{"cells":[{"metadata":{},"cell_type":"markdown","source":"# Multisession K-fold with checkpoints + Annotation Files Dataset","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"### What's the idea?\nThis notebook is a modification of [the following notebook](https://www.kaggle.com/cdeotte/triple-stratified-kfold-with-tfrecords) by @cdeotte. Using this one I realized that 3 hour runtime limit in Kaggle doesn't allow make the most of this notebook. I had to create a way which allows to bypass runtime limits. If we use K-fold CV we can't just save the weights. Instead of this we should save annotation files defining train, validation and test data for each fold. That's what I've done in this notebook. \n\n### What's the diffence?\nThe entire configuration remains the same. You have all the parameters by the same way. Than you should run a new block of code which creates an annotation files for each fold according to your number of folds and saves them to the ouptut of the notebook for further use in different sessions. The next block allows you to train folds one by one in diffent sessions.\n\n### Dataset \nFor your convenience, I have already created a dataset with annotation files for the case when we train 5 folds. It's based on TF Records Images with 512 * 512 resolution and additional data like ISIC 2019. [Here it is](https://www.kaggle.com/vadimtimakin/kfold-checkpoints). You can simply add it to your notebook's input and use it.\n\n### Comments\nI left most of the comments made by @cdeotte. However, I've deleted some of the them and added my own ones. I remind you that you can find the original version of this notebook [here](https://www.kaggle.com/cdeotte/triple-stratified-kfold-with-tfrecords). I've changed the number of steps as well for more ease of explanation instructions.\n\n### How to use it?\n\n1) Make the set up. **(Step 1)**;\n\n2) Create your own annotation files or just use my dataset (it must be added in input). In the last case set the 512 resolution of images and set the positive flag for each fold at INC2018 and INC2019 variables cause my dataset is based on 512 TFRecords images and on additional data like ISIC 2019. **(Step 5)**\n\n3) After that you can comment this block of code cause you won't need it anymore. **(Step 5)**\n\n4) If you decided to create your own annotation files then find the way to add them here and restart the runtime. You should change the paths of loading annotation files as well. **(Step 7)**\n\n5) Set the number of fold which are you going to train in this session (IT STARTS FROM ZERO). You can find it in the begging of the block of code. **(Step 7)**\n\n6) Train a fold. **(Step 7)**\n\n7) Create a submission file. **(Step 8)**\n\n8) Save the submission file from the output manually. MAKE SURE IN ADVANCE THAT YOU HAVE THE SAVE OUTPUT FLAG IN YOUR NOTEBOOK'S \"SAVE VERSION\"SETTINGS.\n\n9) Restart runtime.\n\n*After that you can repeat the steps specified in the instructions with another fold starting from the step 5. Finally, you will have the submission file for each fold and you will only have to average them.*\n\n`Feel free to discuss.`\n\n`Feel free to upvote this notebool if it was helpful for you.`\n\n`Feel free to upvote my dataset if it was helpful for you as well.`\n","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"# Triple Stratified KFold CV with TFRecords\nThis is a simple starter notebook for Kaggle's Melanoma Comp showing triple stratifed KFold with TFRecords. Triple stratified KFold is explained [here][2]. There are many configuration variables below to allow you to experiment. Use either GPU or TPU. You can control which size images are loaded, which efficientNets are used, and whether external data is used. You can experiment with different data augmentation, model architecture, loss, optimizers, and learning schedules. The TFRecords contain meta data, so you can input that into your CNN too. \n\n**NOTE:** this notebook lets you run a different experiment in each fold if you want to run lots of experiments. (Then it is like running multiple holdout validaiton experiments but in that case note that the overall CV score is meaningless because LB will be much higher when the multiple experiments are ensembled to predict test). **If you want a proper CV with a reliable overall CV score you need to choose the same configuration for each fold.