ESM Fast Tokenizer

Is there a FastTokenizer implementation for ESM models?