Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
awesome-aiml-blog
These are excerpts from various blog posts related to AI & ML that deal with important research and business cases.
https://github.com/youngmki/awesome-aiml-blog
- Token Merging: Your *ViT* but Faster (Feb 2023, Meta)
- Autoencoders and Diffusers: A Brief Comparison (Dec 2022)
- *Data2vec 2.0*: Highly Efficient Self-Supervised Learning for Vision, Speech and Text (Dec 2022, Meta)
- Text-to-Image: Diffusion, Text Conditioning, Guidance, Latent Space (Nov 2022)
- How Should AI Systems Behave, and Who Should Decide? (Feb 2023, OpenAI)
- *Atlas*: Few-Shot Learning with Retrieval Augmented LMs (Jan 2023, Meta)
- Overcoming The Limitations Of LLMs (Jan 2023)
- Some Remarks on LLMs (Jan 2023)
- Accelerating Text Generation with Confident Adaptive Language Modeling (*CALM*) (Dec 2022, Google)
- Illustrating Reinforcement Learning from Human Feedback (RLHF) (Dec 2022, HuggingFace)
- Better LMs without Massive Compute (Nov 2022, Google)
- *ChatGPT*: Optimizing LMs for Dialogue (Nov 2022, OpenAI)
- Choosing the Right LM For Your NLP Use Case (Nov 2022)
- MoE with Expert Choice Routing (Nov 2022, Google)
- Robots That Write Their Own Code (Nov 2022, Google)
- The State of Multilingual AI (Nov 2002)
- Aligning LMs to Follow Instructions (Jan 2022, OpenAI)
- Mastering Stratego, the Classic Game of Imperfect Information (Dec 2022, DeepMind)
- Research @ MS 2022: A Look Back at a Year of Accelerating Progress in AI (Dec 2022, Microsoft)
- Swin Transformer v2 - us/research/blog/godel-combining-goal-oriented-dialog-with-real-world-conversations/), 멀티모달 [BEiT](https://github.com/microsoft/unilm/tree/master/beit), 언어 모델 디버깅 도구 [AdaTest](https://www.microsoft.com/en-us/research/blog/partnering-people-with-large-language-models-to-find-and-fix-bugs-in-nlp-systems/), 유해 콘텐츠 데이터 생성 도구 [ToxiGen](https://www.microsoft.com/en-us/research/blog/detoxigen-leveraging-large-language-models-to-build-more-robust-hate-speech-detection-tools/), NN HPO기법 [µTransfer](https://www.microsoft.com/en-us/research/blog/%c2%b5transfer-a-technique-for-hyperparameter-tuning-of-enormous-neural-networks/), 모델 압축과 양자화를 위한 [DeepSpeed Compression](https://www.microsoft.com/en-us/research/blog/deepspeed-compression-a-composable-library-for-extreme-compression-and-zero-cost-quantization/) 등
- Evolutionary-Scale Prediction of Atomic Level Protein Structure with a LM (Nov 2022, Meta)
- Teaching AI Advanced Mathematical Reasoning (Nov 2022, Meta)
- Building on Top of Black Magic (Oct 2022)
- Accelerated *Stable Diffusion* with **PyTorch 2** (Jan 2023)
- SD 2.1
- Discovering Creative Insights in Promotional Artwork (Jan 2023, Netflix)
- Real-Time Tracking of Wildfire Boundaries Using Satellite Imagery (Jan 2023, Google)
- When a Picture is Worth More Than Words (Dec 2022, Airbnb)
- AI 명함 촬영 인식 '리오(RIO)' 적용기 - ML Model Converter와 안드로이드 앱 적용기 (Nov 2022, 드라마앤컴퍼니)
- 1편
- For Your Eyes Only: Improving Netflix Video Quality with Neural Networks (Nov 2022, Netflix)
- FFmpeg
- Match Cutting at Netflix: Finding Cuts with Smooth Visual Transitions (Nov 2022, Netflix)
