{ "cells": [ { "cell_type": "markdown", "id": "38cc24e9-7de9-4fbd-81be-249f6a3b3e21", "metadata": {}, "source": [ "# CNN Rice Classification " ] }, { "cell_type": "markdown", "id": "4bc3c1fc-0da2-4017-9085-07336919daaa", "metadata": {}, "source": [ "" ] }, { "cell_type": "markdown", "id": "17f2e01a-515e-4b85-bb7d-4ad6d5a22c9c", "metadata": {}, "source": [ "## 1. Introduction\n", "\n", "In this project, I built a Convolutional Neural Network (CNN) from scratch to classify rice grain images into five classes:\n", "\n", "- Arborio \n", "- Basmati \n", "- Ipsala \n", "- Jasmine \n", "- Karacadag \n", "\n", "The dataset contains **75,000 images**. I used an **80% training and 20% validation split**. All images were resized to **160×160 pixels**. \n", "The goal of this project is to train a deep CNN without using transfer learning and evaluate its performance." ] }, { "cell_type": "markdown", "id": "8f5c4f91-641e-416a-a63b-2eb24b6b18d5", "metadata": {}, "source": [ "### Import Libraries" ] }, { "cell_type": "code", "execution_count": 1, "id": "3fc21004-42f4-4c3f-adb5-4a0367204772", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "2025-12-06 11:49:58.836655: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.\n", "2025-12-06 11:49:59.297423: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n", "2025-12-06 11:49:59.443572: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8473] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n", "2025-12-06 11:49:59.507143: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1471] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n", "2025-12-06 11:50:00.177702: I tensorflow/core/platform/cpu_feature_guard.cc:211] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n", "To enable the following instructions: SSE3 SSE4.1 SSE4.2 AVX, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n" ] } ], "source": [ "from tensorflow import keras\n", "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n", "from tensorflow.keras.models import Sequential\n", "from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, Rescaling, RandomFlip, RandomRotation, RandomZoom, Flatten, Dense, Dropout, BatchNormalization\n", "from tensorflow.keras.optimizers import Adam\n", "from tensorflow.keras import regularizers\n", "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau\n", "from tensorflow.keras.layers import GlobalAveragePooling2D\n", "from pathlib import Path\n", "from collections import Counter\n", "from tensorflow.keras.models import load_model\n", "import matplotlib.pyplot as plt\n", "from sklearn.metrics import confusion_matrix, classification_report\n", "import seaborn as sns\n", "import numpy as np\n", "import tensorflow as tf\n", "import pickle\n", "import warnings\n", "warnings.filterwarnings(\"ignore\") " ] }, { "cell_type": "markdown", "id": "a868f065-997e-407c-a0e3-04e84f2a5121", "metadata": {}, "source": [ "### Load Data" ] }, { "cell_type": "code", "execution_count": 2, "id": "a08782a9-4206-47f4-9892-19d9d8e854fc", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Path to dataset files: /root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1\n" ] } ], "source": [ "import kagglehub\n", "\n", "# Download latest version\n", "path = kagglehub.dataset_download(\"muratkokludataset/rice-image-dataset\")\n", "\n", "print(\"Path to dataset files:\", path)" ] }, { "cell_type": "code", "execution_count": 3, "id": "e80488d5-b9c4-4c42-96d3-25cd25e25ff0", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset')]" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "base_dir = Path(path)\n", "list(base_dir.iterdir())" ] }, { "cell_type": "code", "execution_count": 4, "id": "942a3390-72d7-48e9-8b29-264ce763b74c", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset\n", "[PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Ipsala'), PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Karacadag'), PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Arborio'), PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Basmati'), PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Jasmine'), PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Rice_Citation_Request.txt')]\n" ] } ], "source": [ "data_dir = base_dir / \"Rice_Image_Dataset\"\n", "print(data_dir)\n", "print(list(data_dir.iterdir()))" ] }, { "cell_type": "code", "execution_count": 5, "id": "a2056314-e087-4ca5-8bf4-504374238fc1", "metadata": { "collapsed": true, "jupyter": { "outputs_hidden": true } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Found 75000 files belonging to 5 classes.\n", "Using 60000 files for training.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n", "I0000 00:00:1764962779.059957 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.343428 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.343467 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.345486 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.345511 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.345518 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.580371 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1764962779.580418 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "2025-12-05 19:26:19.580427: I tensorflow/core/common_runtime/gpu/gpu_device.cc:2112] Could not identify NUMA node of platform GPU id 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n", "I0000 00:00:1764962779.580461 3691 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "2025-12-05 19:26:19.580482: I tensorflow/core/common_runtime/gpu/gpu_device.cc:2021] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 10861 MB memory: -> device: 0, name: NVIDIA GeForce RTX 5080 Laptop GPU, pci bus id: 0000:02:00.0, compute capability: 12.0\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85'' is not a recognized feature for this target (ignoring feature)\n", "+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Found 75000 files belonging to 5 classes.\n", "Using 15000 files for validation.\n", "Classes: ['Arborio', 'Basmati', 'Ipsala', 'Jasmine', 'Karacadag']\n" ] } ], "source": [ "img_size=(160, 160)\n", "batch_size=16\n", "seed=32\n", "\n", "train_ds = tf.keras.utils.image_dataset_from_directory(\n", " data_dir,\n", " validation_split=0.2,\n", " subset=\"training\",\n", " seed=seed,\n", " image_size=img_size,\n", " batch_size=batch_size\n", ")\n", "\n", "val_ds = tf.keras.utils.image_dataset_from_directory(\n", " data_dir,\n", " validation_split=0.2,\n", " subset=\"validation\",\n", " seed=seed,\n", " image_size=img_size,\n", " batch_size=batch_size\n", ")\n", "\n", "class_names = train_ds.class_names\n", "num_classes = len(class_names)\n", "print(\"Classes:\", class_names)" ] }, { "cell_type": "markdown", "id": "d8a00e44-a120-469f-acdf-cc8ee9119c8c", "metadata": {}, "source": [ "## 2. Analysis\n", "\n", "The dataset is **balanced**, which means each class contains almost the same number of images. Therefore, no class weighting was needed.\n", "\n", "I also visualized some example images from the dataset. All rice grains appear on a **black background** and are well centered. Some rice types look very similar, especially **Jasmine, Basmati and Arborio**, which explains the few misclassifications.\n" ] }, { "cell_type": "code", "execution_count": 15, "id": "380afbb5-ace9-4211-b284-6203cbabf7fb", "metadata": { "scrolled": true }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "2025-12-05 20:12:33.602335: I tensorflow/core/framework/local_rendezvous.cc:404] Local rendezvous is aborting with status: OUT_OF_RANGE: End of sequence\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "counter = Counter()\n", "\n", "for images, labels in train_ds:\n", " counter.update(labels.numpy().tolist())\n", "\n", "counts = [counter[i] for i in range(len(class_names))]\n", "\n", "plt.figure(figsize=(10, 5))\n", "plt.bar(class_names, counts)\n", "plt.xticks(rotation=45)\n", "plt.ylabel(\"Number of Picture\")\n", "plt.title(\"Class distribution (Train-Dataset)\")\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 16, "id": "3228923e-ed5f-4f2f-bd8f-74b66a815066", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sample_ds = train_ds.unbatch().take(9)\n", "\n", "images = []\n", "labels = []\n", "\n", "for img, lbl in sample_ds:\n", " images.append(img.numpy().astype(\"uint8\"))\n", " labels.append(lbl.numpy())\n", "\n", "plt.figure(figsize=(10, 10))\n", "for i in range(9):\n", " ax = plt.subplot(3, 3, i + 1)\n", " plt.imshow(images[i])\n", " plt.title(class_names[labels[i]])\n", " plt.axis(\"off\")\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 6, "id": "cc1fd8a0-d955-47d5-af9a-d578ac24f9fd", "metadata": {}, "outputs": [], "source": [ "AUTOTUNE=tf.data.AUTOTUNE\n", "train_ds=(train_ds.shuffle(100).prefetch(1))\n", "val_ds=val_ds.prefetch(1)" ] }, { "cell_type": "code", "execution_count": 8, "id": "a69aec77-88d8-4ab0-ad5c-b9455a3edad7", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"sequential_1\"\n", "_________________________________________________________________\n", " Layer (type) Output Shape Param # \n", "=================================================================\n", " rescaling_1 (Rescaling) (None, 160, 160, 3) 0 \n", " \n", " random_flip_1 (RandomFlip) (None, 160, 160, 3) 0 \n", " \n", " random_rotation_1 (RandomR (None, 160, 160, 3) 0 \n", " otation) \n", " \n", " random_zoom_1 (RandomZoom) (None, 160, 160, 3) 0 \n", " \n", " conv2d_5 (Conv2D) (None, 160, 160, 32) 896 \n", " \n", " batch_normalization_5 (Bat (None, 160, 160, 32) 128 \n", " chNormalization) \n", " \n", " max_pooling2d_3 (MaxPoolin (None, 80, 80, 32) 0 \n", " g2D) \n", " \n", " conv2d_6 (Conv2D) (None, 80, 80, 64) 18496 \n", " \n", " batch_normalization_6 (Bat (None, 80, 80, 64) 256 \n", " chNormalization) \n", " \n", " max_pooling2d_4 (MaxPoolin (None, 40, 40, 64) 0 \n", " g2D) \n", " \n", " conv2d_7 (Conv2D) (None, 40, 40, 128) 73856 \n", " \n", " batch_normalization_7 (Bat (None, 40, 40, 128) 512 \n", " chNormalization) \n", " \n", " conv2d_8 (Conv2D) (None, 40, 40, 128) 147584 \n", " \n", " batch_normalization_8 (Bat (None, 40, 40, 128) 512 \n", " chNormalization) \n", " \n", " max_pooling2d_5 (MaxPoolin (None, 20, 20, 128) 0 \n", " g2D) \n", " \n", " conv2d_9 (Conv2D) (None, 20, 20, 256) 295168 \n", " \n", " batch_normalization_9 (Bat (None, 20, 20, 