The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.

Android Deployment Guide

Verified: SmolLM2-135M on Samsung S20 FE

Speed: 46.0 tokens/sec

File size: 101MB

Output: Correct (mentions Paris for capital of France)

Steps

  1. Download GGUF:
hf download dispatchAI/SmolLM2-135M-Instruct-mobile model.gguf
  1. Push to phone (Windows):
MSYS_NO_PATHCONV=1 adb push model.gguf /data/local/tmp/
  1. Run inference:
MSYS_NO_PATHCONV=1 adb shell "cd /data/local/tmp && LD_LIBRARY_PATH=/data/local/tmp ./llama-cli -m model.gguf -p 'Hello' -n 30 -t 4 -st"
  1. Parse speed from output:
[ Prompt: 202 t/s | Generation: 46 t/s ]

All 3 Phone-Verified Models

Model Size Speed Fits 2GB RAM?
SmolLM2-135M 101MB 46.0 t/s ✅
Qwen2.5-0.5B-int4 469MB 23.2 t/s ✅
Llama-3.2-1B-Q4 770MB 5.4 t/s ✅

🚀 dispatchAI

Downloads last month
-