|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- precision |
|
- recall |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: afro-xlmr-base-hausa-seed-30 |
|
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. --> |
|
|
|
# afro-xlmr-base-hausa-seed-30 |
|
|
|
This model is a fine-tuned version of [Davlan/afro-xlmr-base](https://huggingface.co/Davlan/afro-xlmr-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.1635 |
|
- Precision: 0.7407 |
|
- Recall: 0.5630 |
|
- F1: 0.6398 |
|
- Accuracy: 0.9599 |
|
|
|
## 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: 16 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 5 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
| 0.1599 | 1.0 | 1312 | 0.1431 | 0.7178 | 0.4516 | 0.5544 | 0.9536 | |
|
| 0.1198 | 2.0 | 2624 | 0.1364 | 0.7155 | 0.5470 | 0.6200 | 0.9581 | |
|
| 0.0932 | 3.0 | 3936 | 0.1381 | 0.7165 | 0.5708 | 0.6354 | 0.9588 | |
|
| 0.0705 | 4.0 | 5248 | 0.1564 | 0.7529 | 0.5461 | 0.6330 | 0.9600 | |
|
| 0.0559 | 5.0 | 6560 | 0.1635 | 0.7407 | 0.5630 | 0.6398 | 0.9599 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.30.2 |
|
- Pytorch 2.0.1+cu118 |
|
- Datasets 2.13.1 |
|
- Tokenizers 0.13.3 |
|
|
