Instructions to use england-lobster/zhang-xuefeng-27b-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use england-lobster/zhang-xuefeng-27b-gguf with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M # Run inference directly in the terminal: llama cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M # Run inference directly in the terminal: llama cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Use Docker
docker model run hf.co/england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use england-lobster/zhang-xuefeng-27b-gguf with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "england-lobster/zhang-xuefeng-27b-gguf" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "england-lobster/zhang-xuefeng-27b-gguf", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
- Ollama
How to use england-lobster/zhang-xuefeng-27b-gguf with Ollama:
ollama run hf.co/england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
- Unsloth Studio
How to use england-lobster/zhang-xuefeng-27b-gguf with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for england-lobster/zhang-xuefeng-27b-gguf to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for england-lobster/zhang-xuefeng-27b-gguf to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for england-lobster/zhang-xuefeng-27b-gguf to start chatting
- Pi
How to use england-lobster/zhang-xuefeng-27b-gguf with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- OpenClaw new
How to use england-lobster/zhang-xuefeng-27b-gguf with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
- Docker Model Runner
How to use england-lobster/zhang-xuefeng-27b-gguf with Docker Model Runner:
docker model run hf.co/england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
- Lemonade
How to use england-lobster/zhang-xuefeng-27b-gguf with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Run and chat with the model
lemonade run user.zhang-xuefeng-27b-gguf-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use england-lobster/zhang-xuefeng-27b-gguf with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M
Run Hermes
hermes
- Atomic Chat
Install from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M# Run inference directly in the terminal:
llama cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_MUse pre-built binary
# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M# Run inference directly in the terminal:
./llama-cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_MBuild from source code
git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M# Run inference directly in the terminal:
./build/bin/llama-cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_MUse Docker
docker model run hf.co/england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M张雪峰风格 27B(Qwen3.6-27B LoRA · GGUF Q4_K_M)
在 Qwen3.6-27B 上做 LoRA 风格微调、合并权重后量化为 GGUF Q4_K_M 的模型, 用 llama.cpp / Ollama / LM Studio 等可直接本地运行。训练语料见 england-lobster/zhang-xuefeng-data。
这是风格/人设模型,不是事实问答模型。 目标是"说话像张雪峰"(口语节奏、口头禅、直播语气), 作为数字人直播底座;事实应由下游 RAG/台本注入。模型输出不代表张雪峰本人观点,也不保证事实准确。
⚠️ 使用与合规声明(务必先读)
- 本模型能以张雪峰的语气/人设生成中文文本。底层人设/肖像权益属于本人。
- 仅供学术研究与技术交流。请勿用于:冒充本人、虚假代言/带货、诈骗、误导性宣传,或任何可能侵犯 其人格权/造成公众混淆的场景。是否可用于商用请自行评估并取得授权,因使用本模型产生的责任由使用者承担。
- 输出可能包含 ASR/训练引入的错误信息,不应作为升学/报考的事实依据。
- 如权利人提出异议,请联系仓库维护者下线。
文件
| 文件 | 量化 | 大小 | 说明 |
|---|---|---|---|
qwen36_27b_style_lora_merged.Q4_K_M.gguf |
Q4_K_M | ~16.8 GB | 合并权重的 4-bit K-quant,质量/体积平衡,主流选择 |
模型规格
- 基座:Qwen/Qwen3.6-27B(27B,65 层,hidden 5120,GQA 24/4 头,上下文 262144)
- 微调:LoRA(r=32, α=64, lr 8e-5, 1 epoch, max_len 4096;仅对最后一个 assistant turn 计 loss)
- 量化:GGUF v3,Q4_K_M(
general.file_type=15) - 基座许可:Apache-2.0(见
license_link)
快速使用
llama.cpp
# 下载
huggingface-cli download england-lobster/zhang-xuefeng-27b-gguf \
qwen36_27b_style_lora_merged.Q4_K_M.gguf --local-dir .
# 运行(Qwen3 chat 模板,llama.cpp 内置)
./llama-cli -m qwen36_27b_style_lora_merged.Q4_K_M.gguf \
-p "孩子分数刚过一本线,想报计算机,但担心就业,老师怎么看?" \
-c 8192 -ngl 99
Ollama
# Modelfile
printf 'FROM ./qwen36_27b_style_lora_merged.Q4_K_M.gguf\nPARAMETER temperature 0.8\nSYSTEM 你是张雪峰,按张雪峰本人风格回答升学规划、专业选择、就业和考研相关问题。\n' > Modelfile
ollama create zxf-27b -f Modelfile
ollama run zxf-27b "新高考选科,物化生和物化地怎么选?"
建议 system prompt:你是张雪峰,按张雪峰本人风格回答升学规划、专业选择、就业和考研相关问题。
(与训练时一致);采样 temperature≈0.7–0.9 能更好还原直播口语节奏。
已知局限
- 会保留直播带货语气(训练集约 57% 为带货语域):纯问答场景可能出现"报链接/催单"式表达。
- 事实不可靠:ASR 转写 + 风格优先,会有同音/事实错误,请勿据此做报考决策。
- 口语 ≠ 规整:口头禅、重复、中英夹杂是有意保留的风格特征,不是 bug。
训练数据
england-lobster/zhang-xuefeng-data (5411 条,merged_turn 视图,按直播场次切分无泄漏)。
引用
@misc{zhang_xuefeng_style_27b_gguf,
title = {张雪峰风格 27B (Qwen3.6-27B LoRA, GGUF Q4_K_M)},
author = {england-lobster},
year = {2026},
howpublished = {\url{https://huggingface.co/england-lobster/zhang-xuefeng-27b-gguf}}
}
- Downloads last month
- 310
4-bit
Model tree for england-lobster/zhang-xuefeng-27b-gguf
Base model
Qwen/Qwen3.6-27B
Install (macOS, Linux)
# Start a local OpenAI-compatible server with a web UI: llama serve -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M# Run inference directly in the terminal: llama cli -hf england-lobster/zhang-xuefeng-27b-gguf:Q4_K_M