restore Usage section; fix column name (id -> sample_id) and category values

regenerate parquet from metadata.csv + wav (274 rows, adds mixed)

remove stray Mandarin/1 1.wav (not listed in metadata.csv)

Update dataset card: mention Mixed language, correct total to 274

add mixed Hokkien/Mandarin waiting clips (8)

Retrigger dataset-server processing

Upload dataset

Add structured ER FAQ ASR eval set (Hokkien/Mandarin, metadata, dataset card) (part 2)

Add structured ER FAQ ASR eval set (Hokkien/Mandarin, metadata, dataset card)

Remove old flat sample files, replaced by structured dataset

Add Hokkien/Mandarin ASR sample audio

initial commit