|
--- |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
base_model: asafaya/bert-base-arabic |
|
model-index: |
|
- name: sagemaker-bert-base-arabic-ArSAS |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# sagemaker-bert-base-arabic-ArSAS |
|
|
|
This model is a fine-tuned version of [asafaya/bert-base-arabic](https://huggingface.co/asafaya/bert-base-arabic) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.2592 |
|
- Accuracy: 0.9038 |
|
|
|
## 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: 32 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- num_epochs: 2 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| No log | 1.0 | 291 | 0.3027 | 0.8806 | |
|
| 0.3749 | 2.0 | 582 | 0.2592 | 0.9038 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.12.3 |
|
- Pytorch 1.9.1 |
|
- Datasets 1.15.1 |
|
- Tokenizers 0.10.3 |
|
|
