Update README.md
Browse files
README.md
CHANGED
|
@@ -4,23 +4,46 @@ language:
|
|
| 4 |
- en
|
| 5 |
- de
|
| 6 |
- fr
|
|
|
|
| 7 |
- pt
|
| 8 |
-
- es
|
| 9 |
-
- it
|
| 10 |
- nl
|
| 11 |
- ru
|
| 12 |
-
- zh
|
| 13 |
- ko
|
|
|
|
|
|
|
| 14 |
metrics:
|
| 15 |
- comet
|
| 16 |
-
pipeline_tag:
|
| 17 |
---
|
| 18 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
-
##
|
| 21 |
|
| 22 |
-
|
| 23 |
|
| 24 |
-
##
|
| 25 |
|
| 26 |
-
|
|
|
|
| 4 |
- en
|
| 5 |
- de
|
| 6 |
- fr
|
| 7 |
+
- zh
|
| 8 |
- pt
|
|
|
|
|
|
|
| 9 |
- nl
|
| 10 |
- ru
|
|
|
|
| 11 |
- ko
|
| 12 |
+
- it
|
| 13 |
+
- es
|
| 14 |
metrics:
|
| 15 |
- comet
|
| 16 |
+
pipeline_tag: translation
|
| 17 |
---
|
| 18 |
+
# Model Card for TowerBase-7B-v0.1
|
| 19 |
+
|
| 20 |
+
## Model Details
|
| 21 |
+
|
| 22 |
+
### Model Description
|
| 23 |
+
|
| 24 |
+
TowerBase-7B is a language model that results from continuing the pretraining of Llama 2 on a mix of 20 billion tokens of monolingual data from non-English languages, and bilingual data. TowerBase-7B-v0.1 is the first model in the series.
|
| 25 |
+
The resulting model shows improved performance on the supported languages, while maintaining Llama's capabilities on English. It is particularly well-suited for fine-tuning on translation and related tasks: check out [TowerInstruct](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1)
|
| 26 |
+
We will release more details in the upcoming technical report.
|
| 27 |
+
|
| 28 |
+
- **Developed by:** Unbabel, Instituto Superior Técnico, CentraleSupélec University of Paris-Saclay
|
| 29 |
+
- **Model type:** A 7B parameter model built on top of Llama 2 by continuing pretraining on multilingual data.
|
| 30 |
+
- **Language(s) (NLP):** English, Portuguese, Spanish, French, German, Dutch, Italian, Korean, Chinese, Russian
|
| 31 |
+
- **License:** CC-BY-NC-4.0
|
| 32 |
+
|
| 33 |
+
## Intended uses & limitations
|
| 34 |
+
|
| 35 |
+
The model is intended for research purposes in the 10 languages it supports.
|
| 36 |
+
The model is able to perform well on translation and related tasks (e.g., APE, GEC) on a few-shot regime.
|
| 37 |
+
It can also be fine-tuned to perform these tasks in a zero-shot fashion (see [TowerInstruct](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1), as well as other multilingual tasks.
|
| 38 |
+
|
| 39 |
+
### Out-of-Scope Use
|
| 40 |
+
|
| 41 |
+
The model is not guaranteed to perform well for languages other than the 10 languages it supports.
|
| 42 |
|
| 43 |
+
## Bias, Risks, and Limitations
|
| 44 |
|
| 45 |
+
TowerBase-v0.1 has not been aligned to human preferences, so the model may generate problematic outputs (e.g., hallucinations, harmful content, or false statements).
|
| 46 |
|
| 47 |
+
## Citation
|
| 48 |
|
| 49 |
+
To be completed.
|