Update README.md
Browse files
README.md
CHANGED
|
@@ -4,7 +4,22 @@ language:
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
-
This model is finetuend on "mistralai/Mixtral-8x7B-v0.1" with Firefly
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
|
| 9 |
## Run the model
|
| 10 |
|
|
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
+
This model is finetuend based on "mistralai/Mixtral-8x7B-v0.1" with [Firefly](https://github.com/yangjianxin1/Firefly) and 48k data from ultrachat.
|
| 8 |
+
|
| 9 |
+
## Evaluation
|
| 10 |
+
Though we finetune with only 48k data, our model can also achieve excellent performance.
|
| 11 |
+
|
| 12 |
+
| Model | Open LLM Leaderboard |
|
| 13 |
+
|------------------------------------------------------------------------------------------------|---------------------------------------------|
|
| 14 |
+
| Qwen-72B | 73.6 |
|
| 15 |
+
| Mixtral-8x7B-Instruct-v0.1 | 72.62 |
|
| 16 |
+
|**Firefly-Mixtral-8x7B**|**70.34**|
|
| 17 |
+
|Yi-34B|69.42|
|
| 18 |
+
|Mixtral-8x7B-v0.1|68.42|
|
| 19 |
+
|Llama2-65B-Chat|67.87|
|
| 20 |
+
|Qwen-14B|65.86|
|
| 21 |
+
|Vicuna-33B-v1.3 |58.54|
|
| 22 |
+
|
| 23 |
|
| 24 |
## Run the model
|
| 25 |
|