Quantization vs convaiinnovations/laya. cavi-ai/laya-MLX-8bit is a quantized version of the other (U32/F16).
On the Hub · Downloads on Model Pulse