UnifiedGptVoice!

codes/data/audio/grand_conjoined_dataset.py

@@ -99,7 +99,7 @@ class GrandConjoinedDataset(torch.utils.data.Dataset):
                 'paired_text_tokens': snt['padded_text'],
                 'paired_file': snt['filenames'],
                 'speech_audio': snt['wav'],
-                'speech_lengths': snt['wav_lengths'],
+                'speech_audio_lengths': snt['wav_lengths'],
                 'speech_file': snt['filenames'],
                 'text_text': snt['real_text'],
                 'text_tokens': snt['padded_text'],
@@ -114,7 +114,7 @@ class GrandConjoinedDataset(torch.utils.data.Dataset):
                 'paired_text_tokens': snt['padded_text'],
                 'paired_file': snt['filenames'],
                 'speech_audio': sp['clip'],
-                'speech_lengths': clamp(sp['clip_lengths'], 0, self.max_solo_audio_length),
+                'speech_audio_lengths': clamp(sp['clip_lengths'], 0, self.max_solo_audio_length),
                 'speech_file': sp['path'],
                 'text_text': txt,
                 'text_tokens': txt_tok,

codes/models/gpt_voice/unified_voice.py

@@ -38,7 +38,15 @@ class ConditioningEncoder(nn.Module):
         return h[:, :, 0]
 
 
-class GptTtsHf(nn.Module):
+class UnifiedGptVoice(nn.Module):
+    """
+    Derived from GptTtsHf, but offers multiple modes of operation:
+    - Text only
+    - Voice only
+    - Text conditioned on voice
+    - Voice conditioned on text
+    """
+
     NUMBER_TEXT_TOKENS = 10000  # The number of tokens produced by our bespoke BPE tokenizer.
     START_TEXT_TOKEN = 9999
     STOP_TEXT_TOKEN = 0
@@ -79,87 +87,120 @@ class GptTtsHf(nn.Module):
         tar = F.pad(input, (0,1), value=stop_token)
         return inp, tar
 
-    def get_logits(self, text_inputs, cond_input, mel_inputs, get_attns=False):
+    def set_mel_padding(self, mel_input_tokens, wav_lengths):
-        text_emb = self.text_embedding(text_inputs)
-        cond = self.conditioning_encoder(cond_input).unsqueeze(1)
-        mel_emb = self.gpt.get_input_embeddings()(mel_inputs)
-
-        emb = torch.cat([text_emb, cond, mel_emb], dim=1)
-        gpt_out = self.gpt(inputs_embeds=emb, return_dict=True, output_attentions=get_attns)
-        if get_attns:
-            return gpt_out.attentions
-        enc = gpt_out.last_hidden_state
-
-        text_logits = self.final_norm(enc[:, :text_emb.shape[1]])
-        text_logits = self.text_head(text_logits)
-        text_logits = text_logits.permute(0,2,1)
-        mel_logits = self.final_norm(enc[:, -mel_emb.shape[1]:])
-        mel_logits = self.mel_head(mel_logits)
-        mel_logits = mel_logits.permute(0,2,1)
-
-        return text_logits, mel_logits
-
-    def forward(self, text_inputs, cond_input, mel_targets, wav_lengths, return_attentions=False):
         """
-        Forward pass
+        Given mel tokens that are derived from a padded audio clip and the actual lengths of each batch element in
-        text_inputs: long tensor, (b,t)
+        that audio clip, reformats the tokens with STOP_MEL_TOKEN in place of the zero padding. This is required
-        cond_inputs: MEL float tensor, (b,c,80,s)
+        preformatting to create a working TTS model.
-        mel_targets: long tensor, (b,m)
-        mel_lengths: long tensor, (b,)
         """
         # Set padding areas within MEL (currently it is coded with the MEL code for <zero>).
         mel_lengths = wav_lengths // self.mel_length_compression
         for b in range(len(mel_lengths)):
-            if mel_lengths[b] < mel_targets.shape[-1]:
+            actual_end = mel_lengths[b] + 1  # Due to the convolutional nature of how these tokens are generated, it would be best if the model predicts a token past the actual last token.
-                mel_targets[b, mel_lengths[b]:] = self.STOP_MEL_TOKEN
+            if actual_end < mel_input_tokens.shape[-1]:
+                mel_input_tokens[b, actual_end:] = self.STOP_MEL_TOKEN
+        return mel_input_tokens
 
