Update README.md
Browse files
README.md
CHANGED
@@ -23,8 +23,17 @@ The Cogito v2 LLMs are instruction tuned generative models. All models are relea
|
|
23 |
- This model is trained in over 30 languages and supports a context length of 128k.
|
24 |
|
25 |
# Evaluations
|
26 |
-
|
27 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
|
29 |
# Usage
|
30 |
Here is a snippet below for usage with Transformers:
|
|
|
23 |
- This model is trained in over 30 languages and supports a context length of 128k.
|
24 |
|
25 |
# Evaluations
|
26 |
+
Here is the model performance on some standard industry benchmarks:
|
27 |
|
28 |
+
<p align="left">
|
29 |
+
<img src="images/cogito-v2-671b-non-reasoning-benchmarks.png" alt="Logo" width="90%">
|
30 |
+
</p>
|
31 |
+
|
32 |
+
<p align="left">
|
33 |
+
<img src="images/cogito-v2-671b-reasoning-benchmarks.png" alt="Logo" width="90%">
|
34 |
+
</p>
|
35 |
+
|
36 |
+
For detailed evaluations, please refer to the [Blog Post](https://www.deepcogito.com/research/cogito-v2-preview).
|
37 |
|
38 |
# Usage
|
39 |
Here is a snippet below for usage with Transformers:
|