Install from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_M# Run inference directly in the terminal:
llama cli -hf whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_MBuild 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 whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_M# Run inference directly in the terminal:
./build/bin/llama cli -hf whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_MUse Docker
docker model run hf.co/whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_MLycheeAI-coder-1b-II-GGUF
ๅฝๅ็ๆฌ๏ผV4๏ผไปฃ็ ๆ็ฃจ็๏ผ GGUF ้ๅ็ๆฌ ยท ๅบไบ MiniCPM5-1B ๅพฎ่ฐ ยท Apache 2.0
่ฟๆฏไธชไบบๅผๅ่ ็ๅๆฌกๅฐ่ฏ๏ผๆจกๅ่ฝๅๆ้๏ผ1B ๅๆฐๆณจๅฎๅฎๅชๆฏไธช"่ฝป้ๅฐๅฉๆ"ใ ่ฏฆ็ป่ฝๅ่ฏดๆไธๅฑ้๏ผ่ฏท็ ไธปไปๅบใ
๐ฆ ไธไธช็ๆฌ๏ผๆ่ฎพๅค้
| ๆไปถ | ๅคงๅฐ | ๅ ๅญๅ ็จ | ็จๅจๅช | ่ดจ้ |
|---|---|---|---|---|
lychee-coder2-v3-q4_k_m.gguf |
656 MB | ~1 GB | ๐ฑ ่ๆๆบ๏ผ4GB ๅ ๅญ๏ผใๆ ่ๆดพ | โ ๏ธ ๆๆ |
lychee-coder2-v3-q8_0.gguf |
1.1 GB | ~1.5 GB | ๐ฑ ไธญ็ซฏๆๆบใ่ฝป่ๆฌ | โ ๆฅ่ฟๆ ๆ |
mlx-LycheeAI-coder-1b-II.gguf๏ผf16๏ผ |
2.0 GB | ~2.5 GB | ๐ป Mac / PCใ้ซ็ซฏๆๆบ | โ ๆ ๆ |
ๅๅ๏ผ่ฎพๅคๆๅพไฝๅฐฑ็จ f16๏ผ็ฉบ้ด็ดงๅผ ๅๅพไธ้ใ
๐ฑ ๆๆบ่ฝ่ทๅ๏ผ่ฝใ
1B ๆจกๅ็ไผๅฟๅฐฑๆฏๅฐใq4 ็ๅชๆ 656MB๏ผไธๅฐ 4GB ๅ ๅญ็่ๅฎๅๆบๅฐฑ่ฝ่ท๏ผ
้ๅบฆๅคง็บฆๆฏ็งๅๅ ไธชๅญ๏ผๅฎๅ จ็ฆป็บฟใไธ่็ฝใไธ่ฑ API ่ดน็จใ
๐ฅ ๆไนไธ่ฝฝๆๅฎ็ๆฌ
ๆนๆณ 1๏ผๅฝไปค่ก๏ผๆจ่๏ผๅฏๅชไธๅไธชๆไปถ๏ผ
# HuggingFace๏ผ้ๅ
pip install huggingface_hub๏ผ
huggingface-cli download whcl412/LycheeAI-coder-1b-II-GGUF \
lychee-coder2-v3-q4_k_m.gguf --local-dir ./
# ModelScope๏ผๅฝๅ
ๆจ่๏ผ้ๅ
pip install modelscope๏ผ
modelscope download --model whcl412/LycheeAI-coder-1b-II-GGUF \
