|
--- |
|
license: mit |
|
base_model: emilyalsentzer/Bio_ClinicalBERT |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- sem_eval_2024_task_2 |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: run1 |
|
results: |
|
- task: |
|
name: Text Classification |
|
type: text-classification |
|
dataset: |
|
name: sem_eval_2024_task_2 |
|
type: sem_eval_2024_task_2 |
|
config: sem_eval_2024_task_2_source |
|
split: validation |
|
args: sem_eval_2024_task_2_source |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.575 |
|
- name: Precision |
|
type: precision |
|
value: 0.5800000000000001 |
|
- name: Recall |
|
type: recall |
|
value: 0.575 |
|
- name: F1 |
|
type: f1 |
|
value: 0.5682539682539682 |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# run1 |
|
|
|
This model is a fine-tuned version of [emilyalsentzer/Bio_ClinicalBERT](https://huggingface.co/emilyalsentzer/Bio_ClinicalBERT) on the sem_eval_2024_task_2 dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.2723 |
|
- Accuracy: 0.575 |
|
- Precision: 0.5800 |
|
- Recall: 0.575 |
|
- F1: 0.5683 |
|
|
|
## 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: 64 |
|
- 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: 20 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| No log | 0.99 | 53 | 0.6960 | 0.52 | 0.5339 | 0.52 | 0.4652 | |
|
| 0.7039 | 2.0 | 107 | 0.6921 | 0.545 | 0.5451 | 0.5450 | 0.5447 | |
|
| 0.7039 | 2.99 | 160 | 0.6956 | 0.515 | 0.5436 | 0.515 | 0.4198 | |
|
| 0.6979 | 4.0 | 214 | 0.6857 | 0.555 | 0.5587 | 0.555 | 0.5479 | |
|
| 0.6979 | 4.99 | 267 | 0.6924 | 0.51 | 0.7525 | 0.51 | 0.3552 | |
|
| 0.6831 | 6.0 | 321 | 0.6603 | 0.575 | 0.5750 | 0.575 | 0.5750 | |
|
| 0.6831 | 6.99 | 374 | 0.6572 | 0.61 | 0.6116 | 0.6100 | 0.6086 | |
|
| 0.6346 | 8.0 | 428 | 0.6517 | 0.57 | 0.5700 | 0.5700 | 0.5700 | |
|
| 0.6346 | 8.99 | 481 | 0.7185 | 0.58 | 0.5849 | 0.58 | 0.5739 | |
|
| 0.5337 | 10.0 | 535 | 0.8220 | 0.565 | 0.5767 | 0.565 | 0.5478 | |
|
| 0.5337 | 10.99 | 588 | 0.8002 | 0.595 | 0.5958 | 0.595 | 0.5942 | |
|
| 0.4262 | 12.0 | 642 | 0.8661 | 0.595 | 0.5994 | 0.595 | 0.5905 | |
|
| 0.4262 | 12.99 | 695 | 0.9989 | 0.555 | 0.5608 | 0.5550 | 0.5440 | |
|
| 0.3379 | 14.0 | 749 | 1.0688 | 0.56 | 0.5651 | 0.5600 | 0.5512 | |
|
| 0.3379 | 14.99 | 802 | 1.0439 | 0.585 | 0.5865 | 0.585 | 0.5832 | |
|
| 0.2846 | 16.0 | 856 | 1.1091 | 0.575 | 0.5809 | 0.575 | 0.5671 | |
|
| 0.2846 | 16.99 | 909 | 1.2667 | 0.57 | 0.5791 | 0.5700 | 0.5572 | |
|
| 0.228 | 18.0 | 963 | 1.2367 | 0.58 | 0.5858 | 0.58 | 0.5728 | |
|
| 0.228 | 18.99 | 1016 | 1.2373 | 0.585 | 0.5889 | 0.585 | 0.5804 | |
|
| 0.2137 | 19.81 | 1060 | 1.2723 | 0.575 | 0.5800 | 0.575 | 0.5683 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|
