama 2 models, Adaptation=MNTP, SSCL, and MSG, Evaluation Mode=Linear probing2025.01 | 92.64 | |
| LLM2Vec [MNTP]Model Category=Llama 2 models, Adaptation=MNTP only, Evaluation Mode=Linear probing2025.01 | 91.61 | |
| StructBERT-LargeModel Category=Encoder models, Evaluation Mode=Linear probing, Representation=Token averaging2025.01 | 89.99 | |
| LLM2VecModel Category=Llama 2 models, Adaptation=MNTP and SimCSE, Evaluation Mode=Linear probing2025.01 | 89.66 | |
| Llama-2-7BModel Category=Llama 2 models, Evaluation Mode=Linear probing, Representation=Token averaging2025.01 | 88.23 | |
| DeBERTa-LargeModel Category=Encoder models, Evaluation Mode=Linear probing, Representation=Token averaging2025.01 | 85.74 | |
| XLNet-LargeModel Category=Encoder models, Evaluation Mode=Linear probing, Representation=Token averaging2025.01 | 79.7 | |
| BERT-LargeModel Category=Encoder models, Evaluation Mode=Linear probing, Representation=Token averaging2025.01 | 71.77 |