Move everything into the tortoise/ subdirectory

For eventual packaging.

tortoise/api.py

@@ -1,4 +1,3 @@
-import argparse
 import os
 import random
 from urllib import request
@@ -8,19 +7,18 @@ import torch.nn.functional as F
 import progressbar
 import torchaudio
 
-from models.classifier import AudioMiniEncoderWithClassifierHead
-from models.cvvp import CVVP
-from models.diffusion_decoder import DiffusionTts
-from models.autoregressive import UnifiedVoice
+from tortoise.models.classifier import AudioMiniEncoderWithClassifierHead
+from tortoise.models.cvvp import CVVP
+from tortoise.models.diffusion_decoder import DiffusionTts
+from tortoise.models.autoregressive import UnifiedVoice
 from tqdm import tqdm
 
-from models.arch_util import TorchMelSpectrogram
-from models.clvp import CLVP
-from models.vocoder import UnivNetGenerator
-from utils.audio import load_audio, wav_to_univnet_mel, denormalize_tacotron_mel
-from utils.diffusion import SpacedDiffusion, space_timesteps, get_named_beta_schedule
-from utils.tokenizer import VoiceBpeTokenizer, lev_distance
-
+from tortoise.models.arch_util import TorchMelSpectrogram
+from tortoise.models.clvp import CLVP
+from tortoise.models.vocoder import UnivNetGenerator
+from tortoise.utils.audio import wav_to_univnet_mel, denormalize_tacotron_mel
+from tortoise.utils.diffusion import SpacedDiffusion, space_timesteps, get_named_beta_schedule
+from tortoise.utils.tokenizer import VoiceBpeTokenizer
 
 pbar = None
 

tortoise/do_tts.py

@@ -4,7 +4,7 @@ import os
 import torchaudio
 
 from api import TextToSpeech
-from utils.audio import load_audio, get_voices
+from tortoise.utils.audio import load_audio, get_voices
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()

tortoise/eval_multiple.py

@@ -3,7 +3,7 @@ import os
 import torchaudio
 
 from api import TextToSpeech
-from utils.audio import load_audio
+from tortoise.utils.audio import load_audio
 
 if __name__ == '__main__':
     fname = 'Y:\\clips\\books2\\subset512-oco.tsv'

tortoise/is_this_from_tortoise.py

@@ -1,7 +1,7 @@
 import argparse
 
 from api import classify_audio_clip
-from utils.audio import load_audio
+from tortoise.utils.audio import load_audio
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()

tortoise/models/arch_util.py

@@ -5,7 +5,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import torchaudio
-from models.xtransformers import ContinuousTransformerWrapper, RelativePositionBias
+from tortoise.models.xtransformers import ContinuousTransformerWrapper, RelativePositionBias
 
 
 def zero_module(module):

tortoise/models/autoregressive.py

@@ -6,8 +6,8 @@ import torch.nn.functional as F
 from transformers import GPT2Config, GPT2PreTrainedModel, LogitsProcessorList
 from transformers.modeling_outputs import CausalLMOutputWithCrossAttentions
 from transformers.utils.model_parallel_utils import get_device_map, assert_device_map
-from models.arch_util import AttentionBlock
-from utils.typical_sampling import TypicalLogitsWarper
+from tortoise.models.arch_util import AttentionBlock
+from tortoise.utils.typical_sampling import TypicalLogitsWarper
 
 
 def null_position_embeddings(range, dim):

tortoise/models/classifier.py

@@ -1,9 +1,8 @@
 import torch
 import torch.nn as nn
-import torch.nn.functional as F
 from torch.utils.checkpoint import checkpoint
 
-from models.arch_util import Upsample, Downsample, normalization, zero_module, AttentionBlock
+from tortoise.models.arch_util import Upsample, Downsample, normalization, zero_module, AttentionBlock
 
 
 class ResBlock(nn.Module):

tortoise/models/clvp.py

@@ -3,9 +3,9 @@ import torch.nn as nn
 import torch.nn.functional as F
 from torch import einsum
 
-from models.arch_util import CheckpointedXTransformerEncoder
-from models.transformer import Transformer
-from models.xtransformers import Encoder
+from tortoise.models.arch_util import CheckpointedXTransformerEncoder
+from tortoise.models.transformer import Transformer
+from tortoise.models.xtransformers import Encoder
 
 
 def exists(val):

tortoise/models/cvvp.py

@@ -4,8 +4,8 @@ import torch.nn.functional as F
 from torch import einsum
 from torch.utils.checkpoint import checkpoint
 
-from models.arch_util import AttentionBlock
-from models.xtransformers import ContinuousTransformerWrapper, Encoder
+from tortoise.models.arch_util import AttentionBlock
+from tortoise.models.xtransformers import ContinuousTransformerWrapper, Encoder
 
 
 def exists(val):

tortoise/models/diffusion_decoder.py

@@ -7,7 +7,7 @@ import torch.nn as nn
 import torch.nn.functional as F
 from torch import autocast
 
-from models.arch_util import normalization, AttentionBlock
+from tortoise.models.arch_util import normalization, AttentionBlock
 
 
 def is_latent(t):

tortoise/read.py

@@ -2,12 +2,10 @@ import argparse
 import os
 
 import torch
-import torch.nn.functional as F
 import torchaudio
 
-from api import TextToSpeech, format_conditioning
-from utils.audio import load_audio, get_voices
-from utils.tokenizer import VoiceBpeTokenizer
+from api import TextToSpeech
+from tortoise.utils.audio import load_audio, get_voices
 
 
 def split_and_recombine_text(texts, desired_length=200, max_len=300):

tortoise/samples_generator.py

@@ -4,7 +4,7 @@ import os
 
 if __name__ == '__main__':
     result = "<html><head><title>These words were never spoken.</title></head><body><h1>Handpicked results</h1>"
-    for fv in os.listdir('results/favorites'):
+    for fv in os.listdir('../results/favorites'):
         url = f'https://github.com/neonbjb/tortoise-tts/raw/main/results/favorites/{fv}'
         result = result + f'<audio controls="" style="width: 600px;"><source src="{url}" type="audio/mp3"></audio><br>\n'
 
@@ -30,7 +30,7 @@ if __name__ == '__main__':
         line = line + f'<td><audio controls="" style="width: 150px;"><source src="{url}" type="audio/mp3"></audio></td>'
     line = line + "</tr>"
     lines.append(line)
-    for txt in os.listdir('results/various/'):
+    for txt in os.listdir('../results/various/'):
         if 'desktop' in txt:
             continue
         line = f'<tr><td>{txt}</td>'
@@ -42,7 +42,7 @@ if __name__ == '__main__':
     result = result + '\n'.join(lines) + "</table>"
 
     result = result + "<h1>Longform result for all voices:</h1>"
-    for lf in os.listdir('results/riding_hood'):
+    for lf in os.listdir('../results/riding_hood'):
         url = f'https://github.com/neonbjb/tortoise-tts/raw/main/results/riding_hood/{lf}'
         result = result + f'<audio controls="" style="width: 600px;"><source src="{url}" type="audio/mp3"></audio><br>\n'
 

tortoise/sweep.py

@@ -4,7 +4,7 @@ from random import shuffle
 import torchaudio
 
 from api import TextToSpeech
-from utils.audio import load_audio
+from tortoise.utils.audio import load_audio
 
 
 def permutations(args):

tortoise/utils/audio.py

@@ -6,7 +6,7 @@ import torchaudio
 import numpy as np
 from scipy.io.wavfile import read
 
-from utils.stft import STFT
+from tortoise.utils.stft import STFT
 
 
 def load_wav_to_torch(full_path):