**\n\nThis notebook follows the 5 (8 using checkpoints) step process presented in my \"How to compete with GPUs Workshop\" [here][1]. Some code sections have been reused from AgentAuers' great notebook [here][3]\n\n[1]: https://www.kaggle.com/cdeotte/how-to-compete-with-gpus-workshop\n[2]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/165526\n[3]: https://www.kaggle.com/agentauers/incredible-tpus-finetune-effnetb0-b6-at-once","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"# Initialize Environment","execution_count":null},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"!pip install -q efficientnet >> /dev/null","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import json\nimport pandas as pd, numpy as np\nfrom kaggle_datasets import KaggleDatasets\nimport tensorflow as tf, re, math\nimport tensorflow.keras.backend as K\nimport efficientnet.tfkeras as efn\nfrom sklearn.model_selection import KFold\nfrom sklearn.metrics import roc_auc_score\nimport matplotlib.pyplot as plt","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 1: Configuration\nIn order to be a proper cross validation with a meaningful overall CV score (aligned with LB score), **you need to choose the same** `IMG_SIZES`, `INC2019`, `INC2018`, and `EFF_NETS` **for each fold**. If your goal is to just run lots of experiments, then you can choose to have a different experiment in each fold. Then each fold is like a holdout validation experiment. When you find a configuration you like, you can use that configuration for all folds. \n* DEVICE - is GPU or TPU\n* SEED - a different seed produces a different triple stratified kfold split.\n* FOLDS - number of folds. Best set to 3, 5, or 15 but can be any number between 2 and 15\n* IMG_SIZES - is a Python list of length FOLDS. These are the image sizes to use each fold\n* INC2019 - This includes the new half of the 2019 competition data. The second half of the 2019 data is the comp data from 2018 plus 2017\n* INC2018 - This includes the second half of the 2019 competition data which is the comp data from 2018 plus 2017\n* BATCH_SIZES - is a list of length FOLDS. These are batch sizes for each fold. For maximum speed, it is best to use the largest batch size your GPU or TPU allows.\n* EPOCHS - is a list of length FOLDS. These are maximum epochs. Note that each fold, the best epoch model is saved and used. So if epochs is too large, it won't matter.\n* EFF_NETS - is a list of length FOLDS. These are the EfficientNets to use each fold. The number refers to the B. So a number of `0` refers to EfficientNetB0, and `1` refers to EfficientNetB1, etc.\n* WGTS - this should be `1/FOLDS` for each fold. This is the weight when ensembling the folds to predict the test set. If you want a weird ensemble, you can use different weights.\n* TTA - test time augmentation. Each test image is randomly augmented and predicted TTA times and the average prediction is used. TTA is also applied to OOF during validation.","execution_count":null},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"DEVICE = \"TPU\" #or \"GPU\"\n\n# USE DIFFERENT SEED FOR DIFFERENT STRATIFIED KFOLD\nSEED = 27\n\n# NUMBER OF FOLDS. USE 3, 5, OR 15 \nFOLDS = 5\n\n# WHICH IMAGE SIZES TO LOAD EACH FOLD\n# CHOOSE 128, 192, 256, 384, 512, 768 \nIMG_SIZES = [512,512,512,512,512]\n\n# INCLUDE OLD COMP DATA? YES=1 NO=0\nINC2019 = [1,1,1,1,1]\nINC2018 = [1,1,1,1,1]\n\n# BATCH SIZE AND EPOCHS\nBATCH_SIZES = [32]*FOLDS\nEPOCHS = [20]*FOLDS\n\n# WHICH EFFICIENTNET B? TO USE\nEFF_NETS = [6,6,6,6,6]\n\n# WEIGHTS FOR FOLD MODELS WHEN PREDICTING TEST\nWGTS = [1/FOLDS]*FOLDS\n\n# TEST TIME AUGMENTATION STEPS\nTTA = 11","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"if DEVICE == \"TPU\":\n    print(\"connecting to TPU...