{
 "cells": [
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "from ultralytics import YOLO\n",
    "\n",
    "original_model_name = 'yolov8s'\n",
    "original_model_nickname = 'fashionpedia'\n",
    "yolo_model = YOLO(original_model_name)"
   ],
   "id": "688b209f5c7b833e",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "results = yolo_model.train(data='fashionpedia.yaml', epochs=300, batch=32, save_period=1, plots=True,\n",
    "                           name=f'{original_model_name}-{original_model_nickname}', exist_ok=True, resume=False,\n",
    "                           deterministic=False, close_mosaic=50, val=False,\n",
    "                           label_smoothing=0.05, workers=20, cache=False, task=\"detect\",\n",
    "                           translate=0.2, degrees=30, shear=10, mosaic=0.8, perspective=0.000_01\n",
    "                           )"
   ],
   "id": "dd14c466280421c4",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [],
   "id": "8884f7e383561fc2",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "# load saved model\n",
    "yolo_model_loaded = YOLO(\"shared/yolov8n-fashionpedia-1.torchscript\", task='detect')"
   ],
   "id": "f427e4ba2cff784",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "# Run batched inference on a list of images\n",
    "inference_results = yolo_model_loaded(['./test/courte.png',\n",
    "                                       './test/longue.png',\n",
    "                                       './test/rayure.png',\n",
    "                                       './test/sans-manche.png',\n",
    "                                       './test/trois-quart.png'\n",
    "                                       ], verbose=False)\n",
    "for result in inference_results:\n",
    "    boxes = result.boxes  # Boxes object for bounding box outputs\n",
    "    probs = result.probs  # Probs object for classification outputs\n",
    "    result.show()  # display to screen"
   ],
   "id": "f0815af787b52825",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [],
   "id": "793f625e9139f76d",
   "outputs": [],
   "execution_count": null
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}