BestHub
Sign in
Tagged articles
33 articles
Page 1 of 1
Old Zhang's AI Learning
Old Zhang's AI Learning
Jun 1, 2026 · Artificial Intelligence

Opus‑Distilled Qwen3.5‑Coder Scores 100/100 Tool Calls, 1.4‑2.2× Faster with MTP, 128K Context on Consumer GPU

The article introduces Qwopus3.5‑4B‑Coder‑MTP‑GGUF, a 4‑billion‑parameter agent model fine‑tuned for code debugging, tool calling, and structured reasoning, explains its novel Trace Inversion, high‑quality trajectory data, and Curriculum SFT training, details MTP acceleration, benchmark results, quantization options, and step‑by‑step local deployment instructions.

AgentGGUFMTP
0 likes · 10 min read
Opus‑Distilled Qwen3.5‑Coder Scores 100/100 Tool Calls, 1.4‑2.2× Faster with MTP, 128K Context on Consumer GPU
Data STUDIO
Data STUDIO
Jun 1, 2026 · Artificial Intelligence

Build a CLI AI Agent in Just 250 Python Lines

This tutorial walks through seven incremental stages—starting with a simple while‑True loop and adding tool‑calling, dynamic skill loading, slash commands, JSON persistence, automatic context compression, and a background timed loop—to create a fully functional CLI AI Agent using Ollama and the local qwen3.5 model without GPU or API keys.

AI AgentCLIOllama
0 likes · 15 min read
Build a CLI AI Agent in Just 250 Python Lines
Old Zhang's AI Learning
Old Zhang's AI Learning
Apr 19, 2026 · Artificial Intelligence

From Zero to Deployment: A Complete Qwen3.5 Fine‑Tuning Guide

This guide shows how to fine‑tune Qwen3.5 models—from 0.8B to 122B—using Unsloth Studio or pure code, covering text SFT, vision fine‑tuning, MoE models, reinforcement‑learning (GRPO), extensive GGUF quantization benchmarks, hardware requirements, export formats, and deployment tips.

LLMUnslothVision-Language
0 likes · 12 min read
From Zero to Deployment: A Complete Qwen3.5 Fine‑Tuning Guide
Old Zhang's AI Learning
Old Zhang's AI Learning
Apr 14, 2026 · Artificial Intelligence

Qwen3.5-27B-DFlash Delivers Up to 5× Faster Inference Without Quality Loss

The DFlash approach replaces speculative decoding’s autoregressive drafter with a block diffusion model and injects target‑model hidden features into every KV‑cache layer, achieving up to 5× speed‑up for Qwen3.5‑27B on single‑GPU and 1.5–1.9× on high‑concurrency workloads while preserving output quality.

DFlashSGLangdiffusion model
0 likes · 12 min read
Qwen3.5-27B-DFlash Delivers Up to 5× Faster Inference Without Quality Loss
Old Zhang's AI Learning
Old Zhang's AI Learning
Apr 10, 2026 · Artificial Intelligence

How a 9B‑parameter Qwen3.5 model achieves full‑auto data analysis on a consumer GPU

The open‑source CoPaw‑Flash‑9B‑DataAnalyst‑LoRA model, fine‑tuned via LoRA, can autonomously load, explore, statistically analyze, visualize, and generate structured reports for CSV/Excel/JSON datasets, achieving a 90% success rate with an average of 26 iteration rounds, and it runs on a single consumer‑grade GPU using vLLM and the Data Analyst framework.

AgentData AnalystGPU
0 likes · 10 min read
How a 9B‑parameter Qwen3.5 model achieves full‑auto data analysis on a consumer GPU
Node.js Tech Stack
Node.js Tech Stack
Apr 6, 2026 · Artificial Intelligence

Run Full AI Models Directly in the Browser with Transformers.js v4

Transformers.js v4 rewrites its WebGPU runtime in C++ and compiles to WASM, delivering ten‑fold faster build times, 10% smaller bundles, and up to four‑fold speedups for BERT‑style models, while supporting over 20 new architectures such as Qwen3.5 and enabling offline, privacy‑preserving AI inference directly in the browser.

Transformers.jsWasmWebGPU
0 likes · 8 min read
Run Full AI Models Directly in the Browser with Transformers.js v4
Old Zhang's AI Learning
Old Zhang's AI Learning
Apr 3, 2026 · Artificial Intelligence

Qwopus3.5‑v3: From Reason‑Then‑Act to Act‑Then‑Refine – Claude‑Opus Distillation Turns Qwen3.5 into a Tool‑Using Agent

The newly released Qwopus3.5‑v3 model combines higher‑quality reasoning chains, dedicated tool‑calling reinforcement learning, and an act‑then‑refine paradigm, delivering a 5‑point HumanEval boost, a 1.43‑point MMLU‑Pro gain, 31.7% faster inference and 24% lower token cost, while remaining runnable on a 3090 or a 16 GB MacBook, with easy deployment via GGUF, LM Studio, Ollama or llama.cpp.

Claude OpusDistillationHumanEval
0 likes · 12 min read
Qwopus3.5‑v3: From Reason‑Then‑Act to Act‑Then‑Refine – Claude‑Opus Distillation Turns Qwen3.5 into a Tool‑Using Agent
AI Engineering
AI Engineering
Apr 1, 2026 · Artificial Intelligence

Holo3 AI Model Beats GPT‑5.4 at One‑Tenth the Cost for Computer Use

H Company’s new Holo3 series delivers a visual language model that outperforms GPT‑5.4 on the OSWorld‑Verified benchmark with a 78.85% score while costing only about one‑tenth as much, offering both a flagship API‑only version and an open‑source lightweight variant optimized for GUI agents.

AI benchmarkGUI AgentHolo3
0 likes · 4 min read
Holo3 AI Model Beats GPT‑5.4 at One‑Tenth the Cost for Computer Use
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 31, 2026 · Artificial Intelligence

Turning a Bluetooth Speaker into a Smart Assistant with Qwen 3.5‑Omni

The author demonstrates a proof‑of‑concept that combines Qwen 3.5‑Omni's real‑time internet search and audio output with a locally hosted voice‑wake‑up model to transform a Bluetooth speaker into an always‑on smart assistant, while noting latency challenges and the potential of a sub‑10B open‑source alternative.

AI integrationBluetoothlarge language model
0 likes · 2 min read
Turning a Bluetooth Speaker into a Smart Assistant with Qwen 3.5‑Omni
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 28, 2026 · Artificial Intelligence

Qwen3.5-27B Outperforms the 397B Model in Tool Calling – Q6 Quantization Is Optimal

Using the open‑source ToolCall‑15 benchmark, the author shows that the 27‑billion‑parameter Qwen3.5 model consistently scores full marks while the 397‑billion‑parameter version fails on several tasks, and that the Q6 quantized variant offers the best trade‑off between size and tool‑calling accuracy.

AILLM BenchmarkModel Evaluation
0 likes · 7 min read
Qwen3.5-27B Outperforms the 397B Model in Tool Calling – Q6 Quantization Is Optimal
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 19, 2026 · Artificial Intelligence

Testing the Hot oMLX on Mac: Claude‑Opus‑4.6 Distilled and Qwen3.5‑9B Performance Review

The article evaluates oMLX, a Mac‑only LLM runtime built on Apple Silicon and MLX, by walking through installation, UI features, memory usage, single‑request speed, benchmark results for Claude‑Opus‑4.6 and Qwen3.5‑9B, continuous batch processing gains, Claude Code optimizations, multi‑model support, and the failure to run a 27B model.

Apple SiliconClaude OpusMLX
0 likes · 9 min read
Testing the Hot oMLX on Mac: Claude‑Opus‑4.6 Distilled and Qwen3.5‑9B Performance Review
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 18, 2026 · Artificial Intelligence

Running Claude‑Opus‑4.6‑Distilled Qwen3.5 27B on a Single RTX 4090 with llama.cpp: 46 tokens/s Performance

The article details a hands‑on test of the Claude‑Opus‑4.6‑distilled Qwen3.5 27B model running on a single RTX 4090 via llama.cpp, showing a steady 46 tokens per second generation speed, a 64K context window, and a step‑by‑step Docker‑based setup while comparing it to GLM‑4.7‑Flash‑AWQ‑4bit and discussing llama.cpp’s limitations for multi‑GPU inference.

Claude OpusDockerLLM inference
0 likes · 5 min read
Running Claude‑Opus‑4.6‑Distilled Qwen3.5 27B on a Single RTX 4090 with llama.cpp: 46 tokens/s Performance
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 16, 2026 · Artificial Intelligence

Testing Claude‑Opus‑4.6 Distilled Qwen3.5 9B Model Locally via LM Studio and Claude Code

The article evaluates the GGUF‑quantized Claude‑Opus‑4.6 distilled Qwen3.5 9B model on a 16 GB Mac Mini M4 using LM Studio, detailing model sizes, performance metrics, deployment steps, API integration with Claude Code, and concluding that while the 9B version is usable, its capabilities remain limited compared to larger models.

Claude OpusGGUFLM Studio
0 likes · 12 min read
Testing Claude‑Opus‑4.6 Distilled Qwen3.5 9B Model Locally via LM Studio and Claude Code
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 7, 2026 · Artificial Intelligence

vLLM 0.17.0 Release: Full Qwen 3.5 Support and Anthropic API Compatibility

The vLLM 0.17.0 release brings FlashAttention 4 integration, a mature Model Runner V2, complete Qwen 3.5 series support, a one‑click performance‑mode flag, Anthropic API compatibility, advanced weight‑offloading, broader hardware support beyond NVIDIA, ASR model integration, and detailed upgrade and installation guidance.

ASRAnthropic APIFlashAttention
0 likes · 12 min read
vLLM 0.17.0 Release: Full Qwen 3.5 Support and Anthropic API Compatibility
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 3, 2026 · Artificial Intelligence

How to Deploy and Fine‑Tune Qwen3.5 Small Models (0.8B‑9B) Locally

This guide walks you through deploying Qwen3.5's 0.8B, 2B, 4B and 9B models on CPUs or modest GPUs using Unsloth's GGUF quantization, explains hardware requirements, shows how to run them with llama.cpp, llama‑server, vLLM or SGLang, and provides a free Colab fine‑tuning workflow with export options.

AI modelsGGUFUnsloth
0 likes · 19 min read
How to Deploy and Fine‑Tune Qwen3.5 Small Models (0.8B‑9B) Locally
AI Engineering
AI Engineering
Mar 3, 2026 · Artificial Intelligence

Alibaba Qwen‑3.5 Small Models: 0.8B Parameters Enable Video on Edge Devices

Alibaba released four Qwen‑3.5 models (0.8B‑9B) that use a Gated DeltaNet hybrid‑attention architecture and native multimodal training to achieve 262k‑token contexts, outperform larger rivals on visual, reasoning, and math benchmarks, and run video analysis on phones and laptops, though they still demand significant VRAM.

Gated DeltaNetbenchmarkedge AI
0 likes · 6 min read
Alibaba Qwen‑3.5 Small Models: 0.8B Parameters Enable Video on Edge Devices
Old Zhang's AI Learning
Old Zhang's AI Learning
Mar 2, 2026 · Artificial Intelligence

Qwen3.5 Small Models Unveiled: From 0.8B to 9B with Full Capabilities

The article introduces the newly released Qwen3.5 small model series (0.8B, 2B, 4B, 9B), explains their shared Gated Delta Networks architecture, early multimodal token fusion, 201‑language support and up to 1 million‑token context, and presents benchmark data that show the 9B model rivaling much larger LLMs, followed by practical guidance on model selection and deployment.

Gated Delta Networksbenchmarklocal deployment
0 likes · 10 min read
Qwen3.5 Small Models Unveiled: From 0.8B to 9B with Full Capabilities
Old Zhang's AI Learning
Old Zhang's AI Learning
Feb 26, 2026 · Artificial Intelligence

How to Disable Thinking Output in Qwen3.5 Models Using LM Studio

This guide explains how to turn off the reasoning (thinking) output of Qwen3.5 series large language models in LM Studio by creating a virtual “-no‑thinking” model directory, editing a model.yaml file, and handling common pitfalls and error messages.

AI model configurationLM Studiodisable thinking
0 likes · 8 min read
How to Disable Thinking Output in Qwen3.5 Models Using LM Studio
Old Zhang's AI Learning
Old Zhang's AI Learning
Feb 26, 2026 · Artificial Intelligence

Ultimate Guide to Local Deployment of Qwen3.5 Models (27B‑397B)

This guide reviews the Qwen3.5 model lineup, explains mixed‑inference and MoE architecture, presents benchmark comparisons with GPT‑5.2, Claude 4.5 and Gemini‑3 Pro, evaluates 4‑bit and 3‑bit quantization loss, outlines hardware requirements, and provides step‑by‑step deployment options using llama.cpp or llama‑server.

