Instructions to use LGAI-EXAONE/EXAONE-Deep-2.4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LGAI-EXAONE/EXAONE-Deep-2.4B with Transformers:

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

pipe = pipeline("text-generation", model="LGAI-EXAONE/EXAONE-Deep-2.4B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("LGAI-EXAONE/EXAONE-Deep-2.4B", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use LGAI-EXAONE/EXAONE-Deep-2.4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LGAI-EXAONE/EXAONE-Deep-2.4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LGAI-EXAONE/EXAONE-Deep-2.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LGAI-EXAONE/EXAONE-Deep-2.4B

SGLang

How to use LGAI-EXAONE/EXAONE-Deep-2.4B 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 "LGAI-EXAONE/EXAONE-Deep-2.4B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LGAI-EXAONE/EXAONE-Deep-2.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "LGAI-EXAONE/EXAONE-Deep-2.4B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LGAI-EXAONE/EXAONE-Deep-2.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LGAI-EXAONE/EXAONE-Deep-2.4B with Docker Model Runner:
```
docker model run hf.co/LGAI-EXAONE/EXAONE-Deep-2.4B
```

Chat template difference with 32b

by nbroad - opened Mar 21, 2025

Discussion

nbroad

Mar 21, 2025

Hi @LG-AI-EXAONE ,

I noticed that the tokenizer chat template for 2.4B is not the same as for 32B. Was this intentional or should they have the same chat template?

nuxlear

LG AI Research org Mar 22, 2025

Hello, @nbroad !

We checked the tokenizer configuration and found it uses '\n' instead of '\n', which was not intentional. However, we've tested both versions of the chat template and haven't observed any difference in behavior.
I believe you can either update the chat template to match the 32B version or leave it unchanged - both options should work fine.

nuxlear

LG AI Research org Mar 22, 2025

For consistency and to avoid confusion, we have updated the chat template of 2.4B to align with our 32B model and other models.
This update would not change the model's behavior or functionality.

nbroad

Mar 24, 2025

Great! Thank you 🙂

nbroad changed discussion status to closed Mar 24, 2025

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment