Deploy MobileNet variants on Google Coral Edge TPU

Overview

Run MobileNet variants directly on Google Coral Edge TPU for IoT vision AI. Local inference offers privacy, zero latency, and no ongoing API costs.

Specs: Edge TPU · 1W power

Installation

bash
Install Ollama — easiest local inference runtime
curl -fsSL https://ollama.com/install.sh | sh
Verify installation
ollama --version

Download Model

bash
Pull MobileNet variants (downloads GGUF quantized weights automatically)
ollama pull mobilenet-variants
Run interactive chat
ollama run mobilenet-variants
Start API server
ollama serve
API available at http://localhost:11434

Python Integration

python
import httpx
from typing import Iterator
class LocalAI:
    """Interface to local MobileNet variants running on Google Coral Edge TPU."""
    
    BASE_URL = "http://localhost:11434"
    MODEL = "mobilenet-variants"
    
    def chat(self, message: str, system: str = "") -> str:
        """Single-turn chat."""
        resp = httpx.post(
            f"{self.BASE_URL}/api/chat",
            json={
                "model": self.MODEL,
                "messages": [
                    {"role": "system", "content": system},
                    {"role": "user", "content": message}
                ],
                "stream": False
            },
            timeout=120
        )
        resp.raise_for_status()
        return resp.json()["message"]["content"]
    
    def stream(self, message: str) -> Iterator[str]:
        """Streaming chat for real-time output."""
        with httpx.stream(
            "POST",
            f"{self.BASE_URL}/api/chat",
            json={"model": self.MODEL, "messages": [{"role": "user", "content": message}], "stream": True},
            timeout=120
        ) as r:
            for line in r.iter_lines():
                if line:
                    import json
                    chunk = json.loads(line)
                    if not chunk.get("done"):
                        yield chunk["message"]["content"]
Usage
ai = LocalAI()
response = ai.chat("Help me with IoT vision AI")
print(response)
Streaming
for token in ai.stream("Explain IoT vision AI step by step"):
    print(token, end="", flush=True)

Custom Modelfile

bash
Create optimized configuration for IoT vision AI
cat > Modelfile << 'MODELEOF'
FROM mobilenet-variants
PARAMETER num_ctx 4096
PARAMETER temperature 0.7
PARAMETER top_p 0.9
SYSTEM "You are an AI assistant specialized in IoT vision AI. You run locally on Google Coral Edge TPU. Be concise, accurate, and helpful."
MODELEOF
ollama create IoT-vision-AI-assistant -f Modelfile
ollama run IoT-vision-AI-assistant

Performance Profile

MetricValue

HardwareEdge TPU Memory1W power Speed10-40 tokens/sec (CPU) / 40-100+ tok/s (GPU) First token<200ms (GPU) / <1s (CPU) Context4096-32768 tokens Cost$0 (after hardware)

Production Setup with FastAPI

python
from fastapi import FastAPI
from pydantic import BaseModel
app = FastAPI(title="Google Coral Edge TPU AI API")
ai = LocalAI()
class ChatRequest(BaseModel):
    message: str
    system: str = ""
class ChatResponse(BaseModel):
    response: str
    model: str
    device: str
@app.post("/chat", response_model=ChatResponse)
async def chat_endpoint(req: ChatRequest):
    response = ai.chat(req.message, req.system)
    return ChatResponse(response=response, model="MobileNet variants", device="Google Coral Edge TPU")
@app.get("/health")
async def health():
    return {"status": "ok", "model": "MobileNet variants", "device": "Google Coral Edge TPU"}

Troubleshooting

Slow inference: Switch to Q4_K_M quantization, reduce context window Out of memory: Use smaller model or Q3_K_S quant GPU not used: Install CUDA/Metal drivers, check ollama logs High latency: Warm up model by sending a dummy request on startup

Resources

Ollama library: https://ollama.com/library

GGUF format: https://github.com/ggerganov/llama.cpp

Hardware guide: https://ollama.com/blog/hardware-recommendations