{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.12.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"sourceType":"competition","sourceId":14774,"databundleVersionId":875431}],"dockerImageVersionId":31329,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# ==========================================================\n# Early Detection & Severity Classification of DR\n# ==========================================================\n\nimport os, cv2, numpy as np, pandas as pd, matplotlib.pyplot as plt, seaborn as sns, tensorflow as tf\nfrom sklearn.model_selection import StratifiedKFold\nfrom sklearn.metrics import confusion_matrix, classification_report, roc_auc_score, roc_curve, accuracy_score\nfrom sklearn.utils.class_weight import compute_class_weight\nfrom tensorflow.keras import layers, models\nfrom tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, ReduceLROnPlateau\nfrom tensorflow.keras.utils import to_categorical\nfrom tqdm import tqdm\n\n# ----------------------------------------------------------\n# GPU CHECK\n# ----------------------------------------------------------\nprint(\"TensorFlow:\", tf.__version__)\nprint(\"GPUs:\", tf.config.list_physical_devices('GPU'))\n\n# Multi-GPU Strategy\nstrategy = tf.distribute.MirroredStrategy()\nprint(\"Number of devices:\", strategy.num_replicas_in_sync)\n\n# ----------------------------------------------------------\n# PATHS\n# ----------------------------------------------------------\nCSV_PATH = \"/kaggle/input/competitions/aptos2019-blindness-detection/train.csv\"\nIMG_DIR  = \"/kaggle/input/competitions/aptos2019-blindness-detection/train_images\"\n\nIMG_SIZE = 224\nSEED = 42\nnp.random.seed(SEED)\ntf.random.set_seed(SEED)\n\n# ----------------------------------------------------------\n# LOAD DATA\n# ----------------------------------------------------------\ndf = pd.read_csv(CSV_PATH)\ndf[\"path\"] = df[\"id_code\"].apply(lambda x: f\"{IMG_DIR}/{x}.png\")\n\nprint(df.head())\nprint(df[\"diagnosis\"].value_counts())\n\n# ----------------------------------------------------------\n# PREPROCESSING\n# ----------------------------------------------------------\ndef preprocess_image(path):\n    img = cv2.imread(path)\n    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n    img = cv2.resize(img, (IMG_SIZE, IMG_SIZE))\n\n    # CLAHE\n    lab = cv2.cvtColor(img, cv2.COLOR_RGB2LAB)\n    l, a, b = cv2.split(lab)\n    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))\n    l = clahe.apply(l)\n    lab = cv2.merge((l,a,b))\n    img = cv2.cvtColor(lab, cv2.COLOR_LAB2RGB)\n\n    # Gaussian Blur\n    img = cv2.GaussianBlur(img, (3,3), 0)\n\n    # Normalize\n    img = img.astype(\"float32\") / 255.0\n    return img\n\n# ----------------------------------------------------------\n# LOAD IMAGES INTO MEMORY\n# ----------------------------------------------------------\nX = np.zeros((len(df), IMG_SIZE, IMG_SIZE, 3), dtype=np.float32)\ny = df[\"diagnosis\"].values\n\nfor i, path in enumerate(tqdm(df[\"path\"])):\n    X[i] = preprocess_image(path)\n\n# ----------------------------------------------------------\n# LABELS\n# ----------------------------------------------------------\n# Binary: 0 vs DR\ny_binary = np.where(y == 0, 0, 1)\n\n# Early Detection: 0=No DR, 1=Early(1,2), 2=Severe(3,4)\ndef map_early(v):\n    if v == 0:\n        return 0\n    elif v in [1,2]:\n        return 1\n    else:\n        return 2\n\ny_early = np.array([map_early(v) for v in y])\n\n# Severity: Original 5 classes\ny_severity = y.copy()\n\n# ----------------------------------------------------------\n# AUGMENTATION\n# ----------------------------------------------------------\naugmenter = tf.keras.Sequential([\n    layers.RandomFlip(\"horizontal\"),\n    layers.RandomRotation(0.08),\n    layers.RandomZoom(0.08),\n    layers.RandomContrast(0.1)\n])\n\n# ----------------------------------------------------------\n# PAPER CNN ARCHITECTURE\n# ----------------------------------------------------------\ndef build_model(num_classes):\n    inputs = layers.Input(shape=(IMG_SIZE, IMG_SIZE, 3))\n\n    x = layers.Conv2D(32, (5,5), padding=\"same\")(inputs)\n    x = layers.PReLU()(x)\n    x = layers.MaxPooling2D()(x)\n\n    x = layers.Conv2D(64, (5,5), activation=\"relu\", padding=\"same\")(x)\n    x = layers.MaxPooling2D()(x)\n\n    x = layers.Conv2D(128, (5,5), activation=\"relu\", padding=\"same\")(x)\n    x = layers.MaxPooling2D()(x)\n\n    x = layers.Conv2D(256, (5,5), activation=\"relu\", padding=\"same\")(x)\n    x = layers.MaxPooling2D()(x)\n\n    x = layers.GlobalAveragePooling2D()(x)\n\n    x = layers.Dense(256, activation=\"relu\")(x)\n    x = layers.Dropout(0.3)(x)\n\n    x = layers.Dense(128, activation=\"relu\")(x)\n    x = layers.Dropout(0.3)(x)\n\n    x = layers.Dense(64, activation=\"relu\")(x)\n\n    outputs = layers.Dense(num_classes, activation=\"softmax\")(x)\n\n    model = models.Model(inputs, outputs)\n    return model\n\n# ----------------------------------------------------------\n# TRAIN FUNCTION\n# ----------------------------------------------------------\ndef run_experiment(X, y_labels, num_classes, epochs, batch_size, lr, name):\n\n    print(f\"\\n========== {name} ==========\")\n\n    skf = StratifiedKFold(n_splits=5, shuffle=True, random_state=SEED)\n    scores = []\n\n    all_true = []\n    all_pred = []\n\n    for fold, (train_idx, val_idx) in enumerate(skf.split(X, y_labels), 1):\n\n        print(f\"\\n--- Fold {fold} ---\")\n\n        X_train, X_val = X[train_idx], X[val_idx]\n        y_train, y_val = y_labels[train_idx], y_labels[val_idx]\n\n        y_train_cat = to_categorical(y_train, num_classes)\n        y_val_cat   = to_categorical(y_val, num_classes)\n\n        # Class weights\n        weights = compute_class_weight(\n            class_weight=\"balanced\",\n            classes=np.unique(y_train),\n            y=y_train\n        )\n        class_weights = dict(enumerate(weights))\n\n        # tf.data\n        train_ds = tf.data.Dataset.from_tensor_slices((X_train, y_train_cat))\n        train_ds = train_ds.shuffle(1024).batch(batch_size)\n        train_ds = train_ds.map(lambda a,b: (augmenter(a, training=True), b))\n        train_ds = train_ds.prefetch(tf.data.AUTOTUNE)\n\n        val_ds = tf.data.Dataset.from_tensor_slices((X_val, y_val_cat))\n        val_ds = val_ds.batch(batch_size).prefetch(tf.data.AUTOTUNE)\n\n        # Multi-GPU model build\n        with strategy.scope():\n            model = build_model(num_classes)\n            model.compile(\n                optimizer=tf.keras.optimizers.Adam(learning_rate=lr),\n                loss=\"categorical_crossentropy\",\n                metrics=[\"accuracy\"]\n            )\n\n        callbacks = [\n            ModelCheckpoint(f\"{name}_fold{fold}.keras\", save_best_only=True),\n            EarlyStopping(patience=5, restore_best_weights=True),\n            ReduceLROnPlateau(patience=2, factor=0.5)\n        ]\n\n        model.fit(\n            train_ds,\n            validation_data=val_ds,\n            epochs=epochs,\n            class_weight=class_weights,\n            callbacks=callbacks,\n            verbose=1\n        )\n\n        preds = model.predict(val_ds, verbose=0)\n        pred_labels = np.argmax(preds, axis=1)\n\n        acc = accuracy_score(y_val, pred_labels)\n        scores.append(acc)\n\n        all_true.extend(y_val)\n        all_pred.extend(pred_labels)\n\n        print(\"Fold Accuracy:\", acc)\n\n        # Confusion Matrix\n        cm = confusion_matrix(y_val, pred_labels)\n        plt.figure(figsize=(6,5))\n        sns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\")\n        plt.title(f\"{name} Fold {fold}\")\n        plt.show()\n\n        # ROC for binary\n        if