- PySceneDetect
- Build a Robust Text-Based Toxicity Predictor (Dec 2022, Amazon)
- TextAttack
- Luda Gen 1, 더 재미있고 자연스러운 대화로 돌아온 루다 1편 - 생성 기반 챗봇 (Dec 2022, 스캐터랩)
- 루다, 눈을 뜨다! 포토챗 베타의 멀티모달 기술 소개 (Nov 2022, 스캐터랩)
- AI Hub
- Augment Your HuggingFace Model with Human in the Loop (Nov 2022)
- How AI Text Generation Models are Reshaping Customer Support at Airbnb (Nov 2022, Airbnb)
- Prioritizing Home Attributes Based on Guest Interest (Feb 2023, Airbnb)
- Unsupervised and Semi-Supervised Anomaly Detection with Data-Centric ML (Feb 2023, Google)
- Improving the Customer’s Experience via ML-Driven Payment Routing (Jan 2023, LinkedIn)
- Learning To Rank Diversely (Jan 2023, Airbnb)
- Powering Millions of Real-Time Decisions with **LyftLearn** Serving (Jan 2023, Lyft)
- **PyTorch** Trace Analysis for the Masses (Jan 2023)
- Scalable Annotation Service — **Marken** (Jan 2023, Netflix)
- Open Distro
- Unleashing ML Innovation at Spotify with **Ray** (Jan 2023, Spotify)
- How to Evaluate the Quality of the Synthetic Data – Measuring from the Perspective of Fidelity, Utility, and Privacy (Dec 2022, Amazon)
- 이전 글 - synthetic](https://github.com/ydataai/ydata-synthetic) 라이브러리의 WGAN-GP로 태뷸러 데이터 합성 → 충실도(통계량), 유용성(예측 성능), 정보 보호(암기 X) 기준으로 평가
- 그 많던 벡터는 다 어디로 갔을까? **Milvus** 활용기 (Nov 2022, 라이너)
- Building Airbnb Categories with ML and Human in the Loop (Nov 2022, Airbnb)
- Efficient Multi-Objective Neural Architecture Search with **Ax** (Nov 2022)
- How Pinterest Leverages Realtime User Actions in Recommendation to Boost Homefeed Engagement Volume (Nov 2022, Pinterest)
- How Prime Video Distills Time Series Anomalies into Actionable Alarms (Nov 2022, Amazon)
- ML for Fraud Detection in Streaming Services (Nov 2022, Netflix)
- Scaling Media ML at Netflix (Feb 2023, Netflix)
- DL Tuning Playbook (Jan 2023, Google)
- Open Source **Vizier**: Towards Reliable and Flexible Hyperparameter and Blackbox Optimization (Jan 2023, Google)
- The **Flan** Collection: Advancing Open Source Methods for Instruction Tuning (Jan 2023, Google)
- 새로운 루다를 지탱하는 모델 서빙 아키텍처 — 1편: A/B 테스트를 위한 구조 설계 (Jan 2023, 스캐터랩)
- Self-Serve Feature Platforms: Architectures and APIs (Jan 2023)
- Feathr
- FMS(차량 관제 시스템) 데이터 파이프라인 구축기 2편. 신뢰성 높은 데이터를 위한 테스트 환경 구축기 (Jan 2023, 쏘카)
- 1편
- Automatically Retrain NNs with **Renate** (Dec 2022, Amazon)
- Renate
- CLOVA MD 상품추천 솔루션 (Dec 2022, 네이버)
- Introducing **Fortuna**: A Library for Uncertainty Quantification (Dec 2022, Amazon)
- Fortuna
- ML 모델 서빙 비용 1/4로 줄이기 (Dec 2022, 하이퍼커넥트)
- Ready-to-go Sample Data Pipelines with **Dataflow** (Dec 2022, Netflix)
- 전사 구성원들이 사용하는 배치 데이터 플랫폼 만들기 - **Airflow** Advanced (Nov 2022, 쏘카)
- 쿠팡 로켓그로스의 ML 플랫폼: 20개 이상의 모델 서비스 및 트래픽 처리 비용 효율화 (Nov 2022, 쿠팡)
- Accelerating Large *GPT* Training with Sparse Pre-Training and Dense Fine-Tuning (Nov 2022)
- Apache Beam으로 ML 데이터 파이프라인 구축하기 (Nov 2022, 스캐터랩)
- 1편 - beam-2/)도 읽어보세요.
- CLOps - CLOps가 바꿔 놓은 CLOVA의 모델 상용화 프로세스 (Nov 2022, 네이버)
- 1편 - the-heart-of-the-platform), [3편](https://engineering.clova.ai/posts/2022/09/clops-resource-management), [4편](https://engineering.clova.ai/posts/2022/10/clops-model-management)도 읽어보세요.
