Updated README

README.md

@@ -28,5 +28,19 @@ the_mask_pipe("Ejo ndikwiga nagize <mask> baje kunsura.")
 {'sequence': 'Ejo ndikwiga nagize ngo baje kunsura.', 'score': 0.032475441694259644, 'token': 396, 'token_str': ' ngo'}, 
 {'sequence': 'Ejo ndikwiga nagize abana baje kunsura.', 'score': 0.029481062665581703, 'token': 739, 'token_str': ' abana'}, 
 {'sequence': 'Ejo ndikwiga nagize abantu baje kunsura.', 'score': 0.016263306140899658, 'token': 500, 'token_str': ' abantu'}]
+```
+2) Direct use from the transformer library to get features using AutoModel
+
+```
+from transformers import AutoTokenizer, AutoModelForMaskedLM
+  
+tokenizer = AutoTokenizer.from_pretrained("jean-paul/kinyaRoberta-small")
+
+model = AutoModelForMaskedLM.from_pretrained("jean-paul/kinyaRoberta-small")
+
+input_text = "Ejo ndikwiga nagize abashyitsi baje kunsura."
+encoded_input = tokenizer(input_text, return_tensors='pt')
+output = model(**encoded_input)
+
 ```
 __Note__: We used the huggingface implementations for pretraining RoBerta from scratch, both the RoBerta model and the classes needed to do it.