-        # Randomly permute the conditioning spectrogram, to destroy any structure present.
+    def randomly_permute_conditioning_input(self, speech_conditioning_input):
-        cond_input = cond_input[:,:,torch.randperm(cond_input.shape[-1])]
+        """
+        Randomly permute the conditioning spectrogram, to destroy any structure present. Note that since the
+        conditioning input is derived from a discrete spectrogram, it does actually retain structure, but only a little
+        bit (actually: exactly how much we want; enough to discriminate different vocal qualities, but nothing about
+        what is being said).
+        """
+        cond_input = speech_conditioning_input[:,:,torch.randperm(speech_conditioning_input.shape[-1])]
         if cond_input.shape[-1] > self.max_conditioning_length:
             cond_input = cond_input[:,:,:self.max_conditioning_length]
+        return cond_input
+
+    def get_logits(self, speech_conditioning_input, first_inputs, first_head, second_inputs=None, second_head=None, get_attns=False):
+        if second_inputs is not None:
+            emb = torch.cat([speech_conditioning_input, first_inputs, second_inputs], dim=1)
+        else:
+            emb = torch.cat([speech_conditioning_input, first_inputs], dim=1)
+
+        gpt_out = self.gpt(inputs_embeds=emb, return_dict=True, output_attentions=get_attns)
+        if get_attns:
+            return gpt_out.attentions
+
+        enc = gpt_out.last_hidden_state[:, 1:]  # The first logit is tied to the speech_conditioning_input
+        first_logits = self.final_norm(enc[:, :first_inputs.shape[1]])
+        first_logits = first_head(first_logits)
+        first_logits = first_logits.permute(0,2,1)
+        if second_inputs is not None:
+            second_logits = self.final_norm(enc[:, -second_inputs.shape[1]:])
+            second_logits = second_head(second_logits)
+            second_logits = second_logits.permute(0,2,1)
+            return first_logits, second_logits
+        else:
+            return first_logits
+
+    def forward(self, speech_conditioning_input, text_inputs, mel_inputs, wav_lengths, text_first=True, return_attentions=False):
+        """
+        Forward pass that uses both text and voice in either text conditioning mode or voice conditioning mode
+        (actuated by `text_first`).
+
+        speech_conditioning_input: MEL float tensor, (b,80,s)
+        text_inputs: long tensor, (b,t)
+        mel_inputs:  long tensor, (b,m)
+        wav_lengths: long tensor, (b,)
+        """
+        mel_inputs = self.set_mel_padding(mel_inputs, wav_lengths)
+        speech_conditioning_input = self.randomly_permute_conditioning_input(speech_conditioning_input)
+        speech_conditioning_input = self.conditioning_encoder(speech_conditioning_input).unsqueeze(1)
 
         text_inputs, text_targets = self.build_aligned_inputs_and_targets(text_inputs, self.START_TEXT_TOKEN, self.STOP_TEXT_TOKEN)
-        mel_inputs, mel_targets = self.build_aligned_inputs_and_targets(mel_targets, self.START_MEL_TOKEN, self.STOP_MEL_TOKEN)
+        text_emb = self.text_embedding(text_inputs)
-        text_logits, mel_logits = self.get_logits(text_inputs, cond_input, mel_inputs, get_attns=return_attentions)
+        mel_inputs, mel_targets = self.build_aligned_inputs_and_targets(mel_inputs, self.START_MEL_TOKEN, self.STOP_MEL_TOKEN)
+        mel_emb = self.gpt.get_input_embeddings()(mel_inputs)
+        if text_first:
+            text_logits, mel_logits = self.get_logits(speech_conditioning_input, text_emb, self.text_head, mel_emb, self.mel_head, get_attns=return_attentions)
+        else:
+            mel_logits, text_logits = self.get_logits(speech_conditioning_input, mel_emb, self.mel_head, text_emb, self.text_head, get_attns=return_attentions)
+
         if return_attentions:
             return mel_logits
         loss_text = F.cross_entropy(text_logits, text_targets.long())
         loss_mel = F.cross_entropy(mel_logits, mel_targets.long())
         return loss_text.mean(), loss_mel.mean(), mel_logits
 