\")\n    try:\n        tpu = tf.distribute.cluster_resolver.TPUClusterResolver()\n        print('Running on TPU ', tpu.master())\n    except ValueError:\n        print(\"Could not connect to TPU\")\n        tpu = None\n\n    if tpu:\n        try:\n            print(\"initializing  TPU ...\")\n            tf.config.experimental_connect_to_cluster(tpu)\n            tf.tpu.experimental.initialize_tpu_system(tpu)\n            strategy = tf.distribute.experimental.TPUStrategy(tpu)\n            print(\"TPU initialized\")\n        except _:\n            print(\"failed to initialize TPU\")\n    else:\n        DEVICE = \"GPU\"\n\nif DEVICE != \"TPU\":\n    print(\"Using default strategy for CPU and single GPU\")\n    strategy = tf.distribute.get_strategy()\n\nif DEVICE == \"GPU\":\n    print(\"Num GPUs Available: \", len(tf.config.experimental.list_physical_devices('GPU')))\n    \n\nAUTO     = tf.data.experimental.AUTOTUNE\nREPLICAS = strategy.num_replicas_in_sync\nprint(f'REPLICAS: {REPLICAS}')","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 2: Preprocess\nPreprocess has already been done and saved to TFRecords. Here we choose which size to load. We can use either 128x128, 192x192, 256x256, 384x384, 512x512, 768x768 by changing the `IMG_SIZES` variable in the preceeding code section. These TFRecords are discussed [here][1]. The advantage of using different input sizes is discussed [here][2]\n\n[1]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/155579\n[2]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/160147","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"GCS_PATH = [None]*FOLDS; GCS_PATH2 = [None]*FOLDS\nfor i,k in enumerate(IMG_SIZES):\n    GCS_PATH[i] = KaggleDatasets().get_gcs_path('melanoma-%ix%i'%(k,k))\n    GCS_PATH2[i] = KaggleDatasets().get_gcs_path('isic2019-%ix%i'%(k,k))\nprint(GCS_PATH)\nprint()\nprint(GCS_PATH2)\nfiles_train = np.sort(np.array(tf.io.gfile.glob(GCS_PATH[0] + '/train*.tfrec')))\nfiles_test  = np.sort(np.array(tf.io.gfile.glob(GCS_PATH[0] + '/test*.tfrec')))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 3: Data Augmentation\nThis notebook uses rotation, sheer, zoom, shift augmentation first shown in this notebook [here][1] and successfully used in Melanoma comp by AgentAuers [here][2]. This notebook also uses horizontal flip, hue, saturation, contrast, brightness augmentation similar to last years winner and also similar to AgentAuers' notebook.\n\nAdditionally we can decide to use external data by changing the variables `INC2019` and `INC2018` in the preceeding code section. These variables respectively indicate whether to load last year 2019 data and/or year 2018 + 2017 data. These datasets are discussed [here][3]\n\nConsider experimenting with different augmenation and/or external data. The code to load TFRecords is taken from AgentAuers' notebook [here][2]. Thank you AgentAuers, this is great work.\n\n[1]: https://www.kaggle.com/cdeotte/rotation-augmentation-gpu-tpu-0-96\n[2]: https://www.kaggle.com/agentauers/incredible-tpus-finetune-effnetb0-b6-at-once\n[3]: https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion/164910","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"ROT_ = 180.0\nSHR_ = 2.0\nHZOOM_ = 8.0\nWZOOM_ = 8.0\nHSHIFT_ = 8.0\nWSHIFT_ = 8.0","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_mat(rotation, shear, height_zoom, width_zoom, height_shift, width_shift):\n    # returns 3x3 transformmatrix which transforms indicies\n        \n    # CONVERT DEGREES TO RADIANS\n    rotation = math.pi * rotation / 180.\n    shear    = math.pi * shear    / 180.