dont project maske tokens for mlm loss (#859)

Summary:
This saves ~4-5gb gpu memory while training roberta large with `seq_len=512`.

I am able to fit `--max-sentences=16` on `volta32gb` for `roberta-large`
Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/859

Differential Revision: D17435814

fbshipit-source-id: 2663909768fac0ef0102107613770ee01b1f8c00

fairseq/criterions/masked_lm.py

@@ -30,8 +30,11 @@ class MaskedLmLoss(FairseqCriterion):
         3) logging outputs to display while training
         """
         # compute MLM loss
-        logits = model(**sample['net_input'], return_all_hiddens=False)[0]
+        masked_tokens = sample['target'].ne(self.padding_idx)
+        logits = model(**sample['net_input'], masked_tokens=masked_tokens)[0]
         targets = model.get_targets(sample, [logits])
+        targets = targets[masked_tokens]
+
         loss = F.nll_loss(
             F.log_softmax(
                 logits.view(-1, logits.size(-1)),
@@ -43,7 +46,7 @@ class MaskedLmLoss(FairseqCriterion):
             ignore_index=self.padding_idx,
         )
 
-        sample_size = targets.ne(self.padding_idx).int().sum().item()
+        sample_size = masked_tokens.int().sum().item()
 
         logging_output = {
             'loss': utils.item(loss.data) if reduce else loss.data,
@@ -64,6 +67,7 @@ class MaskedLmLoss(FairseqCriterion):
 
         agg_output = {
             'loss': loss / sample_size / math.log(2),
+            'nll_loss': sum(log.get('nll_loss', 0) for log in logging_outputs) / sample_size / math.log(2) if ntokens > 0 else 0.,
             'ntokens': ntokens,
             'nsentences': nsentences,
             'sample_size': sample_size,

fairseq/models/roberta/model.py

@@ -201,14 +201,17 @@ class RobertaLMHead(nn.Module):
         self.weight = weight
         self.bias = nn.Parameter(torch.zeros(output_dim))
 
-    def forward(self, features, **kwargs):
+    def forward(self, features, masked_tokens=None, **kwargs):
+        # Only project the unmasked tokens while training,
+        # saves both memory and computation
+        if masked_tokens is not None:
+            features = features[masked_tokens, :]
+
         x = self.dense(features)
         x = self.activation_fn(x)
         x = self.layer_norm(x)
-
         # project back to size of vocabulary with bias
         x = F.linear(x, self.weight) + self.bias
-
         return x
 
 
@@ -265,7 +268,7 @@ class RobertaEncoder(FairseqDecoder):
             weight=self.sentence_encoder.embed_tokens.weight,
         )
 
-    def forward(self, src_tokens, features_only=False, return_all_hiddens=False, **unused):
+    def forward(self, src_tokens, features_only=False, return_all_hiddens=False, masked_tokens=None, **unused):
         """
         Args:
             src_tokens (LongTensor): input tokens of shape `(batch, src_len)`
@@ -283,7 +286,7 @@ class RobertaEncoder(FairseqDecoder):
         """
         x, extra = self.extract_features(src_tokens, return_all_hiddens)
         if not features_only:
-            x = self.output_layer(x)
+            x = self.output_layer(x, masked_tokens=masked_tokens)
         return x, extra
 
     def extract_features(self, src_tokens, return_all_hiddens=False, **unused):
@@ -293,8 +296,8 @@ class RobertaEncoder(FairseqDecoder):
         features = inner_states[-1]
         return features, {'inner_states': inner_states if return_all_hiddens else None}
 
-    def output_layer(self, features, **unused):
-        return self.lm_head(features)
+    def output_layer(self, features, masked_tokens=None, **unused):
+        return self.lm_head(features, masked_tokens)
 
     def max_positions(self):
         """Maximum output length supported by the encoder."""