SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World
Back to feed
Programming

Alibaba to Release Qwen3.8 as Open-Weight AI Model

0
·1 views

Alibaba's Qwen team has announced that Qwen3.8, a new AI language model, is launching and will soon be released as an open-weight model. The announcement was made via the official Alibaba Qwen Twitter account. Open-weight models allow developers and researchers to access and use the model's parameters freely. The release is expected to expand accessibility to Alibaba's latest AI capabilities for the broader developer community.

Read the full story at Hacker News

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Developer builds open-source, ad-free, self-hosted music streaming app for Android

A developer has released Starl, a free and open-source music streaming client for Android, after growing frustrated with feature-heavy commercial music apps. The app is designed to work offline, run in the background, and give users full control over their listening data and recommendations. Starl supports content imports from other platforms, customizable mixes, and can connect to either a public server or a self-hosted instance. It carries no ads and tracks nothing, with all user statistics remaining exportable or deletable. The project is available on GitHub under the developer's custom 'Pure Freedom Forever' license, with active development ongoing.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Engineering Team Cuts RAG Pipeline Latency 40% Using Bayesian Search and Hybrid Retrieval

A development team overhauled their Retrieval-Augmented Generation (RAG) pipeline after standard fixed-token chunking and basic vector search proved inadequate in production environments. The team replaced one-size-fits-all 512-token chunking with document-specific strategies, including recursive, semantic, and agentic chunking, achieving recall@10 scores between 91% and 97% across different content types. They also replaced pure vector search with a hybrid retrieval system combining vector search, BM25, and cross-encoder reranking, which improved relevance correlation from roughly 0.75 to 0.92. A Bayesian-informed search approach and parallel retrieval architecture helped reduce overall query latency by 40%, bringing response times down from over one second. The team published their methodology and code on DEV Community to help other engineers build more reliable, tunable RAG systems at scale.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How One Team Built Automated LLM Evaluation That Catches 92% of Hallucinations

A development team discovered critical gaps in their AI quality process after deploying a RAG-based customer support assistant that served hallucinated responses to over 500 users before detection. The failures included fabricated billing policies and rate-limit figures pulled from a competitor's documentation, exposing the team's reliance on informal, manual testing. In response, they built a production-grade evaluation pipeline integrating a suite of automated judges covering faithfulness, instruction-following, JSON schema validation, and domain accuracy. The system is designed to run within CI/CD workflows, blocking code merges that degrade response quality and enabling rapid regression detection. The team published their architecture and code patterns to help other engineers move from subjective review toward measurable, automated LLM quality assurance.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer Builds Kubernetes RBAC Security CLI Tool While Learning Go From Scratch

A Siemens infrastructure engineer with no prior Go experience built kube-radar, an open-source command-line tool designed to detect dangerous RBAC wildcard permission violations in Kubernetes clusters. The developer manages production-grade Kubernetes infrastructure at Siemens, where a misconfigured RoleBinding can expose critical systems to full cluster compromise. Wildcard RBAC rules granting unrestricted access to all resources and verbs are a known security risk, yet they often evade standard cloud-native security scanners because they are technically valid configurations. To build the tool, the developer taught themselves Go through trial and error, working through challenges including SQLite integration, Go's error-as-values model, and CLI architecture using the Cobra framework. The finished tool supports cross-platform builds, SARIF reporting for GitHub Advanced Security, CI/CD exit-code gating, and snapshot-based diff tracking of RBAC findings over time.

0 comments Read more at DEV Community
Alibaba to Release Qwen3.8 as Open-Weight AI Model · ShortSingh