=Filter instances by sequence length

config.json

@@ -12,9 +12,9 @@
   "text_cleaner": "english_cleaners",
 
   "epochs": 2000,
-  "lr": 0.0003,
+  "lr": 0.0006 / 32,
   "warmup_steps": 4000,
-  "batch_size": 8,
+  "batch_size": 1,
   "r": 5,
 
   "griffin_lim_iters": 60,
@@ -25,5 +25,6 @@
   "checkpoint": false,
   "save_step": 69,
   "data_path": "/run/shm/erogol/LJSpeech-1.0",
+  "min_seq_len": 90, 
   "output_path": "result"
 }

datasets/LJSpeech.py

@@ -14,7 +14,8 @@ class LJSpeechDataset(Dataset):
 
     def __init__(self, csv_file, root_dir, outputs_per_step, sample_rate,
                 text_cleaner, num_mels, min_level_db, frame_shift_ms,
-                frame_length_ms, preemphasis, ref_level_db, num_freq, power):
+                frame_length_ms, preemphasis, ref_level_db, num_freq, power,
+                min_seq_len=0):
         
         with open(csv_file, "r") as f:
             self.frames = [line.split('|') for line in f]
@@ -22,6 +23,7 @@ class LJSpeechDataset(Dataset):
         self.outputs_per_step = outputs_per_step
         self.sample_rate = sample_rate
         self.cleaners = text_cleaner
+        self.min_seq_length = min_seq_length
         self.ap = AudioProcessor(sample_rate, num_mels, min_level_db, frame_shift_ms,
                                  frame_length_ms, preemphasis, ref_level_db, num_freq, power)
         print(" > Reading LJSpeech from - {}".format(root_dir))
@@ -45,8 +47,14 @@ class LJSpeechDataset(Dataset):
         
         idxs = np.argsort(lengths)
         new_frames = [None] * len(lengths)
+        ignored = []
         for i, idx in enumerate(idxs):
-            new_frames[i] = self.frames[idx]
+            length = lengths[idx]
+            if length < self.min_seq_length:
+                ignored.append(idx)
+            else
+                new_frames[i] = self.frames[idx]
+        print(" | > {} instances are ignored by min_seq_len ({})".format(len(ignored), self.min_seq_len))
         self.frames = new_frames
         
     def __len__(self):

train.py

@@ -302,7 +302,8 @@ def main(args):
                               c.preemphasis,
                               c.ref_level_db,
                               c.num_freq,
-                              c.power
+                              c.power,
+                              min_seq_len=c.min_seq_len
                              )
 
     train_loader = DataLoader(train_dataset, batch_size=c.batch_size,