Good enough. The comparison is silly though: time series data is not anything like tokenized text data.