File size: 4,604 Bytes
bd3a23c |
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 |
import os
import torch
import soundfile as sf
import pandas as pd
from tqdm import tqdm
from utils import minmax_norm_diff, reverse_minmax_norm_diff
from spk_ext import se_extractor
def rescale_noise_cfg(noise_cfg, noise_pred_text, guidance_rescale=0.0):
"""
Rescale `noise_cfg` according to `guidance_rescale`. Based on findings of [Common Diffusion Noise Schedules and
Sample Steps are Flawed](https://arxiv.org/pdf/2305.08891.pdf). See Section 3.4
"""
std_text = noise_pred_text.std(dim=list(range(1, noise_pred_text.ndim)), keepdim=True)
std_cfg = noise_cfg.std(dim=list(range(1, noise_cfg.ndim)), keepdim=True)
# rescale the results from guidance (fixes overexposure)
noise_pred_rescaled = noise_cfg * (std_text / std_cfg)
# mix with the original results from guidance by factor guidance_rescale to avoid "plain looking" images
noise_cfg = guidance_rescale * noise_pred_rescaled + (1 - guidance_rescale) * noise_cfg
return noise_cfg
@torch.no_grad()
def inference_timbre(gen_shape, text,
model, scheduler,
guidance_scale=5, guidance_rescale=0.7,
ddim_steps=50, eta=1, random_seed=2023,
device='cuda',
):
text, text_mask = text
model.eval()
generator = torch.Generator(device=device).manual_seed(random_seed)
scheduler.set_timesteps(ddim_steps)
# init noise
noise = torch.randn(gen_shape, generator=generator, device=device)
latents = noise
for t in scheduler.timesteps:
latents = scheduler.scale_model_input(latents, t)
if guidance_scale:
output_text = model(latents, t, text, text_mask, train_cfg=False)
output_uncond = model(latents, t, text, text_mask, train_cfg=True, cfg_prob=1.0)
output_pred = output_uncond + guidance_scale * (output_text - output_uncond)
if guidance_rescale > 0.0:
output_pred = rescale_noise_cfg(output_pred, output_text,
guidance_rescale=guidance_rescale)
else:
output_pred = model(latents, t, text, text_mask, train_cfg=False)
latents = scheduler.step(model_output=output_pred, timestep=t, sample=latents,
eta=eta, generator=generator).prev_sample
# pred = reverse_minmax_norm_diff(latents, vmin=0.0, vmax=0.5)
# pred = torch.clip(pred, min=0.0, max=0.5)
return latents
@torch.no_grad()
def eval_plugin_light(vc_model, text_model,
timbre_model, timbre_scheduler, timbre_shape,
val_meta, val_folder,
guidance_scale=3, guidance_rescale=0.7,
ddim_steps=50, eta=1, random_seed=2024,
device='cuda',
epoch=0, save_path='logs/eval/', val_num=10, sr=24000):
tokenizer, text_encoder = text_model
df = pd.read_csv(val_meta)
save_path = save_path + str(epoch) + '/'
os.makedirs(save_path, exist_ok=True)
step = 0
for i in range(len(df)):
row = df.iloc[i]
source_path = val_folder + row['path']
prompt = [row['prompt']]
with torch.no_grad():
text_batch = tokenizer(prompt,
max_length=32,
padding='max_length', truncation=True, return_tensors="pt")
text, text_mask = text_batch.input_ids.to(device), \
text_batch.attention_mask.to(device)
text = text_encoder(input_ids=text, attention_mask=text_mask)[0]
spk_embed = inference_timbre(timbre_shape, [text, text_mask],
timbre_model, timbre_scheduler,
guidance_scale=guidance_scale, guidance_rescale=guidance_rescale,
ddim_steps=ddim_steps, eta=eta, random_seed=random_seed,
device=device)
source_se = se_extractor(source_path, vc_model).to(device)
# print(source_se.shape)
# print(spk_embed.shape)
encode_message = "@MyShell"
vc_model.convert(
audio_src_path=source_path,
src_se=source_se,
tgt_se=spk_embed,
output_path=save_path + f'{step}_{prompt[0]}' + '.wav',
message=encode_message)
step += 1
if step >= val_num:
break
|