{"cells":[{"metadata":{},"cell_type":"markdown","source":"# **Introduction**","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Hey Everyone, My Name is **Nacir Bouazizi** and in this notebook I am going to show how simply add coco pretrained weights and get a **score > 0.735** without Pseudo Labeling and **> 0.76** with Pseudo Labeling . I am going to make this as easy as possible for everyone to make their submission.\n\nDon't Forget to **Upvote** this notebook if you find it helpful.\n\nPs : this notebook is based on the great work of [OrKatz](https://www.kaggle.com/orkatz2) which you can find it [here](https://www.kaggle.com/orkatz2/yolov5-train) + Some useful tips to make training with coco pretrained weights easier for everyone + Useful tips to train on Colab without having runtime issues. So I hope you enjoy reading it as much as I did writing it.\n\nLet's start with importing some useful Modules :","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport os\nfrom tqdm.auto import tqdm\nimport shutil as sh","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Let's copy the content of yolov5 to our current directory.","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"!cp -r ../input/yolov5/* .","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#uncomment this line \n#!pip install -r requirements.txt","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Dataframe Creation : ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"We have to make our data in a format that yolov5 accepts, and that's what we are doing below :","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"df = pd.read_csv('../input/global-wheat-detection/train.csv')\nbboxs = np.stack(df['bbox'].apply(lambda x: np.fromstring(x[1:-1], sep=',')))\nfor i, column in enumerate(['x', 'y', 'w', 'h']):\n    df[column] = bboxs[:,i]\ndf.drop(columns=['bbox'], inplace=True)\ndf['x_center'] = df['x'] + df['w']/2\ndf['y_center'] = df['y'] + df['h']/2\ndf['classes'] = 0\nfrom tqdm.auto import tqdm\nimport shutil as sh\ndf = df[['image_id','x', 'y', 'w', 'h','x_center','y_center','classes']]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"df.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"df.image_id.nunique()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"index =  list(sorted(set(df.image_id)))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"We are going to choose **fold 0** to be our **OOF**. **Fold 0** gave the **best** score compared to the other folds but we can't trust it as we may overfit. Maybe I'll post another notebook to show how to train on different folds and make an ensemble out of them if some of you asked for it. But now let's just take **fold 0** to show how much it is simple to use **Yolo5**.","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"🛑**PS** : this validation strategy is not that good but here we just want to show how to train using coco pretrained weights.\n","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"source = 'train'\nif True:\n    for fold in [0]:\n        val_index = index[len(index)*fold//5:len(index)*(fold+1)//5]\n        for name,mini in tqdm(df.groupby('image_id')):\n            if name in val_index:\n                path2save = 'val2017/'\n            else:\n                path2save = 'train2017/'\n            if not os.path.exists('convertor/fold{}/labels/'.format(fold)+path2save):\n                os.makedirs('convertor/fold{}/labels/'.format(fold)+path2save)\n\n            if not os.path.exists('convertor/fold{}/labels/'.format(fold)+path2save+name+\".txt\"):   \n              with open('convertor/fold{}/labels/'.format(fold)+path2save+name+\".txt\", 'w+') as f:\n                  row = mini[['classes','x_center','y_center','w','h']].astype(float).values\n                  row = row/1024\n                  row = row.astype(str)\n                  for j in range(len(row)):\n                      text = ' '.join(row[j])\n                      f.write(text)\n                      f.write(\"\\n\")\n            if not os.path.exists('convertor/fold{}/images/{}'.format(fold,path2save)):\n                os.makedirs('convertor/fold{}/images/{}'.format(fold,path2save))\n\n            if not os.path.exists('convertor/fold{}/images/{}/{}.jpg'.format(fold,path2save,name)):    \n              sh.copy(\"../input/global-wheat-detection/{}/{}.jpg\".format(source,name),'convertor/fold{}/images/{}/{}.jpg'.format(fold,path2save,name))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Training: ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Now comes the good and easy part. Let's install **apex** first which will make our training faster. ","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"# Apex