distilroberta-base-ner-wikiann-conll2003-4-class
This model is a fine-tuned version of distilroberta-base on the wikiann and conll2003 dataset. It consists out of the classes of conll2003.
O (0), B-PER (1), I-PER (2), B-ORG (3), I-ORG (4) B-LOC (5), I-LOC (6) B-MISC (7), I-MISC (8).
eval F1-Score: 95,39 (merged dataset)
test F1-Score: 90,75 (merged dataset)
Model Usage
from transformers import AutoTokenizer, AutoModelForTokenClassification
from transformers import pipeline
tokenizer = AutoTokenizer.from_pretrained("philschmid/distilroberta-base-ner-wikiann-conll2003-4-class")
model = AutoModelForTokenClassification.from_pretrained("philschmid/distilroberta-base-ner-wikiann-conll2003-4-class")
nlp = pipeline("ner", model=model, tokenizer=tokenizer, grouped_entities=True)
example = "My name is Philipp and live in Germany"
nlp(example)
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 4.9086903597787154e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
It achieves the following results on the evaluation set:
- Loss: 0.0705
- Precision: 0.9492
- Recall: 0.9585
- F1: 0.9539
- Accuracy: 0.9882
It achieves the following results on the test set:
- Loss: 0.239
- Precision: 0.8984
- Recall: 0.9168
- F1: 0.9075
- Accuracy: 0.9741
Framework versions
- Transformers 4.6.1
- Pytorch 1.8.1+cu101
- Datasets 1.6.2
- Tokenizers 0.10.2
- Downloads last month
- 20
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
๐
Ask for provider support
Evaluation results
- Precision on wikiann-conll2003self-reported0.949
- Recall on wikiann-conll2003self-reported0.959
- F1 on wikiann-conll2003self-reported0.954
- Accuracy on wikiann-conll2003self-reported0.988