Enviar aquest missatge de text: SyncFlow: Toward Temporally Aligned Joint Audio-Video Generation from Text