Closed lllllliuji closed 2 months ago
It depends on how fast your CPU is, whether you have enough memory, and how long you are willing to wait for the transcription to finsh.
@lllllliuji , hello. We now support multi quantization precision: int8, int16, fp16, bf16, vv. For more details, you can refer to this doc
If I use a CPU for speech conversion, what is the highest precision that can be supported?