{"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_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"![image.png](attachment:image.png)","metadata":{},"attachments":{"image.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"The aim of this notebook is to install and run a training pipeline using MMDET Framework.","metadata":{}},{"cell_type":"markdown","source":"Dataset Source: https://www.kaggle.com/sreevishnudamodaran/vinbigdata-fusing-bboxes-coco-dataset","metadata":{}},{"cell_type":"markdown","source":"# MMDet Framework","metadata":{}},{"cell_type":"markdown","source":"MMDetection is an open source object detection toolbox based on PyTorch. It is a part of the OpenMMLab project developed by Multimedia Laboratory, CUHK.","metadata":{}},{"cell_type":"markdown","source":"# Features","metadata":{}},{"cell_type":"markdown","source":"1. **Modular Design**\n    1. The detection framework is decomposed into different components. This gives the flexiblity to construct a customized object detection framework using different backbones and models.\n    \n    2. The framework mainly contains following parts:\n    \n        1. Config: This is the place where you get to set the configurations for the framwork like data dirs, num of epochs, gpus to use etc.\n        2. mmdet: This module contains the files related to backbones, necks, heads and losses etc.\n        \n        3. Tools: This is the directory that contains utilities for training, testing and computing the evaluation metric.\n","metadata":{}},{"cell_type":"markdown","source":"2. **Multiple Frameworks**(https://github.com/open-mmlab/mmdetection)\n    1. List of Supported Backbones\n    \n        1. ResNet (CVPR'2016)\n        2. ResNeXt (CVPR'2017)\n        3. VGG (ICLR'2015)\n        4. HRNet (CVPR'2019)\n        5. RegNet (CVPR'2020)\n        6. Res2Net (TPAMI'2020)\n        7. ResNeSt (ArXiv'2020)\n    2. Supported Frameworks\n        1. [RPN (NeurIPS'2015)](https://github.com/open-mmlab/mmdetection/blob/master/configs/rpn)\n        2. [Fast R-CNN (ICCV'2015)](https://github.com/open-mmlab/mmdetection/blob/master/configs/fast_rcnn)\n        3. [Faster R-CNN (NeurIPS'2015)](https://github.com/open-mmlab/mmdetection/blob/master/configs/faster_rcnn)\n        4. [Mask R-CNN (ICCV'2017)](https://github.com/open-mmlab/mmdetection/blob/master/configs/mask_rcnn)\n        5. [Cascade R-CNN (CVPR'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/cascade_rcnn)\n        6. [Cascade Mask R-CNN (CVPR'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/cascade_rcnn)\n        7. [SSD (ECCV'2016)](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd)\n        8. [RetinaNet (ICCV'2017)](https://github.com/open-mmlab/mmdetection/blob/master/configs/retinanet)\n        9. [GHM (AAAI'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/ghm)\n        10. [Mask Scoring R-CNN (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/ms_rcnn)\n        11. [Double-Head R-CNN (CVPR'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/double_heads)\n        12. [Hybrid Task Cascade (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/htc)\n        13. [Libra R-CNN (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/libra_rcnn)\n        14. [Guided Anchoring (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/guided_anchoring)\n        15. [FCOS (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/fcos)\n        16. [RepPoints (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/reppoints)\n        17. [Foveabox (TIP'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/foveabox)\n        18. [FreeAnchor (NeurIPS'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/free_anchor)\n        19. [NAS-FPN (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/nas_fpn)\n        20. [ATSS (CVPR'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/atss)\n        21. [FSAF (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/fsaf)\n        22. [PAFPN (CVPR'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/pafpn)\n        23. [Dynamic R-CNN (ECCV'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/dynamic_rcnn)\n        24. [PointRend (CVPR'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/point_rend)\n        25. [CARAFE (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/carafe/README.md)\n        26. [DCNv2 (CVPR'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/dcn/README.md)\n        27. [Group Normalization (ECCV'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/gn/README.md)\n        28. [Weight Standardization (ArXiv'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/gn+ws/README.md)\n        29. [OHEM (CVPR'2016)](https://github.com/open-mmlab/mmdetection/blob/master/configs/faster_rcnn/faster_rcnn_r50_fpn_ohem_1x_coco.py)\n        30. [Soft-NMS (ICCV'2017)](https://github.com/open-mmlab/mmdetection/blob/master/configs/faster_rcnn/faster_rcnn_r50_fpn_soft_nms_1x_coco.py)\n        31. [Generalized Attention (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/empirical_attention/README.md)\n        32. [GCNet (ICCVW'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/gcnet/README.md)\n        33. [Mixed Precision (FP16) Training (ArXiv'2017)](https://github.com/open-mmlab/mmdetection/blob/master/configs/fp16/README.md)\n        34. [InstaBoost (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/instaboost/README.md)\n        35. [GRoIE (ICPR'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/groie/README.md)\n        36. [DetectoRS (ArXix'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/detectors/README.md)\n        37. [Generalized Focal Loss (NeurIPS'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/gfl/README.md)\n        38. [CornerNet (ECCV'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/cornernet/README.md)\n        39. [Side-Aware Boundary Localization (ECCV'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/sabl/README.md)\n        40. [YOLOv3 (ArXiv'2018)](https://github.com/open-mmlab/mmdetection/blob/master/configs/yolo/README.md)\n        41. [PAA (ECCV'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/paa/README.md)\n        42. [YOLACT (ICCV'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/yolact/README.md)\n        43. [CentripetalNet (CVPR'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/centripetalnet/README.md)\n        44. [VFNet (ArXix'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/vfnet/README.md)\n        45. [DETR (ECCV'2020)](https://github.com/open-mmlab/mmdetection/blob/master/configs/detr/README.md)\n        46. [CascadeRPN (NeurIPS'2019)](https://github.com/open-mmlab/mmdetection/blob/master/configs/cascade_rpn/README.md)\n        47. [SCNet (AAAI'2021)](https://github.com/open-mmlab/mmdetection/blob/master/configs/scnet/README.md)","metadata":{}},{"cell_type":"markdown","source":"3. **High Efficiency**\n    1. All basic bbox and mask operations run on GPUs. The training speed is faster than or comparable to other codebases, including Detectron2, maskrcnn-benchmark and SimpleDet.","metadata":{}},{"cell_type":"markdown","source":"# Installing MMDet Framework","metadata":{}},{"cell_type":"code","source":"# Check nvcc version\n!nvcc -V\n# Check GCC version\n!gcc --version","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:05.169616Z","iopub.execute_input":"2022-07-06T07:13:05.170236Z","iopub.status.idle":"2022-07-06T07:13:06.566582Z","shell.execute_reply.started":"2022-07-06T07:13:05.170114Z","shell.execute_reply":"2022-07-06T07:13:06.565511Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:06.570546Z","iopub.execute_input":"2022-07-06T07:13:06.574236Z","iopub.status.idle":"2022-07-06T07:13:06.580537Z","shell.execute_reply.started":"2022-07-06T07:13:06.574191Z","shell.execute_reply":"2022-07-06T07:13:06.579432Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Prereq: MMCV(~9 mins)","metadata":{}},{"cell_type":"code","source":"%%time\n\nprint(\"this will take around 9 mins\")\n# install dependencies: (use cu101 because colab has CUDA 10.1)\n# !pip install -U torch==1.7.0+cu101 torchvision==0.6.1+cu101 -f https://download.pytorch.org/whl/torch_stable.html\n\n# install mmcv-full thus we could use CUDA operators\n!pip install mmcv-full==1.2\n\n\n","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:06.582651Z","iopub.execute_input":"2022-07-06T07:13:06.583148Z","iopub.status.idle":"2022-07-06T07:13:12.092632Z","shell.execute_reply.started":"2022-07-06T07:13:06.582973Z","shell.execute_reply":"2022-07-06T07:13:12.091239Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Building MMDet From Source","metadata":{}},{"cell_type":"code","source":"!rm -rf mmdetection\n!git clone --branch v2.7.0 https://github.com/open-mmlab/mmdetection.git\n%cd mmdetection\n\n!pip install -e .\n\n# install Pillow 7.0.0 back in order to avoid bug in colab\n!pip install Pillow==7.0.0","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:12.094226Z","iopub.execute_input":"2022-07-06T07:13:12.094604Z","iopub.status.idle":"2022-07-06T07:13:41.469727Z","shell.execute_reply.started":"2022-07-06T07:13:12.094564Z","shell.execute_reply":"2022-07-06T07:13:41.468841Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Verifying Installation","metadata":{}},{"cell_type":"code","source":"# Check Pytorch installation\nimport torch, torchvision\nprint(torch.__version__, torch.cuda.is_available())\n\n# Check MMDetection installation\nimport mmdet\nprint(mmdet.__version__)\n\n# Check mmcv installation\nfrom mmcv.ops import get_compiling_cuda_version, get_compiler_version\nprint(get_compiling_cuda_version())\nprint(get_compiler_version())","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:41.473188Z","iopub.execute_input":"2022-07-06T07:13:41.473472Z","iopub.status.idle":"2022-07-06T07:13:41.479932Z","shell.execute_reply.started":"2022-07-06T07:13:41.473444Z","shell.execute_reply":"2022-07-06T07:13:41.478944Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Sample Inference Demo","metadata":{}},{"cell_type":"code","source":"!ls configs/fcos","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:41.480991Z","iopub.execute_input":"2022-07-06T07:13:41.481321Z","iopub.status.idle":"2022-07-06T07:13:42.164687Z","shell.execute_reply.started":"2022-07-06T07:13:41.481287Z","shell.execute_reply":"2022-07-06T07:13:42.163476Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from mmdet.apis import inference_detector, init_detector, show_result_pyplot\n\n# Choose to use a config and initialize the detector\n# config = 'configs/mask_rcnn/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco.py'\n\nconfig = 'configs/fcos/fcos_x101_64x4d_fpn_gn-head_mstrain_640-800_4x2_2x_coco.py'\n# Setup a checkpoint file to load\n# checkpoint = 'checkpoints/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco_bbox_mAP-0.408__segm_mAP-0.37_20200504_163245-42aa3d00.pth'\n# checkpoint = 'checkpoints/cascade_rcnn_x101_32x4d_fpn_1x_20190501-af628be5.pth'\n# initialize the detector\nmodel = init_detector(config, device='cuda:0')","metadata":{"execution":{"iopub.status.busy":"2022-07-06T07:13:42.169988Z","iopub.execute_input":"2022-07-06T07:13:42.170384Z","iopub.status.idle":"2022-07-06T07:13:50.155611Z","shell.execute_reply.started":"2022-07-06T07:13:42.170314Z","shell.execute_reply":"2022-07-06T07:13:50.154778Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# MMDet on VinBigData","metadata":{}},{"cell_type":"markdown","source":"I am using here faster rcnn for the demo purpose but as listed above the training pipeline can be customized using different framework and backbones. Just need to change the config settings down here. The config dir of mmdet framework contains various implmentation. Do checkit out.","metadata":{}},{"cell_type":"code","source":"from mmcv import Config\ncfg = Config.fromfile('./configs/fcos/fcos_x101_64x4d_fpn_gn-head_mstrain_640-800_4x2_2x_coco.py')\n\n","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:27.073387Z","iopub.execute_input":"2022-07-06T08:17:27.073723Z","iopub.status.idle":"2022-07-06T08:17:27.096243Z","shell.execute_reply.started":"2022-07-06T08:17:27.073693Z","shell.execute_reply":"2022-07-06T08:17:27.09543Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cfg","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:27.323422Z","iopub.execute_input":"2022-07-06T08:17:27.323837Z","iopub.status.idle":"2022-07-06T08:17:27.334374Z","shell.execute_reply.started":"2022-07-06T08:17:27.323797Z","shell.execute_reply":"2022-07-06T08:17:27.332988Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cfg.model.bbox_head","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:27.548332Z","iopub.execute_input":"2022-07-06T08:17:27.548719Z","iopub.status.idle":"2022-07-06T08:17:27.557979Z","shell.execute_reply.started":"2022-07-06T08:17:27.548681Z","shell.execute_reply":"2022-07-06T08:17:27.557108Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cfg.evaluation.metric","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:27.744565Z","iopub.execute_input":"2022-07-06T08:17:27.74482Z","iopub.status.idle":"2022-07-06T08:17:27.74976Z","shell.execute_reply.started":"2022-07-06T08:17:27.744795Z","shell.execute_reply":"2022-07-06T08:17:27.748793Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Configuration Settings On BaseConfig","metadata":{}},{"cell_type":"code","source":"from mmdet.apis import set_random_seed\n\ncfg.dataset_type = 'CocoDataset'\n","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:28.473101Z","iopub.execute_input":"2022-07-06T08:17:28.473442Z","iopub.status.idle":"2022-07-06T08:17:28.477729Z","shell.execute_reply.started":"2022-07-06T08:17:28.473411Z","shell.execute_reply":"2022-07-06T08:17:28.476851Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cfg.classes = (\"gametocyte\", \"schizont\", \"trophozoite\", \"ring\")\n\ncfg.data.train.img_prefix = '/kaggle/input/hcmdataset/datasets/HCM/train/100x/'\ncfg.data.train.classes = cfg.classes\ncfg.data.train.ann_file = '/kaggle/input/hcmdataset/datasets/Annotations/train_100x.json'\ncfg.data.train.type='CocoDataset'\n\n\ncfg.data.val.img_prefix = '/kaggle/input/hcmdataset/datasets/HCM/val/100x/'\ncfg.data.val.classes = cfg.classes\ncfg.data.val.ann_file = '/kaggle/input/hcmdataset/datasets/Annotations/val_100x.json'\ncfg.data.val.type='CocoDataset'\n\n\n\ncfg.data.test.img_prefix = '/kaggle/input/hcmdataset/datasets/HCM/test/100x/'\ncfg.data.test.classes = cfg.classes\ncfg.data.test.ann_file = '/kaggle/input/hcmdataset/datasets/Annotations/test_100x.json'\ncfg.data.test.type='CocoDataset'\n\n\n\n\n# cfg.model.roi_head.bbox_head.num_classes = 14\n\ncfg.model.bbox_head.num_classes = 4\n\n\ncfg.data.train.type = 'CocoDataset'\ncfg.data.val.type = 'CocoDataset'\ncfg.data.test.type = 'CocoDataset'\n\ncfg.optimizer.lr = 0.001\ncfg.lr_config.warmup = None\ncfg.log_config.interval = 5\n\n# Change the evaluation metric since we use customized dataset.\ncfg.evaluation.metric = 'bbox'\n# We can set the evaluation interval to reduce the evaluation times\ncfg.evaluation.interval = 2\n# We can set the checkpoint saving interval to reduce the storage cost\ncfg.checkpoint_config.interval = 5\n\n# Set seed thus the results are more reproducible\ncfg.seed = 0\nset_random_seed(0, deterministic=False)\ncfg.gpu_ids = range(1)\n\n# we can use here mask_rcnn.