File size: 4,102 Bytes
96fd80c
 
 
 
09857e0
 
 
96fd80c
 
09857e0
 
69bf1ce
96fd80c
 
 
09857e0
 
96fd80c
 
 
09857e0
 
96fd80c
 
09857e0
 
96fd80c
 
 
09857e0
96fd80c
09857e0
 
96fd80c
 
 
09857e0
 
96fd80c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
09857e0
96fd80c
 
586bb9c
 
 
 
 
bbea063
586bb9c
96fd80c
 
586bb9c
96fd80c
586bb9c
 
 
09857e0
96fd80c
 
586bb9c
96fd80c
586bb9c
 
 
09857e0
96fd80c
 
 
 
586bb9c
96fd80c
586bb9c
96fd80c
 
586bb9c
 
 
09857e0
586bb9c
 
09857e0
586bb9c
 
 
 
 
 
 
 
 
 
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
import csv
import datasets
from datasets import BuilderConfig, GeneratorBasedBuilder, DatasetInfo, SplitGenerator, Split

_PROMPTS_URLS = {
    "dev": "automatic/validation.csv",
    "train": "automatic/train.csv",
}

_PROMPTS_FILTERED_URLS = {
    "dev": "automatic/validation.csv",
    "train": "automatic/train.csv",
}

_ARCHIVES = {
    "dev": "automatic.tar.gz",
    "train": "automatic.tar.gz",
}

_PATH_TO_CLIPS = {
    "dev": "validation",
    "train": "train",
}

class NurcSPConfig(BuilderConfig):
    def __init__(self, prompts_type="original", **kwargs):
        super().__init__(**kwargs)
        self.prompts_type = prompts_type

class NurcSPDataset(GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        NurcSPConfig(name="original", description="Original audio prompts", prompts_type="original"),
        NurcSPConfig(name="filtered", description="Filtered audio prompts", prompts_type="filtered"),
    ]

    def _info(self):
        return DatasetInfo(
            features=datasets.Features(
                {
                    "audio_name": datasets.Value("string"),
                    "file_path": datasets.Value("string"),
                    "text": datasets.Value("string"),
                    "start_time": datasets.Value("string"),
                    "end_time": datasets.Value("string"),
                    "duration": datasets.Value("string"),
                    "quality": datasets.Value("string"),
                    "speech_genre": datasets.Value("string"),
                    "speech_style": datasets.Value("string"),
                    "variety": datasets.Value("string"),
                    "accent": datasets.Value("string"),
                    "sex": datasets.Value("string"),
                    "age_range": datasets.Value("string"),
                    "num_speakers": datasets.Value("string"),
                    "speaker_id": datasets.Value("string"),
                    "audio": datasets.Audio(sampling_rate=16_000),
                }
            )
        )

    def _split_generators(self, dl_manager):
        prompts_urls = _PROMPTS_URLS if self.config.prompts_type == "original" else _PROMPTS_FILTERED_URLS
        
        # Download the prompts CSV files and audio archive
        prompts_path = dl_manager.download_and_extract(prompts_urls)
        archive = dl_manager.download_and_extract(_ARCHIVES)

        # Return split generators
        return [
            SplitGenerator(
                name=Split.TRAIN,
                gen_kwargs={
                    "prompts_path": prompts_path["train"],
                    "path_to_clips": _PATH_TO_CLIPS["train"],
                    "audio_files": archive["train"],
                }
            ),
            SplitGenerator(
                name=Split.VALIDATION,  # Changed from Split.VALIDATION to match error message
                gen_kwargs={
                    "prompts_path": prompts_path["dev"],
                    "path_to_clips": _PATH_TO_CLIPS["dev"],
                    "audio_files": archive["dev"],
                }
            ),
        ]

    def _generate_examples(self, prompts_path, path_to_clips, audio_files):
        # Load CSV data
        examples = {}
        with open(prompts_path, "r", encoding='utf-8') as f:
            csv_reader = csv.DictReader(f)
            for row in csv_reader:
                examples[row['file_path']] = {
                    key: row[key] 
                    for key in row.keys()
                }

        # Process audio files
        id_ = 0
        for root, _, files in datasets.utils.py_utils.walk(audio_files):
            if path_to_clips in root:
                for fname in files:
                    file_path = f"{path_to_clips}/{fname}"
                    if file_path in examples:
                        full_path = f"{root}/{fname}"
                        with open(full_path, "rb") as audio_file:
                            audio = {"path": file_path, "bytes": audio_file.read()}
                            yield id_, {**examples[file_path], "audio": audio}
                            id_ += 1