Instructions to use LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct", 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-3.5-2.4B-Instruct", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct"
# 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-3.5-2.4B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct

SGLang

How to use LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct 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-3.5-2.4B-Instruct" \
    --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-3.5-2.4B-Instruct",
		"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-3.5-2.4B-Instruct" \
        --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-3.5-2.4B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct with Docker Model Runner:
```
docker model run hf.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct
```

high memory use

by rootw0rm - opened Dec 10, 2024

Discussion

rootw0rm

Dec 10, 2024

EXAONE uses a lot more memory for context compared to Qwen 2.5. Is this inherent to the model or is it something wrong with llama.cpp?

yireun

Dec 10, 2024

•

edited Dec 10, 2024

Hi, electroglyph.

Would you give us more information (e.g., gguf type and llama-cli parameters) for testing?
When compared EXAONE-3.5-2.4B-Instruct-BF16.gguf and qwen2.5-3b-instruct-fp16.gguf with the same parameters (llama-cli -cnv -m '...' -p '...') on CPU, EXAONE used less memory.

electroglyph

Dec 12, 2024

•

edited Dec 14, 2024

i've tested using some of the GPU backends, i.e. SYCL, Vulkan, etc. my context limit is around 50% of what it is with Qwen 2.5 3B. i've tested several versions of llama.cpp so far. i'm going to do some more testing and i'll be back with more detailed information.

...my context limit is somewhere around 60K with EXAONE 2.4B, but I can hit 120K with Qwen 2.5 3B (no quantization). these small models are great for running in parallel, so my actual context is divided by how many parallel tasks I'm running. the lower context limit means i have to lower how many i run in parallel

electroglyph

Dec 14, 2024

After more testing I can update this to say the context limit is nearly exactly 50% that of Qwen 2.5 3B.

I've opened an issue here if you want to weigh in:
https://github.com/ggerganov/llama.cpp/issues/10823

yireun

Dec 16, 2024

Hi, 0xDEADFED5

It is due to the differences of architecture between EXAONE 3.5 2.4B and Qwen 2.5 3B. To be specific, num_attention_heads and num_key_value_heads are difference between them.

Thank you.

electroglyph

Dec 16, 2024

thanks!

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