Please give some context. What this error means is that the product of your tensor’s shape[0] & shape[1] must meet the expected size of 8 but is 64. You may debug the output of tensor operations that’s shown in the error log.
