Update README.md

README.md

@@ -8,6 +8,12 @@ This model is the Mistral-7b model finetuned for 1k steps with a combined lm los
 
 (update: not very good)
 
+| model | avg | arc | hellaswag | mmlu | truthfulqa | winogrande | gsm8k |
+| --- | --- | --- | --- | --- | --- | --- | --- |
+| apricot-wildflower-20 | 59.74 | 59.64 | 81.76 | 63.38 | 41.76 | 77.9 | 33.97 |
+| mistralai/Mistral-7B-v0.1 | 60.97 | 59.98 | 83.31 | 64.16 | 42.15 | 78.37 | 37.83 |
+
+
 ### use
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer