xeno
GitHub

xeno

@xenodeve

View GitHub Profile

About

Dev (Developer) - T4 Labs | C#, HTML, CSS, Python, JavaScript, TypeScript, Lua

Location Nakhon Si Thammarat, Thailand
Company MM Team, @Slow-Inc, @KMUTNB
Followers 16
Following 35
Public Repositories 41
Expertise

Skills & Technologies

C++ PowerShell Python Shell
GitHub Work

Projects & Repositories

Recent public projects and repositories from this profile.

Strata-xeno

C++

fork of Niko1221/Strata: Qwen3.8-Flash-Next 177B on RTX 5060 Ti x4 + RTX 4070 SUPER x16 + i5-13500 + 48 GB. Tracking: xenodeve/Qwen3.8-Flash-Next-Tuning

⭐ 3 Forks: 0

Qwen3.8-Flash-Next-Tuning

PowerShell

Qwen3.8-Flash-Next (MoE) 177B on one desktop: RTX 5060 Ti 16GB + RTX 4070 SUPER 12GB + i5-13500 + 48GB DDR5. Engine work in xenodeve/strata-xeno; measurements and decisions here.

⭐ 1 Forks: 0
PowerShell View Project

exllamav3-xeno

Reference only — Flash-Next work moved to xenodeve/Strata-xeno (2026-09-27). ExLlamaV3 fork: three-tier remote expert slice, router-stats placement, AVX-VNNI mul1 tier; see branch xeno/phase-b-remote-slice and issues #1-#6.

⭐ 1 Forks: 0
Repository View Project

openclink

Python

MCP ที่สามารถเชื่อมต่อ CLI ต่างๆเข้าด้วยกันเพื่อประสิทธิภาพสูงสุดของ Workflow [Claude Code, Codex, Antigravity]

⭐ 1 Forks: 0

xeno-skills

Shell

Multi-agent manual-consensus brainstorm skill for PAL clink (Claude Code / compatible agent runtimes)

⭐ 2 Forks: 0

Qwen-3.8-27B-Tuning

Python

Qwen3.8-27B on two consumer cards in one desktop - RTX 5060 Ti 16GB (Blackwell) + RTX 4070 SUPER 12GB (Ada), tensor-split. Residency, speculative decoding, context to 262k

⭐ 2 Forks: 0