metadata
base_model: demdecuong/vihealthbert-base-word
tags:
- generated_from_trainer
datasets:
- tmnam20/pretrained-vn-med-nli
metrics:
- accuracy
model-index:
- name: vihealthbert-w_unsup-SynPD
results:
- task:
name: Masked Language Modeling
type: fill-mask
dataset:
name: tmnam20/pretrained-vn-med-nli all
type: tmnam20/pretrained-vn-med-nli
args: all
metrics:
- name: Accuracy
type: accuracy
value: 0.686153705209395
vihealthbert-w_unsup-SynPD
This model is a fine-tuned version of demdecuong/vihealthbert-base-word on the tmnam20/pretrained-vn-med-nli all dataset. It achieves the following results on the evaluation set:
- Loss: 1.5768
- Accuracy: 0.6862
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 21363
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 7.0234 | 0.8616 | 5000 | 2.5909 | 0.5576 |
| 5.2736 | 1.7232 | 10000 | 2.1890 | 0.5962 |
| 4.9126 | 2.5849 | 15000 | 1.9095 | 0.6381 |
| 4.791 | 3.4465 | 20000 | 1.8286 | 0.6469 |
| 4.6538 | 4.3081 | 25000 | 1.7144 | 0.6644 |
| 4.5846 | 5.1697 | 30000 | 1.6779 | 0.6704 |
| 4.5568 | 6.0314 | 35000 | 1.6362 | 0.6766 |
| 4.5079 | 6.8930 | 40000 | 1.6008 | 0.6814 |
| 4.469 | 7.7546 | 45000 | 1.6064 | 0.6805 |
| 4.4514 | 8.6162 | 50000 | 1.5800 | 0.6852 |
| 4.4317 | 9.4779 | 55000 | 1.5540 | 0.6880 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.0.1+cu118
- Datasets 2.19.1
- Tokenizers 0.19.1