Update README.md

Author: Kyriection

recipes/experimental/long-context/H2O/README.md

@@ -36,7 +36,7 @@ Expected results on XSUM (Rouge-2 score, ther higher the better) from the above
 
 ### Evaluation on "Needle in a Haystack" Analysis
 
-The following example runs inference of Llama-3-8b-instruct on "Needle in a haystack" test. The test is modified from [https://github.com/gkamradt/LLMTest_NeedleInAHaystack](). Please follow the original repository for installing necessary packages. We're using `--enable_h2o_generation` to enable H2O algorithm that only keeps heavy-hitter and the local KV pairs. Use `--num_heavy_hitter_tokens` to decide the number of heavy-hitter KV pairs and `--num_window_length `for the KV cache size. The number of local KV pairs equals num_window_length - num_heavy_hitter_tokens. Also, use --enable_position_rolling to enable position rolling in the KV cache size that assign the positions in the KV cache instead of the ones in original sequences. Enabling postional rolling is important when sequence length exceeds the pretrained context windows, e.g., 4K in Llama-2.
+The following example runs inference of Llama-3-8b-instruct on "Needle in a haystack" test. The test is modified from [https://github.com/gkamradt/LLMTest_NeedleInAHaystack](). Please follow the original repository for installing necessary packages. We're using `--enable_h2o_generation` to enable H2O algorithm that only keeps heavy-hitter and the local KV pairs. Use `--num_window_length ` to decide the KV cache size. Also, use --enable_position_rolling to enable position rolling in the KV cache size that assign the positions in the KV cache instead of the ones in original sequences. Enabling postional rolling is important when sequence length exceeds the pretrained context windows, e.g., 8K in Llama-3.
 
 ```
 # step 1: obtain prompts for evaluation