Metriq Update: Open Quantum Benchmarking Platform

Today we are releasing a major update to Metriq, the platform for open, community-driven quantum computer benchmarking. It's been built by Unitary Foundation with a lot of help from the external community of open source contributors. Blogpost: https://lnkd.in/ePiPx7sX This release introduces metriq-gym, a new open-source toolkit for defining and running benchmarks across hardware providers, metriq-data, a public dataset of benchmark results, along with a new Metriq website, where results can be tracked and shared: http://metriq.info/ We invite the quantum community to suggest improvements, extend the benchmark suite, run experiments, and upload new results. As quantum computers evolve over time, the Metriq platform will evolve with them. Check out our new paper describing the platform, and see you on GitHub!

The future of media production is local compute. Professional environments require precise hardware memory management today. Scaling AI infrastructure starts with optimized Python pipelines. Eliminate third party dependencies to ensure data security. This technical approach maximizes return on hardware investment. Lead your industry with high performance local intelligence. Blog: https://lnkd.in/eUU43jcg Video: https://lnkd.in/ezCAcpe6 Books: https://lnkd.in/gqejw-pc Blueprints: https://ojamboshop.com Tutorials: https://ojambo.com/contact Consultations: https://lnkd.in/eWRXWP6E #ArtificialIntelligence #DataSovereignty #SoftwareArchitecture

Tired of complex infrastructure setup? The expanded #EOSCEUNode Tools Hub is your one-stop shop for research software, ready for instant deployment. From data processing to advanced analytics, access powerful tools like Galaxy for biomedical research and DaskHub for parallel computing in Python. ✅ One-Stop Access: A curated catalogue of reliable applications. ✅ For All Skill Levels: Deploy easily as a beginner or use TOSCA templates for advanced, reproducible workflows. ✅ Cost-Covered: Computational resources are provided by the European Commission. Ready to get started? We have everything you need: Watch the Demo Video: See how to allocate a Virtual Machine and set up tools in your User Space. Follow the Tutorial: "Tools Hub: Introduction for Researchers" Take the Course: "How to use the EOSC EU Node Tools Hub: A Complete Guide" 🔗 Explore the Tools Hub: https://lnkd.in/epvSbBEj

Most RAG tutorials oversimplify the problem. They make it look like: LLM + Vector DB = Done. ❌ In reality, the real engineering challenge is retrieval architecture. When I tried scaling a RAG system to 10,000+ documents, I ran into problems I didn’t expect: ▶ Slow retrieval ▶ Large context windows ▶ High token costs ▶ Context retention issues So I built a pipeline using LlamaIndex + FAISS and optimized the retrieval layer. Result: ✅ <100ms vector search ✅ 10K+ documents indexed ✅ Lower token usage ✅ More grounded responses I documented the full architecture and workflow in this PDF. Curious to hear from engineers building production RAG systems: What’s been the hardest bottleneck in your pipeline? 1. Retrieval latency 2. Chunking strategy 3. Vector database scaling ...or something else? #GenAI #RAG #LlamaIndex #LangChain #BackendEngineering #SystemDesign #MachineLearning #OpenSource #VectorDatabase #Python

PyStormTracker v0.5.0.dev0 is available on TestPyPI. 🎯 New Tracking Algorithm: Initial Implementation of the Hodges (TRACK) algorithm. 🧮 Spectral Backends: Established ducc0 as the primary spectral backend, validated against NCL/SPHEREPACK, and experimental GPU-accelerated support via JAX. ☁️ Data I/O Upgrades: Upgraded the data loader with automatic format detection, adding native support for remote Zarr datasets alongside NetCDF and GRIB. 🗺️ Track Visualization: Web-based interactive track explorer with real-time filtering, supported by a new utility for converting JSON track data. #Python #AtmosphericScience #ScientificComputing #OpenSource

Build Custom Image Segmentation Model Using YOLOv8 and SAM For anyone studying image segmentation and the Segment Anything Model (SAM), the following resources explain how to build a custom segmentation model by leveraging the strengths of YOLOv8 and SAM. The tutorial demonstrates how to generate high-quality masks and datasets efficiently, focusing on the practical integration of these two architectures for computer vision tasks. Link to the post for Medium users : https://lnkd.in/dRZMcKyM You can find more computer vision tutorials in my blog page : https://eranfeit.net/blog/ Video explanation: https://lnkd.in/dH5u-5Tp Written explanation with code: https://lnkd.in/dpvriqfs This content is for educational purposes only. Constructive feedback is welcome. Eran Feit #Ultralytics #SegmentAnything #ImageSegmentation