lychee-coder2-v3-q4_k_m.gguf --local_dir ./
ๆๆไปถๅๆขๆไฝ ๆณ่ฆ็ๅณๅฏ๏ผๅปๆๆไปถๅๅฐฑๆฏไธ่ฝฝๆดไธชไปๅบ๏ผใ
ๆนๆณ 2๏ผ็ด้พไธ่ฝฝ
# HuggingFace
wget https://huggingface.co/whcl412/LycheeAI-coder-1b-II-GGUF/resolve/main/lychee-coder2-v3-q4_k_m.gguf
# ModelScope
wget https://www.modelscope.cn/models/whcl412/LycheeAI-coder-1b-II-GGUF/resolve/master/lychee-coder2-v3-q4_k_m.gguf
ๆนๆณ 3๏ผ็ฝ้กต็นไธ่ฝฝ
ๆๅผไปๅบ้กต้ข็ Files ๆ ็ญพ โ ็นๆไปถๅ โ ๅณไธ่ง Downloadใ
โ ๏ธ ๅ ณไบ้ๅ็ๅฎ่ฏ
1B ๅๆฐๆฌๆฅๅฐฑๅฐ๏ผๅฏน้ๅๆฏๅคงๆจกๅๆๆๅพๅคใๅฎๆต๏ผ
| ้ฎ้ข | f16๏ผๆ ๆ๏ผ | q8_0 | q4_k_m |
|---|---|---|---|
| 1+1=3 ๅฏนๅ | ไธๆฏ๏ผ1+1=2 โ | ็ญๆ 2+2=4 โ ๏ธ | ็ญๆ 2+2=4 โ |
| Python ๆฏ็ผ่ฏๅ | ไธๆฏ๏ผ่งฃ้ๅจๆง่ก โ | ๆๅฐ GIL โ ๏ธ | ่ฏด GIL ๆฏ็ผ่ฏๅจ โ |
้ๅ่ถไฝ๏ผไบๅฎ็ป่่ถๅฎนๆๅบ้๏ผไฝๆฅๅธธ่ๅคฉใๅไปฃ็ ใ็ฎๅ้ฎ็ญไพ็ถๅฏ็จใ ่ฆๅจ่ๆๆบไธ่ท๏ผq4 ๆฏๅฏไธ็ฐๅฎ็้ๆฉโโ่ฝ่ท่ตทๆฅๆฏไปไน้ฝๅผบใ
ๅฟซ้ๅผๅง
Ollama
ๅ
ๅๅค Modelfile๏ผ้่ฆ๏ผTEMPLATE ๅฟ
้กป้ขๅกซ <think>๏ผๆจกๅไพ่ตๅฎ่ฟๅ
ฅๆ่ๆจกๅผ๏ผ๏ผ
FROM ./lychee-coder2-v3-f16.gguf
SYSTEM """ไฝ ๆฏ LycheeAI-coder-1b-II๏ผ็ฑ MiniCPM5-1B ้่ฟ LoRA ๅพฎ่ฐ่ๆฅ็็ผ็จๅฉๆใ"""
TEMPLATE """{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}<|im_start|>user
{{ .Prompt }}<|im_end|>
<|im_start|>assistant
<think>
"""
PARAMETER stop <|im_end|>
PARAMETER stop <|im_start|>
PARAMETER temperature 0.4
PARAMETER num_ctx 4096
PARAMETER num_predict 2048
็ถๅ๏ผ
ollama create lychee-coder2 -f Modelfile
ollama run lychee-coder2
llama.cpp
./llama-cli -m lychee-coder2-v3-f16.gguf \
-p "็จ Python ๅไธชๅคๆญ่ดจๆฐ็ๅฝๆฐ" \
-c 4096 -n 2048 --temp 0.4
ๅ ณไบๆ่ๆจกๅผ
่ฟไธชๆจกๅๅ็ๆฏๆๆ่ๅผๅ
ณ๏ผๅบๅฑๆฏ enable_thinking ๅๆฐ๏ผ้่ฟ chat_template ๆงๅถ๏ผ๏ผ
