{"blocks": [{"key": "c8b41d98", "text": "Scenario", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "2663b335", "text": "Designing an NLP system for long-sequence text classification under tight inference latency constraints.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "b9e50536", "text": "Question", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "c02034aa", "text": "Contrast RNNs and Transformers in terms of architecture, parallelism, context handling, and training dynamics. Describe bagging and boosting ensemble techniques and when each is preferable.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "88999d1f", "text": "Hints", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "0bf05bd8", "text": "Discuss attention, parallelism, sequence length limits, variance reduction in bagging and bias reduction in boosting.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}], "entityMap": {}}