We have been talking about how great the web-based notebooks are… But they do hit real computational limits. Try something simple like computing eigenvalues of a 1000×1000 matrix. You’re looking at seconds of runtime and ~40 MB of memory consumption just for this single experiment. And now imagine stacking multiple experiments on top of that. That’s where things start to break.   Now with the new compute notebooks you can: • Run multiple notebooks in parallel • Handle heavy numerical workloads • Terminate sessions instantly and reclaim memory • Keep your interactive apps separate from computation   In the short video-demo below: • Interactive notebook apps • Eigenvalue computation of a1000x1000 matrix   #Python #Engineering #DataScience #WebAssembly #Simulation #Computing #concurrency

Hugging Face tokenizers now has riscv64 wheel support merged upstream, thanks to BayLibre's work on PR #1951. Once the next release ships, anyone on RISC-V can just pip install tokenizers without building from source. That's one of the most requested ML packages ticked off the list. On the WebAssembly front, two more PRs landed in Ocre Runtime this week (the Cloud Native Computing Foundation (CNCF) edge container project). A wasi-sysroot refactor and a follow-up fix, both tested on a banana pi Open Source Project F3. WebAssembly on RISC-V is getting real CI coverage now. 34 releases went out across my forks. The Docker stack hit v29.3.1 with containerd 2.2.1 and runc 1.4.0. DuckDB v1.5.1 was built on a native RISE (RISC-V Software Ecosystem) riscv64 runner on Scaleway. And Mistral AI Vibe v2.7.0 brings the Mistral AI coding CLI to riscv64 with a new conversation rewind feature. The riscv64 Python wheel index keeps growing: 50+ packages now, including PyTorch 2.10, cryptography 47.0, tiktoken 0.12, pydantic-core, numpy 2.5, and llama-cpp-python. All built natively on a BananaPi F3. OpenSCAD pushed 6 daily builds with .deb and .rpm packages for riscv64, arm64, and amd64. SDKMAN hit a snag: two PRs for riscv64 state support were closed due to an ongoing backend migration. Tracking the issue for when that settles. What's your experience with pip install on RISC-V? Curious how many packages just work for you out of the box. #RISCV #RISCVEverywhere #OpenSource #Python #Docker #EdgeAI #WebAssembly #DockerCaptain #devEco

The Demo Gap Your demo works perfectly. On your laptop. On your GPU. On your data. In your Jupyter notebook. With your Python version. Ship it to production and it falls apart in ways that take engineers days to debug. This is the demo gap. And it exists because the development and deployment environment are two completely different technology stacks held together by Docker and hope. #NeuralEcosystems closes the gap. Same language. Same runtime. Development to production. No translation layer. What works on your laptop works everywhere. https://lnkd.in/d2EqRgzc #NeuralOS #NeuralSCRIPT #NeuralSCRIPT++ #NeuralCPU #NeuralGPU  #NeuralFUSE #NeuralRV #NeuralEDGE #NeuralDB #NeuralPIPE  #NeuralSENSE #NeuralAUTO #NeuralFUZZY #NeuralIP #NeuralSDR  #NeuralMESH #NeuralUI #NeuralZONE #NeuralGAURD #NeuralSHARE #NeuralGHOST #NeuralBIO #NeuralHEALTH #NeuralNAV #NeuralWEB  #UAE #Innovation

Are you even a math student if you don't build a project that is literally just pure, unadulterated linear algebra and probability theory running under the hood? Python is incredible for data ingestion, but when it comes to running 10,000+ Monte Carlo simulations for portfolio risk, the Global Interpreter Lock (GIL) becomes a massive bottleneck. To fix this, I made a Risk Engine from scratch that completely bypasses Python for the heavy lifting. Here is the V1 architecture: The Math: Geometric Brownian Motion (GBM) and Cholesky Decomposition to ensure simulated asset paths respect real-world market covariance. The Core: A native C++ backend executing the matrix transformations at raw machine speed. The Bridge: Pybind11 to safely pass massive 2D Pandas arrays across languages with zero-copy memory transfers. The UI: Python and Streamlit for live market ingestion via yfinance and reactive visualizations. The system dynamically calculates 95% Value at Risk (VaR),99% Value at Risk (VaR) and Expected Shortfall. Moving from standard ctypes to pybind11 to handle the cross-language memory handoffs was a steep learning curve, but the performance leap of running compiled C++ is undeniable. Next up for V2: Injecting OpenMP for multi-core parallel processing and swapping standard PRNGs for Sobol Sequences to speed up mathematical convergence. GitHub Repo in the comments! 👇 #Cpp #Python #LinearAlgebra #RiskManagement #SoftwareEngineering