Update README.md
Browse files
README.md
CHANGED
@@ -104,6 +104,12 @@ This classification model is built on top of [EUBERT](https://huggingface.co/Eur
|
|
104 |
|
105 |
With less than 100 million parameters, it can be deployed on commodity hardware without GPU acceleration (around 200 ms per inference for 2000 characters).
|
106 |
|
|
|
|
|
|
|
|
|
|
|
|
|
107 |
## Usage
|
108 |
|
109 |
|
|
|
104 |
|
105 |
With less than 100 million parameters, it can be deployed on commodity hardware without GPU acceleration (around 200 ms per inference for 2000 characters).
|
106 |
|
107 |
+
Parameters :
|
108 |
+
- Number of epochs 10
|
109 |
+
- Batch size 10
|
110 |
+
- Max lenght 512
|
111 |
+
- Learning Rate 5e-05
|
112 |
+
|
113 |
## Usage
|
114 |
|
115 |
|