Instructions to use Boollm/Boollm-Small-135M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Boollm/Boollm-Small-135M with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Boollm/Boollm-Small-135M:Q8_0
# Run inference directly in the terminal:
llama cli -hf Boollm/Boollm-Small-135M:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Boollm/Boollm-Small-135M:Q8_0
# Run inference directly in the terminal:
llama cli -hf Boollm/Boollm-Small-135M:Q8_0

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Boollm/Boollm-Small-135M:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf Boollm/Boollm-Small-135M:Q8_0

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Boollm/Boollm-Small-135M:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Boollm/Boollm-Small-135M:Q8_0

Use Docker

docker model run hf.co/Boollm/Boollm-Small-135M:Q8_0

LM Studio
Jan

vLLM

How to use Boollm/Boollm-Small-135M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Boollm/Boollm-Small-135M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Boollm/Boollm-Small-135M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Boollm/Boollm-Small-135M:Q8_0

Ollama
How to use Boollm/Boollm-Small-135M with Ollama:
```
ollama run hf.co/Boollm/Boollm-Small-135M:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use Boollm/Boollm-Small-135M with Docker Model Runner:
```
docker model run hf.co/Boollm/Boollm-Small-135M:Q8_0
```

Lemonade

How to use Boollm/Boollm-Small-135M with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Boollm/Boollm-Small-135M:Q8_0

Run and chat with the model

lemonade run user.Boollm-Small-135M-Q8_0

List all available models

lemonade list

Atomic Chat

Boollm Small 135M

Boollm Small 135M is a tiny, local, text-only proof-of-concept release for testing Boollm model discovery, download, and local-app integration. It is based on HuggingFaceTB/SmolLM2-135M-Instruct and is not a fine-tuned education model.

What it can do

It can demonstrate short conversational answers, simple writing assistance, and basic tutoring prompts on a computer with limited memory. Because it has only 135 million parameters, answers can be shallow, inaccurate, or confused.

Do not use this model as the sole source for nursing, medical, legal, safety, technology certification, mathematics, or other high-stakes information. It does not accept images. Boollm-Edu will use a larger vision-capable model for those tasks.

Files

Boollm-Small-135M.Q8_0.gguf — GGUF Q8_0 local-inference model
LICENSE — Apache-2.0 license for this release and base-model notice
release-manifest.json — file digest and provenance record

Local usage

With llama.cpp:

llama-server -hf Boollm/Boollm-Small-135M -c 2048

In Boollm, download the GGUF file into the local models folder and select it from the local model picker.

Provenance

This release contains the original SmolLM2-135M-Instruct weights in GGUF format. No Boollm fine-tuning has been applied. The Boollm name identifies this packaging and integration proof-of-concept, not a claim of improved model capability.

License

The base model is released under Apache-2.0. See LICENSE and the upstream model card for the complete notice and restrictions.

Downloads last month: 6

GGUF

Model size

0.1B params

Architecture

llama

Hardware compatibility

8-bit

Model tree for Boollm/Boollm-Small-135M

Base model

HuggingFaceTB/SmolLM2-135M

Quantized

HuggingFaceTB/SmolLM2-135M-Instruct

Quantized

(127)

this model