drbert-7gb-finedtuned-ner
This model is a fine-tuned version of Dr-BERT/DrBERT-7GB on the quaero dataset. It achieves the following results on the evaluation set:
- Loss: 1.2330
- Precision: 0.5055
- Recall: 0.5696
- F1: 0.5357
- Accuracy: 0.8004
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 105 | 0.7430 | 0.4129 | 0.4775 | 0.4428 | 0.7671 |
No log | 2.0 | 210 | 0.6968 | 0.4888 | 0.5042 | 0.4964 | 0.7888 |
No log | 3.0 | 315 | 0.8218 | 0.5059 | 0.5323 | 0.5188 | 0.7952 |
No log | 4.0 | 420 | 0.9307 | 0.4869 | 0.5563 | 0.5193 | 0.7913 |
0.4134 | 5.0 | 525 | 0.9970 | 0.4688 | 0.5581 | 0.5095 | 0.7870 |
0.4134 | 6.0 | 630 | 1.0503 | 0.4992 | 0.5541 | 0.5252 | 0.7930 |
0.4134 | 7.0 | 735 | 1.1364 | 0.5034 | 0.5607 | 0.5305 | 0.7994 |
0.4134 | 8.0 | 840 | 1.1994 | 0.4865 | 0.5701 | 0.5250 | 0.7937 |
0.4134 | 9.0 | 945 | 1.2287 | 0.4948 | 0.5683 | 0.5290 | 0.7982 |
0.028 | 10.0 | 1050 | 1.2330 | 0.5055 | 0.5696 | 0.5357 | 0.8004 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.1.0+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2
- Downloads last month
- 13
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for ytonthat/drbert-7gb-finedtuned-ner
Base model
Dr-BERT/DrBERT-7GB
Finetuned
this model
Evaluation results
- Precision on quaerovalidation set self-reported0.506
- Recall on quaerovalidation set self-reported0.570
- F1 on quaerovalidation set self-reported0.536
- Accuracy on quaerovalidation set self-reported0.800