Instructions to use microsoft/bitnet-embedding-270m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/bitnet-embedding-270m with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("microsoft/bitnet-embedding-270m", dtype="auto")

llama-cpp-python

How to use microsoft/bitnet-embedding-270m with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="microsoft/bitnet-embedding-270m",
	filename="bitnet-embeddings-270m-bf16-i2_s.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use microsoft/bitnet-embedding-270m 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 microsoft/bitnet-embedding-270m:BF16
# Run inference directly in the terminal:
llama cli -hf microsoft/bitnet-embedding-270m:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf microsoft/bitnet-embedding-270m:BF16
# Run inference directly in the terminal:
llama cli -hf microsoft/bitnet-embedding-270m:BF16

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 microsoft/bitnet-embedding-270m:BF16
# Run inference directly in the terminal:
./llama-cli -hf microsoft/bitnet-embedding-270m:BF16

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 microsoft/bitnet-embedding-270m:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf microsoft/bitnet-embedding-270m:BF16

Use Docker

docker model run hf.co/microsoft/bitnet-embedding-270m:BF16

LM Studio
Jan
Ollama
How to use microsoft/bitnet-embedding-270m with Ollama:
```
ollama run hf.co/microsoft/bitnet-embedding-270m:BF16
```

Unsloth Studio

How to use microsoft/bitnet-embedding-270m 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 microsoft/bitnet-embedding-270m 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 microsoft/bitnet-embedding-270m to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for microsoft/bitnet-embedding-270m to start chatting

Atomic Chat new
Docker Model Runner
How to use microsoft/bitnet-embedding-270m with Docker Model Runner:
```
docker model run hf.co/microsoft/bitnet-embedding-270m:BF16
```

Lemonade

How to use microsoft/bitnet-embedding-270m with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull microsoft/bitnet-embedding-270m:BF16

Run and chat with the model

lemonade run user.bitnet-embedding-270m-BF16

List all available models

lemonade list

frontierai commited on 1 day ago

Commit

5f1c2fd

verified ·

1 Parent(s): 03b7f59

Update paper reference

Browse files

Files changed (1) hide show

README.md +9 -6

README.md CHANGED Viewed

@@ -118,7 +118,7 @@ They achieve competitive performance on public benchmarks while maintaining exce
 ## Model Sources
 - **Repository:** [https://github.com/microsoft/BitNet](https://github.com/microsoft/BitNet)
-- **Paper:** [The Era of 1-bit LLMs: BitNet b1.58 and its Inference Optimization](https://arxiv.org/abs/2402.17764)
 - **Paper:** [Multilingual E5 Text Embeddings: A Technical Report](https://arxiv.org/abs/2402.05672)
 | Model | Weights | Parameters | Embedding Dimension | Max Tokens | MTEB v2 Mean |
@@ -342,11 +342,14 @@ The embedding is then L2-normalized.
 ## Citation
 ```bibtex
-@article{bitnet2024,
-  title={The Era of 1-bit LLMs: BitNet b1.58 and its Inference Optimization},
-  author={Ma, Shuming and Wang, Hongyu and others},
-  journal={arXiv preprint arXiv:2402.17764},
-  year={2024}
 }
 @inproceedings{wang2025bitnet,

 ## Model Sources
 - **Repository:** [https://github.com/microsoft/BitNet](https://github.com/microsoft/BitNet)
+- **Paper:** [BitNet Text Embeddings](https://arxiv.org/abs/2606.25674)
 - **Paper:** [Multilingual E5 Text Embeddings: A Technical Report](https://arxiv.org/abs/2402.05672)
 | Model | Weights | Parameters | Embedding Dimension | Max Tokens | MTEB v2 Mean |
 ## Citation
 ```bibtex
+@misc{li2026bitnettextembeddings,
+  title={BitNet Text Embeddings},
+  author={Zhen Li and Xin Huang and Liang Wang and Nan Yang and Ting Song and Yan Xia and Xun Wu and Shaohan Huang and Huishuai Zhang and Furu Wei and Dongyan Zhao},
+  year={2026},
+  eprint={2606.25674},
+  archivePrefix={arXiv},
+  primaryClass={cs.CL},
+  url={https://arxiv.org/abs/2606.25674}
 }
 @inproceedings{wang2025bitnet,