- ๅผๆ่๏ผ้ป่ฎค๏ผ๏ผ
<think>ๆ ็ญพๅ ๆจ็๏ผ็ถๅ็ป็ญๆกใๆฐๅญฆใ้ป่พ้ข่ฏทๅกๅฟ ๅผๆ่ใ - ๅ ณๆ่๏ผ็ดๆฅ็ป็ญๆก๏ผๆด็ฎๆดใไฝๆฐๅญฆๅ็กฎ็ไผๆๆพไธ้๏ผ่ฟๆฏๆฌ่ดจๆ่กกใ
ๅจ Ollama ้๏ผๆจกๅๅบๅฎ่ตฐๆ่ๆจกๅผ๏ผOllama ไธๆฏๆ enable_thinking ๅๆฐ๏ผใๅฆๆ้่ฆ็ด็ญ๏ผ่ฏท็จ MLX ็ๆฌๅนถไผ enable_thinking=Falseใ
ๅฎๆต่ธฉๅ๏ผOllama ไธๅ"็ด็ญๅไฝ"๏ผ้ขๅกซ็ฉบ think ๆไธ้ขๅกซ๏ผ้ฝๆ ๆโโๆจกๅไปไผ่ชๅทฑ็ๆๆ่ๅ ๅฎน๏ผ่ไธไปฃ็ ่ดจ้ๅ่ๅๅทฎใๆไปฅๅชๆไพๆ่็ใ
็ๆฌๅๅฒ
| ็ๆฌ | ้็น |
|---|---|
| V1 | ๅบ็ก่ฝๅ๏ผไปฃ็ ใๅค่ฏญ่จใๆฅๅธธใๆ ๆใ่บซไปฝ |
| V2 | ่กฅ้ฝ"ๅ ณๆ่ๆถ็ญ้พ้ข"๏ผไฟฎๆๆ่้ไนฑๅคๆญ่ฏญ่จ็ bug |
| V3 | ๅ่ฐๅช๏ผ็บ ๆญฃ้่ฏฏๅๆ๏ผๅๆถ้ฒๆญข็ซๆ่ฟๆญฃ |
| V4 | ไปฃ็ ๆ็ฃจ๏ผPython ่ฏญไน้ท้ฑใrequests ๆๆฐ้้ฟใPydantic v2ใไธ็ผ้ APIใ็ฎๆณ่พน็ใๆฐๅญฆๅบ็จ้ข็่งฃ |
ๆจ่ๅๆฐ
| ้กน็ฎ | ๅผ |
|---|---|
| ไธไธๆ้ฟๅบฆ | 4096 |
| ๆๅคง่พๅบ | 2048 |
| ๆธฉๅบฆ | 0.4๏ผๅไปฃ็ ๅฏ้ๅฐ 0.2~0.3๏ผ |
ๅฎๅไธๅฐไปไน
- ๅคๆ็ฎๆณไผ็ฟป่ฝฆ โโ ่ฏญๆณๅฏ่ฝๅฏน๏ผ้ป่พๅฏ่ฝ้๏ผๅๅฎ่ฏท่ชๅทฑ่ทใ
- ไบๅฎ็ป่ๅถๅฐ่ก่ฏด โโ ๅฎ็ฅ้"่ฟ่ฏไธๅฏน"๏ผไฝๆชๅฟ ่ฏดๅพๆธ ๅชๅฟไธๅฏนใ
- ่ถ ็บฒ้ป่พ้ขไผ็ปๅ โโ ๅคไบบ็ๅ่ฏๆจ็่ฟ็ฑป๏ผๅฏ่ฝๆจไธๅบๆฅใ
- ๅ ณๆ่ๆถๆฐๅญฆไผ็ฎ้ โโ ๆฌ่ดจๆ่กก๏ผไธๆฏ bugใ
่ฎธๅฏไธ่ด่ฐข
- ่ฎธๅฏ๏ผApache 2.0
- ๅบๅบง๏ผMiniCPM5-1B by ้ขๅฃๆบ่ฝ๏ผOpenBMB๏ผ
- ่ฎญ็ป๏ผMLX-LM ๏ฝ ้ๅ๏ผllama.cpp
ๅ ณๆณจไฝ่
๐บ B็ซ๏ผspace.bilibili.com/3493128967293256
ไธๅฎๆๅไบซ AI ๆจกๅ่ฎญ็ปไธๆ่ พ่ฎฐๅฝ๏ผๆฌข่ฟๆฅ็ฉ๏ฝ
- Downloads last month
- 140
Model tree for whcl412/LycheeAI-coder-1b-II-GGUF
Base model
openbmb/MiniCPM5-1B
Install (macOS, Linux)
# Start a local OpenAI-compatible server with a web UI: llama serve -hf whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_M# Run inference directly in the terminal: llama cli -hf whcl412/LycheeAI-coder-1b-II-GGUF:Q4_K_M