| { | |
| "model_type": "gslm-speech-encoder", | |
| "architectures": ["SpeechEncoder"], | |
| "hubert_backend": "fairseq", | |
| "hubert_ckpt": "hubert_base_ls960.pt", | |
| "hubert_hf_name": "facebook/hubert-base-ls960", | |
| "hubert_layer": 9, | |
| "expected_sample_rate": 16000, | |
| "code_hop_size": 320, | |
| "quantizer_file": "kmeans_100.pt", | |
| "quantizer_key": "", | |
| "deduplicate": true, | |
| "add_bos_eos": false, | |
| "need_f0": false, | |
| "feature_norm": null | |
| } | |