Update README.md

README.md

@@ -4,4 +4,4 @@ datasets:
 - yizhongw/self_instruct
 ---
 
-Finetuned LLaMA-7B with LoRa using 82K [Self-Instruct](https://arxiv.org/abs/2212.10560) instances. The instances were filtered using relevancy classifier targeting alignment between instruction + input and output.
+Finetuned LLaMA-7B with LoRa using [Self-Instruct](https://arxiv.org/abs/2212.10560) instances. The instances were filtered using relevancy classifier targeting alignment between instruction + input and output.