Instructions to use apple/OpenELM-450M-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use apple/OpenELM-450M-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="apple/OpenELM-450M-Instruct", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("apple/OpenELM-450M-Instruct", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use apple/OpenELM-450M-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "apple/OpenELM-450M-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "apple/OpenELM-450M-Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/apple/OpenELM-450M-Instruct

SGLang

How to use apple/OpenELM-450M-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 "apple/OpenELM-450M-Instruct" \
    --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": "apple/OpenELM-450M-Instruct",
		"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 "apple/OpenELM-450M-Instruct" \
        --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": "apple/OpenELM-450M-Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use apple/OpenELM-450M-Instruct with Docker Model Runner:
```
docker model run hf.co/apple/OpenELM-450M-Instruct
```

need gguf support

by huntz47 - opened May 4, 2024

Discussion

huntz47

May 4, 2024

anyone from apple team are seeing this. please add a gguf format for this model .
thank you

sdyy

Jul 27, 2024

I want that too

reach-vb

Aug 1, 2024

Hey hey @huntz47 & @sdyy - sorry for the delay in response. OpenELM is supported in llama.cpp!

I created some quants for the instruct models here:
450M - https://huggingface.co/reach-vb/OpenELM-450M-Instruct-Q8_0-GGUF
1.1B - https://huggingface.co/reach-vb/OpenELM-1_1B-Instruct-Q8_0-GGUF
3B - https://huggingface.co/reach-vb/OpenELM-3B-Instruct-Q8_0-GGUF

Note: I found quite a bit of degradation below Q8, but if you want to create other quants then feel free to use GGUF-my-repo space: https://huggingface.co/spaces/ggml-org/gguf-my-repo for it.

The inference instructions are in the model cards. Enjoy! and do let me know if you have any questions!

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