Instructions to use Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M with Transformers:

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

pipe = pipeline("text-generation", model="Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M")
model = AutoModelForCausalLM.from_pretrained("Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M

SGLang

How to use Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M 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 "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M" \
    --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": "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M",
		"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 "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M" \
        --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": "Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M with Docker Model Runner:
```
docker model run hf.co/Ex0bit/OLMo-3-7B-Instruct-NVFP4-1M
```

OLMo-3-7B-Instruct-NVFP4-1M / quantization_metadata.json

Ex0bit

Upload folder using huggingface_hub

ad2b339 verified 11 months ago

Raw History Blame Contribute Delete

1 kB

	{
	"model_name": "allenai/Olmo-3-7B-Instruct",
	"quantization_format": "NVFP4",
	"calibration_samples": 2048,
	"original_memory_gb": "14.60",
	"quantized_memory_gb": "14.61",
	"memory_reduction_percent": "-0.1",
	"total_parameters_billions": "7.30",
	"target_device": "DGX Spark (GB10)",
	"export_path": "/workspace/quantized-output/Olmo-3-7B-NVFP4",
	"quantization_config": {
	"model_name": "allenai/Olmo-3-7B-Instruct",
	"quantization_format": "NVFP4",
	"calibration_samples": 2048,
	"calibration_max_length": 2048,
	"calibration_min_length": 100,
	"calibration_dataset": "allenai/c4",
	"calibration_dataset_config": "en",
	"calibration_dataset_split": "train",
	"target_device": "DGX Spark (GB10)",
	"hf_cache_dir": "~/.cache/huggingface",
	"output_dir": "./quantized-output/Olmo-3-7B-NVFP4",
	"log_dir": "./logs",
	"torch_dtype": "bfloat16",
	"device_map": "auto",
	"trust_remote_code": true
	},
	"created_at": "2025-11-22T04:10:03.593289"
	}