-    def inference(self, text_inputs, cond_input, **hf_generate_kwargs):
+    def text_forward(self, speech_conditioning_input, text_inputs):
-        if not hasattr(self, 'inference_model'):
+        """
-            self.inference_model = GPT2InferenceModel(self.gpt_config, self.gpt, None, self.final_norm, self.mel_head)
+        Performs autoregressive modeling on only text. Still requires a speech_conditioning_input due to the way the
+        model inputs are formatted. Just provide any audio clip (arguably, zeros could be provided).
+        """
+        speech_conditioning_input = self.randomly_permute_conditioning_input(speech_conditioning_input)
+        speech_conditioning_input = self.conditioning_encoder(speech_conditioning_input).unsqueeze(1)
 
-        text_inputs = F.pad(text_inputs, (0, self.max_symbols_per_phrase - text_inputs.shape[1]), value=self.STOP_TEXT_TOKEN)
         text_inputs, text_targets = self.build_aligned_inputs_and_targets(text_inputs, self.START_TEXT_TOKEN, self.STOP_TEXT_TOKEN)
         text_emb = self.text_embedding(text_inputs)
+        text_logits = self.get_logits(speech_conditioning_input, text_emb, self.text_head)
+        loss_text = F.cross_entropy(text_logits, text_targets.long())
+        return loss_text.mean()
 
-        # Randomly permute the conditioning spectrogram, to destroy any structure present.
+    def speech_forward(self, speech_conditioning_input, mel_inputs, wav_lengths):
-        cond_input = cond_input[:,:,torch.randperm(cond_input.shape[-1])]
+        """
-        if cond_input.shape[-1] > self.max_conditioning_length:
+        Performs autoregressive modeling on only speech data.
-            cond_input = cond_input[:,:,:self.max_conditioning_length]
+        """
-        cond = self.conditioning_encoder(cond_input).unsqueeze(1)
+        mel_inputs = self.set_mel_padding(mel_inputs, wav_lengths)
+        speech_conditioning_input = self.randomly_permute_conditioning_input(speech_conditioning_input)
+        speech_conditioning_input = self.conditioning_encoder(speech_conditioning_input).unsqueeze(1)
 
-        emb = torch.cat([text_emb, cond], dim=1)
+        mel_inputs, mel_targets = self.build_aligned_inputs_and_targets(mel_inputs, self.START_MEL_TOKEN, self.STOP_MEL_TOKEN)
-        self.inference_model.store_mel_emb(emb)
+        mel_emb = self.gpt.get_input_embeddings()(mel_inputs)
-
+        mel_logits = self.get_logits(speech_conditioning_input, mel_emb, self.mel_head)
-        fake_inputs = torch.full((emb.shape[0],emb.shape[1]+1,), fill_value=1, dtype=torch.long, device=text_inputs.device)
+        loss_mel = F.cross_entropy(mel_logits, mel_targets.long())
-        fake_inputs[:,-1] = self.START_MEL_TOKEN
+        return loss_mel.mean()
-
-        gen = self.inference_model.generate(fake_inputs, bos_token_id=self.START_MEL_TOKEN, pad_token_id=self.STOP_MEL_TOKEN, eos_token_id=self.STOP_MEL_TOKEN,
-                          max_length=emb.shape[1]+self.max_mel_tokens, **hf_generate_kwargs)
-        return gen[:, fake_inputs.shape[1]:]
 
 
 @register_model
-def register_gpt_tts_hf(opt_net, opt):
+def register_unified_gpt_voice(opt_net, opt):
-    return GptTtsHf(**opt_get(opt_net, ['kwargs'], {}))
+    return UnifiedGptVoice(**opt_get(opt_net, ['kwargs'], {}))
 
 
 if __name__ == '__main__':
-    gpt = GptTtsHf(model_dim=1024, heads=16)
+    gpt = UnifiedGptVoice(model_dim=256, heads=4)
-    l = gpt(torch.randint(high=len(symbols), size=(2,200)),
+    l = gpt(torch.randn(2, 80, 800),
-            torch.arange(0, 80, 1, dtype=torch.float).view(1,80,1).repeat(2,1,800),
+            torch.randint(high=len(symbols), size=(2,80)),
             torch.randint(high=8192, size=(2,250)),
             torch.tensor([150*256,195*256]))

codes/train.py

@@ -286,7 +286,7 @@ class Trainer:
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()
-    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='../options/train_gpt_tts.yml')
+    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='../options/train_gpt_unified_voice.yml')
     parser.add_argument('--launcher', choices=['none', 'pytorch'], default='none', help='job launcher')
     parser.add_argument('--local_rank', type=int, default=0)
     args = parser.parse_args()