Instructions to use nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit with Transformers:

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

pipe = pipeline("text-generation", model="nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit")
model = AutoModelForCausalLM.from_pretrained("nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit

SGLang

How to use nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit 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 "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit" \
    --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": "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit",
		"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 "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit" \
        --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": "nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit with Docker Model Runner:
```
docker model run hf.co/nanzhang/QuantLRM-Olmo-3-7B-Think-3-bit
```

Add pipeline_tag, library_name, and reasoning tag

by nielsr HF Staff - opened Feb 10

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

-4

nielsr

Feb 10

This PR enhances the model card by adding relevant metadata and minor content improvements:

pipeline_tag: text-generation: Categorizes the model for common LRM tasks, improving discoverability.
library_name: transformers: Specifies compatibility with the transformers library, enabling the automated "Use in Transformers" code snippet on the Hub. Evidence for this is found in the config.json (architectures: ["Olmo3ForCausalLM"], transformers_version: "4.57.3").
tags: - reasoning: Adds a relevant tag to reflect the model's focus on "Large Reasoning Models," as indicated by the paper title.
Content enhancements: Updated the "Developed by" section to include a link to Nan Zhang's Hugging Face profile and added a "Finetuned from model" entry with a link to allenai/Olmo-3-7B-Think for better context.

These updates will improve the model's discoverability and utility for users on the Hugging Face Hub.

Add pipeline_tag, library_name, and reasoning tageb241fc0

nanzhang changed pull request status to merged Feb 11

nanzhang

Owner Feb 11

Thanks for this PR!

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