{ "format": "onnxruntime_matmul_nbits_int4", "bits": 4, "block_size": 128, "scope": "vocoder_decode_supported_ops_only", "quantized_ops": [ "MatMul", "Gather" ], "unquantized_ops_include": [ "Conv", "ConvTranspose" ], "base_model": "bosonai/higgs-audio-v3-tts-4b", "base_model_relation": "quantized" }