Instructions to use baseten/smol_llama-101M-GQA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use baseten/smol_llama-101M-GQA with Transformers:

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

pipe = pipeline("text-generation", model="baseten/smol_llama-101M-GQA")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("baseten/smol_llama-101M-GQA")
model = AutoModelForCausalLM.from_pretrained("baseten/smol_llama-101M-GQA", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use baseten/smol_llama-101M-GQA with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "baseten/smol_llama-101M-GQA"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "baseten/smol_llama-101M-GQA",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/baseten/smol_llama-101M-GQA

SGLang

How to use baseten/smol_llama-101M-GQA 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 "baseten/smol_llama-101M-GQA" \
    --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": "baseten/smol_llama-101M-GQA",
		"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 "baseten/smol_llama-101M-GQA" \
        --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": "baseten/smol_llama-101M-GQA",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use baseten/smol_llama-101M-GQA with Docker Model Runner:
```
docker model run hf.co/baseten/smol_llama-101M-GQA
```

Model Card for Model ID

Model Details

Created using

from transformers import LlamaTokenizer, LlamaForCausalLM

# Specify the model name or path from the Hugging Face Hub
org_in = "BEE-spoke-data/"
org_out = "baseten/"
model_name = "smol_llama-101M-GQA"
embedding = False
embedding_name = "embedding-" if embedding else ""
tokenizer = LlamaTokenizer.from_pretrained(org_in+model_name)
model = LlamaForCausalLM.from_pretrained(org_in+model_name)

prompt = "Hello, this is a test."
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

model_llama = model
if embedding:
    model_llama = model_llama.model

# 2. Push to a new Hugging Face repository
# Make sure you have run `huggingface-cli login` beforehand to authenticate
model_llama.push_to_hub(org_out+embedding_name+model_name, token="xxx")
tokenizer.push_to_hub(org_out+embedding_name+model_name, token="xx")

Downloads last month: 13

Safetensors

Model size

0.1B params

Tensor type

F32