- Kurly만의 MLOps 구축하기 - KubeFlow 도입기 (Nov 2022, 컬리)
- 1편
- Explain Text Classification Model Predictions Using **Amazon SageMaker Clarify** (Jan 2023)
- Measure the Business Impact of **Amazon Personalize** Recommendations (Feb 2023)
- **Amazon SageMaker** Built-In **LightGBM** Now Offers Distributed Training Using **Dask** (Jan 2023)
- **Amazon SageMaker** Automatic Model Tuning Now Supports Three New Completion Criteria for HPO (Feb 2023)
- Optimize Your ML Deployments with Auto Scaling on **Amazon SageMaker** (Feb 2023)
- Scaling LLM Training with **Amazon EC2 Trn1 UltraClusters** (Feb 2023)
- 농심의 **Amazon SageMaker**를 활용한 원자재 가격예측과 MLOps 여정 (Feb 2023)
- AWS를 이용한 MLOps 구축 사례 살펴보기 (Jan 2023)
- Best Practices for Load Testing **Amazon SageMaker** Real-Time Inference Endpoints (Jan 2023)
- Locust
- Enriching Real-Time News Streams with the Refinitiv Data Library, AWS Services, and **Amazon SageMaker** (Jan 2023)
- How Thomson Reuters Built an AI platform Using **Amazon SageMaker** to Accelerate Delivery of ML Projects (Jan 2023)
- Model Hosting Patterns in **Amazon SageMaker**, Part 1: Common Design Patterns for Building ML Applications on **Amazon SageMaker** (Jan 2023)
- Best Practices for **Amazon SageMaker Training Managed Warm Pools** (Dec 2022)
- Scaling Distributed Training with **AWS Trainium** and **Amazon EKS** (Jan 2023)
- TorchX 범용 작업 시작 관리자 - 값 저장소) + TorchX DDP 모듈
- XLA 장치
- Define Customized Permissions in Minutes with **Amazon SageMaker Role Manager** (Dec 2022)
- Deploy **Amazon SageMaker Autopilot** Models to **Serverless Inference Endpoints** (Dec 2022)
- Improve Governance of Your ML Models with **Amazon SageMaker** (Dec 2022)
- Minimize the Production Impact of ML Model Updates with **Amazon SageMaker Shadow Testing** (Dec 2022)
- New Performance Improvements in **Amazon SageMaker Model Parallel Library** (Dec 2022)
- FlashAttention - NeoX 100B의 경우 이전 버전 대비 훈련 속도 30% 향상
- Next Generation **Amazon SageMaker Experiments** – Organize, Track, and Compare Your ML Trainings at Scale (Dec 2022)
- Speed ML Development Using **SageMaker Feature Store** and **Apache Iceberg** Offline Store Compaction (Dec 2022)
- Build a Cross-Account MLOps Workflow Using the **Amazon SageMaker Model Registry** (Nov 2022)
- ByteDance Saves Up to 60% on Inference Costs While Reducing Latency and Increasing Throughput Using **AWS Inferentia** (Nov 2022)
- Deploy *BLOOM-176B* and *OPT-30B* on **Amazon SageMaker** with Large Model Inference **DLC**s and **DeepSpeed** (Nov 2022)
- 컨테이너 패키지
- Enabling Hybrid ML Workflows on **Amazon EKS** and **Amazon SageMaker** with One-Click **Kubeflow** on AWS Deployment (Nov 2022)
- Getting Started with Deploying Real-Time Models on **Amazon SageMaker** (Nov 2022)
- Launch **Amazon SageMaker Autopilot** Experiments Directly from within **Amazon SageMaker Pipelines** to Easily Automate MLOps Workflows (Nov 2022)
- Model Hosting Patterns in **SageMaker**: Best Practices in Testing and Updating Models on **SageMaker** (Nov 2022)
- Run ML Inference Workloads on **AWS Graviton**-Based Instances with **Amazon SageMaker** (Nov 2022)
- Serve Multiple Models with **Amazon SageMaker** and **Triton Inference Server** (Nov 2022)
- Choose the Best Data Source for Your **Amazon SageMaker** Training Job (Feb 2022)
- Using Streaming Ingestion with **Amazon SageMaker Feature Store** to Make ML-Backed Decisions in Near-Real Time (Dec 2020)
- Faker
Keywords
machine-learning
2
adversarial-attacks
1
adversarial-examples
1
adversarial-machine-learning
1
data-augmentation
1
natural-language-processing
1
nlp
1
security
1
apache-spark
1
artificial-intelligence
1
azure
1
data-engineering
1
data-quality
1
data-science
1
feature-engineering
1
feature-governance
1
feature-management
1
feature-marketplace
1
feature-metadata
1
feature-platform
1
feature-store
1
mlops
1
benchmarking
1
http
1
load-generator
1
load-test
1
load-testing
1
load-tests
1
locust
1
performance
1
performance-testing
1
python
1