Update readme for version 4

README.md

@@ -13,13 +13,6 @@
   <img src="https://github.com/aimat-lab/gcnn_keras/blob/master/docs/source/_static/icon.svg" height="80"/>
 </p>
 
-> [!IMPORTANT]  
-> The versions of kgcnn<=3.1.0 were focused on ragged tensors of tensorflow.
-> The current main branch is developed for Keras 3.0 and kgcnn version 4.0. 
-> Please use last release of 3.1.0 for previous version of kgcnn.
-> 
-> We hope that all hyperparamter for models will also transfer to kgcnn 4.0.
-> You only have to add `input_tensor_type: "ragged"` to your model config.
 
 [General](#general) | [Requirements](#requirements) | [Installation](#installation) | [Documentation](#documentation) | [Implementation details](#implementation-details)
  | [Literature](#literature) | [Data](#data)  | [Datasets](#datasets) | [Training](#training) | [Issues](#issues) | [Citing](#citing) | [References](#references)
@@ -35,6 +28,9 @@ A [documentation](https://kgcnn.readthedocs.io/en/latest/index.html) is generate
 Focus of [kgcnn](kgcnn) is (batched) graph learning for molecules [kgcnn.molecule](kgcnn/molecule) and materials [kgcnn.crystal](kgcnn/crystal).
 If you want to get in contact, feel free to [discuss](https://github.com/aimat-lab/gcnn_keras/discussions). 
 
+Note that kgcnn>=4.0.0 requires keras>=3.0.0. Previous versions of kgcnn were focused on ragged tensors of tensorflow, for which
+hyperparameter for models should also transfer to kgcnn 4.0 by adding `input_tensor_type: "ragged"` and checking the order and *dtype* of inputs.
+
 <a name="requirements"></a>
 # Requirements