Play the clip inline in a notebook, and say what Colab needs

Fix the usage snippet: speaker.npy needs unsqueeze(0), and the processor ships in this repo

Score the card against all four ASR judges, including the Cebuano-finetuned one

Rewrite the card: total clip count, bench results, and the speaker-conditioning limit

Drop the duplicated final/ folder (identical sha to the root weights)

Add sample_1.wav

Add sample_0.wav

Continue-finetune on 13807 ceb clips

Model save

Training in progress, step 4000, checkpoint

Training in progress, step 4000

Training in progress, step 3500, checkpoint

Training in progress, step 3500

Training in progress, step 3000, checkpoint

Training in progress, step 3000

Training in progress, step 2500, checkpoint

Training in progress, step 2500

Training in progress, step 2000, checkpoint

Training in progress, step 2000

Training in progress, step 1500, checkpoint

Training in progress, step 1500

Training in progress, step 1000, checkpoint

Training in progress, step 1000

Training in progress, step 500, checkpoint

Training in progress, step 500

initial commit