Instructions to use ddidacus/olmoe-controller with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ddidacus/olmoe-controller with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("allenai/OLMoE-1B-7B-0125-Instruct")
model = PeftModel.from_pretrained(base_model, "ddidacus/olmoe-controller")

Transformers

How to use ddidacus/olmoe-controller with Transformers:

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

pipe = pipeline("text-generation", model="ddidacus/olmoe-controller")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("ddidacus/olmoe-controller", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ddidacus/olmoe-controller with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/ddidacus/olmoe-controller

SGLang

How to use ddidacus/olmoe-controller 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 "ddidacus/olmoe-controller" \
    --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": "ddidacus/olmoe-controller",
		"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 "ddidacus/olmoe-controller" \
        --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": "ddidacus/olmoe-controller",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ddidacus/olmoe-controller with Docker Model Runner:
```
docker model run hf.co/ddidacus/olmoe-controller
```

olmoe-controller / controller.pt

ddidacus

Upload folder using huggingface_hub

89ab82a verified about 10 hours ago

Download

History Blame Contribute Delete

54.5 MB

This file is stored with Xet . It is too big to display, but you can still download it.

Xet Pointer Details

( Raw pointer file )

Xet hash:: 88c699c34c1f38d3766a15a05689c2e3d27c1163c292555dd420552dbfaf9511
Size of remote file:: 54.5 MB
SHA256:: 5bf3ec2df7ea133b65c800dcb7c16df420b18f569b2d4151afb48e47541a6923

Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.