Update README.md

README.md

@@ -29,6 +29,8 @@ There are two different sized Llama models, all pretrained from scratch for 131B
 | [llama-3b-finnish](https://huggingface.co/Finnish-NLP/llama-3b-finnish)         | 2048           | 26     | 3200 | 32    | 3.6B   |
 | [llama-7b-finnish](https://huggingface.co/Finnish-NLP/llama-7b-finnish)         | 2048           | 32     | 4096 | 32    | 7.0B   |
 
+Note: this model is pretrained, not fine-tuned for instruction following or chat use. For instruction tuned models, you can check models listed in this collection: https://huggingface.co/collections/Finnish-NLP/instruction-tuned-models-65c08db79fa020161be2e942
+
 ## Model description
 
 Finnish Llama is a GPT-style model pretrained on a very large corpus of Finnish data in a self-supervised fashion. This