diff --git a/lib/rvc/train.py b/lib/rvc/train.py
index babe473..d3beaad 100644
--- a/lib/rvc/train.py
+++ b/lib/rvc/train.py
@@ -22,46 +22,56 @@
 from . import commons, utils
 from .checkpoints import save
 from .config import DatasetMetadata, TrainConfig
-from .data_utils import (DistributedBucketSampler, TextAudioCollate,
-                         TextAudioCollateMultiNSFsid, TextAudioLoader,
-                         TextAudioLoaderMultiNSFsid)
+from .data_utils import (
+    DistributedBucketSampler,
+    TextAudioCollate,
+    TextAudioCollateMultiNSFsid,
+    TextAudioLoader,
+    TextAudioLoaderMultiNSFsid,
+)
 from .losses import discriminator_loss, feature_loss, generator_loss, kl_loss
 from .mel_processing import mel_spectrogram_torch, spec_to_mel_torch
-from .models import (MultiPeriodDiscriminator, SynthesizerTrnMs256NSFSid,
-                     SynthesizerTrnMs256NSFSidNono)
+from .models import (
+    MultiPeriodDiscriminator,
+    SynthesizerTrnMs256NSFSid,
+    SynthesizerTrnMs256NSFSidNono,
+)
 
 
 def glob_dataset(glob_str: str, speaker_id: int):
+    globs = glob_str.split(",")
     datasets_speakers = []
-    if os.path.isdir(glob_str):
-        files = os.listdir(glob_str)
-        # pattern: {glob_str}/{decimal}[_]* and isdir
-        dirs = [
-            (os.path.join(glob_str, f), int(f.split("_")[0]))
-            for f in files
-            if os.path.isdir(os.path.join(glob_str, f)) and f.split("_")[0].isdecimal()
-        ]
-
-        if len(dirs) > 0:
-            # multi speakers at once train
-            match_files_re = re.compile(r".+\.(wav|flac)")  # matches .wav and .flac
-            datasets_speakers = [
-                (file, dir[1])
-                for dir in dirs
-                for file in glob.iglob(os.path.join(dir[0], "*"), recursive=True)
-                if match_files_re.search(file)
+    for glob_str in globs:
+        if os.path.isdir(glob_str):
+            files = os.listdir(glob_str)
+            # pattern: {glob_str}/{decimal}[_]* and isdir
+            dirs = [
+                (os.path.join(glob_str, f), int(f.split("_")[0]))
+                for f in files
+                if os.path.isdir(os.path.join(glob_str, f))
+                and f.split("_")[0].isdecimal()
             ]
-            # for dir in dirs:
-            #     for file in glob.iglob(dir[0], recursive=True):
-            #         if match_files_re.search(file):
-            #             datasets_speakers.append((file, dirs[1]))
-            # return sorted(datasets_speakers, key=operator.itemgetter(0))
 
-        glob_str = os.path.join(glob_str, "*.wav")
-
-    datasets_speakers.extend(
-        [(file, speaker_id) for file in glob.iglob(glob_str, recursive=True)]
-    )
+            if len(dirs) > 0:
+                # multi speakers at once train
+                match_files_re = re.compile(r".+\.(wav|flac)")  # matches .wav and .flac
+                datasets_speakers = [
+                    (file, dir[1])
+                    for dir in dirs
+                    for file in glob.iglob(os.path.join(dir[0], "*"), recursive=True)
+                    if match_files_re.search(file)
+                ]
+                # for dir in dirs:
+                #     for file in glob.iglob(dir[0], recursive=True):
+                #         if match_files_re.search(file):
+                #             datasets_speakers.append((file, dirs[1]))
+                # return sorted(datasets_speakers, key=operator.itemgetter(0))
+
+            glob_str = os.path.join(glob_str, "*.wav")
+
+        datasets_speakers.extend(
+            [(file, speaker_id) for file in glob.iglob(glob_str, recursive=True)]
+        )
 
     return sorted(datasets_speakers, key=operator.itemgetter(0))
 
diff --git a/modules/tabs/training.py b/modules/tabs/training.py
index ae74d14..0d3f398 100644
--- a/modules/tabs/training.py
+++ b/modules/tabs/training.py
@@ -7,7 +7,7 @@
 from lib.rvc.preprocessing import extract_f0, extract_feature, split
 from lib.rvc.train import create_dataset_meta, glob_dataset, train_index, train_model
 from modules import models, utils
-from modules.shared import MODELS_DIR, device
+from modules.shared import MODELS_DIR
 from modules.ui import Tab
 
 SR_DICT = {
@@ -31,6 +31,7 @@ def train_index_only(
             f0,
             dataset_glob,
             speaker_id,
+            gpu_id,
             num_cpu_process,
             norm_audio_when_preprocess,
             pitch_extraction_algo,
@@ -42,6 +43,7 @@ def train_index_only(
             f0 = f0 == "Yes"
             norm_audio_when_preprocess = norm_audio_when_preprocess == "Yes"
             training_dir = os.path.join(MODELS_DIR, "training", "models", model_name)
+            gpu_ids = [int(x.strip()) for x in gpu_id.split(",")]
             yield f"Training directory: {training_dir}"
 
             if os.path.exists(training_dir) and ignore_cache:
@@ -78,8 +80,7 @@ def train_index_only(
                 embedder_filepath,
                 embedder_load_from,
                 embedding_channels == 768,
-                None,
-                device,
+                gpu_ids,
             )
 
             out_dir = os.path.join(MODELS_DIR, "checkpoints")
@@ -304,6 +305,7 @@ def train_all(
                 f0,
                 dataset_glob,
                 speaker_id,
+                gpu_id,
                 num_cpu_process,
                 norm_audio_when_preprocess,
                 pitch_extraction_algo,