X has 8 features, but MLPClassifier is expecting 64 features as input.

If your data has only 8 features and your model expects 64 features as input, then there may be a mismatch between the dimensions of your input data and the model's input layer. One possible solution is to add additional layers to your model that can learn higher-level representations of the input features, which can help to map the 8 features to the 64 expected by the model. Another option is to use a technique called dimensionality reduction, such as Principal Component Analysis (PCA), to reduce the number of input features while still preserving the important information. This can help to reduce the computational complexity of the model and improve its performance.