What is distant supervision (NER from KB)?
Machine Learning / AI
Have you had experience working with NLP tasks?
What is query rewriting in RAG?
What is Megatron-LM and its tensor parallel approach?
How did you split the sample into training and testing?
Why does regularization help fight overfitting?
Why can model weights become large and why is it necessary to reduce them with regularization?
Explain the scaled dot-product attention formula. Why divide by sqrt(d_k)?
What is model-based RL and what are its advantages over model-free?
How did you choose the model and hyperparameters, and how did you select the classification threshold?
Как должна выглядеть разметка для обучения трансформера распознавать спецтерминологию?
How does PR-AUC differ from ROC-AUC?
How are MAP and MRR metrics calculated in ranking tasks?
What is the difference between NaN and None in Python?
What algorithm is the model based on?
What task does transformer model tuning solve?
What should be contained in a training dataset for training with cross-entropy?
What is NTK-aware scaling RoPE for extending the context window?
What does the result of CNN decomposition look like visually?
What are recovery prompts in low-confidence?