Problème avec passage tokenizer

[Rollybre]

tokenizer = AutoTokenizer.from_pretrained("camembert-base", use_fast=True)

fb_emotion = Fabula(scorer=TransformersScorer(model="astrosbd/french_emotion_camembert"), 
                    segmenter=SlidingWindowTokenSegmenter(tokenizer = tokenizer,window_tokens=512, stride_tokens=256),
                    analysis="emotion")

Output :
Token indices sequence length is longer than the specified maximum sequence length for this model (14046 > 512). Running this sequence through the model will result in indexing errors

Je suppose un problème au niveau du self.tokeize dans segment.py, j'investigue