optimizeModel
Whether to allow model optimization techniques such as quantization, speculative decoding, and kernel tuning. The default is true.
Whether to allow model optimization techniques such as quantization, speculative decoding, and kernel tuning. The default is true.