num_classes == 2:\n            auc = roc_auc_score(y_val, preds[:,1])\n            fpr, tpr, _ = roc_curve(y_val, preds[:,1])\n\n            plt.figure(figsize=(6,5))\n            plt.plot(fpr, tpr, label=f\"AUC = {auc:.3f}\")\n            plt.plot([0,1],[0,1],'--')\n            plt.legend()\n            plt.title(\"ROC Curve\")\n            plt.show()\n\n    print(\"\\nAverage Accuracy:\", np.mean(scores))\n    print(\"\\nClassification Report:\")\n    print(classification_report(all_true, all_pred))\n\n# ----------------------------------------------------------\n# RUN ALL 3 EXPERIMENTS\n# ----------------------------------------------------------\n\n# 1 Binary Classification\nrun_experiment(\n    X, y_binary,\n    num_classes=2,\n    epochs=30,\n    batch_size=32,\n    lr=1e-4,\n    name=\"Binary\"\n)\n\n# 2 Early Detection\nrun_experiment(\n    X, y_early,\n    num_classes=3,\n    epochs=40,\n    batch_size=16,\n    lr=1e-3,\n    name=\"Early\"\n)\n\n# 3 Severity Classification\nrun_experiment(\n    X, y_severity,\n    num_classes=5,\n    epochs=40,\n    batch_size=8,\n    lr=1e-4,\n    name=\"Severity\"\n)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2026-04-15T20:29:11.246476Z","iopub.execute_input":"2026-04-15T20:29:11.247306Z","iopub.status.idle":"2026-04-16T00:32:41.748035Z","shell.execute_reply.started":"2026-04-15T20:29:11.24727Z","shell.execute_reply":"2026-04-16T00:32:41.7472Z"}},"outputs":[{"name":"stdout","text":"TensorFlow: 2.19.0\nGPUs: [PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU'), PhysicalDevice(name='/physical_device:GPU:1', device_type='GPU')]\nINFO:tensorflow:Using MirroredStrategy with devices ('/job:localhost/replica:0/task:0/device:GPU:0', '/job:localhost/replica:0/task:0/device:GPU:1')\nNumber of devices: 2\n        id_code  diagnosis                                               path\n0  000c1434d8d7          2  /kaggle/input/competitions/aptos2019-blindness...\n1  001639a390f0          4  /kaggle/input/competitions/aptos2019-blindness...\n2  0024cdab0c1e          1  /kaggle/input/competitions/aptos2019-blindness...\n3  002c21358ce6          0  /kaggle/input/competitions/aptos2019-blindness...\n4  005b95c28852          0  /kaggle/input/competitions/aptos2019-blindness...\ndiagnosis\n0    1805\n2     999\n1     370\n4     295\n3     193\nName: count, dtype: int64\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3662/3662 [07:27<00:00,  8.18it/s]\n","output_type":"stream"},{"name":"stdout","text":"\n========== Binary ==========\n\n--- Fold 1 ---\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nINFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\nEpoch 1/30\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n","output_type":"stream"},{"name":"stderr","text":"I0000 00:00:1776285417.897487     129 cuda_dnn.cc:529] Loaded cuDNN version 91002\nI0000 00:00:1776285417.900761     130 cuda_dnn.cc:529] Loaded cuDNN version 91002\n","output_type":"stream"},{"name":"stdout","text":"\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 382ms/step - accuracy: 0.5303 - loss: 0.6911INFO:tensorflow:Reduce to /job:localhost/replica:0/task:0/device:CPU:0 then broadcast to ('/job:localhost/replica:0/task:0/device:CPU:0',).\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m53s\u001b[0m 449ms/step - accuracy: 0.5303 - loss: 0.6911 - val_accuracy: 0.4925 - val_loss: 0.6973 - learning_rate: 1.0000e-04\nEpoch 2/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.6345 - loss: 0.6341 - val_accuracy: 0.8786 - val_loss: 0.3227 - learning_rate: 1.0000e-04\nEpoch 3/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.8468 - loss: 0.3707 - val_accuracy: 0.9004 - val_loss: 0.2430 - learning_rate: 1.0000e-04\nEpoch 4/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9108 - loss: 0.2488 - val_accuracy: 0.8922 - val_loss: 0.2525 - learning_rate: 1.0000e-04\nEpoch 5/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9327 - loss: 0.2166 - val_accuracy: 0.9332 - val_loss: 0.2114 - learning_rate: 1.0000e-04\nEpoch 6/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 475ms/step - accuracy: 0.9284 - loss: 0.2195 - val_accuracy: 0.9318 - val_loss: 0.2025 - learning_rate: 1.0000e-04\nEpoch 7/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 466ms/step - accuracy: 0.9337 - loss: 0.2019 - val_accuracy: 0.9332 - val_loss: 0.2106 - learning_rate: 1.0000e-04\nEpoch 8/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9403 - loss: 0.1968 - val_accuracy: 0.9332 - val_loss: 0.2114 - learning_rate: 1.0000e-04\nEpoch 9/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9411 - loss: 0.1963 - val_accuracy: 0.9304 - val_loss: 0.2139 - learning_rate: 5.0000e-05\nEpoch 10/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9303 - loss: 0.2017 - val_accuracy: 0.9332 - val_loss: 0.2057 - learning_rate: 5.0000e-05\nEpoch 11/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9406 - loss: 0.1911 - val_accuracy: 0.9345 - val_loss: 0.2023 - learning_rate: 2.5000e-05\nEpoch 12/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 465ms/step - accuracy: 0.9377 - loss: 0.1936 - val_accuracy: 0.9345 - val_loss: 0.2047 - learning_rate: 2.5000e-05\nEpoch 13/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9400 - loss: 0.1940 - val_accuracy: 0.9345 - val_loss: 0.2010 - learning_rate: 2.5000e-05\nEpoch 14/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 465ms/step - accuracy: 0.9393 - loss: 0.1865 - val_accuracy: 0.9304 - val_loss: 0.2055 - learning_rate: 2.5000e-05\nEpoch 15/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9493 - loss: 0.1695 - val_accuracy: 0.9345 - val_loss: 0.2028 - learning_rate: 2.5000e-05\nEpoch 16/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 476ms/step - accuracy: 0.9409 - loss: 0.1792 - val_accuracy: 0.9345 - val_loss: 0.1980 - learning_rate: 1.2500e-05\nEpoch 17/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9425 - loss: 0.1840 - val_accuracy: 0.9291 - val_loss: 0.2047 - learning_rate: 1.2500e-05\nEpoch 18/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 474ms/step - accuracy: 0.9450 - loss: 0.1729 - val_accuracy: 0.9304 - val_loss: 0.2099 - learning_rate: 1.2500e-05\nEpoch 19/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9431 - loss: 0.1810 - val_accuracy: 0.9332 - val_loss: 0.2002 - learning_rate: 6.2500e-06\nEpoch 20/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9398 - loss: 0.1879 - val_accuracy: 0.9345 - val_loss: 0.1985 - learning_rate: 6.2500e-06\nEpoch 21/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9425 - loss: 0.1776 - val_accuracy: 0.9345 - val_loss: 0.1999 - learning_rate: 3.1250e-06\nFold Accuracy: 0.9345156889495225\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 2 ---\nEpoch 1/30\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 483ms/step - accuracy: 0.5186 - loss: 0.6927 - val_accuracy: 0.6276 - val_loss: 0.6622 - learning_rate: 1.0000e-04\nEpoch 2/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 476ms/step - accuracy: 0.6608 - loss: 0.6329 - val_accuracy: 0.8881 - val_loss: 0.3342 - learning_rate: 1.0000e-04\nEpoch 3/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.8368 - loss: 0.3892 - val_accuracy: 0.8813 - val_loss: 0.2650 - learning_rate: 1.0000e-04\nEpoch 4/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.8851 - loss: 0.2834 - val_accuracy: 0.8990 - val_loss: 0.2374 - learning_rate: 1.0000e-04\nEpoch 5/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9165 - loss: 0.2346 - val_accuracy: 0.9359 - val_loss: 0.2020 - learning_rate: 1.0000e-04\nEpoch 6/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9180 - loss: 0.2398 - val_accuracy: 0.9250 - val_loss: 0.2139 - learning_rate: 1.0000e-04\nEpoch 7/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9338 - loss: 0.1937 - val_accuracy: 0.9345 - val_loss: 0.1988 - learning_rate: 1.0000e-04\nEpoch 8/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 476ms/step - accuracy: 0.9383 - loss: 0.2048 - val_accuracy: 0.9359 - val_loss: 0.2053 - learning_rate: 1.0000e-04\nEpoch 9/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9376 - loss: 0.2080 - val_accuracy: 0.9400 - val_loss: 0.1940 - learning_rate: 1.0000e-04\nEpoch 10/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9357 - loss: 0.1979 - val_accuracy: 0.9400 - val_loss: 0.1961 - learning_rate: 1.0000e-04\nEpoch 11/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9393 - loss: 0.1988 - val_accuracy: 0.9332 - val_loss: 0.2029 - learning_rate: 1.0000e-04\nEpoch 12/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 465ms/step - accuracy: 0.9350 - loss: 0.1942 - val_accuracy: 0.9386 - val_loss: 0.1981 - learning_rate: 5.0000e-05\nEpoch 13/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 466ms/step - accuracy: 0.9407 - loss: 0.1956 - val_accuracy: 0.9372 - val_loss: 0.1955 - learning_rate: 5.0000e-05\nEpoch 14/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9426 - loss: 0.1831 - val_accuracy: 0.9400 - val_loss: 0.1891 - learning_rate: 2.5000e-05\nEpoch 15/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9419 - loss: 0.1815 - val_accuracy: 0.9468 - val_loss: 0.1882 - learning_rate: 2.5000e-05\nEpoch 16/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9374 - loss: 0.1839 - val_accuracy: 0.9482 - val_loss: 0.1858 - learning_rate: 2.5000e-05\nEpoch 17/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 475ms/step - accuracy: 0.9417 - loss: 0.1781 - val_accuracy: 0.9468 - val_loss: 0.1949 - learning_rate: 2.5000e-05\nEpoch 18/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9438 - loss: 0.1868 - val_accuracy: 0.9372 - val_loss: 0.1886 - learning_rate: 2.5000e-05\nEpoch 19/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9388 - loss: 0.1974 - val_accuracy: 0.9468 - val_loss: 0.1895 - learning_rate: 1.2500e-05\nEpoch 20/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9393 - loss: 0.1787 - val_accuracy: 0.9413 - val_loss: 0.1899 - learning_rate: 1.2500e-05\nEpoch 21/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9371 - loss: 0.1838 - val_accuracy: 0.9441 - val_loss: 0.1868 - learning_rate: 6.2500e-06\nFold Accuracy: 0.9481582537517054\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 3 ---\nEpoch 1/30\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 496ms/step - accuracy: 0.5200 - loss: 0.6922 - val_accuracy: 0.5150 - val_loss: 0.6300 - learning_rate: 1.0000e-04\nEpoch 2/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.7329 - loss: 0.5386 - val_accuracy: 0.8798 - val_loss: 0.3036 - learning_rate: 1.0000e-04\nEpoch 3/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 467ms/step - accuracy: 0.8488 - loss: 0.3617 - val_accuracy: 0.8757 - val_loss: 0.3026 - learning_rate: 1.0000e-04\nEpoch 4/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 477ms/step - accuracy: 0.8932 - loss: 0.2588 - val_accuracy: 0.9153 - val_loss: 0.2285 - learning_rate: 1.0000e-04\nEpoch 5/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 477ms/step - accuracy: 0.9050 - loss: 0.2596 - val_accuracy: 0.9303 - val_loss: 0.2194 - learning_rate: 1.0000e-04\nEpoch 6/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 474ms/step - accuracy: 0.9178 - loss: 0.2299 - val_accuracy: 0.9481 - val_loss: 0.1943 - learning_rate: 1.0000e-04\nEpoch 7/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9238 - loss: 0.2207 - val_accuracy: 0.9413 - val_loss: 0.1941 - learning_rate: 1.0000e-04\nEpoch 8/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9197 - loss: 0.2215 - val_accuracy: 0.9563 - val_loss: 0.1931 - learning_rate: 1.0000e-04\nEpoch 9/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 477ms/step - accuracy: 0.9310 - loss: 0.2155 - val_accuracy: 0.9522 - val_loss: 0.1906 - learning_rate: 1.0000e-04\nEpoch 10/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 481ms/step - accuracy: 0.9268 - loss: 0.2063 - val_accuracy: 0.9563 - val_loss: 0.1856 - learning_rate: 1.0000e-04\nEpoch 11/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9276 - loss: 0.2024 - val_accuracy: 0.9454 - val_loss: 0.1790 - learning_rate: 1.0000e-04\nEpoch 12/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 465ms/step - accuracy: 0.9130 - loss: 0.2185 - val_accuracy: 0.9577 - val_loss: 0.1846 - learning_rate: 1.0000e-04\nEpoch 13/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9314 - loss: 0.2038 - val_accuracy: 0.9358 - val_loss: 0.1985 - learning_rate: 1.0000e-04\nEpoch 14/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9285 - loss: 0.1975 - val_accuracy: 0.9549 - val_loss: 0.1780 - learning_rate: 5.0000e-05\nEpoch 15/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9354 - loss: 0.1989 - val_accuracy: 0.9467 - val_loss: 0.1860 - learning_rate: 5.0000e-05\nEpoch 16/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9276 - loss: 0.2026 - val_accuracy: 0.9454 - val_loss: 0.1909 - learning_rate: 5.0000e-05\nEpoch 17/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9359 - loss: 0.1938 - val_accuracy: 0.9577 - val_loss: 0.1766 - learning_rate: 2.5000e-05\nEpoch 18/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9340 - loss: 0.1969 - val_accuracy: 0.9372 - val_loss: 0.1847 - learning_rate: 2.5000e-05\nEpoch 19/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9392 - loss: 0.1838 - val_accuracy: 0.9426 - val_loss: 0.1815 - learning_rate: 2.5000e-05\nEpoch 20/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9319 - loss: 0.1989 - val_accuracy: 0.9590 - val_loss: 0.1770 - learning_rate: 1.2500e-05\nEpoch 21/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9352 - loss: 0.1937 - val_accuracy: 0.9536 - val_loss: 0.1723 - learning_rate: 1.2500e-05\nEpoch 22/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9368 - loss: 0.1804 - val_accuracy: 0.9563 - val_loss: 0.1708 - learning_rate: 1.2500e-05\nEpoch 23/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 465ms/step - accuracy: 0.9344 - loss: 0.1967 - val_accuracy: 0.9549 - val_loss: 0.1760 - learning_rate: 1.2500e-05\nEpoch 24/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9338 - loss: 0.1904 - val_accuracy: 0.9577 - val_loss: 0.1785 - learning_rate: 1.2500e-05\nEpoch 25/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9346 - loss: 0.2044 - val_accuracy: 0.9536 - val_loss: 0.1708 - learning_rate: 6.2500e-06\nEpoch 26/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 477ms/step - accuracy: 0.9367 - loss: 0.1903 - val_accuracy: 0.9577 - val_loss: 0.1741 - learning_rate: 6.2500e-06\nEpoch 27/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9368 - loss: 0.1876 - val_accuracy: 0.9577 - val_loss: 0.1698 - learning_rate: 3.1250e-06\nEpoch 28/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9424 - loss: 0.1845 - val_accuracy: 0.9577 - val_loss: 0.1696 - learning_rate: 3.1250e-06\nEpoch 29/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9310 - loss: 0.1948 - val_accuracy: 0.9577 - val_loss: 0.1712 - learning_rate: 3.1250e-06\nEpoch 30/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 478ms/step - accuracy: 0.9369 - loss: 0.1884 - val_accuracy: 0.9549 - val_loss: 0.1695 - learning_rate: 3.1250e-06\nFold Accuracy: 0.9549180327868853\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 4 ---\nEpoch 1/30\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 486ms/step - accuracy: 0.4927 - loss: 0.6933 - val_accuracy: 0.6257 - val_loss: 0.6633 - learning_rate: 1.0000e-04\nEpoch 2/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.7166 - loss: 0.5686 - val_accuracy: 0.8675 - val_loss: 0.3382 - learning_rate: 1.0000e-04\nEpoch 3/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.8608 - loss: 0.3687 - val_accuracy: 0.8852 - val_loss: 0.2540 - learning_rate: 1.0000e-04\nEpoch 4/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.8737 - loss: 0.3103 - val_accuracy: 0.9208 - val_loss: 0.2182 - learning_rate: 1.0000e-04\nEpoch 5/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9081 - loss: 0.2594 - val_accuracy: 0.9262 - val_loss: 0.2124 - learning_rate: 1.0000e-04\nEpoch 6/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9236 - loss: 0.2308 - val_accuracy: 0.9303 - val_loss: 0.2121 - learning_rate: 1.0000e-04\nEpoch 7/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9227 - loss: 0.2174 - val_accuracy: 0.9262 - val_loss: 0.2097 - learning_rate: 1.0000e-04\nEpoch 8/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9271 - loss: 0.2095 - val_accuracy: 0.9303 - val_loss: 0.2051 - learning_rate: 1.0000e-04\nEpoch 9/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 466ms/step - accuracy: 0.9300 - loss: 0.2022 - val_accuracy: 0.9276 - val_loss: 0.2115 - learning_rate: 1.0000e-04\nEpoch 10/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9364 - loss: 0.2014 - val_accuracy: 0.9290 - val_loss: 0.2039 - learning_rate: 1.0000e-04\nEpoch 11/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9350 - loss: 0.2030 - val_accuracy: 0.9331 - val_loss: 0.2059 - learning_rate: 1.0000e-04\nEpoch 12/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 466ms/step - accuracy: 0.9377 - loss: 0.1956 - val_accuracy: 0.9372 - val_loss: 0.2062 - learning_rate: 1.0000e-04\nEpoch 13/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 481ms/step - accuracy: 0.9379 - loss: 0.1790 - val_accuracy: 0.9331 - val_loss: 0.2030 - learning_rate: 5.0000e-05\nEpoch 14/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 480ms/step - accuracy: 0.9303 - loss: 0.2011 - val_accuracy: 0.9317 - val_loss: 0.2012 - learning_rate: 5.0000e-05\nEpoch 15/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9333 - loss: 0.1911 - val_accuracy: 0.9303 - val_loss: 0.2014 - learning_rate: 5.0000e-05\nEpoch 16/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9416 - loss: 0.1803 - val_accuracy: 0.9385 - val_loss: 0.2000 - learning_rate: 5.0000e-05\nEpoch 17/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m44s\u001b[0m 463ms/step - accuracy: 0.9348 - loss: 0.1905 - val_accuracy: 0.9317 - val_loss: 0.2013 - learning_rate: 5.0000e-05\nEpoch 18/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m44s\u001b[0m 463ms/step - accuracy: 0.9312 - loss: 0.2027 - val_accuracy: 0.9399 - val_loss: 0.2033 - learning_rate: 5.0000e-05\nEpoch 19/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9414 - loss: 0.1703 - val_accuracy: 0.9372 - val_loss: 0.1967 - learning_rate: 2.5000e-05\nEpoch 20/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 467ms/step - accuracy: 0.9338 - loss: 0.1873 - val_accuracy: 0.9385 - val_loss: 0.1975 - learning_rate: 2.5000e-05\nEpoch 21/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 466ms/step - accuracy: 0.9327 - loss: 0.1808 - val_accuracy: 0.9358 - val_loss: 0.1969 - learning_rate: 2.5000e-05\nEpoch 22/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9402 - loss: 0.1733 - val_accuracy: 0.9372 - val_loss: 0.1959 - learning_rate: 1.2500e-05\nEpoch 23/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9437 - loss: 0.1808 - val_accuracy: 0.9358 - val_loss: 0.1941 - learning_rate: 1.2500e-05\nEpoch 24/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9412 - loss: 0.1791 - val_accuracy: 0.9372 - val_loss: 0.1945 - learning_rate: 1.2500e-05\nEpoch 25/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9421 - loss: 0.1654 - val_accuracy: 0.9372 - val_loss: 0.1945 - learning_rate: 1.2500e-05\nEpoch 26/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9410 - loss: 0.1759 - val_accuracy: 0.9372 - val_loss: 0.1953 - learning_rate: 6.2500e-06\nEpoch 27/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9332 - loss: 0.1814 - val_accuracy: 0.9372 - val_loss: 0.1944 - learning_rate: 6.2500e-06\nEpoch 28/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9417 - loss: 0.1705 - val_accuracy: 0.9344 - val_loss: 0.1943 - learning_rate: 3.1250e-06\nFold Accuracy: 0.9357923497267759\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 5 ---\nEpoch 1/30\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 480ms/step - accuracy: 0.5372 - loss: 0.6885 - val_accuracy: 0.8525 - val_loss: 0.5487 - learning_rate: 1.0000e-04\nEpoch 2/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.7420 - loss: 0.5403 - val_accuracy: 0.8538 - val_loss: 0.3467 - learning_rate: 1.0000e-04\nEpoch 3/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 467ms/step - accuracy: 0.8151 - loss: 0.4209 - val_accuracy: 0.8538 - val_loss: 0.2916 - learning_rate: 1.0000e-04\nEpoch 4/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.8621 - loss: 0.3222 - val_accuracy: 0.9344 - val_loss: 0.2269 - learning_rate: 1.0000e-04\nEpoch 5/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9160 - loss: 0.2380 - val_accuracy: 0.9331 - val_loss: 0.2056 - learning_rate: 1.0000e-04\nEpoch 6/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9224 - loss: 0.2376 - val_accuracy: 0.9317 - val_loss: 0.2118 - learning_rate: 1.0000e-04\nEpoch 7/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 473ms/step - accuracy: 0.9295 - loss: 0.2234 - val_accuracy: 0.9344 - val_loss: 0.1973 - learning_rate: 1.0000e-04\nEpoch 8/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9374 - loss: 0.1947 - val_accuracy: 0.9372 - val_loss: 0.2018 - learning_rate: 1.0000e-04\nEpoch 9/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 477ms/step - accuracy: 0.9303 - loss: 0.2046 - val_accuracy: 0.9467 - val_loss: 0.1885 - learning_rate: 1.0000e-04\nEpoch 10/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 472ms/step - accuracy: 0.9165 - loss: 0.2297 - val_accuracy: 0.9399 - val_loss: 0.2006 - learning_rate: 1.0000e-04\nEpoch 11/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9475 - loss: 0.1649 - val_accuracy: 0.9426 - val_loss: 0.1906 - learning_rate: 1.0000e-04\nEpoch 12/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 468ms/step - accuracy: 0.9361 - loss: 0.1950 - val_accuracy: 0.9454 - val_loss: 0.1869 - learning_rate: 5.0000e-05\nEpoch 13/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 468ms/step - accuracy: 0.9384 - loss: 0.2001 - val_accuracy: 0.9372 - val_loss: 0.1968 - learning_rate: 5.0000e-05\nEpoch 14/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 470ms/step - accuracy: 0.9377 - loss: 0.1958 - val_accuracy: 0.9372 - val_loss: 0.2076 - learning_rate: 5.0000e-05\nEpoch 15/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 474ms/step - accuracy: 0.9347 - loss: 0.2008 - val_accuracy: 0.9413 - val_loss: 0.1931 - learning_rate: 2.5000e-05\nEpoch 16/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 471ms/step - accuracy: 0.9355 - loss: 0.1924 - val_accuracy: 0.9399 - val_loss: 0.1926 - learning_rate: 2.5000e-05\nEpoch 17/30\n\u001b[1m92/92\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m45s\u001b[0m 469ms/step - accuracy: 0.9392 - loss: 0.1826 - val_accuracy: 0.9426 - val_loss: 0.1873 - learning_rate: 1.2500e-05\nFold Accuracy: 0.9453551912568307\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 1 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\nAverage Accuracy: 0.943747903294344\n\nClassification Report:\n              precision    recall  f1-score   support\n\n           0       0.92      0.97      0.94      1805\n           1       0.97      0.92      0.94      1857\n\n    accuracy                           0.94      3662\n   macro avg       0.94      0.94      0.94      3662\nweighted avg       0.94      0.94      0.94      3662\n\n\n========== Early ==========\n\n--- Fold 1 