2025-12¶
- mise backend
- GitHub - cocoindex-io/cocoindex: Data transformation framework for AI. Ultra performant, with incremental processing. 🌟 Star if you like it!
- mkdocs の blogging plugin で creation を指定した時にうまく作成日を取得できていなかった。原因を調査すると、git の follow オプションが True になっているのが原因に思えた。code
- GitHub - d-kuro/gwq: 🌳 Git worktree manager with fuzzy finder - Work on multiple branches simultaneously, perfect for parallel AI coding workflows
- What is GPU Memory and Why it Matters for LLM Inference
- Amazon OpenSearch Service が GPU アクセラレーションと自動最適化でベクトルデータベースのパフォーマンスとコストを改善 | Amazon Web Services ブログ
- Yes, PyTorch is all for easy Tabular Deep Learning with new PyTorch-Frame
- torch_frame.gbdt — pytorch-frame documentation
- GitHub - pyg-team/pytorch-frame: Tabular Deep Learning Library for PyTorch
- メッセージキューを使い倒した私が語る、RabbitMQ vs Kafka vs Redis - CLYR
- メッセージ駆動が可能にする結合の最適化 - Speaker Deck
- GitHub - Cookee24/PairTranslate: 开源的沉浸式翻译，支持选段/查词/选词翻译 Open-source immersive translation, supporting segment/word lookup and selection translation
- GitHub - artie-labs/transfer: Database replication platform that leverages change data capture. Stream production data from databases to your data warehouse (Snowflake, BigQuery, Redshift, Databricks) in real-time.
- GitHub - TobikoData/sqlmesh: Scalable and efficient data transformation framework - backwards compatible with dbt.
- Self-host and scale web apps without Kubernetes complexity
- GitHub - mehrantsi/feox-server: High-Performance, Drop-In Replacement for Redis
- GitHub - ClickHouse/ClickHouse: ClickHouse® is a real-time analytics database management system
- GitHub - kevin1024/vcrpy: Automatically mock your HTTP interactions to simplify and speed up testing
- GitHub - NVIDIA/cutile-python: cuTile is a programming model for writing parallel kernels for NVIDIA GPUs
- cuTile Python — cuTile Python
- マルチノードで動かす vLLM のハマりどころ（LLM 開発コンペ 2025 参加記) #Ray
- Model picker for Copilot coding agent for Copilot Pro and Pro+ subscribers - GitHub Changelog
- GitHub - lakesoul-io/LakeSoul: LakeSoul is an end-to-end, realtime and cloud native Lakehouse framework with fast data ingestion, concurrent update and incremental data analytics on cloud storages for both BI and AI applications.
- GitHub - vdaas/vald: Vald. A Highly Scalable Distributed Vector Search Engine
- Polar — Monetize your software with ease
- GitHub - hydro-project/hydro: A Rust framework for correct and performant distributed systems
- Pydantic Logfire Pricing is Changing
- GitHub - danoon2/Boxedwine: Emulator that can run 32-bit Windows programs/games on multiple platforms via Wine
- GitHub - vllm-project/guidellm: Evaluate and Enhance Your LLM Deployments for Real-World Inference Needs
- pueue でｼﾞｮﾌﾞをｶｼﾞｭｱﾙに管理する
- AI スパコン「さくら ONE」で挑む LLM・HPC ベンチマーク (2) MLPerf GPT-3 175B 事前学習性能検証 | さくらのナレッジ
- 【Python】脱 dotenv！次世代の環境変数管理ツール「dotenvx」を使ってみた | 可茂 IT 塾
- GitHub - thedotmack/claude-mem: A Claude Code plugin that automatically captures everything Claude does during your coding sessions, compresses it with AI (using Claude's agent-sdk), and injects relevant context back into future sessions.
- SkyPilot 0.11: Multi-Cloud Pools for Batch Inference, Fast Managed Jobs, Enterprise-Ready at Scale, Programmability | SkyPilot Blog
- Ansible の WinRM と PSRP の違いをみてみよう(パフ �� ーマンス計測編） - 赤帽エンジニアブログ
- GitHub - samanhappy/mcphub: A unified hub for centrally managing and dynamically orchestrating multiple MCP servers/APIs into separate endpoints with flexible routing strategies
- 9 Python Libraries That Make You Look Like a Data Scientist
- GitHub - eli5-org/eli5: A library for debugging/inspecting machine learning classifiers and explaining their predictions
- GitHub - alteryx/featuretools: An open source python library for automated feature engineering
- polars 対応はしていない
- GitHub - shakedzy/dython: A set of data tools in Python
- GitHub - skrub-data/skrub: Machine learning with dataframes
- polars 対応はしていない
- GitHub - eli5-org/eli5: A library for debugging/inspecting machine learning classifiers and explaining their predictions
- GitHub - drivendataorg/cloudpathlib: Python pathlib-style classes for cloud storage services such as Amazon S3, Azure Blob Storage, and Google Cloud Storage.
- GitHub Copilot でカスタムエージェントを runSubagent できるようになったらしい
- subAgent からさらに subAgent を呼ぶのは不可
- GitHub - CogitatorTech/infera: A DuckDB extension for in-database inference
- GitHub - armink/FlashDB: An ultra-lightweight database that supports key-value and time series data | 一款支持 KV 数据和时序数据的超轻量级数据库
- pandas vs Polars vs DuckDB: A Data Scientist's Guide to Choosing the Right Tool
- マイクロサービスで避けられない分散トランザクション問題を Saga パターンで解決する
- Chrome DevTools MCP での既存セッションへの接続が楽になっていきそう #Chrome
- 【Omarchy】Intel Mac + 外付け SSD でデュアルブート開発環境 #Mac
- 【Go 言語】死なないジョブワーカー処理～数万並行処理でも落ちない設計～
- 今年一番 Python で関心した機能 #Python
- 分散サービスのトランザクションでアンチパターンを設計してしまった話 〜とその改善〜
- pixi で CUDA を管理する
- LLM 量子化 Day 4: 量子化手法の全体像 #システム開発
- LLM 量子化 Day 8: llama.cpp と GGUF 形式の世界 #システム開発
- GitHub Copilot now supports Agent Skills - GitHub Changelog
- The Institute for Ethical AI & Machine Learning
- Post
-
TorchVision, PIL, OpenCV のそれぞれを使用して画像をダウンサンプリングした場合の画像そのもののピクセルレベルの差分を計測した結果を追記した。破滅する。判断として、Python 環境でしか絶対使わない場合以外は PIL を使用するな。モデルの汎用性が下がる。
-
- GitHub - salsa-rs/salsa: A generic framework for on-demand, incrementalized computation. Inspired by adapton, glimmer, and rustc's query system.
- GitHub - wpnpeiris/nats-s3: S3 API compatible NATS Object Store
- GitHub - simstudioai/sim: Open-source platform to build and deploy AI agent workflows.
- GitHub - endorhq/rover: A manager for AI coding agents that works with Claude Code, Cursor, Gemini, Codex, and Qwen.
- GitHub - litebase/litebase: Open source distributed database built on SQLite and backed by distributed file systems and object storage
- GitHub - just-every/code: Every Code - push frontier AI to it limits. A fork of the Codex CLI with validation, automation, browser integration, multi-agents, theming, and much more. Orchestrate agents from OpenAI, Claude, Gemini or any provider.
- GitHub - SanjulaGanepola/github-local-actions: ✨ Run your GitHub Actions locally in VS Code!
- GitHub - sigoden/proxyfor: A powerful and flexible proxy CLI for capturing and inspecting HTTP(S) and WS(S) traffic, with TUI and WebUI.
- GitHub - wshobson/agents: Intelligent automation and multi-agent orchestration for Claude Code
- GitHub - drehelis/gcp-emulator-ui: A modern web interface for managing Google Cloud Platform emulator services 🎮
- Cloudflare Tunnel で内部公開の Cloud Run にアクセスしてみる | CyberAgent Developers Blog
- Claude Code Hooks ガイド：AI コーディングを自動化・カスタマイズする方法 | SIOS Tech Lab
- アウトボックスパターンとはなにか | SIOS Tech Lab
- 【コピペ OK】GitHub Copilot を"DeepWiki 化"するドキュメント指示書を作成してみた #AI
- Dockerfile を触らず GUI 環境と VNC 環境を後から入れる方法 #Linux
- dstack 0.20 GA: Fleet-first UX and other important changes
- 意外と知らない！？Claude Code で使える小技集 #Terminal
- GitHub - platform-engineering-labs/formae: Infrastructure-as-Code Platform Built for the Future
- SageMaker × vLLM で LLM 推論基盤を構築する方法【AsyncLLMEngine・LMI コンテナ設定ガイド】 - asken テックブログ
- Copilot memory early access for Pro and Pro+ - GitHub Changelog
- GitHub - rue-language/rue: A programming language that is higher level than Rust but lower level than Go. Someday, anyways.
