Instructions to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp 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 AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp 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 AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0 # Run inference directly in the terminal: llama cli -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0 # Run inference directly in the terminal: llama cli -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0 # Run inference directly in the terminal: ./build/bin/llama cli -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Use Docker
docker model run hf.co/AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
- Llama new
- LM Studio
- Jan
- Ollama
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with Ollama:
ollama run hf.co/AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
- Unsloth Desktop
- Pi
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with Docker Model Runner:
docker model run hf.co/AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
- Lemonade
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Run and chat with the model
lemonade run user.MiniCPM5-2B-AX8860-llama.cpp-Q8_0
List all available models
lemonade list
- Hermes Agent
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
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://127.0.0.1:9931/v1 hermes config set model.default AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0
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://127.0.0.1:9931/v1 \ --custom-model-id "AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp:Q8_0" \ --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"
MiniCPM5-2B on AX8860 (llama.cpp + ggml-axera)
ๅจ็ฑ่ฏ AX8860 ็ NPU ไธ็จ llama.cpp ่ฟ่ก MiniCPM5-2B ๆๆฌๆจกๅ๏ผๆ่ๅผๅ ณใๅทฅๅ ท่ฐ็จ๏ผOpenAI ๅ ผๅฎนๆๅกใ MiniCPM5-2B on the Axera AX8860 NPU with llama.cpp: text model with a thinking switch, tool calling and an OpenAI-compatible server.
ๆง่ฝ / Performance
| decode | 59.5 tokens/s |
| 4k prompt prefill | 6219 tokens/s๏ผ3907 token 0.63 s๏ผ |
| ไธไธๆ / context | ๆ้ฟ / up to 131072 |
sh bench.sh ๅค็ฐ / reproduce๏ผ็ๆ 50 token / 50 generated tokens๏ผ๏ผ
| prompt | -c |
prompt ่ๆถ / time | decode tokens/s |
|---|---|---|---|
| 5 token | 8192 | 0.04 s | 59.5 |
| 46 token | 8192 | 0.05 s | 59.2 |
| 3907 token | 8192 | 0.63 s | 51.9 |
| 5 token | 131072 | 0.04 s | 59.2 |
| 3907 token | 131072 | 0.63 s | 52.1 |
ๅฟซ้ๅผๅง / Quick start
sh run.sh "The capital of France is" 64 # ๆๆฌ่กฅๅ
จ / completion
sh run.sh -c # ๅฝไปค่กๅฏน่ฏ / chat
sh run.sh -s 8080 # OpenAI ๅ
ผๅฎนๆๅก / server
sh bench.sh # ๆง่ฝๅค็ฐ / benchmark
sh bench.sh ppl # ๅฐๆๅบฆ / perplexity
้่ฆ AX8860๏ผSDK V0.5.1 ๆๆดๆฐ๏ผHF ไธ่ฝฝ็ๆไปถๆฒกๆๆง่กๆ้๏ผ็จ sh ่ฐ็จๅณๅฏ / needs an AX8860 with SDK V0.5.1+; call the scripts with sh.
ๆๅก่ฏทๆฑ / Server request:
curl http://127.0.0.1:8080/v1/chat/completions -H 'Content-Type: application/json' -d '{
"messages": [{"role": "user", "content": "็จไธๅฅ่ฏไป็ป้ฟๆฑ"}],
"chat_template_kwargs": {"enable_thinking": false}
}'
ๆไปถ / Files
| ๆไปถ / File | ่ฏดๆ / Description |
|---|---|
MiniCPM5-2B-Q8_0.gguf |
ๆจกๅๆ้๏ผๅฎๆน Q8_0๏ผ/ weights (official Q8_0)๏ผ2.68 GB๏ผmd5 57e37e3fd7758a9e4f7c687cf5c0e257 |
bin/ |
llama.cpp ๆฟ็ซฏ็จๅบ๏ผ้ๆ้พๆฅ๏ผ/ board programs (static) |
kernels/ |
NPU kernel 2 ไธช๏ผprefillใdecode๏ผๅซ่พๅบๅคด๏ผ+ manifest.json / two kernels: prefill and decode (with the output head) + manifest.json |
run.sh๏ผchat_template.jinja |
่ฟ่ก่ๆฌไธๆๅก็ซฏๅฏน่ฏๆจกๆฟ / run script and the server's chat template |
bench.sh ๅ prompt |
ๆง่ฝไธๅฐๆๅบฆๅค็ฐ / benchmark and perplexity |
่ฏดๆ / Notes
- ๆ้ๅจๅ ่ฝฝๆถ็ฑ GGUF ่ฝฌๆ MXFP8๏ผKV ไธบ BF16๏ผQ8_0 ไธ F16 GGUF ้ฝๅฏๅ ่ฝฝ / weights are converted from the GGUF to MXFP8 at load, with a BF16 KV cache; both Q8_0 and F16 GGUFs load.
- ไธๅฅ kernel ่ฆ็ๅฐ 131072 ไธไธๆ๏ผKV ๆ
-cๅ้ ๏ผ็จ็ฏๅขๅ้CTX่ฐๆด / one kernel set covers up to 131072 tokens and allocates the KV by-c; setCTXto change it. - ๆ่ๅผๅ
ณ็จ
chat_template_kwargs.enable_thinking๏ผๅทฅๅ ท่ฐ็จไฝฟ็จๅ ๅchat_template.jinja/ toggle thinking withchat_template_kwargs.enable_thinking; tool calling uses the bundledchat_template.jinja. - ๆฒกๆ MTP ๅคด๏ผไธๆฏๆ MTP ๆๆบ่งฃ็ / no MTP head, so no MTP speculative decoding.
run.sh -c้่ฆ็ๅฎ็ป็ซฏ / chat mode needs an interactive terminal.- ่ฟ่กๆถ็ฌๅ NPU 4 ๆ ธ๏ผๆๅก็ซฏ
-np 1/ takes all 4 NPU cores; the server runs-np 1.
ๆญฃ็กฎๆง / Correctness
- ๆๆฌ่กฅๅ จใๆๅก็ซฏๅฏน่ฏไธๆ่ๅผๅ ณๅจๆฟไธๅทฅไฝๆญฃๅธธ / completion, server chat and the thinking switch work on the board.
- PPL๏ผ
sh bench.sh ppl๏ผ-c 512๏ผ/ perplexity:
| ่ฏญ่จ / language | -ub 1 |
-ub 4 |
|---|---|---|
| ไธญๆ / zh | 16.70 | 16.72 |
| ่ฑๆ / en | 13.33 | 13.44 |
ๆบ็ / Sources
- llama.cpp๏ผZHEQIUSHUI/llama.cpp @
1970d6c5d
ๆดๆฐ / Changelog
- 2026-11-15๏ผ้ฆๆฌกๅๅธ / first release.
- Downloads last month
- 20
8-bit
Model tree for AXERA-TECH/MiniCPM5-2B-AX8860-llama.cpp
Base model
openbmb/MiniCPM5-2B