|
--- |
|
license: mit |
|
datasets: |
|
- uvci/Koumankan_mt_dyu_fr |
|
language: |
|
- fr |
|
metrics: |
|
- bleu |
|
--- |
|
|
|
# Model Card for a Machine Translation Model for Dyula to French |
|
|
|
<!-- Provide a quick summary of what the model is/does. --> |
|
|
|
This modelcard aims to provide information on how the Dyula to French machine translation built using Tf-Keras |
|
|
|
## Model Details |
|
|
|
### Model Description |
|
|
|
For more information concerning the model check the following [notebook implementation](https://github.com/Ronnie-Leon76/Dyula-French-Neural-Machine-Translation/blob/main/dyula_to_french_neural_machine_translation_model_colab_amini_gpu.ipynb) |
|
|
|
## Usage |
|
|
|
### Load and use for inference |
|
|
|
- Create a directory `translator` which should contain the Huggingface repo model weights and assets. |
|
|
|
```python |
|
import tensorflow as tf |
|
reloaded = tf.saved_model.load('translator') |
|
reloaded("a y'a dabla ka yɛlɛ n na").numpy() |
|
``` |
|
|
|
- **Developed by:** Adeptschneider |
|
- **Model type:** Transformer |
|
- **Language(s) (NLP):** Python |
|
- **License:** MIT |
