BioNLP13CG_bioBERT_NER
This model is a fine-tuned version of dmis-lab/biobert-v1.1 on the None dataset. It achieves the following results on the evaluation set:
Loss: 0.1928
Seqeval classification report: precision recall f1-score support
Amino_acid 0.89 0.88 0.88 576 Anatomical_system 0.96 0.82 0.89 317 Cancer 0.92 0.91 0.91 1649 Cell 0.00 0.00 0.00 25 Cellular_component 0.00 0.00 0.00 12
Developing_anatomical_structure 0.75 0.85 0.80 438 Gene_or_gene_product 0.87 0.18 0.29 74 Immaterial_anatomical_entity 0.84 0.84 0.84 4142 Multi-tissue_structure 0.85 0.84 0.84 451 Organ 0.51 0.23 0.31 80 Organism 0.52 0.66 0.58 182 Organism_subdivision 0.81 0.80 0.81 314 Organism_substance 0.73 0.66 0.69 96 Pathological_formation 0.75 0.68 0.71 262 Simple_chemical 0.55 0.44 0.49 112 Tissue 0.82 0.91 0.87 300
micro avg 0.84 0.82 0.83 9030
macro avg 0.67 0.60 0.62 9030
weighted avg 0.84 0.82 0.83 9030
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Seqeval classification report |
|---|---|---|---|---|
| No log | 0.99 | 95 | 0.2929 | precision recall f1-score support |
Amino_acid 0.68 0.81 0.73 576
Anatomical_system 0.93 0.74 0.82 317
Cancer 0.89 0.89 0.89 1649
Cell 0.00 0.00 0.00 25
Cellular_component 0.00 0.00 0.00 12
Developing_anatomical_structure 0.56 0.79 0.65 438 Gene_or_gene_product 0.00 0.00 0.00 74 Immaterial_anatomical_entity 0.79 0.76 0.77 4142 Multi-tissue_structure 0.84 0.75 0.79 451 Organ 0.00 0.00 0.00 80 Organism 0.62 0.08 0.15 182 Organism_subdivision 0.64 0.78 0.70 314 Organism_substance 0.00 0.00 0.00 96 Pathological_formation 0.63 0.44 0.52 262 Simple_chemical 0.79 0.13 0.23 112 Tissue 0.82 0.45 0.58 300
micro avg 0.78 0.72 0.75 9030
macro avg 0.51 0.41 0.43 9030
weighted avg 0.76 0.72 0.73 9030
| | No log | 2.0 | 191 | 0.2053 | precision recall f1-score support
Amino_acid 0.87 0.87 0.87 576
Anatomical_system 0.98 0.80 0.88 317
Cancer 0.89 0.92 0.91 1649
Cell 0.00 0.00 0.00 25
Cellular_component 0.00 0.00 0.00 12
Developing_anatomical_structure 0.74 0.84 0.79 438 Gene_or_gene_product 1.00 0.05 0.10 74 Immaterial_anatomical_entity 0.83 0.83 0.83 4142 Multi-tissue_structure 0.85 0.82 0.83 451 Organ 0.48 0.15 0.23 80 Organism 0.49 0.66 0.56 182 Organism_subdivision 0.79 0.80 0.80 314 Organism_substance 0.75 0.58 0.65 96 Pathological_formation 0.76 0.66 0.71 262 Simple_chemical 0.48 0.42 0.45 112 Tissue 0.80 0.90 0.85 300
micro avg 0.82 0.82 0.82 9030
macro avg 0.67 0.58 0.59 9030
weighted avg 0.82 0.82 0.81 9030
| | No log | 2.98 | 285 | 0.1928 | precision recall f1-score support
Amino_acid 0.89 0.88 0.88 576
Anatomical_system 0.96 0.82 0.89 317
Cancer 0.92 0.91 0.91 1649
Cell 0.00 0.00 0.00 25
Cellular_component 0.00 0.00 0.00 12
Developing_anatomical_structure 0.75 0.85 0.80 438 Gene_or_gene_product 0.87 0.18 0.29 74 Immaterial_anatomical_entity 0.84 0.84 0.84 4142 Multi-tissue_structure 0.85 0.84 0.84 451 Organ 0.51 0.23 0.31 80 Organism 0.52 0.66 0.58 182 Organism_subdivision 0.81 0.80 0.81 314 Organism_substance 0.73 0.66 0.69 96 Pathological_formation 0.75 0.68 0.71 262 Simple_chemical 0.55 0.44 0.49 112 Tissue 0.82 0.91 0.87 300
micro avg 0.84 0.82 0.83 9030
macro avg 0.67 0.60 0.62 9030
weighted avg 0.84 0.82 0.83 9030
|
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 1