Instructions to use echarlaix/distilgpt2-openvino-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use echarlaix/distilgpt2-openvino-4bit with Transformers:

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

pipe = pipeline("text-generation", model="echarlaix/distilgpt2-openvino-4bit")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("echarlaix/distilgpt2-openvino-4bit")
model = AutoModelForCausalLM.from_pretrained("echarlaix/distilgpt2-openvino-4bit", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use echarlaix/distilgpt2-openvino-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "echarlaix/distilgpt2-openvino-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "echarlaix/distilgpt2-openvino-4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/echarlaix/distilgpt2-openvino-4bit

SGLang

How to use echarlaix/distilgpt2-openvino-4bit 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 "echarlaix/distilgpt2-openvino-4bit" \
    --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": "echarlaix/distilgpt2-openvino-4bit",
		"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 "echarlaix/distilgpt2-openvino-4bit" \
        --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": "echarlaix/distilgpt2-openvino-4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use echarlaix/distilgpt2-openvino-4bit with Docker Model Runner:
```
docker model run hf.co/echarlaix/distilgpt2-openvino-4bit
```

distilgpt2-openvino-4bit

File size: 1,041 Bytes

f24ea7c

---
base_model: echarlaix/distilgpt2-openvino
datasets:
- openwebtext
language: en
license: apache-2.0
tags:
- exbert
- openvino
- openvino
co2_eq_emissions: 149200
model-index:
- name: distilgpt2
  results:
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: WikiText-103
      type: wikitext
    metrics:
    - type: perplexity
      value: 21.1
      name: Perplexity
---

This model is a quantized version of [`echarlaix/distilgpt2-openvino`](https://huggingface.co/echarlaix/distilgpt2-openvino) and was exported to the OpenVINO format using [optimum-intel](https://github.com/huggingface/optimum-intel) via the [nncf-quantization](https://huggingface.co/spaces/echarlaix/nncf-quantization) space.

First make sure you have optimum-intel installed:

```bash
pip install optimum[openvino]
```

To load your model you can do as follows:

```python
from optimum.intel import OVModelForCausalLM

model_id = "echarlaix/distilgpt2-openvino-int4"
model = OVModelForCausalLM.from_pretrained(model_id)
```