Update README.md
Browse files
README.md
CHANGED
|
@@ -22,6 +22,16 @@ this model is only 1B but you can call it somehow an SOTA
|
|
| 22 |
|
| 23 |
this model can also run on 4 GB GPU RAM and know dialogs as well
|
| 24 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
## Usage Code
|
| 26 |
|
| 27 |
```python
|
|
|
|
| 22 |
|
| 23 |
this model can also run on 4 GB GPU RAM and know dialogs as well
|
| 24 |
|
| 25 |
+
|
| 26 |
+
### Train Parametes
|
| 27 |
+
|
| 28 |
+
- learning-rate : 2e-4
|
| 29 |
+
- sc : cosine lr
|
| 30 |
+
- device : T4 GPU * 4
|
| 31 |
+
- batch-size: AutoFind
|
| 32 |
+
- train time 12 H
|
| 33 |
+
- max sequence length: 1024
|
| 34 |
+
- epochs : 2
|
| 35 |
## Usage Code
|
| 36 |
|
| 37 |
```python
|