{"id":13645091,"url":"https://github.com/slyautomation/osrs_yolov5","last_synced_at":"2025-10-23T11:31:18.155Z","repository":{"id":63179092,"uuid":"428122789","full_name":"slyautomation/osrs_yolov5","owner":"slyautomation","description":"Yolov5 Object Detection In OSRS using Python code, Detecting Cows - Botting","archived":false,"fork":false,"pushed_at":"2024-01-13T06:21:37.000Z","size":82493,"stargazers_count":33,"open_issues_count":2,"forks_count":12,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-01-30T18:04:38.996Z","etag":null,"topics":["botting","cuda","machine-learning","mlbot","osrs","pycharm","python","pytorch","runescape","yolov5"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/slyautomation.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2021-11-15T04:31:05.000Z","updated_at":"2025-01-26T21:25:56.000Z","dependencies_parsed_at":"2024-08-02T01:38:22.353Z","dependency_job_id":null,"html_url":"https://github.com/slyautomation/osrs_yolov5","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/slyautomation%2Fosrs_yolov5","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/slyautomation%2Fosrs_yolov5/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/slyautomation%2Fosrs_yolov5/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/slyautomation%2Fosrs_yolov5/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/slyautomation","download_url":"https://codeload.github.com/slyautomation/osrs_yolov5/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":237821403,"owners_count":19371749,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["botting","cuda","machine-learning","mlbot","osrs","pycharm","python","pytorch","runescape","yolov5"],"created_at":"2024-08-02T01:02:26.628Z","updated_at":"2025-10-23T11:31:13.919Z","avatar_url":"https://github.com/slyautomation.png","language":"Python","funding_links":[],"categories":["Object Detection Applications"],"sub_categories":[],"readme":"\noriginal work created by https://github.com/ultralytics/yolov5\n\nYOLOv5 🚀 is the world's most loved vision AI, representing \u003ca href=\"https://ultralytics.com/\" rel=\"nofollow\"\u003eUltralytics\u003c/a\u003e open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development.\n\nTo request a commercial license please complete the form at \u003ca href=\"https://ultralytics.com/license\" rel=\"nofollow\"\u003eUltralytics Licensing\u003c/a\u003e.\n\n# Yolov5 Object Detection In OSRS using Python code, Detecting Cows - Botting\n\nname: GeForce GTX 1060 6GB (average fps 11 on monitor display using screenshots)\n\n```diff \n- note: There's issues as at July 2022 with newer gpus namely GeForce RTX 3090 \u0026 2080 with the Pytorch Framework, hopefully in the future these issues can be resolved and a stable release can be made.\n```\n\nhttps://user-images.githubusercontent.com/81003470/148143834-97b237c1-205c-4e95-b2c7-fb16e9938184.mp4\n\nFor a video with commentary: https://youtu.be/rqk0kq4Vu3M\n\nFull Video Tutorial: https://youtu.be/pjIEI27DImA\n\n# Quick Start\n\n### installing python\n\ninstalling python programming language\n\n\u003ca href=\"https://www.python.org/downloads/\" rel=\"nofollow\"\u003epython downloads website\u003c/a\u003e\n\nPytorch is only compatiable with versions of python 3.7 or earlier, to get a previous version:\n\nhttps://www.python.org/downloads/windows/\n\nhttps://www.python.org/downloads/macos/\n\nhttps://www.python.org/ftp/python/3.7.9/python-3.7.9-amd64.exe\n\n![image](https://user-images.githubusercontent.com/81003470/197636886-08a85a8c-7a65-46fc-a83e-865972bff6bb.png)\n\n\n### installing pycharm\n\ninstalling pycharm \u003ca href=\"https://www.jetbrains.com/pycharm/download/\" rel=\"nofollow\"\u003epycharm\u003c/a\u003e.\n\n\u003ca href=\"https://www.jetbrains.com/pycharm/download/download-thanks.html?platform=windows\u0026code=PCC\" rel=\"nofollow\"\u003epycharm windows\u003c/a\u003e\n\nStep walkthrough Installation of pycharm: https://github.com/slyautomation/osrs_yolov5/wiki/How-to-Install-Pycharm\n\n### Check your cuda version\n\ntype in terminal: ```nvidia-smi```\n\n![image](https://user-images.githubusercontent.com/81003470/147712277-5b1fae1d-33b2-4ff0-a4de-19ef762e1b14.png)\n\nmy version that i can use is up to: 11.5 but for simplicity i can use previous versions namely 10.2\n\nCheck if your gpu will work: https://developer.nvidia.com/cuda-gpus and use the cuda for your model and the latest cudnn for the cuda version.