File size: 8,689 Bytes
2ed72d6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
import pytorch_lightning as pl
import sys, gc
import random
import torch
import torchaudio
import typing as tp
import wandb

from aeiou.viz import pca_point_cloud, audio_spectrogram_image, tokens_spectrogram_image
from ema_pytorch import EMA
from einops import rearrange
from torch import optim
from torch.nn import functional as F
from pytorch_lightning.utilities.rank_zero import rank_zero_only

from audiocraft.models import MusicGen
from audiocraft.modules.conditioners import ClassifierFreeGuidanceDropout, ConditioningAttributes

from time import time

class Profiler:

    def __init__(self):
        self.ticks = [[time(), None]]

    def tick(self, msg):
        self.ticks.append([time(), msg])

    def __repr__(self):
        rep = 80 * "=" + "\n"
        for i in range(1, len(self.ticks)):
            msg = self.ticks[i][1]
            ellapsed = self.ticks[i][0] - self.ticks[i - 1][0]
            rep += msg + f": {ellapsed*1000:.2f}ms\n"
        rep += 80 * "=" + "\n\n\n"
        return rep


class MusicGenTrainingWrapper(pl.LightningModule):
    def __init__(self, musicgen_model, lr = 1e-4, ema_copy=None):
        super().__init__()

        self.musicgen_model: MusicGen = musicgen_model

        self.musicgen_model.compression_model.requires_grad_(False)

        self.lm = self.musicgen_model.lm

        self.lm.to(torch.float32).train().requires_grad_(True)

        self.lm_ema = EMA(self.lm, ema_model=ema_copy, beta=0.99, update_every=10)

        self.cfg_dropout = ClassifierFreeGuidanceDropout(0.1)

        self.lr = lr

    def configure_optimizers(self):
        optimizer = optim.AdamW([*self.lm.parameters()], lr=self.lr, betas=(0.9, 0.95), weight_decay=0.1)

        return optimizer
    
    # Copied and modified from https://github.com/facebookresearch/audiocraft/blob/main/audiocraft/solvers/musicgen.py under MIT license
    # License can be found in LICENSES/LICENSE_META.txt

    def _compute_cross_entropy(
        self, logits: torch.Tensor, targets: torch.Tensor, mask: torch.Tensor
    ) -> tp.Tuple[torch.Tensor, tp.List[torch.Tensor]]:
        """Compute cross entropy between multi-codebook targets and model's logits.
        The cross entropy is computed per codebook to provide codebook-level cross entropy.
        Valid timesteps for each of the codebook are pulled from the mask, where invalid
        timesteps are set to 0.

        Args:
            logits (torch.Tensor): Model's logits of shape [B, K, T, card].
            targets (torch.Tensor): Target codes, of shape [B, K, T].
            mask (torch.Tensor): Mask for valid target codes, of shape [B, K, T].
        Returns:
            ce (torch.Tensor): Cross entropy averaged over the codebooks
            ce_per_codebook (list of torch.Tensor): Cross entropy per codebook (detached).
        """
        B, K, T = targets.shape
        assert logits.shape[:-1] == targets.shape
        assert mask.shape == targets.shape
        ce = torch.zeros([], device=targets.device)
        ce_per_codebook: tp.List[torch.Tensor] = []
        for k in range(K):
            logits_k = logits[:, k, ...].contiguous().view(-1, logits.size(-1))  # [B x T, card]
            targets_k = targets[:, k, ...].contiguous().view(-1)  # [B x T]
            mask_k = mask[:, k, ...].contiguous().view(-1)  # [B x T]
            ce_targets = targets_k[mask_k]
            ce_logits = logits_k[mask_k]
            q_ce = F.cross_entropy(ce_logits, ce_targets)
            ce += q_ce
            ce_per_codebook.append(q_ce.detach())
        # average cross entropy across codebooks
        ce = ce / K
        return ce, ce_per_codebook

    def training_step(self, batch, batch_idx):
        reals, metadata = batch

        if reals.ndim == 4 and reals.shape[0] == 1:
            reals = reals[0]