\n\n    def get_3x3_mat(lst):\n        return tf.reshape(tf.concat([lst],axis=0), [3,3])\n    \n    # ROTATION MATRIX\n    c1   = tf.math.cos(rotation)\n    s1   = tf.math.sin(rotation)\n    one  = tf.constant([1],dtype='float32')\n    zero = tf.constant([0],dtype='float32')\n    \n    rotation_matrix = get_3x3_mat([c1,   s1,   zero, \n                                   -s1,  c1,   zero, \n                                   zero, zero, one])    \n    # SHEAR MATRIX\n    c2 = tf.math.cos(shear)\n    s2 = tf.math.sin(shear)    \n    \n    shear_matrix = get_3x3_mat([one,  s2,   zero, \n                                zero, c2,   zero, \n                                zero, zero, one])        \n    # ZOOM MATRIX\n    zoom_matrix = get_3x3_mat([one/height_zoom, zero,           zero, \n                               zero,            one/width_zoom, zero, \n                               zero,            zero,           one])    \n    # SHIFT MATRIX\n    shift_matrix = get_3x3_mat([one,  zero, height_shift, \n                                zero, one,  width_shift, \n                                zero, zero, one])\n    \n    return K.dot(K.dot(rotation_matrix, shear_matrix), \n                 K.dot(zoom_matrix,     shift_matrix))\n\n\ndef transform(image, DIM=256):    \n    # input image - is one image of size [dim,dim,3] not a batch of [b,dim,dim,3]\n    # output - image randomly rotated, sheared, zoomed, and shifted\n    XDIM = DIM%2 #fix for size 331\n    \n    rot = ROT_ * tf.random.normal([1], dtype='float32')\n    shr = SHR_ * tf.random.normal([1], dtype='float32') \n    h_zoom = 1.0 + tf.random.normal([1], dtype='float32') / HZOOM_\n    w_zoom = 1.0 + tf.random.normal([1], dtype='float32') / WZOOM_\n    h_shift = HSHIFT_ * tf.random.normal([1], dtype='float32') \n    w_shift = WSHIFT_ * tf.random.normal([1], dtype='float32') \n\n    # GET TRANSFORMATION MATRIX\n    m = get_mat(rot,shr,h_zoom,w_zoom,h_shift,w_shift) \n\n    # LIST DESTINATION PIXEL INDICES\n    x   = tf.repeat(tf.range(DIM//2, -DIM//2,-1), DIM)\n    y   = tf.tile(tf.range(-DIM//2, DIM//2), [DIM])\n    z   = tf.ones([DIM*DIM], dtype='int32')\n    idx = tf.stack( [x,y,z] )\n    \n    # ROTATE DESTINATION PIXELS ONTO ORIGIN PIXELS\n    idx2 = K.dot(m, tf.cast(idx, dtype='float32'))\n    idx2 = K.cast(idx2, dtype='int32')\n    idx2 = K.clip(idx2, -DIM//2+XDIM+1, DIM//2)\n    \n    # FIND ORIGIN PIXEL VALUES           \n    idx3 = tf.stack([DIM//2-idx2[0,], DIM//2-1+idx2[1,]])\n    d    = tf.gather_nd(image, tf.transpose(idx3))\n        \n    return tf.reshape(d,[DIM, DIM,3])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def read_labeled_tfrecord(example):\n    tfrec_format = {\n        'image'                        : tf.io.FixedLenFeature([], tf.string),\n        'image_name'                   : tf.io.FixedLenFeature([], tf.string),\n        'patient_id'                   : tf.io.FixedLenFeature([], tf.int64),\n        'sex'                          : tf.io.FixedLenFeature([], tf.int64),\n        'age_approx'                   : tf.io.FixedLenFeature([], tf.int64),\n        'anatom_site_general_challenge': tf.io.FixedLenFeature([], tf.int64),\n        'diagnosis'                    : tf.io.FixedLenFeature([], tf.int64),\n        'target'                       : tf.io.FixedLenFeature([], tf.int64)\n    }           \n    example = tf.io.parse_single_example(example, tfrec_format)\n    return example['image'], example['target']\n\n\ndef read_unlabeled_tfrecord(example, return_image_name):\n    tfrec_format = {\n        'image'                        : tf.io.FixedLenFeature([], tf.string),\n        'image_name'                   : tf.io.FixedLenFeature([], tf.string),\n    }\n    example = tf.io.parse_single_example(example, tfrec_format)\n    return example['image'], example['image_name'] if return_image_name else 0\n\n \ndef prepare_image(img, augment=True, dim=256):    \n    img = tf.image.decode_jpeg(img, channels=3)\n    img = tf.cast(img, tf.float32) / 255.0\n    \n    if augment:\n        img = transform(img,DIM=dim)\n        img = tf.image.random_flip_left_right(img)\n        #img = tf.image.random_hue(img, 0.01)\n        img = tf.image.random_saturation(img, 0.7, 1.3)\n        img = tf.image.random_contrast(img, 0.8, 1.2)\n        img = tf.image.random_brightness(img, 0.1)\n                      \n    img = tf.reshape(img, [dim,dim, 3])\n            \n    return img\n\ndef count_data_items(filenames):\n    n = [int(re.compile(r\"-([0-9]*)\\.