audiolm-pytorch==1.1.4
fairseq
huggingface-hub
sentencepiece
transformers
encodec
soundfile; platform_system == "Windows"
sox; platform_system != "Windows"