split dataset outside preprocessor

datasets/preprocess.py

@@ -146,7 +146,7 @@ def common_voice(root_path, meta_file):
     return items
 
 
-def libri_tts(root_path, meta_files=None, is_eval=False):
+def libri_tts(root_path, meta_files=None):
     """https://ai.google/tools/datasets/libri-tts/"""
     items = []
     if meta_files is None:
@@ -164,6 +164,4 @@ def libri_tts(root_path, meta_files=None, is_eval=False):
                 items.append([text, wav_file, speaker_name])
     for item in items:
         assert os.path.exists(item[1]), f" [!] wav file is not exist - {item[1]}"
-    if meta_files is None:
-        return items[:500] if is_eval else items[500:]
     return items

train.py

@@ -21,7 +21,8 @@ from utils.generic_utils import (NoamLR, check_update, count_parameters,
                                  load_config, lr_decay,
                                  remove_experiment_folder, save_best_model,
                                  save_checkpoint, sequence_mask, weight_decay,
-                                 set_init_dict, copy_config_file, setup_model)
+                                 set_init_dict, copy_config_file, setup_model,
+                                 split_dataset)
 from utils.logger import Logger
 from utils.speakers import load_speaker_mapping, save_speaker_mapping, \
     get_speakers
@@ -44,15 +45,15 @@ def setup_loader(is_val=False, verbose=False):
     global meta_data_train
     global meta_data_eval
     if "meta_data_train" not in globals():
-        if c.meta_file_train:
+        if c.meta_file_train is not None:
             meta_data_train = get_preprocessor_by_name(c.dataset)(c.data_path, c.meta_file_train)
         else:
-            meta_data_train = get_preprocessor_by_name(c.dataset)(c.data_path, is_eval=False)
+            meta_data_train = get_preprocessor_by_name(c.dataset)(c.data_path)
-    if "meta_data_eval" not in globals():
+    if "meta_data_eval" not in globals() and c.run_eval:
-        if c.meta_file_val:
+        if c.meta_file_val is not None:
             meta_data_eval = get_preprocessor_by_name(c.dataset)(c.data_path, c.meta_file_val)
         else:
-            meta_data_eval = get_preprocessor_by_name(c.dataset)(c.data_path, is_eval=True)
+            meta_data_eval, meta_data_train = split_dataset(meta_data_train)
     if is_val and not c.run_eval:
         loader = None
     else:

utils/generic_utils.py

@@ -10,7 +10,7 @@ import torch
 import subprocess
 import importlib
 import numpy as np
-from collections import OrderedDict
+from collections import OrderedDict, Counter
 from torch.autograd import Variable
 from utils.text import text_to_sequence
 
@@ -287,3 +287,26 @@ def setup_model(num_chars, num_speakers, c):
             location_attn=c.location_attn,
             separate_stopnet=c.separate_stopnet)
     return model
+
+
+def split_dataset(items):
+    is_multi_speaker = False
+    speakers = [item[-1] for item in items]
+    is_multi_speaker = len(set(speakers)) > 1
+    eval_split_size = 500 if 500 < len(items) * 0.01 else int(len(items) * 0.01)
+    np.random.seed(0)
+    np.random.shuffle(items)
+    if is_multi_speaker:
+        items_eval = []
+        # most stupid code ever -- Fix it !
+        while len(items_eval) < eval_split_size:
+            speakers = [item[-1] for item in items]
+            speaker_counter = Counter(speakers) 
+            item_idx = np.random.randint(0, len(items))
+            if speaker_counter[items[item_idx][-1]] > 1:
+                items_eval.append(items[item_idx])
+                del items[item_idx]
+        return items_eval, items
+    else:
+        return items[:eval_split_size], items[eval_split_size:]
+