256) 1024 \n", " chNormalization) \n", " \n", " global_average_pooling2d_1 (None, 256) 0 \n", " (GlobalAveragePooling2D) \n", " \n", " dropout_1 (Dropout) (None, 256) 0 \n", " \n", " dense (Dense) (None, 128) 32896 \n", " \n", " dropout_2 (Dropout) (None, 128) 0 \n", " \n", " dense_1 (Dense) (None, 5) 645 \n", " \n", "=================================================================\n", "Total params: 571973 (2.18 MB)\n", "Trainable params: 570757 (2.18 MB)\n", "Non-trainable params: 1216 (4.75 KB)\n", "_________________________________________________________________\n" ] } ], "source": [ "model = keras.Sequential()\n", "\n", "model.add(Input(shape=(160,160,3)))\n", "model.add(Rescaling(1./255))\n", "\n", "# Augmentation\n", "model.add(RandomFlip(\"horizontal\"))\n", "model.add(RandomRotation(0.05))\n", "model.add(RandomZoom(0.1))\n", "\n", "# 5 Blocks\n", "model.add(Conv2D(32, (3, 3),padding=\"same\",activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(BatchNormalization())\n", "model.add(MaxPooling2D((2, 2)))\n", "\n", "model.add(Conv2D(64, (3, 3),padding=\"same\",activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(BatchNormalization())\n", "model.add(MaxPooling2D((2, 2)))\n", "\n", "model.add(Conv2D(128, (3, 3),padding=\"same\",activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(BatchNormalization())\n", "\n", "model.add(Conv2D(128, (3, 3),padding=\"same\",activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(BatchNormalization())\n", "model.add(MaxPooling2D((2, 2)))\n", "\n", "model.add(Conv2D(256, (3, 3),padding=\"same\",activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(BatchNormalization())\n", "\n", "# Classifier\n", "model.add(GlobalAveragePooling2D())\n", "model.add(Dropout(0.4))\n", "\n", "model.add(Dense(128,activation=\"relu\",kernel_regularizer=regularizers.l2(1e-4)))\n", "model.add(Dropout(0.4))\n", "model.add(Dense(5, activation=\"softmax\"))\n", "\n", "model.summary()" ] }, { "cell_type": "markdown", "id": "24a1d672-a6e6-4670-b6e9-2d55194feeed", "metadata": {}, "source": [ "## 3. Methods\n", "\n", "### Data Loading\n", "\n", "The images were loaded using:\n", "\n", "- `tf.keras.utils.image_dataset_from_directory`\n", "- Image size: **160×160**\n", "- Batch size: **16**\n", "- Validation split: **20%**\n", "\n", "\n", "\n", "### Data Augmentation\n", "\n", "To improve generalization, the following augmentations were applied:\n", "\n", "- Random horizontal flip \n", "- Random rotation \n", "- Random zoom \n", "\n", "### Model Architecture\n", "\n", "The CNN architecture consists of:\n", "\n", "- **5 Convolutional layers**\n", "- **3 MaxPooling layers**\n", "- Batch Normalization after each convolution\n", "- L2 Regularization\n", "- Global Average Pooling\n", "- Dense layer with Dropout\n", "- Final Softmax layer with **5 output classes**\n", "\n", "### Training Setup\n", "\n", "- **Loss Function:** Sparse Categorical Crossentropy \n", "- **Optimizer:** Adam (learning rate = 0.0001) \n", "- **Metrics:** Accuracy \n", "- **Callbacks:**\n", " - EarlyStopping \n", " - ReduceLROnPlateau \n", " - ModelCheckpoint \n", "- **Epochs:** 40 \n" ] }, { "cell_type": "code", "execution_count": 9, "id": "24cf1f54-66c1-4944-977d-95630faa60d1", "metadata": {}, "outputs": [], "source": [ "model.compile(optimizer=keras.optimizers.Adam(learning_rate=1e-4),loss=\"sparse_categorical_crossentropy\",metrics=[\"accuracy\",keras.metrics.TopKCategoricalAccuracy(k=3, name=\"top3_acc\")])" ] }, { "cell_type": "code", "execution_count": 10, "id": "a8888042-f3d1-41d3-b6fb-d31b68bb3938", "metadata": {}, "outputs": [], "source": [ "callbacks = [\n", " EarlyStopping(\n", " monitor=\"val_loss\",\n", " patience=8,\n", " restore_best_weights=True\n", " ),\n", " ReduceLROnPlateau(\n", " monitor=\"val_loss\",\n", " factor=0.5,\n", " patience=4,\n", " min_lr=1e-7\n", " ),\n", " ModelCheckpoint(\n", " \"best_rice_cnn.keras\",\n", " monitor=\"val_loss\",\n", " save_best_only=True\n", " )\n", "]" ] }, { "cell_type": "code", "execution_count": 11, "id": "08414fab-b492-4973-8590-31e38efeaf75", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/40\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "2025-12-05 19:27:00.153460: I external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:531] Loaded cuDNN version 90701\n", "W0000 00:00:1764962820.374314 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.400749 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.404232 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.417039 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.420716 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.424750 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.455004 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.470228 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.473879 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.569174 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.574303 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.587941 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.647971 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.648822 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.649969 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.650386 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.650817 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.651245 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.651710 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.652300 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.654833 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.655496 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.655944 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.656547 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.657210 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.664042 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.664964 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.668308 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.675801 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.676284 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.676585 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.676895 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.677153 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.677555 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.677978 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.678421 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.678707 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.679174 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.679828 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.680392 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.680780 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.681211 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.681717 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.687220 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.687976 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.688445 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.688906 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.689334 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.689905 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.690565 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.691289 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.692118 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.692970 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.693764 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.694605 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.695540 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.696516 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.699414 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.702410 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.702847 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.703170 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.703699 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.703982 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.704243 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.704718 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.705205 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.705465 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.706021 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.706444 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.707498 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.707849 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.708387 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962820.708891 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n", "I0000 00:00:1764962821.158423 3828 service.cc:146] XLA service 0x7e01496f1650 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n", "I0000 00:00:1764962821.158452 3828 service.cc:154] StreamExecutor device (0): NVIDIA GeForce RTX 5080 Laptop GPU, Compute Capability 12.0\n", "2025-12-05 19:27:01.170163: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:268] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "I0000 00:00:1764962821.240868 3828 device_compiler.h:188] Compiled cluster using XLA! This line is logged at most once for the lifetime of the process.