Spaces:
Sleeping
Sleeping
File size: 2,772 Bytes
599236e b70b9c3 a3ea780 b70b9c3 a3ea780 3e97a5d b70b9c3 3e97a5d b70b9c3 a3ea780 b70b9c3 a3ea780 b70b9c3 3e97a5d b70b9c3 a3ea780 031f538 a3ea780 | 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 | import os
from dataclasses import dataclass, field
from pathlib import Path
from typing import List, Optional
@dataclass
class ProcessingConfig:
audio_path: Path = Path("data/audio/0")
augmented_path: Path = Path("data/audio/")
log_mel_path: Path = Path("data/preprocessed")
n_bands: int = 128
n_mels: int = 128
frame_size: int = 1024
hop_size: int = 1024
sample_rate: int = 44100
fft_size: int = 8192
target_seconds: float = 5.0
augmentation_probability_lists = [
[0.0, 1.0, 1.0], [1.0, 1.0, 0.0], [1.0, 0.0, 1.0],
[0.0, 0.0, 0.0], [0.5, 0.5, 0.5],
]
time_stretch_rates = [0.81, 0.93, 1.07, 1.23]
pitch_shift_rates = [-3.5, -2.5, -2, -1, 1, 2.5, 3, 3.5]
drc_types = ["radio", "filmstandard", "musicstandard", "speech"]
@dataclass
class DatasetConfig:
cnn_input_length: int = 128
sample_rate: int = 44100
esc50_labels: List[str] = field(default_factory=lambda: [
'dog', 'rooster', 'pig', 'cow', 'frog',
'cat', 'hen', 'insects', 'sheep', 'crow',
'rain', 'sea_waves', 'crackling_fire', 'crickets', 'chirping_birds',
'water_drops', 'wind', 'pouring_water', 'toilet_flush', 'thunderstorm',
'crying_baby', 'sneezing', 'clapping', 'breathing', 'coughing',
'footsteps', 'laughing', 'brushing_teeth', 'snoring', 'drinking_sipping',
'door_wood_knock', 'mouse_click', 'keyboard_typing', 'door_wood_creaks', 'can_opening',
'washing_machine', 'vacuum_cleaner', 'clock_alarm', 'clock_tick', 'glass_breaking',
'helicopter', 'chainsaw', 'siren', 'car_horn', 'engine',
'train', 'church_bells', 'airplane', 'fireworks', 'hand_saw'
])
@dataclass
class DownloadConfig:
repo_url: str = "https://github.com/karolpiczak/ESC-50/archive/refs/heads/master.zip"
repo_dst_dir: Path = Path("data")
audio_dst_dir: Path = field(init=False)
extracted_dir: Path = os.path.join(repo_dst_dir, "ESC-50-master")
audio_src_dir = os.path.join(extracted_dir, "audio")
paths_to_delete: List[str] = field(default_factory=lambda: [
".gitignore", "esc50.gif", "LICENSE", "pytest.ini", "README.md",
"requirements.txt", "tests", "meta", ".github", ".circleci", "ESC-50-master"
])
def __post_init__(self):
object.__setattr__(self, "audio_dst_dir", self.repo_dst_dir / "audio" / "0")
@dataclass
class TrainConfig:
epochs: int = 50
batch_size: int = 100
lr: int = 0.001
device = "cuda"
use_all_patches: bool = True
samples_per_epoch_fraction: float = 1/8
checkpoint_dir_cnn: str = "models/cnn/checkpoints"
checkpoint_dir_transformer: str = "models/transformer/checkpoints"
save_every_n_epoch: int = 1
resume_from: Optional[str] = None
|