- GitHub - shubh-io/DockMate: Terminal-based Docker manager - monitor CPU/memory, view logs, and manage containers
- GitHub - nicotsx/zerobyte: Backup automation for self-hosters. Built on top of restic
- GitHub - AndyMik90/Auto-Claude: Autonomous multi-session AI coding
- Lython
- GitHub Copilot にも追加された Agent Skills とは | ドクセル
- ローカル実行の常識！ LLM 量子化 ( Quantization ) の仕組みと手法を完全理解する #軽量化
- GitHub - vllm-project/vllm-omni: A framework for efficient model inference with omni-modality models
- GPU memory bottleneck
- GitHub - voidauth/voidauth: Single Sign-On for Your Self-Hosted Universe 🐈⬛🔒
- GitHub - 2mawi2/schaltwerk: The IDE without editor
- Top 5 AI Model Optimization Techniques for Faster, Smarter Inference | NVIDIA Technical Blog
- Python Typing Survey 2025: Code Quality and Flexibility As Top Reasons for Typing Adoption - Engineering at Meta
- “A Philosophy of Software Design” を 30 分でざっと理解する / Understand roughly "Philosophy of Software Design" in 30 minutes - Speaker Deck
- 【Cloud SQL】パスワード認証から IAM 認証へ「安全に」移行する手順
- Unsloth でお手軽に LLM の学習を高速化 #生成 AI
- Docker Sandboxes
- 自社 LLM を超高速でデプロイ！待ち時間を劇的に短縮する GKE 高速化テクニック #GoogleCloud
- image streaming + SSD
- GitHub - satococoa/wtp: 🌳 A powerful Git worktree CLI tool with automated setup, branch tracking, and smart navigation
.envのような git 管理していないファイルも worktree 間で共有できる
- GitHub - pgschema/pgschema: Terraform-style, declarative schema migration for Postgres
- 2026 年の PostgreSQL スキーマ管理は pgschema で #PostgreSQL
- コード差分に関係するテストだけ実行する pytest-testmon
-
SQLite3 のテーブルに、テストとそのテストで実行されたファイルの紐づけを持っており、変更されたファイル（のハッシュ）に紐づくテストを探して実行してるようです。
-
- GitHub - sgl-project/mini-sglang: A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.
- GitHub - kreuzberg-dev/kreuzberg: A polyglot document intelligence framework with a Rust core. Extract text, metadata, and structured information from PDFs, Office documents, images, and 50+ formats. Available for Rust, Python, Ruby, Go, and TypeScript/Node.js—or use via CLI, REST API, or MCP server.
- GitHub - duriantaco/skylos: Skylos is the watchdog for your repository. It maps your code's structure to hunt down dead logic, trace tainted data, and kill security rot
- GitHub - metaist/cosmofy: Cosmopolitan Python Bundler
- 分散推論基盤の基礎技術 〜高火力 PHY で作る分散推論基盤 vol.2〜 | さくらのナレッジ
- GitHub - ozeranskii/httptap: Rich-powered CLI that breaks each HTTP request into DNS, connect, TLS, wait, and transfer phases with waterfall timelines, compact summaries, or metrics-only output.
- GitHub - joouha/modshim: Override and customize Python packages without touching their code
- GitHub - cjpais/Handy: A free, open source, and extensible speech-to-text application that works completely offline.
- モデルによっては英語のみ対応している。Whisper Turbo は日本語にも対応していることを確認。
- 精度的には Aqua Voice > Handy (Whisper Turbo) > VSCode の音声認識 という印象
- Aqua Voice が Linux に対応していないので、その代替として考えていたが CPU だけのマシンでどこまで快適に動くかは不明。
- Handy 起動しておくとそれなりにメモリを消費するのが気になる
- GitHub - rendercv/rendercv: Typst-based CV/resume generator for academics and engineers
- vLLM の 1 年間の進化（v0.6.6→v0.13.0）をソースコードで追う #Python
- NVIDIA CUDA-X Powers the New Sirius GPU Engine for DuckDB, Setting ClickBench Records | NVIDIA Technical Blog
- Docker Desktop が重いから WSL2 で LazyDocker に移行してみた
- OSS を社内で使うときのミラー運用 #Git
- GitHub - antfu-collective/ni: 💡 Use the right package manager
- GitHub - ZingerLittleBee/dockerman.app: A modern, lightweight Docker management UI built with Tauri and React. Focus on simplicity and performance for Docker container management.