\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962821.464875 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.467614 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.468165 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.468640 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.470913 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.474451 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.478425 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.479004 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.480505 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.481153 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.482828 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.483536 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.484279 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.485185 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.490247 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.490852 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.491255 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.511033 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.511544 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.511810 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.512084 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.512355 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.514637 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.514995 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.515247 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.515517 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.531617 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.538126 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.538644 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.539492 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962821.669624 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.670546 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.671179 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.672099 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.672974 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.673444 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.674342 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.675305 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.676438 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.677938 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.679247 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.680375 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.681752 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.683109 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.684692 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.686027 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.688515 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.693375 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.694000 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.694476 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.695000 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.695460 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.695961 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.696633 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.697166 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.697726 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.698742 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.699765 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.700664 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.701910 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962821.791579 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.792046 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.792307 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.792562 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.792816 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.793223 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.793555 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.794009 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.794286 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.794761 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.795255 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.795674 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.796283 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.798418 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.799103 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.799453 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.799824 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.800238 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.800555 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.801034 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.801675 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.802227 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.802742 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.803202 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.803811 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.804140 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.804862 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.805525 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.806390 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.807399 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.809566 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962821.922134 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.922928 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.923442 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.923956 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.924410 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.925010 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.925762 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.926755 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.927378 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.928646 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.929494 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.930195 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.941677 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.944257 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.945080 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.945679 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.946175 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.946760 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.947339 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.948295 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.948929 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.949630 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.950309 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.951223 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.951932 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.953298 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.954535 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.955102 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.957270 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.959907 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962821.968250 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962822.088961 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.089408 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.089760 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.092518 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.093341 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.094111 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.094869 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.096238 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.098165 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.099319 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.100627 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.103713 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.118737 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.120360 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 1/3750 [..............................] - ETA: 4:09:10 - loss: 1.5886 - accuracy: 0.2500 - top3_acc: 0.7500" ] }, { "name": "stderr", "output_type": "stream", "text": [ "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1764962822.200466 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.200979 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.201238 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.201586 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.202057 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.202507 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.203130 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.203743 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.204470 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.208168 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.209592 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.211917 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.214326 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.214837 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.215220 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.215643 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.216018 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.216393 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.216803 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.217200 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.217772 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.218343 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.218699 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.219183 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.219747 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.220214 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.220932 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.223950 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.231237 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.231662 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.231979 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.232306 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.232573 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.232980 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.233430 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.233915 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.234232 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.234722 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.235388 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.235999 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.236401 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.236841 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.237398 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.240406 