Update README.md

README.md

@@ -28,7 +28,7 @@ It is also possible to evaluate the model with https://github.com/EleutherAI/lm-
 ## Training
 - For the Llama2 training data, we refer the user to https://huggingface.co/meta-llama/Llama-2-70b-hf.
 - For continued pretraining, 4.2B climate domain tokens (tokenized by the Llama tokenizer) are used.
-- For instruction finetuning, about 1.1B instruction-finetuning tokens (both in the climate domain but also general domain) are used.
+- For instruction finetuning, about 272K instruction-completion pairs (both in the climate domain but also general domain) are used.
 
 ## Environmental Impact
 - **Hardware Type:** H100