Wav2Gloss: Generating Interlinear Glossed Text from Speech