\n\nfull list of cuda versions: https://developer.nvidia.com/cuda-toolkit-archive\n\ncuda 10.2 = https://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda_10.2.89_441.22_win10.exe\n\n### Install Cudnn\n\ncuDNN = https://developer.nvidia.com/rdp/cudnn-archive#a-collapse765-10\n\nStep walkthrough Installation of cuDNN: https://github.com/slyautomation/osrs_yolov5/wiki/How-to-install-CuDNN\n\n### Download and Use LabelImg for creating annotated (object identified bounding boxes) image data\n\nlabelImg = https://github.com/heartexlabs/labelImg/releases\n\nStep walkthrough capturing screen images, downloading and using LabelImg: https://github.com/slyautomation/osrs_yolov5/wiki/Downloading-and-using-LabelImg\n\n### Install OCR-Tesseract\n\nOCR-Tesseract:\n\n\u003ca href=\"https://digi.bib.uni-mannheim.de/tesseract/tesseract-ocr-w32-setup-v5.0.1.20220118.exe\"\u003etesseract-ocr-w32-setup-v5.0.1.20220118.exe (32 bit)\u003c/a\u003e and\n\n\u003ca href=\"https://digi.bib.uni-mannheim.de/tesseract/tesseract-ocr-w64-setup-v5.0.1.20220118.exe\"\u003etesseract-ocr-w64-setup-v5.0.1.20220118.exe (64 bit)\u003c/a\u003e resp.\n\nStep walkthrough installing OCR-Tesseract: https://github.com/slyautomation/osrs_yolov5/wiki/How-to-Install-OCR-Tesseract\n\n### Using this Repo\n\nAdd this repo to a project in Pycharm\n\nStep walkthrough adding project with Pycharm: https://github.com/slyautomation/osrs_yolov5/wiki/How-to-add-Project-with-Pycharm\n\n### Ensure Project is within venv (virtual environment)\n\nStep walkthrough activating venv: https://github.com/slyautomation/osrs_yolov5/wiki/how-to-ensure-venv-(virtual-environment)-is-active\n\n## Install Module Requirements\n\nin the terminal type:\n\n```pip install -r requirements.txt```\n\n![image](https://user-images.githubusercontent.com/81003470/147746531-aa622ccb-d6a0-4310-85b7-4775f8b0732a.png)\n\n## Check cuda version is compatiable with torch and torchvision\n\ngoto website and check version https://download.pytorch.org/whl/torch_stable.html\n\nTo take advantage of the gpu and cuda refer to the list for your cuda version search for cu\u003cversion of cuda no spaces or fullstops\u003e e.g cu102 for cuda 10.2.\n  \nuse the latest versions found, i at this point in time found: torch 1.9.0 and torchvision 0.10.0 (these 2 module versions so far i have had no issues other versions i get errors when running \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/detect.py\"\u003edetect.py\u003c/a\u003e)\n\n  ![image](https://user-images.githubusercontent.com/81003470/147751626-8be13bfb-e97d-4642-81db-20955f2a41ad.png)\n  \n  in the terminal type the torch version + your cuda version (except for torchaudio no cuda version required):\n  \n  ```pip install torch==1.9.0+cu102 torchvision==0.10.0+cu102 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html```\n  \n Alternatively use this one:\n  \n ```pip install --upgrade --force-reinstall torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118```\n\n```pip install tensorboard seaborn tqdm matplotlib pandas pyyaml opencv-python```\n\n\n\n  \n  ![image](https://user-images.githubusercontent.com/81003470/147749033-c5de2a74-5365-444c-93c1-f5d9f75512c4.png)\n\n  \n  ![image](https://user-images.githubusercontent.com/81003470/147749284-9411be6f-f000-4bf9-a167-b0d214b977f5.png)\n\n  \n- `Make sure when installing torchvision it doesn't try to install another version due to incompatability, try to either find a later version of torch or use a downgraded version of torchvision. there could be issues if another torch version is installed but the cuda version doesn't align with your gpu.`\n\n## Test pytorch and cuda work\n  \nin the project run \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/main.py\"\u003emain.py\u003c/a\u003e, the output should result in the device used as cuda, and the tensor calculations should run without errors:\n  \n  ![image](https://user-images.githubusercontent.com/81003470/147753127-c97b0ce4-e9c6-49d4-a817-f9a71928e240.png)\n\n  This will also download the yolov5 weight files:\n  \n  ![image](https://user-images.githubusercontent.com/81003470/147753307-5c3df94e-206b-4bac-8f2d-8a5e7301c010.png)\n\n# Custom training setup with YAML\n  \n\u003cp\u003e\u003ca href=\"https://www.kaggle.com/ultralytics/coco128\" rel=\"nofollow\"\u003eCOCO128\u003c/a\u003e is an example small tutorial dataset composed of the first 128 images in \u003ca href=\"http://cocodataset.org/#home\" rel=\"nofollow\"\u003eCOCO\u003c/a\u003e train2017. These same 128 images are used for both training and validation to verify our training pipeline is capable of overfitting. \u003ca href=\"https://github.com/ultralytics/yolov5/blob/master/data/coco128.yaml\"\u003edata/coco128.yaml\u003c/a\u003e, shown below, is the dataset config file that defines 1) the dataset root directory \u003ccode\u003epath\u003c/code\u003e and relative paths to \u003ccode\u003etrain\u003c/code\u003e / \u003ccode\u003eval\u003c/code\u003e / \u003ccode\u003etest\u003c/code\u003e image directories (or *.txt files with image paths), 2) the number of classes \u003ccode\u003enc\u003c/code\u003e and 3) a list of class \u003ccode\u003enames\u003c/code\u003e:\u003c/p\u003e\n  \n```\n# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]\npath: ../datasets/coco128  # dataset root dir\ntrain: images/train2017  # train images (relative to 'path') 128 images\nval: images/train2017  # val images (relative to 'path') 128 images\ntest:  # test images (optional)\n\n# Classes\nnc: 80  # number of classes\nnames: [ 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',\n         'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',\n         'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',\n         'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',\n         'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',\n         'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',\n         'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone',\n         'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear',\n         'hair drier', 'toothbrush' ]  # class names\n  ```\nCopying the method above i have done the same with \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/master/data/osrs.yaml\"\u003edata/osrs.yaml\u003c/a\u003e \n  \n```\n# parent\n# ├── yolov5\n# └── datasets\n#     └── osrs ← downloads here\n#       └── cow ← add each class\n  #     └── xxx ← add each class\n\n\n# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]\npath: ./datasets/osrs  # dataset root dir\ntrain: images/ # train images (relative to 'path') 128 images\nval: images/  # val images (relative to 'path') 128 images\ntest:  # test images (optional)\n\n# Classes\nnc: 1  # number of classes\nnames: ['cow']  # class names\n\n```\nTo start using \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/master/data/osrs.yaml\"\u003edata/osrs.yaml\u003c/a\u003e run \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/extract_osrs_zip.py\"\u003eextract_osrs_zip.py\u003c/a\u003e, this will unzip the cow.zip.001, cow.zip.002 and cow.zip.003 files\nand will create a folder in datasets osrs ready to train the osrs cow model.\n\n# Training\n\nEpochs. Start with 300 epochs. If this overfits early then you can reduce epochs. If overfitting does not occur after 300 epochs, train longer, i.e. 600, 1200 etc epochs.\n\nImage size. COCO trains at native resolution of --img 640, though due to the high amount of small objects in the dataset it can benefit from training at higher resolutions such as --img 1280. If there are many small objects then custom datasets will benefit from training at native or higher resolution. Best inference results are obtained at the same --img as the training was run at, i.e. if you train at --img 1280 you should also test and detect at --img 1280.\n\nBatch size. Use the largest --batch-size that your hardware allows for. Small batch sizes produce poor batchnorm statistics and should be avoided.\nIn the terminal type:\n\nFor more information and tips on datasets, model selection and training settings refer to: https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results\n\nThe setting i found useful with a GeForce GTX 1060 6GB gpu are as follows. In the terminal type:\n\n```python train.py --data osrs.yaml --weights yolov5s.pt --img 640``` \n  \n![image](https://user-images.githubusercontent.com/81003470/147907954-cd20c621-d848-49e3-83a5-45032ba768ba.png)\n  \nThis will run \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/train.py\"\u003etrain.py\u003c/a\u003e with the parameters mentioned above.\n\n![image](https://user-images.githubusercontent.com/81003470/147908948-10a1de98-4eb5-449e-8e24-92e421139b49.png)\n\n## Training Finished\n\nOnce finished the resulting model best.pt and last.pt will be saved in the folder runs/train/exp\u003cnumber\u003e\n  \n![image](https://user-images.githubusercontent.com/81003470/147910872-6700f739-232e-42f4-a210-479dd7c12734.png)\n\n# Detecting\n\nThis is where the detecting of objects take place, based on the parameters given, the code will run the default or custom weights and identify objects (inference) in \nimages, videos, directories, streams, etc.\n  \n## Test Dectections\n\nRun a test to ensure all is installed correctly, in the terminal type:\n\n```python detect.py --source data/images/bus.jpg --weights yolov5s.pt --img 640```\n\n![image](https://user-images.githubusercontent.com/81003470/148015379-5c099720-af00-425a-92b0-0d9e05545cd7.png)\n\nThis will run the default yolov5s weight file on the bus image and store the results in runs/detect/exp\n\nThese are the labels (the first integer is the class index and the rest are coordinates and bounding areas of the object)\n\n```\n5 0.502469 0.466204 0.995062 0.547222 # bus\n0 0.917284 0.59213 0.162963 0.450926 # person\n0 0.17284 0.603241 0.222222 0.469444 # person\n0 0.35 0.588889 0.146914 0.424074 # person\n```\n \nHere is the resulting image with bounding boxes identifying the bus and people:\n  \n![bus](https://user-images.githubusercontent.com/81003470/148015666-65439829-1856-435f-a8d0-eea7b9baade0.jpg)\n\n## Test Custom Detections (osrs cows model)\n\nMove the trained model located in runs/train/exp\u003cnumber\u003e to the parent folder (overwrite the previous best.pt):\n\n![image](https://user-images.githubusercontent.com/81003470/148020954-d42a32b0-b741-4791-8b69-300af762966d.png)\n  \nLet's see the results for osrs cow detection, to test in the terminal type:\n  \n``` python detect.py --source data/images/cow_osrs_test.png --weights best.pt --img 640```\n\nThe labels results are:\n```\n  0 0.946552 0.362295 0.062069 0.0754098\n0 0.398276 0.460656 0.106897 0.140984\n0 0.426724 0.572131 0.105172 0.140984\n0 0.352586 0.67377 0.122414 0.167213\n0 0.310345 0.898361 0.117241 0.190164\n0 0.151724 0.411475 0.062069 0.180328\n0 0.705172 0.37541 0.0689655 0.127869\n0 0.812931 0.319672 0.087931 0.127869\n  ```\nAnd here's the image result:\n  \n![image](https://user-images.githubusercontent.com/81003470/148021301-f7c58ad1-bc2e-43af-a82f-71a929d5a0cc.png)\n\n## Test Custom Detections (osrs cows model) on Monitor Display (screenshots with Pillow ImageGrab)\n\nFor a constant stream of the monitor display, in the terminal run:\n  \n```python detect.py --source screen --weights 46classes.pt --img 640```\n  \n\nThis will run \u003ca href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/detect.py\"\u003edetect.py\u003c/a\u003e with the default parameters listed below and can be changed to suit your needs.\n```\ndef run(\n        weights=ROOT / 'yolov5s.pt',  # model path or triton URL\n        source=ROOT / 'data/images',  # file/dir/URL/glob/screen/0(webcam)\n        data=ROOT / 'data/coco128.yaml',  # dataset.yaml path\n        imgsz=(640, 640),  # inference size (height, width)\n        conf_thres=0.25,  # confidence threshold\n        iou_thres=0.45,  # NMS IOU threshold\n        max_det=1000,  # maximum detections per image\n        device='',  # cuda device, i.e. 0 or 0,1,2,3 or cpu\n        view_img=False,  # show results\n        save_txt=False,  # save results to *.txt\n        save_conf=False,  # save confidences in --save-txt labels\n        save_crop=False,  # save cropped prediction boxes\n        nosave=False,  # do not save images/videos\n        classes=None,  # filter by class: --class 0, or --class 0 2 3\n        agnostic_nms=False,  # class-agnostic NMS\n        augment=False,  # augmented inference\n        visualize=False,  # visualize features\n        update=False,  # update all models\n        project=ROOT / 'runs/detect',  # save results to project/name\n        name='exp',  # save results to project/name\n        exist_ok=False,  # existing project/name ok, do not increment\n        line_thickness=3,  # bounding box thickness (pixels)\n        hide_labels=False,  # hide labels\n        hide_conf=False,  # hide confidences\n        half=False,  # use FP16 half-precision inference\n        dnn=False,  # use OpenCV DNN for ONNX inference\n        vid_stride=1,  # video frame-rate stride\n        attack=False,\n        window=0,\n        click_threshold= 0.9,\n):\n ```\nTo run on the captured screen and view the result add ```--view-img``` to the run script command.\n\n```python detect.py --source screen --weights 46classes.pt --img 640 --view-img```\n\nAdjust the size of the viewing window by adding ```--window size``` and the size (single int) and add to source in double quotes \"screen left top right bottom\".