Speech tokeniser benchmarks – 2x compression vs OS SOTA
American accent
hardcoded_0_spk-unknown_spkemb_source_g_01560000 (15).mp3
hardcoded_0_spk-unknown_spkemb_source_g_01560000 (11).mp3
hardcoded_0_spk-unknown_spkemb_source_g_01560000 (14).mp3
hardcoded_0_spk-unknown_spkemb_source_g_01560000 (13).mp3