ont-seqdist-cuda101 0.0.4

Probability distributions over sequences in pytorch and cupy

Seqdist

Probability distributions over sequences in pytorch and cupy.

Install

pip install seqdist

How to use

Comparison against builtin pytorch implementation of the standard CTC loss:

sample_inputs = logits, targets, input_lengths, target_lengths = ctc.generate_sample_inputs(T_min=450, T_max=500, N=128, C=20, L_min=80, L_max=100)
print('pytorch loss: {:.4f}'.format(ctc.loss_pytorch(*sample_inputs)))
print('seqdist loss: {:.4f}'.format(ctc.loss_cupy(*sample_inputs)))

pytorch loss: 12.8080
seqdist loss: 12.8080

Speed comparison

Pytorch:

report(benchmark_fwd_bwd(ctc.loss_pytorch, *sample_inputs))

fwd: 4.79ms (4.17-5.33ms)
bwd: 9.69ms (8.33-10.88ms)
tot: 14.47ms (12.67-16.20ms)

Seqdist:

report(benchmark_fwd_bwd(ctc.loss_cupy, *sample_inputs))

fwd: 7.22ms (6.78-7.85ms)
bwd: 6.21ms (5.82-8.57ms)
tot: 13.43ms (12.63-16.41ms)

Alignments

betas = [0.1, 1.0, 10.]
alignments = {'beta={:.1f}'.format(beta): to_np(ctc.soft_alignments(*sample_inputs, beta=beta)) for beta in betas}
alignments['viterbi'] = to_np(ctc.viterbi_alignments(*sample_inputs))
fig, axs = plt.subplots(2, 2, figsize=(15, 8))
for (ax, (title, data)) in zip(np.array(axs).flatten(), alignments.items()):
    ax.imshow(data[:, 0].T, vmax=0.05);
    ax.set_title(title)