\").search(filename).group(1)) \n         for filename in filenames]\n    return np.sum(n)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_dataset(files, augment = False, shuffle = False, repeat = False, \n                labeled=True, return_image_names=True, batch_size=16, dim=256):\n    \n    ds = tf.data.TFRecordDataset(files, num_parallel_reads=AUTO)\n    ds = ds.cache()\n    \n    if repeat:\n        ds = ds.repeat()\n    \n    if shuffle: \n        ds = ds.shuffle(1024*8)\n        opt = tf.data.Options()\n        opt.experimental_deterministic = False\n        ds = ds.with_options(opt)\n        \n    if labeled: \n        ds = ds.map(read_labeled_tfrecord, num_parallel_calls=AUTO)\n    else:\n        ds = ds.map(lambda example: read_unlabeled_tfrecord(example, return_image_names), \n                    num_parallel_calls=AUTO)      \n    \n    ds = ds.map(lambda img, imgname_or_label: (prepare_image(img, augment=augment, dim=dim), \n                                               imgname_or_label), \n                num_parallel_calls=AUTO)\n    \n    ds = ds.batch(batch_size * REPLICAS)\n    ds = ds.prefetch(AUTO)\n    return ds","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 4: Build Model\nThis is a common model architecute. Consider experimenting with different backbones, custom heads, losses, and optimizers. Also consider inputing meta features into your CNN.","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"EFNS = [efn.EfficientNetB0, efn.EfficientNetB1, efn.EfficientNetB2, efn.EfficientNetB3, \n        efn.EfficientNetB4, efn.EfficientNetB5, efn.EfficientNetB6]\n\ndef build_model(dim=128, ef=6):\n    inp = tf.keras.layers.Input(shape=(dim,dim,3))\n    base = EFNS[ef](input_shape=(dim,dim,3),weights='imagenet',include_top=False)\n    x = base(inp)\n    x = tf.keras.layers.GlobalAveragePooling2D()(x)\n    x = tf.keras.layers.Dense(1,activation='sigmoid')(x)\n    model = tf.keras.Model(inputs=inp,outputs=x)\n    opt = tf.keras.optimizers.Adam(learning_rate=0.001)\n    loss = tf.keras.losses.BinaryCrossentropy(label_smoothing=0.05) \n    model.compile(optimizer=opt,loss=loss,metrics=['AUC'])\n    return model","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 5: Train Schedule\nThis is a common train schedule for transfer learning. The learning rate starts near zero, then increases to a maximum, then decays over time. Consider changing the schedule and/or learning rates. Note how the learning rate max is larger with larger batches sizes. This is a good practice to follow.","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_lr_callback(batch_size=8):\n    lr_start   = 0.000005\n    lr_max     = 0.00000125 * REPLICAS * batch_size\n    lr_min     = 0.000001\n    lr_ramp_ep = 5\n    lr_sus_ep  = 0\n    lr_decay   = 0.8\n   \n    def lrfn(epoch):\n        if epoch < lr_ramp_ep:\n            lr = (lr_max - lr_start) / lr_ramp_ep * epoch + lr_start\n            \n        elif epoch < lr_ramp_ep + lr_sus_ep:\n            lr = lr_max\n            \n        else:\n            lr = (lr_max - lr_min) * lr_decay**(epoch - lr_ramp_ep - lr_sus_ep) + lr_min\n            \n        return lr\n\n    lr_callback = tf.keras.callbacks.LearningRateScheduler(lrfn, verbose=False)\n    return lr_callback","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 6: creating