- Introduction | Apache Fory™
- GitHub - apache/fory: A blazingly fast multi-language serialization framework powered by JIT and zero-copy.
- GitHub - shipshapecode/shepherd: Guide your users through a tour of your app
- GitHub - MemoriLabs/Memori: SQL Native Memory Layer for LLMs, AI Agents & Multi-Agent Systems
- GitHub - davila7/claude-code-templates: CLI tool for configuring and monitoring Claude Code
- Go 1.26 は、軽くなり速くなる
- GitHub - Katakate/k7: Your own self-hosted infra for lightweight VM sandboxes to safely execute untrusted code. CLI, API, Python SDK. ⭐ Star it if you like it! ⭐
- UpScale your XGBoost Model Training for Large Datasets with one simple change: QuantileDMatrix
- XGBoost で大規模データセットを扱う際に、
DMatrixの代わりにQuantileDMatrixを使うとメモリ使用量を削減できる。
- XGBoost で大規模データセットを扱う際に、
- http://github.com/syrupy-project/syrupy
- pytest でスナップショットテスト syrupy
- GitHub - muratcankoylan/Agent-Skills-for-Context-Engineering: A comprehensive collection of Agent Skills for context engineering, multi-agent architectures, and production agent systems. Use when building, optimizing, or debugging agent systems that require effective context management.
- 論理削除という技術的負債、それでも僕たちは使い続ける - じゃあ、おうちで学べる
- 対応策
- 削除されたデータを別テーブルに移動
- 履歴テーブル
- Event Sourcing
- 対応策
- LLM サービングライブラリの LMDeploy と vLLM を比較した - Taste of Tech Topics
- GitHub - InternLM/lmdeploy: LMDeploy is a toolkit for compressing, deploying, and serving LLMs.
- Groq and Nvidia Enter Non-Exclusive Inference Technology Licensing Agreement to Accelerate AI Inference at Global Scale | Groq is fast, low cost inference.
- Docker で開発が遅くなる原因チェックリスト Windows 編 #開発環境
- Starrocks, a Database too Fast for its Own Good
- GitHub - StarRocks/starrocks: The world's fastest open query engine for sub-second analytics both on and off the data lakehouse. With the flexibility to support nearly any scenario, StarRocks provides best-in-class performance for multi-dimensional analytics, real-time analytics, and ad-hoc queries. A Linux Foundation project.
- starrocks benchmark
- 10 Rich/Loguru Patterns for Beautiful, Useful Logs
- fluxzy CLI is 30x to 70x faster than mitmproxy / mitmdump, 4x faster than Squid
- Claude Code アドベントカレンダー: 24 Tips まとめ
- GitHub - ruvnet/claude-flow: 🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, distributed swarm intelligence, RAG integration, and native Claude Code support via MCP protocol. Ranked #1 in agent-based frameworks.
- vLLM V1 の実装 ① : EngineCoreClient
- vLLM V1 の実装 ② : EngineCore
- vLLM V1 の実装 ③ : MultiprocExecutor と Worker プロセスの協調動作
- vLLM V1 の実装 ④ : Scheduler
- vLLM V1 の実装 ⑤ : KVConnector
- GitHub - azukds/tubular: Python package implementing ML feature engineering and pre-processing for polars or pandas dataframes.
- Codon: Python That Runs 100x Faster Without Changing Your Code
- 松尾研 LLM 講座 2025 応用編 Day3「軽量化」 講義資料 - Speaker Deck
- GitHub - AshBuk/speak-to-ai: Speak to AI • Native Linux Speech-to-Text (STT) • Offline, Privacy-Focused
- GitHub - parcadei/Continuous-Claude: Context management for Claude Code. Hooks maintain state via ledgers and handoffs. MCP execution without context pollution. Agent orchestration with isolated context windows.
- GitHub - topoteretes/awesome-ai-memory: A list of AI memory projects
- GitHub - onebirdrocks/ebook-mcp: A MCP server that supports mainstream eBook formats including EPUB, PDF and more. Simplify your eBook user experience with LLM.
- GitHub - zerebos/ghostty-config: A beautiful config generator for Ghostty terminal.