\n\n```python detect.py --source \"screen 0 0 800 800\" --weights 46classes.pt --img 640 --view-img --window 800```\n\nTo enable mouse clicks and for the computer to click on detected object add ```--attack```\n\n```python detect.py --source \"screen 0 0 800 800\" --weights 46classes.pt --img 640 --view-img --attack --window 800```\n\nTo change what confidence threshhold each mouse click will target to only get the highest confidence for detected objects add ```--click-threshold``` then a float number.\n\n```python detect.py --source \"screen 0 0 800 800\" --weights 46classes.pt --img 640 --view-img --attack --window 800 --click-threshold 0.8```\n  \n## Test Custom Detections (osrs cows model) on Data Capture\n  \nTo increase the fps (frames per second) and get a better detection rate, i use a hdmi data capture device. This takes a data stream of your monitor displays and sends the data like a webcam, which results in a significant increase in fps compared to taking screenshots of the screen:\n\nIn the terminal type:\n  \n```python detect.py --source 0 --weights best.pt --img 640```\n  \nSee below examples on amazon/aliexpress under $20:\n  \n-\u003ca target=\"_blank\" href=\"https://amzn.to/3MFoWbO\"\u003eAmazon - $11.07  - Audio Video Capture Cards HDMI to USB 3.0 1080P 4K Record Via DSLR Camcorder Action Cam for High Definition Acquisition, Live Broadcasting\u003c/a\u003e\n\n-\u003ca target=\"_blank\" href=\"https://s.click.aliexpress.com/e/_A43K9L\"\u003eAliexpress - $9.90 - 4K Video Capture Card USB 3.0 USB2.0 HDMI-compatible Grabber Recorder for PS4 Game DVD Camcorder Camera Recording Live Streaming\u003c/a\u003e\n  \n![image](https://user-images.githubusercontent.com/81003470/148019089-7c226fd6-382a-4da6-ab37-e8a66a3e838a.png)\n  \n## Retrain using Detections\n  \nBy default detect.py will not take labels and fullscreen images while the detection is running. \n\nEither change the default settings as mentioned above on detect.py from line 69 to 98 or add to the run script command:\n\nchange in the settings save_txt=True, save_crop=True or add the following to the run script:\n\n```--save-txt --save-crop```\n\nThese will be saved under runs/detect/exp\u003cnumber\u003e, full screenshots and labels in labels folder and cropped images in crops folder.\n\n![image](https://user-images.githubusercontent.com/81003470/148140979-4b7d1bd1-7e5f-4c58-8102-3e860f1b132b.png)\n\nTo save labels and images with detect_screenshots.py set save_text and save_crop to True.\n  \n![image](https://user-images.githubusercontent.com/81003470/148141081-8bfd281d-94fb-460c-9b6f-c34494e43553.png)\n\nMove the txt files (labels) to datasets/osrs/labels or datasets/[name of dataset]/labels.\n  \nMove the image files (crops) to datasets/osrs/images or datasets/[name of dataset]/images.\n\n![image](https://user-images.githubusercontent.com/81003470/148141478-95fcd5a0-f68c-4f88-b85f-5a17a97234b9.png)\n\nAs mentioned above follow the steps for \u003ca target=\"_blank\" href=\"https://github.com/slyautomation/osrs_yolov5/blob/main/README.md#training\"\u003etraining\u003c/a\u003e. This will retrain the object detection model with the detected images potentially increasing the accuracy and effectiveness.\n### Troubleshooting\n  \nRuntimeerror on Train.py: make sure there is enough hard drive storage space, the models will need approx 20 gbs of space to run smoothly.\n  \nRuntimeError on Detect.py: Couldn't load custom C++ ops. This can happen if your PyTorch and torchvision versions are incompatible, or if you had errors while compiling torchvision from source. For further information on the compatible versions, check https://github.com/pytorch/vision#installation for the compatibility matrix. Please check your PyTorch version with torch.__version__ and your torchvision version with torchvision.__version__ and verify if they are compatible and if not please reinstall torchvision so that it matches your PyTorch install.\n\nIf the above error occurs install a different version of pytorch and install the compatiable torchvision module. TBA i'll add a list of compatiable versions with this project with my gpu and cuda version.\n  \nRuntimeError: The size of tensor a (80) must match the size of tensor b (56) at non-singleton dimension 3.\n  \nThe weighted file is corrupted, use the following link and replace the weight file: https://github.com/ultralytics/yolov5/releases/download/v5.0/yolov5s.pt\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fslyautomation%2Fosrs_yolov5","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fslyautomation%2Fosrs_yolov5","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fslyautomation%2Fosrs_yolov5/lists"}