Run the forward pass for a encoder-only model. Feeds a batch of tokens through the encoder to generate features. Args: src_tokens (LongTensor): input tokens of shape `(batch, src_len)` src_lengths (LongTensor): source sentence lengths of shape `(bat
(self, src_tokens, src_lengths, **kwargs)
| 529 | assert isinstance(self.encoder, FairseqEncoder) |
| 530 | |
| 531 | def forward(self, src_tokens, src_lengths, **kwargs): |
| 532 | """ |
| 533 | Run the forward pass for a encoder-only model. |
| 534 | |
| 535 | Feeds a batch of tokens through the encoder to generate features. |
| 536 | |
| 537 | Args: |
| 538 | src_tokens (LongTensor): input tokens of shape `(batch, src_len)` |
| 539 | src_lengths (LongTensor): source sentence lengths of shape `(batch)` |
| 540 | |
| 541 | Returns: |
| 542 | the encoder's output, typically of shape `(batch, src_len, features)` |
| 543 | """ |
| 544 | return self.encoder(src_tokens, src_lengths, **kwargs) |
| 545 | |
| 546 | def get_normalized_probs(self, net_output, log_probs, sample=None): |
| 547 | """Get normalized probabilities (or log probs) from a net's output.""" |