MoEQuantizationinference
0 likes · 14 min read
Ultimate Guide to Local Deployment of Qwen3.5 Models (27B‑397B)
Fun with Large Models
Fun with Large Models
Feb 17, 2026 · Artificial Intelligence

Inside Qwen3.5: The World’s Strongest Open‑Source Multimodal Model and Its Core Features

Qwen3.5‑397B‑A17B, the newly open‑sourced multimodal giant, combines a 400‑billion‑parameter sparse MoE architecture with FP8 pipelines and an asynchronous RL framework to deliver GPT‑5.2‑level capabilities, 60% lower memory usage, up to 19× higher throughput, and extensive image, video, and agent support, while outlining its deployment requirements and API pricing.

AI inferenceFP8Sparse MoE
0 likes · 11 min read
Inside Qwen3.5: The World’s Strongest Open‑Source Multimodal Model and Its Core Features
Old Zhang's AI Learning
Old Zhang's AI Learning
Feb 17, 2026 · Artificial Intelligence

Running Qwen3.5 Locally: Step‑by‑Step Guide with Unsloth Dynamic Quantization

This article explains how to run the 397B Qwen3.5 model on a Mac by using Unsloth Dynamic 2.0 quantization (2‑bit, 3‑bit, or 4‑bit), outlines hardware requirements, provides compilation and download commands for llama.cpp, shows how to launch inference in thinking and non‑thinking modes, and compares several deployment options such as llama‑server, Transformers, SGLang/vLLM, and MLX.

Dynamic QuantizationGGUFLLM deployment
0 likes · 14 min read
Running Qwen3.5 Locally: Step‑by‑Step Guide with Unsloth Dynamic Quantization
Alibaba Cloud Big Data AI Platform
Alibaba Cloud Big Data AI Platform
Feb 17, 2026 · Artificial Intelligence

Deploy Alibaba’s Qwen3.5‑397B‑A17B Model in One Click with PAI‑Model Gallery

Alibaba's open‑source Qwen3.5‑397B‑A17B model, featuring 397 billion parameters and a hybrid Gated Delta Network/MoE architecture, delivers superior performance and reduced memory usage, and can be deployed instantly through the PAI‑Model Gallery with step‑by‑step guidance and enterprise‑grade security.

AI inferenceAlibaba CloudOne-click deployment
0 likes · 5 min read
Deploy Alibaba’s Qwen3.5‑397B‑A17B Model in One Click with PAI‑Model Gallery
Machine Learning Algorithms & Natural Language Processing
Machine Learning Algorithms & Natural Language Processing
Feb 16, 2026 · Artificial Intelligence

Alibaba’s Qwen 3.5‑Plus: 397 B Open‑Source Model Beats Gemini‑3 and GPT‑5.2 at Low Cost

Alibaba released the Qwen 3.5‑Plus open‑source large model (397 B total parameters, 170 B active) that outperforms top closed‑source models such as Gemini‑3‑Pro and GPT‑5.2 on multiple benchmarks, offers native multimodal understanding, supports 201 languages, reduces deployment memory by 60 % and inference latency by up to 19×, and is priced at only 0.8 CNY per million tokens.

AIbenchmarklarge language model
0 likes · 15 min read
Alibaba’s Qwen 3.5‑Plus: 397 B Open‑Source Model Beats Gemini‑3 and GPT‑5.2 at Low Cost
Old Zhang's AI Learning
Old Zhang's AI Learning
Feb 16, 2026 · Artificial Intelligence

Qwen3.5 Deep Dive: Multimodal Architecture, Benchmarks, and Deployment Guide

This article provides a detailed analysis of Qwen3.5, covering its multimodal MoE design, massive inference speedups, extensive benchmark results against GPT‑5.2, Claude 4.5 Opus and Gemini‑3 Pro, RL scaling strategies, training infrastructure innovations, and practical usage via API and local deployment.

FP8 trainingbenchmarklarge language model
0 likes · 13 min read
Qwen3.5 Deep Dive: Multimodal Architecture, Benchmarks, and Deployment Guide
AI Engineering
AI Engineering
Feb 16, 2026 · Artificial Intelligence

Qwen3.5-397B: 397B‑Parameter Multimodal LLM Boosts Inference Speed 8‑19×

Alibaba’s Qwen3.5-397B-A17B, a 397‑billion‑parameter open‑source multimodal LLM, combines mixed linear attention with a sparse MoE architecture to achieve 8.6‑19× higher decoding throughput than Qwen3‑Max, supports 201 languages, and can be deployed via vLLM, Docker, Transformers, or SGLang with various optimization presets.

Sparse MoEinference optimizationlarge language model
0 likes · 8 min read
Qwen3.5-397B: 397B‑Parameter Multimodal LLM Boosts Inference Speed 8‑19×