My model works well during training and validation.
But inference results differ even with similar inputs.
There’s no obvious bug in the code.
It feels like something subtle is off.
Lost your password? Please enter your email address. You will receive a link and will create a new password via email.
Please briefly explain why you feel this question should be reported.
Please briefly explain why you feel this answer should be reported.
Please briefly explain why you feel this user should be reported.
My model works well during training and validation.
But inference results differ even with similar inputs.
There’s no obvious bug in the code.
It feels like something subtle is off.
Training–serving skew occurs when feature transformations differ between training and inference.
This often happens when preprocessing is implemented separately in notebooks and production services. Even small differences in scaling, encoding, or default values can change predictions significantly.
The most reliable fix is to package preprocessing logic as part of the model artifact. Use shared libraries, serialized transformers, or pipeline objects that are reused during inference.
If that’s not possible, enforce strict feature tests that compare transformed outputs between environments.