VietMed_labeled / README.md
doof-ferb's picture
Upload 2 files
3a35a0a verified
metadata
license: cc-by-4.0
task_categories:
  - automatic-speech-recognition
  - text-to-speech
language:
  - vi
pretty_name: VietMed labeled set
size_categories:
  - 1K<n<10K
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: transcription
      dtype: string
    - name: Speaker ID
      dtype: string
  splits:
    - name: train
      num_bytes: 58513440.578
      num_examples: 2858
    - name: validation
      num_bytes: 56714850.712
      num_examples: 2912
    - name: test
      num_bytes: 70051704.606
      num_examples: 3437
  download_size: 183555285
  dataset_size: 185279995.896
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*

unofficial mirror of VietMed (Vietnamese speech data in medical domain) labeled set

official announcement: https://arxiv.org/abs/2404.05659

official download: https://huggingface.co/datasets/leduckhai/VietMed

this repo contains the labeled set: 9.2k samples

i also gather the metadata: see info.csv

my extraction code: https://github.com/phineas-pta/fine-tune-whisper-vi/blob/main/misc/vietmed-labeled.py

need to do: check misspelling, restore foreign words phonetised to vietnamese

usage with HuggingFace:

# pip install -q "datasets[audio]"
from datasets import load_dataset
from huggingface_hub import hf_hub_download
from pandas import read_csv

repo_id = "doof-ferb/VietMed_labeled"
dataset = load_dataset(repo_id, split="train", streaming=True)
info_file = hf_hub_download(repo_id=repo_id, filename="info.csv", repo_type="dataset")
info_dict = read_csv(info_file, index_col=0).to_dict("index")

def merge_info(batch):
    meta = info_dict.get(batch["Speaker ID"], "")
    if meta != "":
        batch["Recording Condition"] = meta["Recording Condition"]
        batch["ICD-10 Code"]         = meta["ICD-10 Code"]
        batch["Role"]                = meta["Role"]
        batch["Gender"]              = meta["Gender"]
        batch["Accent"]              = meta["Accent"]
    else:
        batch["Recording Condition"] = ""
        batch["ICD-10 Code"]         = ""
        batch["Role"]                = ""
        batch["Gender"]              = ""
        batch["Accent"]              = ""
    return batch
dataset = dataset.map(merge_info)