This feature enables the r2inference support for multiple predictions. When received more than 1 output tensor, they will be concatenated into a single 1D array. Currently, no elements should be affected by this since all of them will continue to use the only output available for the supported models.
This feature enables the r2inference support for multiple predictions. When received more than 1 output tensor, they will be concatenated into a single 1D array. Currently, no elements should be affected by this since all of them will continue to use the only output available for the supported models.