\n# cfg.load_from = './checkpoints/cascade_rcnn_x101_32x4d_fpn_1x_20190501-af628be5.pth'\ncfg.work_dir = \"../vinbig_output\"\n\n# One Epoch takes around 18 mins\ncfg.total_epochs = 6","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:50.45151Z","iopub.execute_input":"2022-07-06T08:17:50.451918Z","iopub.status.idle":"2022-07-06T08:17:50.462994Z","shell.execute_reply.started":"2022-07-06T08:17:50.451882Z","shell.execute_reply":"2022-07-06T08:17:50.461603Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# the dataset has been taken : https://www.kaggle.com/sreevishnudamodaran/vinbigdata-fusing-bboxes-coco-dataset\nimport os\n# os.listdir(\"../../input/hcmdataset/datasets/HCM/train/1000x\")","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:51.213797Z","iopub.execute_input":"2022-07-06T08:17:51.214219Z","iopub.status.idle":"2022-07-06T08:17:51.218531Z","shell.execute_reply.started":"2022-07-06T08:17:51.214176Z","shell.execute_reply":"2022-07-06T08:17:51.217381Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# for saving checkpoint and plots\nimport os\nos.makedirs('../vinbig_output')","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:51.759374Z","iopub.execute_input":"2022-07-06T08:17:51.759733Z","iopub.status.idle":"2022-07-06T08:17:51.763709Z","shell.execute_reply.started":"2022-07-06T08:17:51.759702Z","shell.execute_reply":"2022-07-06T08:17:51.762636Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f'Config:\\n{cfg.pretty_text}')","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:51.983227Z","iopub.execute_input":"2022-07-06T08:17:51.98355Z","iopub.status.idle":"2022-07-06T08:17:52.422489Z","shell.execute_reply.started":"2022-07-06T08:17:51.983521Z","shell.execute_reply":"2022-07-06T08:17:52.421563Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from mmdet.datasets import build_dataset\nfrom mmdet.models import build_detector\nfrom mmdet.apis import train_detector","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:52.424037Z","iopub.execute_input":"2022-07-06T08:17:52.424554Z","iopub.status.idle":"2022-07-06T08:17:52.429152Z","shell.execute_reply.started":"2022-07-06T08:17:52.424515Z","shell.execute_reply":"2022-07-06T08:17:52.428093Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Initializing Detector","metadata":{}},{"cell_type":"code","source":"model = build_detector(\n    cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg)\n\n","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:52.690582Z","iopub.execute_input":"2022-07-06T08:17:52.69084Z","iopub.status.idle":"2022-07-06T08:17:54.009256Z","shell.execute_reply.started":"2022-07-06T08:17:52.690815Z","shell.execute_reply":"2022-07-06T08:17:54.008442Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Building Dataset","metadata":{}},{"cell_type":"code","source":"datasets = [build_dataset(cfg.data.train)]\ndatasets","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:54.011448Z","iopub.execute_input":"2022-07-06T08:17:54.011794Z","iopub.status.idle":"2022-07-06T08:17:54.057567Z","shell.execute_reply.started":"2022-07-06T08:17:54.011758Z","shell.execute_reply":"2022-07-06T08:17:54.056653Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Training Detector With BBox Eval","metadata":{}},{"cell_type":"code","source":"# cfg.lr_config.policy='step'\ntrain_detector(model, datasets[0], cfg, distributed=False, validate=True)\n","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:17:55.571837Z","iopub.execute_input":"2022-07-06T08:17:55.572146Z","iopub.status.idle":"2022-07-06T09:03:42.363318Z","shell.execute_reply.started":"2022-07-06T08:17:55.572118Z","shell.execute_reply":"2022-07-06T09:03:42.362466Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"os.chdir('../')","metadata":{"execution":{"iopub.status.busy":"2022-07-06T08:06:08.650287Z","iopub.execute_input":"2022-07-06T08:06:08.650589Z","iopub.status.idle":"2022-07-06T08:06:08.654873Z","shell.execute_reply.started":"2022-07-06T08:06:08.650561Z","shell.execute_reply":"2022-07-06T08:06:08.654042Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Plotting Classification Loss","metadata":{}},{"cell_type":"markdown","source":"Check the output dir for the plot.","metadata":{}},{"cell_type":"code","source":"\n!python mmdetection/tools/analyze_logs.py plot_curve ./vinbig_output/None.log.json --keys loss_cls --legend loss_cls --out \"loss_cls\"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"!rm -rf \"./mmdetection\"","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}