📰 MarkTechPost

In this tutorial, we build a stable workflow around the Fable 5 Traces dataset from Hugging Face. We avoid fragile dependencies and manually parse the merged JS

Liquid AI released LFM2.5-230M, its smallest model yet. The 230M-parameter, open-weight model runs on-device at 213 tok/s on a Galaxy S25 Ultra and 42 on a Rasp

DeepSeek open-sourced DSpark, a speculative decoding framework that attaches a draft module to existing DeepSeek-V4 weights. It pairs a parallel draft backbone

In this tutorial, we work with NVIDIA's Open-SWE-Traces dataset to study agentic software-engineering trajectories for fine-tuning. We stream the data directly

OpenAI's GPT-5.6 family adds tiered models with max and ultra reasoning. Here is what early-level engineers should know. The post OpenAI Previews GPT-5.6 With S

DeepReinforce released Ornith-1.0, an open-source coding model family built on Gemma 4 and Qwen 3.5. Instead of a fixed harness, the model learns its own scaffo

Gradium released two real-time speech translation models, stt-translate and s2s-translate, covering English, French, German, Spanish, and Portuguese across 20 l

Mistral AI released OCR 4 on June 23, 2026, moving from clean text extraction to structured document output. Each block returns a bounding box, a typed classifi

In this tutorial, we build a multilingual ASR and speech translation pipeline with NVIDIA Canary-1B-v2. We load the model on a GPU-enabled runtime, prepare audi

Prime Intellect has released prime-rl 0.6.0, an open framework for asynchronous reinforcement learning on trillion-parameter Mixture-of-Experts models. It train

We build a practical GLM-5.2 workflow using its hosted, OpenAI-compatible API instead of running the model locally. We set up multiple providers, load the API k

Fugu and Fugu Ultra route tasks across a swappable model pool, leading most coding, reasoning, and agentic benchmarks. The post Sakana AI Launches Sakana Fugu:

LLMs are stateless by default. Agent memory fixes that. This guide breaks down all 7 types — working, semantic, episodic, procedural, retrieval, parametric, and

Cisco Foundation AI has open-sourced FAPO (Fully Automated Prompt Optimization), a Claude Code-driven system that autonomously optimizes multi-step LLM pipeline

VibeThinker-3B, a 3B MIT-licensed reasoning model matching DeepSeek V3.2 and Kimi K2.5 on verifiable benchmarks. The post VibeThinker-3B: A 3B Dense Reasoning M

We implement an end-to-end workflow for Salesforce CodeGen, loaded from Hugging Face. We move past basic inference by adding function extraction, syntax checkin

The KV cache now outweighs model weights at long context. Here's how TurboQuant, OSCAR, and EpiCache each attack that memory bottleneck — and why they're more c

OpenAI's LifeSciBench evaluates whether frontier AI can handle real life-science research across 750 expert-authored tasks, seven workflows, and seven biologica

MiniMax released MSA, a sparse attention built on Grouped Query Attention. A lightweight Index Branch selects Top-k key-value blocks per query and GQA group; th

We implement xFormers, a practical toolkit for fast, memory-efficient Transformer models on GPUs. We validate memory-efficient attention against a standard impl

We break down Qwen-RobotSuite, the Qwen team's three new embodied AI models. We cover RobotManip, a Vision-Language-Action model built on Qwen3.5-4B for manipul

Z.ai launched GLM-5.2 on June 13, 2026, across every GLM Coding Plan tier. The headline is a usable 1-million-token context window plus High and Max effort leve

We look at Gemini-SQL2, the text-to-SQL capability Google Research announced on June 12, 2026. Powered by Gemini 3.1 Pro, it posted 80.04% execution accuracy on

Zyphra has released Zamba2-VL, a family of open vision-language models at 1.2B, 2.7B, and 7B parameters. The models use a hybrid Mamba2 state-space and Transfor