Forward-backward algorithm computed in log domain. Args: state_trans_log_probs: tensor of shape [states, states] or if different transition matrix per batch [batch_size, states, states] initial_state_log_probs: tensor of shape [batch_size, states] final_state_log_probs: tensor o
(state_trans_log_probs, initial_state_log_probs,
final_state_log_probs, observed_log_probs,
sequence_length)
source not stored for this graph (policy: none)
no test coverage detected