Tagged articles
Qwen2.5
12 articles · Page 1 of 1
A developer shares the technical journey of building a floating chat assistant for Android that suggests replies using on-device LLMs, detailing accessibility-based screen reading, OCR fallback for WeChat, a local memory system, CPU vs GPU benchmarking surprises, and AI-assisted development with Claude Code.
AI-assisted developmentAccessibilityServiceAndroid
0 likes · 11 min read

This article details integrating local LLMs into HarmonyOS apps, covering use cases like narrative generation and offline privacy, real-world benchmarks on Mate 60 Pro with Qwen2.5-0.5B, architecture using ArkTS and llama.cpp, performance optimizations via O3/LTO/KleidiAI, and key pitfalls like token limits and UI threading.
GGUFHarmonyOSOn-device Inference
0 likes · 15 min read

The paper presents a systematic empirical study that derives a power‑law scaling formula for reinforcement‑learning‑after‑training of large language models, demonstrating accurate inter‑ and intra‑model performance prediction, learning‑efficiency saturation, data‑reuse benefits, and cross‑architecture validity.
Data ReuseLlama-3Model Efficiency
0 likes · 11 min read

TinyLoRA, a Meta‑proposed method that fine‑tunes Qwen2.5‑7B with only 13 trainable parameters (26 bytes), achieves 91% accuracy on GSM8K under reinforcement learning, revealing that ultra‑low‑parameter RL can rival full‑scale supervised fine‑tuning.
GSM8KParameter-Efficient Fine-TuningQwen2.5
0 likes · 7 min read


UbuntuJan 24, 2026 · Artificial Intelligence
This step‑by‑step guide shows how to replace cloud‑based AI coding tools with a private, zero‑cost workflow on Ubuntu by installing Ollama, configuring systemd, adding DeepSeek or Qwen2.5 models, installing Claude, Codex and Gemini CLIs, and routing them through CC Switch.
AI codingClaude CodeDeepSeek
0 likes · 7 min read

This article explains the GRPO reinforcement‑learning algorithm, shows its core idea of internal group competition without a separate evaluator model, and provides a complete, step‑by‑step code walkthrough—including environment setup, dataset preparation, reward‑function design, training configuration, and evaluation—using the Qwen2.5‑0.5B‑Instruct model on the GSM8K math dataset.
GRPOGSM8KQwen2.5
0 likes · 23 min read

This article walks through the full workflow for using Alibaba Cloud's open‑source DistilQwen2.5 models on the PAI platform, covering environment setup, model deployment, fine‑tuning with SFT and DPO, evaluation, and model compression for resource‑constrained scenarios.
DistilQwen2.5PAIQwen2.5
0 likes · 13 min read

This article explains how to replicate Deepseek RI's slow‑thinking inference using the GRPO reinforcement‑learning algorithm on the Qwen2.5‑7B model in a free Colab notebook, covering the underlying COT concept, reward‑function design, data preparation, training configuration, and observed results.
GRPOLLMPython
0 likes · 14 min read

Researchers from Stanford and Washington trained the s1 inference model in just 26 minutes using under $50 of cloud credits, achieving performance comparable to OpenAI's o1 and DeepSeek's R1 by building a curated 1,000‑sample dataset and a budget‑enforced test‑time scaling algorithm.
AI benchmarkingQwen2.5budget enforcement
0 likes · 7 min read

This article compiles and analyzes the post‑training pipelines of Llama 3.1, DeepSeek‑V3, TÜLU 3 and Qwen 2.5, detailing their data compositions, SFT, reward modeling, DPO, GRPO, RLVR methods, hyper‑parameters, and practical tricks for large‑language‑model alignment.
DPODeepSeek-V3Llama3.1
0 likes · 22 min read

This guide explains how to deploy the Qwen2.5 large language model on Alibaba Cloud Function Compute using Ollama and Open WebUI, covering model selection, resource configuration, deployment steps, interface setup, multilingual capabilities, and automatic scaling for high‑concurrency workloads.
AI model deploymentFunction ComputeOllama
0 likes · 10 min read


NewBeeNLPDec 23, 2024 · Artificial Intelligence
The Qwen2.5 Technical Report introduces a new series of large language models with up to 72 B parameters, expanded pre‑training data to 18 trillion tokens, advanced supervised fine‑tuning and reinforcement learning pipelines, and demonstrates strong performance across comprehension, reasoning, coding, and long‑context tasks.
LLMPretrainingQwen2.5
0 likes · 5 min read
