Instructions to use aloobun/Pruned-SmolLM2-1.4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aloobun/Pruned-SmolLM2-1.4B with Transformers:

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

pipe = pipeline("text-generation", model="aloobun/Pruned-SmolLM2-1.4B")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("aloobun/Pruned-SmolLM2-1.4B")
model = AutoModelForCausalLM.from_pretrained("aloobun/Pruned-SmolLM2-1.4B")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use aloobun/Pruned-SmolLM2-1.4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aloobun/Pruned-SmolLM2-1.4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aloobun/Pruned-SmolLM2-1.4B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/aloobun/Pruned-SmolLM2-1.4B

SGLang

How to use aloobun/Pruned-SmolLM2-1.4B 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 "aloobun/Pruned-SmolLM2-1.4B" \
    --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": "aloobun/Pruned-SmolLM2-1.4B",
		"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 "aloobun/Pruned-SmolLM2-1.4B" \
        --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": "aloobun/Pruned-SmolLM2-1.4B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use aloobun/Pruned-SmolLM2-1.4B with Docker Model Runner:
```
docker model run hf.co/aloobun/Pruned-SmolLM2-1.4B
```

Pruned SmolLM2-1.7B model - total parameters 1.47B.

Intermediate step, requires further training - try it yourself.

Eval results using SmolLM evaluation scripts (LightEval):

Task	Version	Metric	Value		Stderr
all		acc_norm	0.4555	±	0.0114
		qem	0.0431	±	0.0022
custom:arc:_average:0		acc_norm	0.5021	±	0.0120
custom:arc:challenge:0	0	acc_norm	0.3686	±	0.0141
custom:arc:easy:0	0	acc_norm	0.6355	±	0.0099
custom:commonsense_qa:0	0	acc_norm	0.3333	±	0.0135
custom:gsm8k:5	0	qem	0.0076	±	0.0024
custom:hellaswag:0	0	acc_norm	0.5568	±	0.0050
custom:mmlu_pro:0	0	acc_norm	0.1287	±	0.0031
custom:openbook_qa:0	0	acc_norm	0.3660	±	0.0216
custom:piqa:0	0	acc_norm	0.7187	±	0.0105
custom:trivia_qa:0	0	qem	0.0787	±	0.0020
custom:winogrande:0	0	acc_norm	0.5367	±	0.0140

Downloads last month: -

Safetensors

Model size

1B params

Tensor type

BF16