Update README + metadata: severity recomputed from canonical speaker mapping (test Mild 197->421, Severe 438->214).

Re-encode LibriSpeech supplement from FLAC to WAV (PCM16, 16 kHz mono) for format uniformity

Update README + metadata: severity recomputed from canonical speaker mapping (test Mild 197->421, Severe 438->214).

Fix per-speaker severity labels to match canonical mapping (TORGO + UA-Speech)

Update README + metadata: LibriSpeech Normal supplement, canonicalized gender/speaker_id, new split counts (44737 train / 1780 test).

Canonicalize metadata across all shards:

Add LibriSpeech train-clean-100 'Normal' supplement (train+test, speaker-disjoint) to break the Normal=TORGO-only corpus-severity confound.

Add dataset metadata JSON

Add train/test splits (test=1500 stratified by severity)

Rename eval_heldout_metadata.json to pathological_speech .json

Add evaluation metadata JSON

Add held-out evaluation set (unseen speakers, severity metadata)

initial commit