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.241168 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.241525 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.241993 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.242460 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.243008 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.243795 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.244529 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.245354 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.246200 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.246989 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.247879 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.248854 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.249856 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.252805 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.255172 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.255648 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.255999 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.256513 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.256805 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.257086 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.257553 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.258028 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.258276 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.258832 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.259219 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.260073 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.260414 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.260942 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.261437 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.265841 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.266396 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.266764 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.267100 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.267370 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.267561 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.267950 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.268322 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.268719 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.269154 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.269635 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.270111 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.270639 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.271246 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.271968 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.272659 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.273242 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.281667 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.286385 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.286678 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.286976 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.287245 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.287497 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.287822 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.288073 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.288337 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.288854 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.289237 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.289651 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.290375 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.292511 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.293350 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.293849 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.294551 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.295201 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.295578 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.296240 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.296966 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.297824 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.299026 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.300033 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.300875 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.301938 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.302961 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.304066 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.305139 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.307273 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.310027 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.310516 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.310944 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.311409 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.311821 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.312268 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.312851 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.313374 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.313941 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.314818 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.315715 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.316515 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.317523 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.320263 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.320486 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.320748 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.321011 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.321262 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.321671 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.322003 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.322437 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.322722 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.323182 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.323656 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.324084 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.324654 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.326736 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.327340 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.327682 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.328093 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.328526 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.328873 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.329339 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.329980 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.330566 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.331099 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.331583 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.332199 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.332434 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.333071 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.333724 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.334569 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.335560 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.337647 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 3/3750 [..............................] - ETA: 10:45 - loss: 1.6209 - accuracy: 0.3750 - top3_acc: 0.7292 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962822.346157 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.346882 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.347466 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.348042 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.348470 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.349007 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.349686 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.350484 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.351023 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.352223 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.352931 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.353534 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.356386 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.358560 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.359352 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.359878 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.360360 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.360915 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.361459 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.362382 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.362982 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.363624 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.364261 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.365106 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.365769 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.367050 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.368167 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.368686 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.370800 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.373324 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.381384 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.392948 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.393296 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.393583 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.394048 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.394544 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.395056 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.395545 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.396375 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.397570 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.398348 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.399203 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.401158 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.412758 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.414093 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.426620 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.426901 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.427111 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.427442 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.427832 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.428186 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.428681 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.429191 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.429747 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.432637 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.433448 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.434550 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.436622 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.436968 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.437351 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.437685 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.438112 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.438554 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.439005 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.439497 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.440055 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.440621 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.441020 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.441569 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.442171 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.442582 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.443366 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.445970 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.452547 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.453049 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.453379 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.453804 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.454202 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.454730 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.455359 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.456020 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.456762 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.457515 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.458256 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.459109 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.459969 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.460854 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.463328 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.467324 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.467679 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.468011 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.468339 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.468594 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.468782 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.469137 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.469497 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.469860 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.470288 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.470807 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.471298 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.471787 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.472371 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.473068 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.473756 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.474328 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.482341 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.482574 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.482874 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.483154 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.483394 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.483642 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.483968 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.484223 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.484487 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.485006 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.485392 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.485797 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.486484 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.488872 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.489291 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.489655 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.490036 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.490460 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.490793 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.491252 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.491863 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.492428 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.492951 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.493425 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.494034 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.494254 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.494929 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.495566 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.496404 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.497305 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.499340 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.507950 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.509172 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.509467 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.509944 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.510469 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.510987 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.511504 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.512381 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.513636 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.514476 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.515397 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.517482 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.529779 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.531182 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.543590 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.543893 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.544123 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.544462 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.544879 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.545269 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.545781 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.546288 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.546854 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 5/3750 [..............................] - ETA: 8:06 - loss: 1.5823 - accuracy: 0.3625 - top3_acc: 0.7000 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962822.549765 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.550601 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.551753 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.554099 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.554503 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.554883 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.555270 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.555680 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.556105 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.556590 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.557138 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.557707 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.558273 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.558671 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.559218 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.559813 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.560290 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.561102 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.563658 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.570058 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.570561 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.570909 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.571341 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.571758 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.572251 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.572878 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.573565 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.574327 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.575087 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.575823 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.576639 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.577525 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.578439 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.580987 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.585568 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.585949 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.586284 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.586624 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.586879 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.587104 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.587458 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.587834 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.588196 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.588637 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.589267 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.589761 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.590331 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.590931 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.591635 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.592315 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.592889 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.601039 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.601261 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.601558 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.601874 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.602123 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.602372 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.602712 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.602958 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.603212 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.603724 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.604117 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.604527 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.605195 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.607623 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.607994 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.608326 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.608693 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.609121 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.609445 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.609904 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.610513 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.611070 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.611591 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.612044 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.612639 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.612852 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.613523 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.614209 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.615031 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.615932 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.617943 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.628540 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.628897 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.629180 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.629648 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.630160 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.630676 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.631187 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.632061 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.633324 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.634188 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.635105 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.637173 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.649361 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.650717 3831 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.662319 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.662751 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.663171 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.663542 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.663981 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.664427 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.664885 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.665412 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.666046 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.666695 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.667146 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.667745 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.668349 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.668907 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.669695 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.672242 3829 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.684427 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.684725 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.685029 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.685515 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.686043 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.686546 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.687060 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.687923 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.689164 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.689994 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.690907 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.692941 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.705305 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.706574 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.719597 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.720025 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.720425 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.720804 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.721231 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.721659 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.722094 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.722545 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.723140 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.723747 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.724169 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.724763 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.725374 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.725794 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.726577 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.729146 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.741672 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.741964 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.742230 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.742661 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.743098 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.743546 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.743994 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.744779 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.746039 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.746858 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.747762 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.749850 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 10/3750 [..............................] - ETA: 5:11 - loss: 1.5352 - accuracy: 0.4125 - top3_acc: 0.6938" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962822.762341 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.763833 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.776299 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.776759 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.777171 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.777542 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.777998 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.778460 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.778934 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.779446 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.780097 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.780775 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.781246 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.781890 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.782595 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.783135 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.784100 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.786863 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.797650 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.799995 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.800273 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.800731 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.801221 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.801710 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.802214 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.803048 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.804230 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.805022 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.805941 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.808012 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.820410 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.821804 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.834830 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.835219 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.835604 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.835949 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.836363 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.836777 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.837193 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.837616 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.838175 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.838768 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.839179 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.839785 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.840399 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.840913 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.841730 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.844292 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.854011 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.856200 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.856478 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.856922 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.857395 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.857875 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.858364 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.859140 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.860249 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.861003 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.861829 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.863688 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.875755 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.877118 3828 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.889757 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.890136 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.890523 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.890869 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.891281 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.891686 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.892133 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.892562 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.893157 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.893790 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.894230 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.894819 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.895417 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.895936 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.896733 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.899289 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.918187 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.918579 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.918979 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.919342 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.919770 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.920205 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.920653 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.921135 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.921746 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.922368 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.922803 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.923441 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.924143 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.924661 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.925538 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962822.928159 3827 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 58s 14ms/step - loss: 0.2132 - accuracy: 0.9550 - top3_acc: 0.6703 - val_loss: 0.1212 - val_accuracy: 0.9783 - val_top3_acc: 0.6175 - lr: 1.0000e-04\n", "Epoch 2/40\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962875.780763 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.780986 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.781121 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.781303 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.781550 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.781763 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.781987 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.782201 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.782519 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.784866 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.785338 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.785933 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.788023 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.788411 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.788586 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.788848 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.789188 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.789392 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.789833 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.790075 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.790412 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.790781 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.791047 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.791312 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.791617 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.791909 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.792325 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.794290 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.798114 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.798346 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.798512 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.798679 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.798872 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.799057 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.799334 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.799559 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.799869 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.800127 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.800371 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.800605 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.800870 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.801146 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.801604 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.802835 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.806704 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.807027 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.807271 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.807496 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.807796 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.808131 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.808645 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.808993 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.809456 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.809891 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.810297 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.810731 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.811231 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.811718 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.812526 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.816365 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.816608 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.816792 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.816966 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.817167 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.817356 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.817562 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.817807 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.818089 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.818392 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.818662 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.818928 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.819296 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.819616 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.820263 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962875.821028 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 55s 15ms/step - loss: 0.1025 - accuracy: 0.9865 - top3_acc: 0.6777 - val_loss: 0.1211 - val_accuracy: 0.9805 - val_top3_acc: 0.8163 - lr: 1.0000e-04\n", "Epoch 3/40\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962930.413974 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.414219 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.414377 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.414599 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.414860 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.415104 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.415378 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.415634 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.416007 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.418606 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.419145 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.419866 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.422132 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.422533 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.422728 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.423021 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.423365 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.423647 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.424169 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.424437 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.424820 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.425214 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.425463 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.425721 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.426024 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.426352 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.427295 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.429492 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.433851 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.434120 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.434290 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.434452 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.434651 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.434851 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.435158 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.435387 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.435722 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.435986 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.436237 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.436499 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.436825 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.437142 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.437635 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.438956 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.443349 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.443791 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.444069 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.444313 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.444865 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.445217 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.445781 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.446162 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.446703 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.447169 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.447607 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.448136 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.448696 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.449261 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.450167 