        # Convert reals to mono if necessary
        if self.musicgen_model.audio_channels == 1:
            reals = reals.mean(dim=1, keepdim=True)

        self.musicgen_model.compression_model.to(self.device).eval()
        self.lm.to(self.device).train()
        self.lm.condition_provider.to(self.device).eval()

        self.lm.condition_provider.conditioners["description"].device = self.device
        self.lm.condition_provider.conditioners["description"].t5.to(self.device).eval()

        with torch.cuda.amp.autocast():

            codes, _ = self.musicgen_model.compression_model.encode(reals) # [b, k, t]

            attributes = [ConditioningAttributes(text={'description': md["prompt"][0][:512]}) for md in metadata]
            attributes = self.lm.cfg_dropout(attributes)
            attributes = self.lm.att_dropout(attributes)
            tokenized = self.lm.condition_provider.tokenize(attributes)
     
            with torch.cuda.amp.autocast(enabled=False):
                condition_tensors = self.lm.condition_provider(tokenized)
                
            lm_output = self.lm.compute_predictions(
                codes=codes,
                conditions = [],
                condition_tensors = condition_tensors,
            )

            logits = lm_output.logits # [b, k, t, c]
            logits_mask = lm_output.mask # [b, k, t]

            cross_entropy, cross_entropy_per_codebook = self._compute_cross_entropy(logits, codes, logits_mask)

            loss = cross_entropy

        log_dict = {
            'train/loss': loss.detach(),
            'train/cross_entropy': cross_entropy.detach(),
            'train/perplexity': torch.exp(cross_entropy).detach(),
        }

        for k, ce_q in enumerate(cross_entropy_per_codebook):
            log_dict[f'cross_entropy_q{k + 1}'] = ce_q
            log_dict[f'perplexity_q{k + 1}'] = torch.exp(ce_q)

        self.log_dict(log_dict, prog_bar=True, on_step=True)
        return loss

    def on_before_zero_grad(self, *args, **kwargs):
        self.lm_ema.update()

    def export_model(self, path):
        self.musicgen_model.lm = self.lm_ema.ema_model
        export_state_dict = {"state_dict": self.musicgen_model.state_dict()}
        
        torch.save(export_state_dict, path)

class MusicGenDemoCallback(pl.Callback):
    def __init__(self, 
                 demo_every=2000,
                 num_demos=8,
                 sample_size=65536,
                 sample_rate=48000,
                 demo_conditioning: tp.Optional[tp.Dict[str, tp.Any]] = None,
                 demo_cfg_scales: tp.Optional[tp.List[int]] = [3, 5, 7],
                 **kwargs
    ):
        super().__init__()

        self.demo_every = demo_every
        self.num_demos = num_demos
        self.demo_samples = sample_size
        self.sample_rate = sample_rate
        self.last_demo_step = -1
        self.demo_conditioning = demo_conditioning
        self.demo_cfg_scales = demo_cfg_scales

    @rank_zero_only
    @torch.no_grad()
    def on_train_batch_end(self, trainer, module: MusicGenTrainingWrapper, outputs, batch, batch_idx):        

        if (trainer.global_step - 1) % self.demo_every != 0 or self.last_demo_step == trainer.global_step:
            return

        module.eval()

        print(f"Generating demo")
        self.last_demo_step = trainer.global_step

        demo_length_sec = self.demo_samples // self.sample_rate

        try:
            print("Getting conditioning")

            prompts = [md["prompt"][:512] for md in self.demo_conditioning]

            for cfg_scale in self.demo_cfg_scales:

                module.musicgen_model.set_generation_params(duration=demo_length_sec, cfg_coef=cfg_scale)

                with torch.cuda.amp.autocast():
                    print(f"Generating demo for cfg scale {cfg_scale}")
                    fakes = module.musicgen_model.generate(prompts, progress=True)

                # Put the demos together
                fakes = rearrange(fakes, 'b d n -> d (b n)')

                log_dict = {}
                
                filename = f'demo_cfg_{cfg_scale}_{trainer.global_step:08}.wav'
                fakes = fakes.to(torch.float32).clamp(-1, 1).mul(32767).to(torch.int16).cpu()
                torchaudio.save(filename, fakes, self.sample_rate)

                log_dict[f'demo_cfg_{cfg_scale}'] = wandb.Audio(filename,
                                                    sample_rate=self.sample_rate,
                                                    caption=f'Reconstructed')
            
                log_dict[f'demo_melspec_left_cfg_{cfg_scale}'] = wandb.Image(audio_spectrogram_image(fakes))

                trainer.logger.experiment.log(log_dict)

        except Exception as e:
            raise e
        finally:
            gc.collect()
            torch.cuda.empty_cache()
            module.train()