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 246ms/step - accuracy: 0.4113 - loss: 1.0450 - val_accuracy: 0.6289 - val_loss: 0.7496 - learning_rate: 0.0010\nEpoch 2/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6055 - loss: 0.7595 - val_accuracy: 0.7149 - val_loss: 0.6219 - learning_rate: 0.0010\nEpoch 3/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6691 - loss: 0.7012 - val_accuracy: 0.7517 - val_loss: 0.6069 - learning_rate: 0.0010\nEpoch 4/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6555 - loss: 0.6893 - val_accuracy: 0.6126 - val_loss: 0.6375 - learning_rate: 0.0010\nEpoch 5/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7075 - loss: 0.6654 - val_accuracy: 0.7408 - val_loss: 0.6239 - learning_rate: 0.0010\nEpoch 6/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 240ms/step - accuracy: 0.6743 - loss: 0.6514 - val_accuracy: 0.7490 - val_loss: 0.6127 - learning_rate: 5.0000e-04\nEpoch 7/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 239ms/step - accuracy: 0.6678 - loss: 0.6607 - val_accuracy: 0.6467 - val_loss: 0.6203 - learning_rate: 5.0000e-04\nEpoch 8/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 240ms/step - accuracy: 0.7091 - loss: 0.6518 - val_accuracy: 0.5880 - val_loss: 0.6125 - learning_rate: 2.5000e-04\nFold Accuracy: 0.7517053206002728\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 2 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 253ms/step - accuracy: 0.3887 - loss: 1.0951 - val_accuracy: 0.7544 - val_loss: 0.6723 - learning_rate: 0.0010\nEpoch 2/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6728 - loss: 0.7453 - val_accuracy: 0.5498 - val_loss: 0.6124 - learning_rate: 0.0010\nEpoch 3/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6083 - loss: 0.7080 - val_accuracy: 0.5512 - val_loss: 0.5934 - learning_rate: 0.0010\nEpoch 4/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6054 - loss: 0.7108 - val_accuracy: 0.7449 - val_loss: 0.5337 - learning_rate: 0.0010\nEpoch 5/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 246ms/step - accuracy: 0.6833 - loss: 0.6882 - val_accuracy: 0.5962 - val_loss: 0.5912 - learning_rate: 0.0010\nEpoch 6/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6537 - loss: 0.6878 - val_accuracy: 0.7913 - val_loss: 0.5181 - learning_rate: 0.0010\nEpoch 7/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7136 - loss: 0.6730 - val_accuracy: 0.5880 - val_loss: 0.5646 - learning_rate: 0.0010\nEpoch 8/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 243ms/step - accuracy: 0.6452 - loss: 0.6936 - val_accuracy: 0.5975 - val_loss: 0.5461 - learning_rate: 0.0010\nEpoch 9/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6490 - loss: 0.6541 - val_accuracy: 0.6412 - val_loss: 0.5394 - learning_rate: 5.0000e-04\nEpoch 10/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6851 - loss: 0.6612 - val_accuracy: 0.7735 - val_loss: 0.5293 - learning_rate: 5.0000e-04\nEpoch 11/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7337 - loss: 0.6315 - val_accuracy: 0.6180 - val_loss: 0.5427 - learning_rate: 2.5000e-04\nFold Accuracy: 0.791268758526603\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 3 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 253ms/step - accuracy: 0.3648 - loss: 1.0565 - val_accuracy: 0.6585 - val_loss: 0.6751 - learning_rate: 0.0010\nEpoch 2/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 246ms/step - accuracy: 0.6495 - loss: 0.7350 - val_accuracy: 0.5697 - val_loss: 0.6190 - learning_rate: 0.0010\nEpoch 3/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 243ms/step - accuracy: 0.6530 - loss: 0.6922 - val_accuracy: 0.6421 - val_loss: 0.5687 - learning_rate: 0.0010\nEpoch 4/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6438 - loss: 0.6911 - val_accuracy: 0.7609 - val_loss: 0.6122 - learning_rate: 0.0010\nEpoch 5/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 243ms/step - accuracy: 0.6729 - loss: 0.7054 - val_accuracy: 0.7937 - val_loss: 0.5286 - learning_rate: 0.0010\nEpoch 6/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6966 - loss: 0.7107 - val_accuracy: 0.7500 - val_loss: 0.6204 - learning_rate: 0.0010\nEpoch 7/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7212 - loss: 0.6765 - val_accuracy: 0.5956 - val_loss: 0.5497 - learning_rate: 0.0010\nEpoch 8/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 248ms/step - accuracy: 0.6635 - loss: 0.6568 - val_accuracy: 0.7978 - val_loss: 0.5311 - learning_rate: 5.0000e-04\nEpoch 9/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 248ms/step - accuracy: 0.7180 - loss: 0.6534 - val_accuracy: 0.7691 - val_loss: 0.5872 - learning_rate: 5.0000e-04\nEpoch 10/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.7399 - loss: 0.6384 - val_accuracy: 0.7923 - val_loss: 0.5386 - learning_rate: 2.5000e-04\nFold Accuracy: 0.7937158469945356\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 4 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 249ms/step - accuracy: 0.3272 - loss: 1.0872 - val_accuracy: 0.8115 - val_loss: 0.5952 - learning_rate: 0.0010\nEpoch 2/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6726 - loss: 0.7188 - val_accuracy: 0.8019 - val_loss: 0.5633 - learning_rate: 0.0010\nEpoch 3/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6955 - loss: 0.6936 - val_accuracy: 0.7746 - val_loss: 0.5644 - learning_rate: 0.0010\nEpoch 4/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.7022 - loss: 0.6917 - val_accuracy: 0.6926 - val_loss: 0.5445 - learning_rate: 0.0010\nEpoch 5/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6536 - loss: 0.6560 - val_accuracy: 0.5929 - val_loss: 0.5536 - learning_rate: 0.0010\nEpoch 6/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 243ms/step - accuracy: 0.6244 - loss: 0.7093 - val_accuracy: 0.8183 - val_loss: 0.5205 - learning_rate: 0.0010\nEpoch 7/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7074 - loss: 0.6486 - val_accuracy: 0.6639 - val_loss: 0.5429 - learning_rate: 0.0010\nEpoch 8/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.7193 - loss: 0.6404 - val_accuracy: 0.6011 - val_loss: 0.5367 - learning_rate: 0.0010\nEpoch 9/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6973 - loss: 0.6353 - val_accuracy: 0.5956 - val_loss: 0.5327 - learning_rate: 5.0000e-04\nEpoch 10/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6851 - loss: 0.6657 - val_accuracy: 0.7254 - val_loss: 0.5481 - learning_rate: 5.0000e-04\nEpoch 11/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6652 - loss: 0.6433 - val_accuracy: 0.8156 - val_loss: 0.5415 - learning_rate: 2.5000e-04\nFold Accuracy: 0.8183060109289617\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 5 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m50s\u001b[0m 247ms/step - accuracy: 0.3182 - loss: 1.1245 - val_accuracy: 0.2910 - val_loss: 1.0233 - learning_rate: 0.0010\nEpoch 2/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.3625 - loss: 1.0573 - val_accuracy: 0.7678 - val_loss: 0.6000 - learning_rate: 0.0010\nEpoch 3/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 242ms/step - accuracy: 0.6507 - loss: 0.7538 - val_accuracy: 0.7432 - val_loss: 0.5974 - learning_rate: 0.0010\nEpoch 4/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.6427 - loss: 0.7108 - val_accuracy: 0.7473 - val_loss: 0.5754 - learning_rate: 0.0010\nEpoch 5/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6907 - loss: 0.6797 - val_accuracy: 0.6011 - val_loss: 0.5796 - learning_rate: 0.0010\nEpoch 6/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 240ms/step - accuracy: 0.6582 - loss: 0.7064 - val_accuracy: 0.6516 - val_loss: 0.5922 - learning_rate: 0.0010\nEpoch 7/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6964 - loss: 0.6745 - val_accuracy: 0.7828 - val_loss: 0.5596 - learning_rate: 5.0000e-04\nEpoch 8/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7321 - loss: 0.6604 - val_accuracy: 0.6380 - val_loss: 0.5572 - learning_rate: 5.0000e-04\nEpoch 9/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7108 - loss: 0.6406 - val_accuracy: 0.6038 - val_loss: 0.5569 - learning_rate: 5.0000e-04\nEpoch 10/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6791 - loss: 0.6491 - val_accuracy: 0.7336 - val_loss: 0.5501 - learning_rate: 5.0000e-04\nEpoch 11/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6971 - loss: 0.6673 - val_accuracy: 0.6066 - val_loss: 0.5595 - learning_rate: 5.0000e-04\nEpoch 12/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 239ms/step - accuracy: 0.6281 - loss: 0.6696 - val_accuracy: 0.7801 - val_loss: 0.5536 - learning_rate: 5.0000e-04\nEpoch 13/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7489 - loss: 0.6375 - val_accuracy: 0.7199 - val_loss: 0.5443 - learning_rate: 2.5000e-04\nEpoch 14/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7383 - loss: 0.6124 - val_accuracy: 0.6858 - val_loss: 0.5498 - learning_rate: 2.5000e-04\nEpoch 15/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6483 - loss: 0.6504 - val_accuracy: 0.7855 - val_loss: 0.5380 - learning_rate: 2.5000e-04\nEpoch 16/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7211 - loss: 0.6542 - val_accuracy: 0.6325 - val_loss: 0.5716 - learning_rate: 2.5000e-04\nEpoch 17/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6509 - loss: 0.6408 - val_accuracy: 0.7732 - val_loss: 0.5374 - learning_rate: 2.5000e-04\nEpoch 18/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7118 - loss: 0.6338 - val_accuracy: 0.6148 - val_loss: 0.5588 - learning_rate: 2.5000e-04\nEpoch 19/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6971 - loss: 0.6303 - val_accuracy: 0.6667 - val_loss: 0.5415 - learning_rate: 2.5000e-04\nEpoch 20/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.6744 - loss: 0.6294 - val_accuracy: 0.6626 - val_loss: 0.5466 - learning_rate: 1.2500e-04\nEpoch 21/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.7130 - loss: 0.6289 - val_accuracy: 0.6503 - val_loss: 0.5427 - learning_rate: 1.2500e-04\nEpoch 22/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6547 - loss: 0.6453 - val_accuracy: 0.6762 - val_loss: 0.5372 - learning_rate: 6.2500e-05\nEpoch 23/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.6900 - loss: 0.6335 - val_accuracy: 0.6776 - val_loss: 0.5365 - learning_rate: 6.2500e-05\nEpoch 24/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7073 - loss: 0.5923 - val_accuracy: 0.6598 - val_loss: 0.5425 - learning_rate: 6.2500e-05\nEpoch 25/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7055 - loss: 0.6128 - val_accuracy: 0.6612 - val_loss: 0.5400 - learning_rate: 6.2500e-05\nEpoch 26/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 248ms/step - accuracy: 0.6838 - loss: 0.6228 - val_accuracy: 0.6708 - val_loss: 0.5354 - learning_rate: 3.1250e-05\nEpoch 27/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6924 - loss: 0.6196 - val_accuracy: 0.6749 - val_loss: 0.5352 - learning_rate: 3.1250e-05\nEpoch 28/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7120 - loss: 0.6334 - val_accuracy: 0.6940 - val_loss: 0.5368 - learning_rate: 3.1250e-05\nEpoch 29/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.7052 - loss: 0.6111 - val_accuracy: 0.7008 - val_loss: 0.5330 - learning_rate: 3.1250e-05\nEpoch 30/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.6985 - loss: 0.6269 - val_accuracy: 0.6858 - val_loss: 0.5323 - learning_rate: 3.1250e-05\nEpoch 31/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.7065 - loss: 0.6258 - val_accuracy: 0.6899 - val_loss: 0.5360 - learning_rate: 3.1250e-05\nEpoch 32/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.7174 - loss: 0.6077 - val_accuracy: 0.6694 - val_loss: 0.5373 - learning_rate: 3.1250e-05\nEpoch 33/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 242ms/step - accuracy: 0.6946 - loss: 0.6340 - val_accuracy: 0.6967 - val_loss: 0.5324 - learning_rate: 1.5625e-05\nEpoch 34/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.6951 - loss: 0.6231 - val_accuracy: 0.7254 - val_loss: 0.5312 - learning_rate: 1.5625e-05\nEpoch 35/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.7023 - loss: 0.6139 - val_accuracy: 0.7309 - val_loss: 0.5309 - learning_rate: 1.5625e-05\nEpoch 36/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 245ms/step - accuracy: 0.6993 - loss: 0.6190 - val_accuracy: 0.7514 - val_loss: 0.5312 - learning_rate: 1.5625e-05\nEpoch 37/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 243ms/step - accuracy: 0.7139 - loss: 0.6184 - val_accuracy: 0.7445 - val_loss: 0.5298 - learning_rate: 1.5625e-05\nEpoch 38/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 244ms/step - accuracy: 0.7220 - loss: 0.6140 - val_accuracy: 0.7650 - val_loss: 0.5277 - learning_rate: 1.5625e-05\nEpoch 39/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 241ms/step - accuracy: 0.7161 - loss: 0.6126 - val_accuracy: 0.7582 - val_loss: 0.5312 - learning_rate: 1.5625e-05\nEpoch 40/40\n\u001b[1m184/184\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 240ms/step - accuracy: 0.7040 - loss: 0.6170 - val_accuracy: 0.7432 - val_loss: 0.5336 - learning_rate: 1.5625e-05\nFold Accuracy: 0.7650273224043715\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\nAverage Accuracy: 0.7840046518909489\n\nClassification Report:\n              precision    recall  f1-score   support\n\n           0       0.90      0.96      0.93      1805\n           1       0.73      0.77      0.75      1369\n           2       0.27      0.17      0.21       488\n\n    accuracy                           0.78      3662\n   macro avg       0.64      0.63      0.63      3662\nweighted avg       0.75      0.78      0.77      3662\n\n\n========== Severity ==========\n\n--- Fold 1 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 127ms/step - accuracy: 0.2407 - loss: 1.6107 - val_accuracy: 0.5130 - val_loss: 1.4321 - learning_rate: 1.0000e-04\nEpoch 2/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.4549 - loss: 1.5433 - val_accuracy: 0.6248 - val_loss: 1.1555 - learning_rate: 1.0000e-04\nEpoch 3/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.4512 - loss: 1.4957 - val_accuracy: 0.6016 - val_loss: 1.0333 - learning_rate: 1.0000e-04\nEpoch 4/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5412 - loss: 1.3549 - val_accuracy: 0.6739 - val_loss: 0.9471 - learning_rate: 1.0000e-04\nEpoch 5/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.6013 - loss: 1.2875 - val_accuracy: 0.5198 - val_loss: 0.9582 - learning_rate: 1.0000e-04\nEpoch 6/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5587 - loss: 1.3061 - val_accuracy: 0.5580 - val_loss: 0.9830 - learning_rate: 1.0000e-04\nEpoch 7/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5811 - loss: 1.2583 - val_accuracy: 0.6494 - val_loss: 0.9196 - learning_rate: 5.0000e-05\nEpoch 8/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5888 - loss: 1.3107 - val_accuracy: 0.6248 - val_loss: 0.9007 - learning_rate: 5.0000e-05\nEpoch 9/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5856 - loss: 1.2598 - val_accuracy: 0.6166 - val_loss: 0.9134 - learning_rate: 5.0000e-05\nEpoch 10/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5845 - loss: 1.2794 - val_accuracy: 0.6671 - val_loss: 0.9165 - learning_rate: 5.0000e-05\nEpoch 11/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5980 - loss: 1.3063 - val_accuracy: 0.6739 - val_loss: 0.9149 - learning_rate: 2.5000e-05\nEpoch 12/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5844 - loss: 1.2900 - val_accuracy: 0.6535 - val_loss: 0.8833 - learning_rate: 