Please fill in the following information:

Issue description:  

 "Granite Model Family" needs some editing to avoid covering the same information twice in the same paragraph:

"IBM increased the context window from the 4k tokens on Granite 3.0, up to 128k tokens on Granite 3.1 and 3.2. This increase in context window size enables models to handle big inputs, prolonged dialogues, and consider more context for each output. Context window sizes not only changes between versions, it can vary across different models of the same Granite version. For example, Granite 3.0 has models that support up to 4k tokens, and both Granite 3.1 and 3.2 have models with 128k token context windows. Bigger context windows enable use cases such as multi-document question-answering or repository-level code understanding." 

Steps to reproduce:

Workaround:

Expected result: