Update README.md

README.md

@@ -13,6 +13,8 @@ tags:
 
 palmer-004 a is merge of models targetting to get the performance of palmer-003 all they way to 32k context window. It was merged using the [TIES](https://arxiv.org/abs/2306.01708) merge method using [palmer-002-32k](https://huggingface.co/appvoid/palmer-002-32k) as a base.
 
+palmer-004 performs better than coven_tiny_1.1b_32k_orpo_alpha which is the current sota at open-llm-leaderboard, making this one the best overall 1b model on huggingface as of 06/01/2024.
+
 The following models were included in the merge:
 * [coven_tiny_1.1b_32k_orpo_alpha](https://huggingface.co/raidhon/coven_tiny_1.1b_32k_orpo_alpha)
 * [palmer-003](https://huggingface.co/appvoid/palmer-003)