Instructions to use Entrit/Mistral-7B-v0.3-trit-uniform-d4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Entrit/Mistral-7B-v0.3-trit-uniform-d4 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Entrit/Mistral-7B-v0.3-trit-uniform-d4")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Entrit/Mistral-7B-v0.3-trit-uniform-d4")
model = AutoModelForCausalLM.from_pretrained("Entrit/Mistral-7B-v0.3-trit-uniform-d4")

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Entrit/Mistral-7B-v0.3-trit-uniform-d4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Entrit/Mistral-7B-v0.3-trit-uniform-d4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Entrit/Mistral-7B-v0.3-trit-uniform-d4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Entrit/Mistral-7B-v0.3-trit-uniform-d4

SGLang

How to use Entrit/Mistral-7B-v0.3-trit-uniform-d4 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Entrit/Mistral-7B-v0.3-trit-uniform-d4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Entrit/Mistral-7B-v0.3-trit-uniform-d4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Entrit/Mistral-7B-v0.3-trit-uniform-d4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Entrit/Mistral-7B-v0.3-trit-uniform-d4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Entrit/Mistral-7B-v0.3-trit-uniform-d4 with Docker Model Runner:
```
docker model run hf.co/Entrit/Mistral-7B-v0.3-trit-uniform-d4
```

Mistral-7B-v0.3-trit-uniform-d4

Balanced ternary quantization of mistralai/Mistral-7B-v0.3 at depth d=4 (81 levels per weight, 6.64 bits per weight).

Produced with the codec from "Balanced Ternary Post-Training Quantization for Large Language Models" (Stentzel, 2026). See Entrit/tritllm-codec for the codec source.

Quick load

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("Entrit/Mistral-7B-v0.3-trit-uniform-d4")
tokenizer = AutoTokenizer.from_pretrained("Entrit/Mistral-7B-v0.3-trit-uniform-d4")

The weights are dequantized to FP16 for stock-transformers compatibility. The on-disk size is therefore the same as the FP16 source. The 6.64-bpw figure refers to the information content of the quantized matrices and is what matters for inference on hardware that consumes the packed trit format directly (see Entrit/tritllm-kernel).

Quantization details

Field	Value
Source model	`mistralai/Mistral-7B-v0.3`
Depth	d=4 (81 levels)
Bits per weight	6.64
Group size	16
Scale codebook	27-entry log-spaced (scale_depth=3)
Method	Uniform PTQ
Quantized layers	all 2D linear matrices
Kept FP16	`lm_head`, token embeddings, all `*_norm` layers
Codec	tritllm v2

Citation

@article{stentzel2026ternaryptq,
  title  = {Balanced Ternary Post-Training Quantization for Large Language Models},
  author = {Stentzel, Eric},
  year   = 2026,
  note   = {Entrit Systems}
}

Reproducibility

git clone https://huggingface.co/Entrit/tritllm-codec
cd tritllm-codec
python quantize_model_v2.py --model mistralai/Mistral-7B-v0.3 --configs uniform-d4 --out ./out

Downloads last month: 152

Safetensors

Model size

7B params

Tensor type

F16

Model tree for Entrit/Mistral-7B-v0.3-trit-uniform-d4

Base model

mistralai/Mistral-7B-v0.3

Finetuned

(394)

this model