Update README.md

README.md

@@ -4,8 +4,19 @@ library_name: peft
 ---
 ## Training procedure
 
-It looks like overfitting has been circumvented. Generalization capabilities based on [this notebook](https://huggingface.co/AmelieSchreiber/esm2_t12_35M_lora_binding_sites_770K_v1/blob/main/esmbind-validation-struct.ipynb)
-are still lacking though.  
+Generalization capabilities based on [this notebook](https://huggingface.co/AmelieSchreiber/esm2_t12_35M_lora_binding_sites_770K_v1/blob/main/esmbind-validation-struct.ipynb)
+are still lacking. The average metrics on the train datasets from the paper linked to in the notebook are:
+
+```
+Average Accuracy: 0.9169
+Average Precision: 0.1416
+Average Recall: 0.2769
+Average F1 Score: 0.1743
+Average AUC: 0.6088
+Average MCC: 0.1522
+```
+
+The average metrics on the test datasets provided by the paper are very similar. 
 
 ### Framework versions