diff --git a/.gitignore b/.gitignore
index d632a14..82504f8 100644
--- a/.gitignore
+++ b/.gitignore
@@ -129,7 +129,6 @@ dmypy.json
 .pyre/
 
 .idea/*
-tortoise/.models/*
-tortoise/random_voices/*
+.models/*
 .custom/*
 results/*
\ No newline at end of file
diff --git a/tortoise/models/arch_util.py b/tortoise/models/arch_util.py
index 3a004af..5d8c36e 100644
--- a/tortoise/models/arch_util.py
+++ b/tortoise/models/arch_util.py
@@ -290,7 +290,7 @@ class AudioMiniEncoder(nn.Module):
 
 class TorchMelSpectrogram(nn.Module):
     def __init__(self, filter_length=1024, hop_length=256, win_length=1024, n_mel_channels=80, mel_fmin=0, mel_fmax=8000,
-                 sampling_rate=22050, normalize=False, mel_norm_file='data/mel_norms.pth'):
+                 sampling_rate=22050, normalize=False, mel_norm_file='tortoise/data/mel_norms.pth'):
         super().__init__()
         # These are the default tacotron values for the MEL spectrogram.
         self.filter_length = filter_length
diff --git a/tortoise/read.py b/tortoise/read.py
index b5d6efe..5212e3b 100644
--- a/tortoise/read.py
+++ b/tortoise/read.py
@@ -28,7 +28,7 @@ def split_and_recombine_text(texts, desired_length=200, max_len=300):
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()
-    parser.add_argument('--textfile', type=str, help='A file containing the text to read.', default="data/riding_hood.txt")
+    parser.add_argument('--textfile', type=str, help='A file containing the text to read.', default="tortoise/data/riding_hood.txt")
     parser.add_argument('--voice', type=str, help='Selects the voice to use for generation. See options in voices/ directory (and add your own!) '
                                                  'Use the & character to join two voices together. Use a comma to perform inference on multiple voices.', default='pat')
     parser.add_argument('--output_path', type=str, help='Where to store outputs.', default='../results/longform/')
diff --git a/tortoise/utils/audio.py b/tortoise/utils/audio.py
index c76f7a6..d857da9 100644
--- a/tortoise/utils/audio.py
+++ b/tortoise/utils/audio.py
@@ -82,10 +82,10 @@ def dynamic_range_decompression(x, C=1):
 
 
 def get_voices():
-    subs = os.listdir('voices')
+    subs = os.listdir('tortoise/voices')
     voices = {}
     for sub in subs:
-        subj = os.path.join('voices', sub)
+        subj = os.path.join('tortoise/voices', sub)
         if os.path.isdir(subj):
             voices[sub] = list(glob(f'{subj}/*.wav')) + list(glob(f'{subj}/*.mp3')) + list(glob(f'{subj}/*.pth'))
     return voices
diff --git a/tortoise/utils/tokenizer.py b/tortoise/utils/tokenizer.py
index ed7e4cd..2f36a06 100644
--- a/tortoise/utils/tokenizer.py
+++ b/tortoise/utils/tokenizer.py
@@ -164,7 +164,7 @@ def lev_distance(s1, s2):
   return distances[-1]
 
 class VoiceBpeTokenizer:
-    def __init__(self, vocab_file='data/tokenizer.json'):
+    def __init__(self, vocab_file='tortoise/data/tokenizer.json'):
         if vocab_file is not None:
             self.tokenizer = Tokenizer.from_file(vocab_file)