Commit
·
3427a45
1
Parent(s):
3b5a709
Update README.md
Browse files
README.md
CHANGED
|
@@ -35,12 +35,12 @@ The performance of the model is the following:
|
|
| 35 |
The model is provided by [vitas.ai](https://www.vitas.ai/).
|
| 36 |
|
| 37 |
## Pipeline description
|
| 38 |
-
This
|
| 39 |
|
| 40 |
- Pretrained Llama2 Tokenizer that transforms words into subwords.
|
| 41 |
- Llama2 to generate the next sentence given the history of the dialogue.
|
| 42 |
|
| 43 |
-
The system is trained with dialogue from MultiWOZ corpus.
|
| 44 |
|
| 45 |
|
| 46 |
## Install SpeechBrain
|
|
@@ -57,7 +57,7 @@ pip install -r recipes/MultiWOZ/response_generation/llama2/extra_requirements.tx
|
|
| 57 |
Please notice that we encourage you to read our tutorials and learn more about
|
| 58 |
[SpeechBrain](https://speechbrain.github.io).
|
| 59 |
|
| 60 |
-
### Generating your Own
|
| 61 |
|
| 62 |
```python
|
| 63 |
from speechbrain.inference.text import Llama2ResponseGenerator
|
|
|
|
| 35 |
The model is provided by [vitas.ai](https://www.vitas.ai/).
|
| 36 |
|
| 37 |
## Pipeline description
|
| 38 |
+
This dialogue system is composed of 2 different but linked blocks:
|
| 39 |
|
| 40 |
- Pretrained Llama2 Tokenizer that transforms words into subwords.
|
| 41 |
- Llama2 to generate the next sentence given the history of the dialogue.
|
| 42 |
|
| 43 |
+
The system is trained with dialogue from the MultiWOZ corpus.
|
| 44 |
|
| 45 |
|
| 46 |
## Install SpeechBrain
|
|
|
|
| 57 |
Please notice that we encourage you to read our tutorials and learn more about
|
| 58 |
[SpeechBrain](https://speechbrain.github.io).
|
| 59 |
|
| 60 |
+
### Generating your Own Dialogue
|
| 61 |
|
| 62 |
```python
|
| 63 |
from speechbrain.inference.text import Llama2ResponseGenerator
|