2.5000e-05\nEpoch 13/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.6201 - loss: 1.2088 - val_accuracy: 0.6426 - val_loss: 0.8856 - learning_rate: 2.5000e-05\nEpoch 14/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 127ms/step - accuracy: 0.5944 - loss: 1.2466 - val_accuracy: 0.6371 - val_loss: 0.8847 - learning_rate: 2.5000e-05\nEpoch 15/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.6007 - loss: 1.2467 - val_accuracy: 0.6535 - val_loss: 0.8746 - learning_rate: 1.2500e-05\nEpoch 16/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6026 - loss: 1.2801 - val_accuracy: 0.6467 - val_loss: 0.8824 - learning_rate: 1.2500e-05\nEpoch 17/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6074 - loss: 1.2287 - val_accuracy: 0.6535 - val_loss: 0.8767 - learning_rate: 1.2500e-05\nEpoch 18/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 127ms/step - accuracy: 0.6171 - loss: 1.2211 - val_accuracy: 0.6439 - val_loss: 0.8833 - learning_rate: 6.2500e-06\nEpoch 19/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5947 - loss: 1.2553 - val_accuracy: 0.6439 - val_loss: 0.8834 - learning_rate: 6.2500e-06\nEpoch 20/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5893 - loss: 1.2636 - val_accuracy: 0.6480 - val_loss: 0.8812 - learning_rate: 3.1250e-06\nFold Accuracy: 0.6534788540245566\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 2 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m51s\u001b[0m 128ms/step - accuracy: 0.2023 - loss: 1.6118 - val_accuracy: 0.4898 - val_loss: 1.4130 - learning_rate: 1.0000e-04\nEpoch 2/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.4800 - loss: 1.5159 - val_accuracy: 0.6166 - val_loss: 0.9386 - learning_rate: 1.0000e-04\nEpoch 3/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5727 - loss: 1.3484 - val_accuracy: 0.5525 - val_loss: 0.9301 - learning_rate: 1.0000e-04\nEpoch 4/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5675 - loss: 1.3320 - val_accuracy: 0.5348 - val_loss: 0.9304 - learning_rate: 1.0000e-04\nEpoch 5/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5651 - loss: 1.3194 - val_accuracy: 0.7094 - val_loss: 0.9016 - learning_rate: 1.0000e-04\nEpoch 6/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5874 - loss: 1.3488 - val_accuracy: 0.6658 - val_loss: 0.9026 - learning_rate: 1.0000e-04\nEpoch 7/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5998 - loss: 1.2719 - val_accuracy: 0.5880 - val_loss: 0.8967 - learning_rate: 1.0000e-04\nEpoch 8/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5883 - loss: 1.2826 - val_accuracy: 0.5853 - val_loss: 0.8913 - learning_rate: 1.0000e-04\nEpoch 9/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5884 - loss: 1.2896 - val_accuracy: 0.5812 - val_loss: 0.8968 - learning_rate: 1.0000e-04\nEpoch 10/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5997 - loss: 1.2419 - val_accuracy: 0.6562 - val_loss: 0.8973 - learning_rate: 1.0000e-04\nEpoch 11/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5882 - loss: 1.2616 - val_accuracy: 0.5812 - val_loss: 0.8872 - learning_rate: 5.0000e-05\nEpoch 12/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 127ms/step - accuracy: 0.5982 - loss: 1.2392 - val_accuracy: 0.5621 - val_loss: 0.8792 - learning_rate: 5.0000e-05\nEpoch 13/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5814 - loss: 1.2756 - val_accuracy: 0.6235 - val_loss: 0.8799 - learning_rate: 5.0000e-05\nEpoch 14/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5864 - loss: 1.2805 - val_accuracy: 0.6180 - val_loss: 0.8783 - learning_rate: 5.0000e-05\nEpoch 15/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 125ms/step - accuracy: 0.6059 - loss: 1.2917 - val_accuracy: 0.6644 - val_loss: 0.8821 - learning_rate: 5.0000e-05\nEpoch 16/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6199 - loss: 1.2907 - val_accuracy: 0.6003 - val_loss: 0.8789 - learning_rate: 5.0000e-05\nEpoch 17/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 128ms/step - accuracy: 0.6064 - loss: 1.2623 - val_accuracy: 0.5866 - val_loss: 0.8777 - learning_rate: 2.5000e-05\nEpoch 18/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.6026 - loss: 1.2178 - val_accuracy: 0.5989 - val_loss: 0.8693 - learning_rate: 2.5000e-05\nEpoch 19/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.6119 - loss: 1.2229 - val_accuracy: 0.6003 - val_loss: 0.8743 - learning_rate: 2.5000e-05\nEpoch 20/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m49s\u001b[0m 127ms/step - accuracy: 0.5881 - loss: 1.2746 - val_accuracy: 0.6016 - val_loss: 0.8642 - learning_rate: 2.5000e-05\nEpoch 21/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.6102 - loss: 1.2024 - val_accuracy: 0.5935 - val_loss: 0.8688 - learning_rate: 2.5000e-05\nEpoch 22/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5914 - loss: 1.2420 - val_accuracy: 0.5825 - val_loss: 0.8756 - learning_rate: 2.5000e-05\nEpoch 23/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6015 - loss: 1.2333 - val_accuracy: 0.5948 - val_loss: 0.8837 - learning_rate: 1.2500e-05\nEpoch 24/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5968 - loss: 1.2870 - val_accuracy: 0.5880 - val_loss: 0.8781 - learning_rate: 1.2500e-05\nEpoch 25/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5961 - loss: 1.2371 - val_accuracy: 0.5921 - val_loss: 0.8788 - learning_rate: 6.2500e-06\nFold Accuracy: 0.6016371077762619\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 3 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 128ms/step - accuracy: 0.1797 - loss: 1.6225 - val_accuracy: 0.5123 - val_loss: 1.5118 - learning_rate: 1.0000e-04\nEpoch 2/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.4798 - loss: 1.5213 - val_accuracy: 0.4904 - val_loss: 1.0256 - learning_rate: 1.0000e-04\nEpoch 3/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.4994 - loss: 1.4300 - val_accuracy: 0.6025 - val_loss: 0.9228 - learning_rate: 1.0000e-04\nEpoch 4/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5844 - loss: 1.3441 - val_accuracy: 0.5710 - val_loss: 0.9295 - learning_rate: 1.0000e-04\nEpoch 5/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5668 - loss: 1.3257 - val_accuracy: 0.5751 - val_loss: 0.9484 - learning_rate: 1.0000e-04\nEpoch 6/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5747 - loss: 1.3007 - val_accuracy: 0.5314 - val_loss: 0.9427 - learning_rate: 5.0000e-05\nEpoch 7/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5829 - loss: 1.2899 - val_accuracy: 0.5642 - val_loss: 0.9352 - learning_rate: 5.0000e-05\nEpoch 8/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5827 - loss: 1.2693 - val_accuracy: 0.5642 - val_loss: 0.9179 - learning_rate: 2.5000e-05\nEpoch 9/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5810 - loss: 1.2733 - val_accuracy: 0.5601 - val_loss: 0.9312 - learning_rate: 2.5000e-05\nEpoch 10/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5897 - loss: 1.2515 - val_accuracy: 0.5724 - val_loss: 0.9136 - learning_rate: 2.5000e-05\nEpoch 11/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5994 - loss: 1.2532 - val_accuracy: 0.5656 - val_loss: 0.9302 - learning_rate: 2.5000e-05\nEpoch 12/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5919 - loss: 1.2607 - val_accuracy: 0.5765 - val_loss: 0.9167 - learning_rate: 2.5000e-05\nEpoch 13/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6113 - loss: 1.2661 - val_accuracy: 0.5669 - val_loss: 0.9222 - learning_rate: 1.2500e-05\nEpoch 14/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5958 - loss: 1.2441 - val_accuracy: 0.5710 - val_loss: 0.9322 - learning_rate: 1.2500e-05\nEpoch 15/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 127ms/step - accuracy: 0.5860 - loss: 1.2734 - val_accuracy: 0.5628 - val_loss: 0.9266 - learning_rate: 