[226]NAVER 광고 deep click prediction: 모델링부터 서빙까지NAVER D2
This document presents a formula for calculating the loss function J(θ) in machine learning models. The formula averages the negative log likelihood of the predicted probabilities being correct over all samples S, and includes a regularization term λ that penalizes predicted embeddings being dissimilar from actual embeddings. It also defines the cosine similarity term used in the regularization.
[214] Ai Serving Platform: 하루 수 억 건의 인퍼런스를 처리하기 위한 고군분투기NAVER D2
The document discusses running a TensorFlow Serving (TFS) container using Docker. It shows commands to:
1. Pull the TFS Docker image from a repository
2. Define a script to configure and run the TFS container, specifying the model path, name, and port mapping
3. Run the script to start the TFS container exposing port 13377
The document discusses linear algebra concepts including:
- Representing a system of linear equations as a matrix equation Ax = b where A is a coefficient matrix, x is a vector of unknowns, and b is a vector of constants.
- Solving for the vector x that satisfies the matrix equation using linear algebra techniques such as row reduction.
- Examples of matrix equations and their component vectors are shown.
This document describes the steps to convert a TensorFlow model to a TensorRT engine for inference. It includes steps to parse the model, optimize it, generate a runtime engine, serialize and deserialize the engine, as well as perform inference using the engine. It also provides code snippets for a PReLU plugin implementation in C++.
The document discusses machine reading comprehension (MRC) techniques for question answering (QA) systems, comparing search-based and natural language processing (NLP)-based approaches. It covers key milestones in the development of extractive QA models using NLP, from early sentence-level models to current state-of-the-art techniques like cross-attention, self-attention, and transfer learning. It notes the speed and scalability benefits of combining search and reading methods for QA.
This document describes the steps to convert a TensorFlow model to a TensorRT engine for inference. It includes steps to parse the model, optimize it, generate a runtime engine, serialize and deserialize the engine, as well as perform inference using the engine. It also provides code snippets for a PReLU plugin implementation in C++.
The document discusses various machine learning clustering algorithms like K-means clustering, DBSCAN, and EM clustering. It also discusses neural network architectures like bidirectional LSTMs, attention mechanisms, and convolutional sequence-to-sequence models. Finally, it shows charts comparing the performance of different chatbot models on validation data.
The document discusses designing hardware chips for artificial intelligence. It emphasizes that microarchitecture exploration is important and takes time, suggesting allocating 6-8 months to build performance models. The document also explains that hardware description languages like Verilog and VHDL are used to specify chip designs at the logic gate level, and that physical compilation translates this into an actual chip design. Overall, the document provides an overview of the chip design process from exploring microarchitectures to physical implementation.
[246]QANet: Towards Efficient and Human-Level Reading Comprehension on SQuADNAVER D2
Adams Wei Yu is a PhD candidate at CMU working on machine reading comprehension and large scale optimization. His advisors are Jaime Carbonell and Alex Smola. He has worked on question answering models and datasets like SQuAD. QANet is one of his contributions, which uses self-attention and convolutional layers instead of RNNs for question answering. It achieves state-of-the-art results while being much faster to train and run than previous models.
This document describes several algorithms for calculating relevance scores between queries and documents. It discusses TF-IDF scoring, where term frequency is multiplied by inverse document frequency. It also describes using cosine similarity to calculate image similarity based on visual features. Finally, it discusses combining base ranking scores with image similarity scores.
[145] printf("Hello, AR"); //세상을 바꾸는 새로운 눈NAVER D2
This document discusses a company that has the world's largest collection of 3D real estate data at over 5.5 million units and 3D product data at over 7,000 units. The company provides intelligent foundations, apps, analytics and things through digital twins, from cloud to edge, on a conversational platform and immersive experiences using mesh blockchain and event-driven, continuous adaptive risk and trust models and digital twins technology.
46. <ul>
<li>The <em>quick</em> <b>brown <big>fox</big></b> jumps over
the <em>lazy</em> <b><big>dog</big></b>.</li>
<li>The <b><big>dog</big></b> is so <em>startled</em> he almost
jumps out of his skin.</li>
</ul>