GitLab

[codemod] c10::optional -> std::optional in pyspeech/experimental/csrc/decoders/TransducerDecoder.h +20

Differential Revision: D57294284

Pull Request resolved: https://github.com/pytorch/audio/pull/3793

Differential Revision: D50086556

Pull Request resolved: https://github.com/pytorch/audio/pull/3648

Differential Revision: D49965263

Pull Request resolved: https://github.com/pytorch/audio/pull/3639

Summary:
Add `fused_log_softmax` argument (default/current behavior = True) to rnnt loss.

If setting it to `False`, call `log_softmax` on the logits prior to passing it in to the rnnt loss function.

The following should produce the same output:
```
rnnt_loss(logits, targets, logit_lengths, target_lengths, fused_log_softmax=True)
```

```
log_probs = torch.nn.functional.log_softmax(logits, dim=-1)
rnnt_loss(log_probs, targets, logit_lengths, target_lengths, fused_log_softmax=False)
```

testing -- unit tests + get same results on the conformer rnnt recipe

Pull Request resolved: https://github.com/pytorch/audio/pull/2798

Reviewed By: xiaohui-zhang

Differential Revision: D41083523

Pulled By: carolineechen

fbshipit-source-id: e15442ceed1f461bbf06b724aa0561ff8827ad61

Replace the prototype RNNT implementation (using warp-transducer) with one without external library dependencies