checkpoints","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"skf = KFold(n_splits=FOLDS,shuffle=True,random_state=SEED)\n\nfor fold,(idxT,idxV) in enumerate(skf.split(np.arange(15))):\n    \n    # DISPLAY FOLD INFO\n    if DEVICE=='TPU':\n        if tpu: tf.tpu.experimental.initialize_tpu_system(tpu)\n    print('#'*25); print('#### FOLD',fold+1)\n    print('#### Image Size %i with EfficientNet B%i and batch_size %i'%\n          (IMG_SIZES[fold],EFF_NETS[fold],BATCH_SIZES[fold]*REPLICAS))\n    \n    # CREATE TRAIN AND VALIDATION SUBSETS\n    files_train = tf.io.gfile.glob([GCS_PATH[fold] + '/train%.2i*.tfrec'%x for x in idxT])\n    if INC2019[fold]:\n        files_train += tf.io.gfile.glob([GCS_PATH2[fold] + '/train%.2i*.tfrec'%x for x in idxT*2+1])\n        print('#### Using 2019 external data')\n    if INC2018[fold]:\n        files_train += tf.io.gfile.glob([GCS_PATH2[fold] + '/train%.2i*.tfrec'%x for x in idxT*2])\n        print('#### Using 2018+2017 external data')\n    np.random.shuffle(files_train); print('#'*25)\n    files_valid = tf.io.gfile.glob([GCS_PATH[fold] + '/train%.2i*.tfrec'%x for x in idxV])\n    files_test = np.sort(np.array(tf.io.gfile.glob(GCS_PATH[fold] + '/test*.tfrec')))\n    \n    # SAVE ANNOTATION FILES\n    with open(f\"train{fold}.json\", \"w\") as write_file:\n        json.dump(files_train, write_file)\n\n    with open(f\"valid{fold}.json\", \"w\") as write_file:\n        json.dump(files_valid, write_file)\n\n    np.save(f\"test{fold}\", files_test)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 7: Train Model\nOur model will be trained for the number of FOLDS and EPOCHS you chose in the configuration above. Each fold the model with lowest validation loss will be saved and used to predict OOF and test. Adjust the variables `VERBOSE` and `DISPLOY_PLOT` below to determine what output you want displayed. The variable `VERBOSE=1 or 2` will display the training and validation loss and auc for each epoch as text. The variable `DISPLAY_PLOT` shows this information as a plot. ","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"# SET THE NUMBER OF YOUR FOLD \nfold = 3\n\n# USE VERBOSE=0 for silent, VERBOSE=1 for interactive, VERBOSE=2 for commit\nVERBOSE = 2\nDISPLAY_PLOT = True\n\nskf = KFold(n_splits=FOLDS,shuffle=True,random_state=SEED)\noof_pred = []; oof_tar = []; oof_val = []; oof_names = []; oof_folds = [] \npreds = np.zeros((count_data_items(files_test),1))\n    \n# DISPLAY FOLD INFO\nif DEVICE=='TPU':\n    if tpu: tf.tpu.experimental.initialize_tpu_system(tpu)\nprint('#'*25); print('#### FOLD',fold+1)\nprint('#### Image Size %i with EfficientNet B%i and batch_size %i'%\n      (IMG_SIZES[fold],EFF_NETS[fold],BATCH_SIZES[fold]*REPLICAS))\n\n# CREATE TRAIN AND VALIDATION SUBSETS\n\nwith open(f\"../input/kfold-checkpoints/train{fold}.json\", \"r\") as write_file:\n    files_train = json.load(write_file)\n\nwith open(f\"../input/kfold-checkpoints/valid{fold}.json\", \"r\") as write_file:\n    files_valid = json.load(write_file)\n\nfiles_test = np.load(f\"../input/kfold-checkpoints/test{fold}.npy\")\n\n\n# BUILD MODEL\nK.clear_session()\nwith strategy.scope():\n    model = build_model(dim=IMG_SIZES[fold],ef=EFF_NETS[fold])\n\n# SAVE BEST MODEL EACH FOLD\nsv = tf.keras.callbacks.ModelCheckpoint(\n    'fold-%i.h5'%fold, monitor='val_loss', verbose=0, save_best_only=True,\n    save_weights_only=True, mode='min', save_freq='epoch')\n\n# TRAIN\nprint('Training...')\nhistory = model.fit(\n    get_dataset(files_train, augment=True, shuffle=True, repeat=True,\n            dim=IMG_SIZES[fold], batch_size = BATCH_SIZES[fold]), \n    epochs=EPOCHS[fold], callbacks = [sv,get_lr_callback(BATCH_SIZES[fold])], \n    steps_per_epoch=count_data_items(files_train)/BATCH_SIZES[fold]//REPLICAS,\n    validation_data=get_dataset(files_valid,augment=False,shuffle=False,\n            repeat=False,dim=IMG_SIZES[fold]), #class_weight = {0:1,1:2},\n    verbose=VERBOSE\n)\n\nprint('Loading best model...')