We’re releasing a new cross-lingual Question Answering dataset, MLQA: Evaluating Cross-lingual Extractive Question Answering Check out the dataset here