install\n#uncomment this line\n#!git clone https://github.com/NVIDIA/apex && cd apex && pip install -v --no-cache-dir --global-option=\"--cpp_ext\" --global-option=\"--cuda_ext\" . --user && cd .. && rm -rf apex","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Now we will train our model using coco pretrained weights so below you can find the command and I'm going to explain important arguments in this command : ","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"#uncomment this line \n#!python ./train.py --img 1024 --batch 2 --epochs 1 --data \"../input/configyolo5v1/configYolo5v1/wheat0.yaml\" --cfg \"../input/yolov5yaml/yolov5xx.yaml\"   --name yolov5x_fold0 --weights yolov5x.pt","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"* **--batch** : the batch size, and here I choose 2 because it gives the best results.\n* **--epochs** : here I choose 1 but to just show you how it works. You can set it to **100**\n* **--data** : it's path to the yaml file which contains the paths to our training data (which we stored in **convertor** directory)\n* **--cfg** : the path to our custom yolov5 yaml config file.\n* **--name** : after training this prefix s added to last and best so we get best_yolov5x_fold0.pt and last_yolov5x_fold0.pt.\n* **--weights** : that's the most important argument, here we tell Yolov5 to use coco pretrained weights to train on our data.  \n","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"rm -rf ./convertor","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# What Next ? ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Training for 100 epochs takes alot of time and we can't do that on kaggle because we have limited ressources. So my advice would be to use Colab ( or Colab Pro if you can pay **10$/month**) \nBut many are complaining about runtime errors and random disconnections on Colab. \nSo here's how you can train on it and go to sleep 😃","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"**If you choose to use Normal Colab :** ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Every epoch takes 10 to 12 minutes so that's 16.6 to 20 hours to train for 100 epochs. \nNormal Colab gives you a session of 12 hours per day. so we can train up to 72 epochs (maximum) and 60 (minimum ) . That's not always the case but it's an approximation. \nSo my advice would be :\n* to train for 50 epochs and save your weights (which you can find in the weights folder) to your drive using this command : ","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"#!cp \"/content/weights/last_yolov5x_fold3.pt\" \"/content/drive/My Drive/your_path_here\"\n#!cp \"/content/weights/best_yolov5x_fold3.pt\" \"/content/drive/My Drive/your_path_here\"","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"**If you choose to use Colab Pro :**","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"You won't have a problem because you will have a 24 hours session 😃","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"**Avoid Runtime Erros and disconnections :** ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Many of you have been asking how to fix this because we loose everything everytime Colab disconnects and no one wants that especially in this competition 😢 So here's how to avoid it using a simple Javascript code. You can copy paste it in your console. **ctrl +Shift + I** to open it.","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"\"\"\"\"\"\n\nfunction ClickConnect(){\nconsole.log(\"Working\"); \ndocument.querySelector(\"colab-toolbar-button\").click() \n}setInterval(ClickConnect,60000)\n\n\"\"\"","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Here's a screenshot to better illustrate my words. ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}},"execution_count":null},{"metadata":{},"cell_type":"markdown","source":"# **Next  ?** ","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"Go to sleep 😴.\nAnd the next day you wakeup, train for the remaining epochs (if you use normal colab). \nWhen you finish training take a look on this great notebooks to make an inference : \n* [YoloV5 Pseudo Labeling](https://www.kaggle.com/nvnnghia/yolov5-pseudo-labeling)\n* [YoloV5 fake or real [Single Model] L.B>0.753](https://www.kaggle.com/orkatz2/yolov5-fake-or-real-single-model-l-b-0-753)\n* [YoloV5 [Single Model] with more TTA L.B>0.745](https://www.kaggle.com/wasupandceacar/yolov5-single-model-with-more-tta-lb-0-745)","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"I think I finished my job here, I hope you like this notebook, it's my first one. Upvote to see more content like this one. And feel free to ask anything below. I can help just ask 😊","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}