log valid loss, time dataset process

deepspeech/exps/deepspeech2/model.py

@@ -46,7 +46,6 @@ class DeepSpeech2Trainer(Trainer):
         super().__init__(config, args)
 
     def train_batch(self, batch_index, batch_data, msg):
-        self.model.train()
         start = time.time()
 
         loss = self.model(*batch_data)
@@ -100,6 +99,8 @@ class DeepSpeech2Trainer(Trainer):
                 self.visualizer.add_scalar("valid/{}".format(k), v,
                                            self.iteration)
 
+        return valid_losses
+
     def setup_model(self):
         config = self.config
         model = DeepSpeech2Model(

deepspeech/exps/u2/bin/test.py

@@ -12,6 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Evaluation for U2 model."""
+import os
+import cProfile
 
 from deepspeech.training.cli import default_argument_parser
 from deepspeech.utils.utility import print_arguments
@@ -48,4 +50,7 @@ if __name__ == "__main__":
         with open(args.dump_config, 'w') as f:
             print(config, file=f)
 
-    main(config, args)
+    # Setting for profiling
+    pr = cProfile.Profile()
+    pr.runcall(main, config, args)
+    pr.dump_stats(os.path.join('.', 'test.profile'))

deepspeech/exps/u2/model.py

@@ -77,8 +77,6 @@ class U2Trainer(Trainer):
 
     def train_batch(self, batch_index, batch_data, msg):
         train_conf = self.config.training
-        self.model.train()
-
         start = time.time()
 
         loss, attention_loss, ctc_loss = self.model(*batch_data)
@@ -134,6 +132,7 @@ class U2Trainer(Trainer):
         self.logger.info(
             f"Train Total Examples: {len(self.train_loader.dataset)}")
         while self.epoch < self.config.training.n_epoch:
+            self.model.train()
             try:
                 data_start_time = time.time()
                 for batch_index, batch in enumerate(self.train_loader):
@@ -149,8 +148,8 @@ class U2Trainer(Trainer):
                 self.logger.error(e)
                 raise e
 
-            self.valid()
+            valid_losses = self.valid()
-            self.save()
+            self.save(infos=valid_losses)
             self.new_epoch()
 
     @mp_tools.rank_zero_only
@@ -182,6 +181,7 @@ class U2Trainer(Trainer):
             for k, v in valid_losses.items():
                 self.visualizer.add_scalar("valid/{}".format(k), v,
                                            self.iteration)
+        return valid_losses
 
     def setup_dataloader(self):
         config = self.config.clone()

deepspeech/io/dataset.py

@@ -290,19 +290,34 @@ class ManifestDataset(Dataset):
                  where transcription part could be token ids or text.
         :rtype: tuple of (2darray, list)
         """
+        start_time = time.time()
         if isinstance(audio_file, str) and audio_file.startswith('tar:'):
             speech_segment = SpeechSegment.from_file(
                 self._subfile_from_tar(audio_file), transcript)
         else:
             speech_segment = SpeechSegment.from_file(audio_file, transcript)
+        load_wav_time = time.time() - start_time
+        logger.debug(f"load wav time: {load_wav_time}")
+
         # audio augment
+        start_time = time.time()
         self._augmentation_pipeline.transform_audio(speech_segment)
+        audio_aug_time = time.time() - start_time
+        logger.debug(f"audio augmentation time: {audio_aug_time}")
+
+        start_time = time.time()
         specgram, transcript_part = self._speech_featurizer.featurize(
             speech_segment, self._keep_transcription_text)
         if self._normalizer:
             specgram = self._normalizer.apply(specgram)
+        feature_time = time.time() - start_time
+        logger.debug(f"audio & test feature time: {feature_time}")
+
         # specgram augment
+        start_time = time.time()
         specgram = self._augmentation_pipeline.transform_feature(specgram)
+        feature_aug_time = time.time() - start_time
+        logger.debug(f"audio feature augmentation time: {feature_aug_time}")
         return specgram, transcript_part
 
     def _instance_reader_creator(self, manifest):

deepspeech/models/u2.py

@@ -821,7 +821,8 @@ class U2Model(U2BaseModel):
             mean, istd = load_cmvn(configs['cmvn_file'],
                                    configs['cmvn_file_type'])
             global_cmvn = GlobalCMVN(
-                paddle.to_tensor(mean).float(), paddle.to_tensor(istd).float())
+                paddle.to_tensor(mean, dtype=paddle.float),
+                paddle.to_tensor(istd, dtype=paddle.float))
         else:
             global_cmvn = None
 

deepspeech/training/trainer.py

@@ -128,15 +128,15 @@ class Trainer():
         dist.init_parallel_env()
 
     @mp_tools.rank_zero_only
-    def save(self, tag=None, infos=None):
+    def save(self, tag=None, infos: dict=None):
         """Save checkpoint (model parameters and optimizer states).
         """
-        if infos is None:
+        infos = infos if infos else dict()
-            infos = {
+        infos.update({
-                "step": self.iteration,
+            "step": self.iteration,
-                "epoch": self.epoch,
+            "epoch": self.epoch,
-                "lr": self.optimizer.get_lr(),
+            "lr": self.optimizer.get_lr()
-            }
+        })
         checkpoint.save_parameters(self.checkpoint_dir, self.iteration
                                    if tag is None else tag, self.model,
                                    self.optimizer, infos)
@@ -185,6 +185,7 @@ class Trainer():
         self.logger.info(
             f"Train Total Examples: {len(self.train_loader.dataset)}")
         while self.epoch < self.config.training.n_epoch:
+            self.model.train()
             try:
                 data_start_time = time.time()
                 for batch_index, batch in enumerate(self.train_loader):
@@ -200,8 +201,8 @@ class Trainer():
                 self.logger.error(e)
                 raise e
 
-            self.valid()
+            valid_losses = self.valid()
-            self.save()
+            self.save(infos=valid_losses)
             self.lr_scheduler.step()
             self.new_epoch()