Evaluation fails when all samples are cached

[baberabb]

When all samples are already cached, the process errors out instead (instead of skipping to the metric calculation) on the subsequent run due to lack of requests to pass on to the LM.