This adds a minimal bigram distribution over strings (#7 ), for now assuming the alphabet consists of printable ASCII characters.
I’m not at all confident in the logp_score implementation (summing the logp_scores of each conditional distribution), so I’d appreciate an extra pair of eyes on that especially.
This adds a minimal bigram distribution over strings (#7 ), for now assuming the alphabet consists of printable ASCII characters.
I’m not at all confident in the
logp_score
implementation (summing the logp_scores of each conditional distribution), so I’d appreciate an extra pair of eyes on that especially.