Update README.md

README.md

@@ -8,7 +8,7 @@ license: llama2
 ---
 # Code Llama for Petals
 
-Resharded Code Llama repository optimized for Petals inference. Instead of having 7 of ~10GiB each, the current repository has 49 shards each one of ~1.5GiB.
+Resharded Code Llama repository optimized for Petals inference. Instead of having 7 shards of ~10GiB each, the current repository has 49 shards each one of ~1.5GiB.
 
 For more information about the model, you can check the official card [here](https://huggingface.co/codellama/CodeLlama-34b-Instruct-hf).