How to specify sequence length when using "feature-extraction"

🤗Transformers
2

what happens if you specify model_max_len=512 when you load the tokenizer? i’d try that and do a sanity check with tokenizer(text) to make sure the truncation is working as expected