\nmodel.load_weights('fold-%i.h5'%fold)\n\n# PREDICT OOF USING TTA\nprint('Predicting OOF with TTA...')\nds_valid = get_dataset(files_valid,labeled=False,return_image_names=False,augment=True,\n        repeat=True,shuffle=False,dim=IMG_SIZES[fold],batch_size=BATCH_SIZES[fold]*4)\nct_valid = count_data_items(files_valid); STEPS = TTA * ct_valid/BATCH_SIZES[fold]/4/REPLICAS\npred = model.predict(ds_valid,steps=STEPS,verbose=VERBOSE)[:TTA*ct_valid,] \noof_pred.append( np.mean(pred.reshape((ct_valid,TTA),order='F'),axis=1) )                 \n\n# GET OOF TARGETS AND NAMES\nds_valid = get_dataset(files_valid, augment=False, repeat=False, dim=IMG_SIZES[fold],\n        labeled=True, return_image_names=True)\noof_tar.append( np.array([target.numpy() for img, target in iter(ds_valid.unbatch())]) )\noof_folds.append( np.ones_like(oof_tar[-1],dtype='int8')*fold )\nds = get_dataset(files_valid, augment=False, repeat=False, dim=IMG_SIZES[fold],\n            labeled=False, return_image_names=True)\noof_names.append( np.array([img_name.numpy().decode(\"utf-8\") for img, img_name in iter(ds.unbatch())]))\n\n# PREDICT TEST USING TTA\nprint('Predicting Test with TTA...')\nds_test = get_dataset(files_test,labeled=False,return_image_names=False,augment=True,\n        repeat=True,shuffle=False,dim=IMG_SIZES[fold],batch_size=BATCH_SIZES[fold]*4)\nct_test = count_data_items(files_test); STEPS = TTA * ct_test/BATCH_SIZES[fold]/4/REPLICAS\npred = model.predict(ds_test,steps=STEPS,verbose=VERBOSE)[:TTA*ct_test,] \npreds[:,0] += np.mean(pred.reshape((ct_test,TTA),order='F'),axis=1) * WGTS[fold]\n\n# REPORT RESULTS\nauc = roc_auc_score(oof_tar[-1],oof_pred[-1])\noof_val.append(np.max( history.history['val_auc'] ))\nprint('#### FOLD %i OOF AUC without TTA = %.3f, with TTA = %.3f'%(fold+1,oof_val[-1],auc))\n\n# PLOT TRAINING\nif DISPLAY_PLOT:\n    plt.figure(figsize=(15,5))\n    plt.plot(np.arange(EPOCHS[fold]),history.history['auc'],'-o',label='Train AUC',color='#ff7f0e')\n    plt.plot(np.arange(EPOCHS[fold]),history.history['val_auc'],'-o',label='Val AUC',color='#1f77b4')\n    x = np.argmax( history.history['val_auc'] ); y = np.max( history.history['val_auc'] )\n    xdist = plt.xlim()[1] - plt.xlim()[0]; ydist = plt.ylim()[1] - plt.ylim()[0]\n    plt.scatter(x,y,s=200,color='#1f77b4'); plt.text(x-0.03*xdist,y-0.13*ydist,'max auc\\n%.2f'%y,size=14)\n    plt.ylabel('AUC',size=14); plt.xlabel('Epoch',size=14)\n    plt.legend(loc=2)\n    plt2 = plt.gca().twinx()\n    plt2.plot(np.arange(EPOCHS[fold]),history.history['loss'],'-o',label='Train Loss',color='#2ca02c')\n    plt2.plot(np.arange(EPOCHS[fold]),history.history['val_loss'],'-o',label='Val Loss',color='#d62728')\n    x = np.argmin( history.history['val_loss'] ); y = np.min( history.history['val_loss'] )\n    ydist = plt.ylim()[1] - plt.ylim()[0]\n    plt.scatter(x,y,s=200,color='#d62728'); plt.text(x-0.03*xdist,y+0.05*ydist,'min loss',size=14)\n    plt.ylabel('Loss',size=14)\n    plt.title('FOLD %i - Image Size %i, EfficientNet B%i, inc2019=%i, inc2018=%i'%\n            (fold+1,IMG_SIZES[fold],EFF_NETS[fold],INC2019[fold],INC2018[fold]),size=18)\n    plt.legend(loc=3)\n    plt.show()  ","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Step 8: save the CSV-file.","execution_count":null},{"metadata":{"trusted":true,"_kg_hide-output":false},"cell_type":"code","source":"ds = get_dataset(files_test, augment=False, repeat=False, dim=IMG_SIZES[fold],\n                 labeled=False, return_image_names=True)\nimage_names = np.array([img_name.numpy().decode(\"utf-8\") \n                        for img, img_name in iter(ds.unbatch())])\n\nsubmission = pd.DataFrame(dict(image_name=image_names, target=preds[:,0]))\nsubmission = submission.sort_values('image_name') \nsubmission.to_csv(f'submission{fold}.csv', index=False)\nsubmission.head()","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}