Update README.md
#1
by
shafire
- opened
README.md
CHANGED
|
@@ -13,6 +13,8 @@ Native chain of thought approach means that Cerebrum is trained to devise a tact
|
|
| 13 |
|
| 14 |
Zero-shot prompted Cerebrum significantly outperforms few-shot prompted Mistral 7b as well as much larger models (such as Llama 2 70b) on a range of tasks that require reasoning, including ARC Challenge, GSM8k, and Math.
|
| 15 |
|
|
|
|
|
|
|
| 16 |
## Benchmarking
|
| 17 |
An overview of Cerebrum 7b performance compared to reported performance Mistral 7b and LLama 2 70b on selected benchmarks that require reasoning:
|
| 18 |
|
|
|
|
| 13 |
|
| 14 |
Zero-shot prompted Cerebrum significantly outperforms few-shot prompted Mistral 7b as well as much larger models (such as Llama 2 70b) on a range of tasks that require reasoning, including ARC Challenge, GSM8k, and Math.
|
| 15 |
|
| 16 |
+
This LLM model works a lot better than any other mistral mixtral models for agent data, tested on 14th March 2024.
|
| 17 |
+
|
| 18 |
## Benchmarking
|
| 19 |
An overview of Cerebrum 7b performance compared to reported performance Mistral 7b and LLama 2 70b on selected benchmarks that require reasoning:
|
| 20 |
|