Instructions to use MiniLLM/MiniPLM-llama3.1-212M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MiniLLM/MiniPLM-llama3.1-212M with Transformers:

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

pipe = pipeline("text-generation", model="MiniLLM/MiniPLM-llama3.1-212M")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MiniLLM/MiniPLM-llama3.1-212M")
model = AutoModelForCausalLM.from_pretrained("MiniLLM/MiniPLM-llama3.1-212M")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use MiniLLM/MiniPLM-llama3.1-212M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MiniLLM/MiniPLM-llama3.1-212M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MiniLLM/MiniPLM-llama3.1-212M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MiniLLM/MiniPLM-llama3.1-212M

SGLang

How to use MiniLLM/MiniPLM-llama3.1-212M 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 "MiniLLM/MiniPLM-llama3.1-212M" \
    --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": "MiniLLM/MiniPLM-llama3.1-212M",
		"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 "MiniLLM/MiniPLM-llama3.1-212M" \
        --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": "MiniLLM/MiniPLM-llama3.1-212M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MiniLLM/MiniPLM-llama3.1-212M with Docker Model Runner:
```
docker model run hf.co/MiniLLM/MiniPLM-llama3.1-212M
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

MiniPLM-llama3.1-212M

paper | code

MiniPLM-llama3.1-212M is a 212M model with the LLaMA3.1 achitecture pre-trained from scratch on the Pile using the MiniPLM knowledge distillation framework with the offcial Qwen1.5-1.8B as the teacher model. This model shows the flexibility of the MiniPLM framework in conducting knowledge distillation across model families.

We also open-source the pre-training corpus refined by Difference Sampling in MiniPLM for reproducibility.

Evaluation

MiniPLM models achieves better performance given the same computation and scales well across model sizes:

Baseline Models

Conventional Pre-Training

Citation

@article{miniplm,
    title={MiniPLM: Knowledge Distillation for Pre-Training Language Models}, 
    author={Yuxian Gu and Hao Zhou and Fandong Meng and Jie Zhou and Minlie Huang},
    journal={arXiv preprint arXiv:2410.17215},
    year={2024}
}