Instructions to use joelhenwang/OdinNext-138M-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use joelhenwang/OdinNext-138M-Base with Transformers:

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

pipe = pipeline("text-generation", model="joelhenwang/OdinNext-138M-Base", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("joelhenwang/OdinNext-138M-Base", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use joelhenwang/OdinNext-138M-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "joelhenwang/OdinNext-138M-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "joelhenwang/OdinNext-138M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/joelhenwang/OdinNext-138M-Base

SGLang

How to use joelhenwang/OdinNext-138M-Base 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 "joelhenwang/OdinNext-138M-Base" \
    --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": "joelhenwang/OdinNext-138M-Base",
		"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 "joelhenwang/OdinNext-138M-Base" \
        --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": "joelhenwang/OdinNext-138M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use joelhenwang/OdinNext-138M-Base with Docker Model Runner:
```
docker model run hf.co/joelhenwang/OdinNext-138M-Base
```

Awesome Model

by Datdanboi25 - opened about 17 hours ago

Discussion

Datdanboi25

about 17 hours ago

Awesome stuff man! Mind if I add it to my leaderboard?
https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard
Cant wait to see whats next!

joelhenwang

Owner about 17 hours ago

@Datdanboi25 Of course! You're more than welcome :D

Datdanboi25

about 15 hours ago

•

edited about 15 hours ago

Both Base + instruct should be up! Impressive math scores!
However couldn't replicate your hellaswag score on the internal harness or lm eval harness.

joelhenwang

Owner about 6 hours ago

•

edited about 6 hours ago

Thanks so much for adding OdinNext to the leaderboard, really appreciate it! On the HellaSwag score, I re-ran it to double-check, and I realised that I ran with the flag to run the first 1000 tests, after running it again to cover the full benchmark, the Hellaswag score was similar as the leaderboard. Thanks for mentioning it!

Datdanboi25

about 2 hours ago

Nah no worries, doing some genuinely impressive stuff!
Keep up the good work and I'm intrigued to see what comes next!

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