Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,29 @@
|
|
1 |
---
|
|
|
2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
language: de
|
3 |
license: mit
|
4 |
+
tags:
|
5 |
+
- pytorch
|
6 |
+
- causal-lm
|
7 |
+
datasets:
|
8 |
+
- c4
|
9 |
---
|
10 |
+
|
11 |
+
# Cedille AI
|
12 |
+
Cedille is a project to bring large language models to non-English languages.
|
13 |
+
|
14 |
+
## fr-boris
|
15 |
+
Anna is a 6B parameter autoregressive language model based on the GPT-J architecture and trained using the [mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax) codebase.
|
16 |
+
|
17 |
+
Anna was trained on German text with a similar methodology to [Boris](https://huggingface.co/Cedille/fr-boris), our French model. We started training from GPT-J, which has been trained on [The Pile](https://pile.eleuther.ai/). As a consequence the model still has good performance in English language. Anna makes use of the unmodified GPT-2 tokenizer.
|
18 |
+
|
19 |
+
# How to run
|
20 |
+
TO DO
|
21 |
+
|
22 |
+
## Contact us
|
23 |
+
For any custom development please contact us at hello@cedille.ai.
|
24 |
+
|
25 |
+
## Links
|
26 |
+
* [Official website](https://en.cedille.ai/)
|
27 |
+
* [Blog](https://en.cedille.ai/blog)
|
28 |
+
* [GitHub](https://github.com/coteries/cedille-ai)
|
29 |
+
* [Twitter](https://twitter.com/CedilleAI)
|