run.sh

#!/bin/bash

# Start the Ollama server in the background
ollama serve &

# Wait a few seconds for the server to be fully operational
sleep 5

# Pull the model from Hugging Face.
# This command downloads the model and makes it available to the API.
# It's the non-interactive version of "ollama run".
ollama pull smollm2:360m-instruct-q5_K_M
ollama pull hf.co/ggml-org/SmolLM3-3B-GGUF:Q4_K_M
ollama pull hf.co/bartowski/nvidia_OpenReasoning-Nemotron-1.5B-GGUF:Q5_K_M
ollama pull hf.co/bartowski/Llama-3.2-3B-Instruct-GGUF:Q4_K_M
ollama pull hf.co/unsloth/gemma-3-4b-it-qat-GGUF:Q4_K_M
ollama pull hf.co/bartowski/Qwen_Qwen3-1.7B-GGUF:Q5_K_M
ollama pull gemma3n:e2b-it-q4_K_M
ollama pull granite3.3:2b

# Start the Gradio web application
# This will connect to the Ollama server which is already running.
python3 app.py