6.2500e-06\nFold Accuracy: 0.5724043715846995\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 4 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 129ms/step - accuracy: 0.3071 - loss: 1.5731 - val_accuracy: 0.4713 - val_loss: 1.4007 - learning_rate: 1.0000e-04\nEpoch 2/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5006 - loss: 1.4579 - val_accuracy: 0.6913 - val_loss: 0.9502 - learning_rate: 1.0000e-04\nEpoch 3/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5828 - loss: 1.3320 - val_accuracy: 0.5055 - val_loss: 1.0082 - learning_rate: 1.0000e-04\nEpoch 4/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5689 - loss: 1.2773 - val_accuracy: 0.5697 - val_loss: 0.9215 - learning_rate: 1.0000e-04\nEpoch 5/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5531 - loss: 1.3178 - val_accuracy: 0.5902 - val_loss: 0.9144 - learning_rate: 1.0000e-04\nEpoch 6/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5745 - loss: 1.2492 - val_accuracy: 0.5615 - val_loss: 0.8939 - learning_rate: 1.0000e-04\nEpoch 7/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5815 - loss: 1.2304 - val_accuracy: 0.5574 - val_loss: 0.9254 - learning_rate: 1.0000e-04\nEpoch 8/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5480 - loss: 1.2433 - val_accuracy: 0.5642 - val_loss: 0.9140 - learning_rate: 1.0000e-04\nEpoch 9/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5843 - loss: 1.2677 - val_accuracy: 0.5943 - val_loss: 0.9103 - learning_rate: 5.0000e-05\nEpoch 10/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5681 - loss: 1.2676 - val_accuracy: 0.5779 - val_loss: 0.8878 - learning_rate: 5.0000e-05\nEpoch 11/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5846 - loss: 1.2006 - val_accuracy: 0.5874 - val_loss: 0.9007 - learning_rate: 5.0000e-05\nEpoch 12/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5934 - loss: 1.2264 - val_accuracy: 0.5915 - val_loss: 0.8917 - learning_rate: 5.0000e-05\nEpoch 13/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5914 - loss: 1.2403 - val_accuracy: 0.5915 - val_loss: 0.9001 - learning_rate: 2.5000e-05\nEpoch 14/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5881 - loss: 1.2441 - val_accuracy: 0.5970 - val_loss: 0.8917 - learning_rate: 2.5000e-05\nEpoch 15/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5916 - loss: 1.2084 - val_accuracy: 0.5970 - val_loss: 0.8851 - learning_rate: 1.2500e-05\nEpoch 16/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5906 - loss: 1.2429 - val_accuracy: 0.5915 - val_loss: 0.8881 - learning_rate: 1.2500e-05\nEpoch 17/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5824 - loss: 1.1979 - val_accuracy: 0.5929 - val_loss: 0.8855 - learning_rate: 1.2500e-05\nEpoch 18/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5838 - loss: 1.2121 - val_accuracy: 0.5929 - val_loss: 0.8811 - learning_rate: 6.2500e-06\nEpoch 19/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5805 - loss: 1.2391 - val_accuracy: 0.5956 - val_loss: 0.8860 - learning_rate: 6.2500e-06\nEpoch 20/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5824 - loss: 1.2469 - val_accuracy: 0.5929 - val_loss: 0.8832 - learning_rate: 6.2500e-06\nEpoch 21/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5781 - loss: 1.2251 - val_accuracy: 0.5943 - val_loss: 0.8802 - learning_rate: 3.1250e-06\nEpoch 22/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5854 - loss: 1.2112 - val_accuracy: 0.5943 - val_loss: 0.8793 - learning_rate: 3.1250e-06\nEpoch 23/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5851 - loss: 1.2406 - val_accuracy: 0.5929 - val_loss: 0.8750 - learning_rate: 3.1250e-06\nEpoch 24/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 127ms/step - accuracy: 0.5822 - loss: 1.2139 - val_accuracy: 0.5915 - val_loss: 0.8787 - learning_rate: 3.1250e-06\nEpoch 25/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5858 - loss: 1.2411 - val_accuracy: 0.5915 - val_loss: 0.8794 - learning_rate: 3.1250e-06\nEpoch 26/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5802 - loss: 1.1982 - val_accuracy: 0.5929 - val_loss: 0.8778 - learning_rate: 1.5625e-06\nEpoch 27/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5896 - loss: 1.2052 - val_accuracy: 0.5943 - val_loss: 0.8773 - learning_rate: 1.5625e-06\nEpoch 28/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5933 - loss: 1.1925 - val_accuracy: 0.5943 - val_loss: 0.8772 - learning_rate: 7.8125e-07\nFold Accuracy: 0.592896174863388\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\n--- Fold 5 ---\nEpoch 1/40\nINFO:tensorflow:Collective all_reduce tensors: 17 all_reduces, num_devices = 2, group_size = 2, implementation = CommunicationImplementation.NCCL, num_packs = 1\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m52s\u001b[0m 129ms/step - accuracy: 0.1290 - loss: 1.6299 - val_accuracy: 0.4945 - val_loss: 1.5587 - learning_rate: 1.0000e-04\nEpoch 2/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5132 - loss: 1.5456 - val_accuracy: 0.7104 - val_loss: 1.0282 - learning_rate: 1.0000e-04\nEpoch 3/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5716 - loss: 1.3826 - val_accuracy: 0.7186 - val_loss: 0.9516 - learning_rate: 1.0000e-04\nEpoch 4/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.6063 - loss: 1.3173 - val_accuracy: 0.6735 - val_loss: 0.9523 - learning_rate: 1.0000e-04\nEpoch 5/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5786 - loss: 1.3235 - val_accuracy: 0.5710 - val_loss: 0.9522 - learning_rate: 1.0000e-04\nEpoch 6/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.5764 - loss: 1.2472 - val_accuracy: 0.5628 - val_loss: 0.9576 - learning_rate: 5.0000e-05\nEpoch 7/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 126ms/step - accuracy: 0.5865 - loss: 1.2651 - val_accuracy: 0.5792 - val_loss: 0.9307 - learning_rate: 5.0000e-05\nEpoch 8/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.6020 - loss: 1.2651 - val_accuracy: 0.6803 - val_loss: 0.9385 - learning_rate: 5.0000e-05\nEpoch 9/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 125ms/step - accuracy: 0.6031 - loss: 1.2552 - val_accuracy: 0.6052 - val_loss: 0.9207 - learning_rate: 5.0000e-05\nEpoch 10/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m48s\u001b[0m 124ms/step - accuracy: 0.6165 - loss: 1.2653 - val_accuracy: 0.5738 - val_loss: 0.9224 - learning_rate: 5.0000e-05\nEpoch 11/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5916 - loss: 1.2583 - val_accuracy: 0.5601 - val_loss: 0.9509 - learning_rate: 5.0000e-05\nEpoch 12/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5907 - loss: 1.2043 - val_accuracy: 0.5669 - val_loss: 0.9521 - learning_rate: 2.5000e-05\nEpoch 13/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5931 - loss: 1.2519 - val_accuracy: 0.5628 - val_loss: 0.9441 - learning_rate: 2.5000e-05\nEpoch 14/40\n\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m47s\u001b[0m 124ms/step - accuracy: 0.5936 - loss: 1.2744 - val_accuracy: 0.5833 - val_loss: 0.9297 - learning_rate: 1.2500e-05\nFold Accuracy: 0.605191256830601\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 600x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"\nAverage Accuracy: 0.6051215530159014\n\nClassification Report:\n              precision    recall  f1-score   support\n\n           0       0.92      0.96      0.94      1805\n           1       0.32      0.64      0.43       370\n           2       0.63      0.12      0.20       999\n           3       0.14      0.32      0.19       193\n           4       0.15      0.21      0.18       295\n\n    accuracy                           0.61      3662\n   macro avg       0.43      0.45      0.39      3662\nweighted avg       0.68      0.61      0.58      3662\n\n","output_type":"stream"}],"execution_count":8}]}