SmolLM2-135M-Instruct-ArduinoUnoQ-GGUF

Quantized and optimized for the Arduino Uno Q using an NVIDIA DGX Spark.

Optimization Specs

Architecture: ARM64 / Adreno 702 GPU
Quantization: Q4_0 with --pure alignment for Adreno kernels.
Base Model: SmolLM2-135M-Instruct

⚙️ Inference Engine

This model is designed to run with the optimized llama-cli binary found in the official Uno Q Edge AI repository: 👉 GitHub: Arduino-UnoQ-Optimized-Llama-CLI

Usage

bash ./llama-cli -m unoq_optimized.gguf -ngl 99 -t 4 -c 2048

Downloads last month: 45

GGUF

Model size

0.1B params

Architecture

llama

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for assix-research/SmolLM2-135M-Instruct-ArduinoUnoQ-GGUF

Base model

HuggingFaceTB/SmolLM2-135M

Quantized

HuggingFaceTB/SmolLM2-135M-Instruct

Quantized

(154)

this model