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.454070 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.454290 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.454464 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.454611 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.454831 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.455021 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.455233 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.455482 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.455779 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.456071 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.456345 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.456600 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.456978 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.457383 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.458070 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962930.458834 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0850 - accuracy: 0.9898 - top3_acc: 0.6903 - val_loss: 0.3051 - val_accuracy: 0.9046 - val_top3_acc: 0.2930 - lr: 1.0000e-04\n", "Epoch 4/40\n", " 5/3750 [..............................] - ETA: 56s - loss: 0.1339 - accuracy: 0.9750 - top3_acc: 0.7375 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764962984.951731 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.952032 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.952254 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.952427 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.952617 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.952806 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.953108 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.953325 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.953647 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.953911 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.954150 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.954397 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.954701 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.955086 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.955560 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764962984.956863 3830 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0758 - accuracy: 0.9916 - top3_acc: 0.6907 - val_loss: 0.7950 - val_accuracy: 0.7987 - val_top3_acc: 0.7580 - lr: 1.0000e-04\n", "Epoch 5/40\n", " 5/3750 [..............................] - ETA: 53s - loss: 0.0573 - accuracy: 1.0000 - top3_acc: 0.6750 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764963039.110609 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.110888 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.111106 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.111269 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.111464 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.111668 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.111981 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.112222 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.112597 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.112861 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.113114 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.113368 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.113671 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.113998 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.114478 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764963039.115830 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0695 - accuracy: 0.9921 - top3_acc: 0.6944 - val_loss: 0.1501 - val_accuracy: 0.9650 - val_top3_acc: 0.8427 - lr: 1.0000e-04\n", "Epoch 6/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0571 - accuracy: 0.9952 - top3_acc: 0.6866 - val_loss: 0.0531 - val_accuracy: 0.9957 - val_top3_acc: 0.6773 - lr: 5.0000e-05\n", "Epoch 7/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0526 - accuracy: 0.9959 - top3_acc: 0.6833 - val_loss: 0.7447 - val_accuracy: 0.7954 - val_top3_acc: 0.4316 - lr: 5.0000e-05\n", "Epoch 8/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0486 - accuracy: 0.9966 - top3_acc: 0.6702 - val_loss: 0.2275 - val_accuracy: 0.9430 - val_top3_acc: 0.7697 - lr: 5.0000e-05\n", "Epoch 9/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0482 - accuracy: 0.9957 - top3_acc: 0.6592 - val_loss: 0.0453 - val_accuracy: 0.9962 - val_top3_acc: 0.5577 - lr: 5.0000e-05\n", "Epoch 10/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0458 - accuracy: 0.9960 - top3_acc: 0.6681 - val_loss: 1.6984 - val_accuracy: 0.6723 - val_top3_acc: 0.8062 - lr: 5.0000e-05\n", "Epoch 11/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0432 - accuracy: 0.9964 - top3_acc: 0.6712 - val_loss: 0.0851 - val_accuracy: 0.9837 - val_top3_acc: 0.6615 - lr: 5.0000e-05\n", "Epoch 12/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0401 - accuracy: 0.9969 - top3_acc: 0.6689 - val_loss: 0.2802 - val_accuracy: 0.9031 - val_top3_acc: 0.4696 - lr: 5.0000e-05\n", "Epoch 13/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0384 - accuracy: 0.9967 - top3_acc: 0.6658 - val_loss: 1.2961 - val_accuracy: 0.6603 - val_top3_acc: 0.9149 - lr: 5.0000e-05\n", "Epoch 14/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0357 - accuracy: 0.9972 - top3_acc: 0.6615 - val_loss: 0.0323 - val_accuracy: 0.9981 - val_top3_acc: 0.7861 - lr: 2.5000e-05\n", "Epoch 15/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0337 - accuracy: 0.9977 - top3_acc: 0.6645 - val_loss: 0.0477 - val_accuracy: 0.9931 - val_top3_acc: 0.7375 - lr: 2.5000e-05\n", "Epoch 16/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0329 - accuracy: 0.9974 - top3_acc: 0.6578 - val_loss: 0.0302 - val_accuracy: 0.9978 - val_top3_acc: 0.7482 - lr: 2.5000e-05\n", "Epoch 17/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0313 - accuracy: 0.9978 - top3_acc: 0.6587 - val_loss: 0.0264 - val_accuracy: 0.9991 - val_top3_acc: 0.6920 - lr: 2.5000e-05\n", "Epoch 18/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0314 - accuracy: 0.9975 - top3_acc: 0.6636 - val_loss: 0.0263 - val_accuracy: 0.9990 - val_top3_acc: 0.7213 - lr: 2.5000e-05\n", "Epoch 19/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0304 - accuracy: 0.9980 - top3_acc: 0.6579 - val_loss: 0.0261 - val_accuracy: 0.9988 - val_top3_acc: 0.6640 - lr: 2.5000e-05\n", "Epoch 20/40\n", "3750/3750 [==============================] - 57s 15ms/step - loss: 0.0301 - accuracy: 0.9976 - top3_acc: 0.6684 - val_loss: 0.0316 - val_accuracy: 0.9969 - val_top3_acc: 0.7527 - lr: 2.5000e-05\n", "Epoch 21/40\n", "3750/3750 [==============================] - 57s 15ms/step - loss: 0.0288 - accuracy: 0.9979 - top3_acc: 0.6588 - val_loss: 0.0268 - val_accuracy: 0.9984 - val_top3_acc: 0.6732 - lr: 2.5000e-05\n", "Epoch 22/40\n", "3750/3750 [==============================] - 58s 15ms/step - loss: 0.0279 - accuracy: 0.9978 - top3_acc: 0.6527 - val_loss: 0.0539 - val_accuracy: 0.9893 - val_top3_acc: 0.7961 - lr: 2.5000e-05\n", "Epoch 23/40\n", "3750/3750 [==============================] - 59s 16ms/step - loss: 0.0275 - accuracy: 0.9980 - top3_acc: 0.6619 - val_loss: 0.0242 - val_accuracy: 0.9988 - val_top3_acc: 0.7345 - lr: 2.5000e-05\n", "Epoch 24/40\n", "3750/3750 [==============================] - 58s 15ms/step - loss: 0.0271 - accuracy: 0.9982 - top3_acc: 0.6580 - val_loss: 0.0323 - val_accuracy: 0.9959 - val_top3_acc: 0.7932 - lr: 2.5000e-05\n", "Epoch 25/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0262 - accuracy: 0.9981 - top3_acc: 0.6639 - val_loss: 0.0238 - val_accuracy: 0.9983 - val_top3_acc: 0.7146 - lr: 2.5000e-05\n", "Epoch 26/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0249 - accuracy: 0.9984 - top3_acc: 0.6568 - val_loss: 0.0879 - val_accuracy: 0.9761 - val_top3_acc: 0.4728 - lr: 2.5000e-05\n", "Epoch 27/40\n", "3750/3750 [==============================] - 57s 15ms/step - loss: 0.0254 - accuracy: 0.9980 - top3_acc: 0.6592 - val_loss: 0.0234 - val_accuracy: 0.9983 - val_top3_acc: 0.7680 - lr: 2.5000e-05\n", "Epoch 28/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0249 - accuracy: 0.9982 - top3_acc: 0.6533 - val_loss: 0.0236 - val_accuracy: 0.9983 - val_top3_acc: 0.7088 - lr: 2.5000e-05\n", "Epoch 29/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0239 - accuracy: 0.9982 - top3_acc: 0.6490 - val_loss: 0.0209 - val_accuracy: 0.9989 - val_top3_acc: 0.7253 - lr: 2.5000e-05\n", "Epoch 30/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0241 - accuracy: 0.9981 - top3_acc: 0.6437 - val_loss: 0.0792 - val_accuracy: 0.9787 - val_top3_acc: 0.8204 - lr: 2.5000e-05\n", "Epoch 31/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0231 - accuracy: 0.9983 - top3_acc: 0.6457 - val_loss: 0.0344 - val_accuracy: 0.9952 - val_top3_acc: 0.7089 - lr: 2.5000e-05\n", "Epoch 32/40\n", "3750/3750 [==============================] - 55s 15ms/step - loss: 0.0230 - accuracy: 0.9981 - top3_acc: 0.6504 - val_loss: 0.0279 - val_accuracy: 0.9968 - val_top3_acc: 0.7577 - lr: 2.5000e-05\n", "Epoch 33/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0230 - accuracy: 0.9981 - top3_acc: 0.6415 - val_loss: 0.0278 - val_accuracy: 0.9969 - val_top3_acc: 0.5799 - lr: 2.5000e-05\n", "Epoch 34/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0209 - accuracy: 0.9988 - top3_acc: 0.6518 - val_loss: 0.0202 - val_accuracy: 0.9986 - val_top3_acc: 0.7493 - lr: 1.2500e-05\n", "Epoch 35/40\n", "3750/3750 [==============================] - 57s 15ms/step - loss: 0.0210 - accuracy: 0.9986 - top3_acc: 0.6360 - val_loss: 0.0195 - val_accuracy: 0.9986 - val_top3_acc: 0.6893 - lr: 1.2500e-05\n", "Epoch 36/40\n", "3750/3750 [==============================] - 56s 15ms/step - loss: 0.0202 - accuracy: 0.9989 - top3_acc: 0.6345 - val_loss: 0.0179 - val_accuracy: 0.9994 - val_top3_acc: 0.5762 - lr: 1.2500e-05\n", "Epoch 37/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0200 - accuracy: 0.9988 - top3_acc: 0.6409 - val_loss: 0.0199 - val_accuracy: 0.9986 - val_top3_acc: 0.7049 - lr: 1.2500e-05\n", "Epoch 38/40\n", "3750/3750 [==============================] - 53s 14ms/step - loss: 0.0196 - accuracy: 0.9988 - top3_acc: 0.6457 - val_loss: 0.0191 - val_accuracy: 0.9985 - val_top3_acc: 0.7480 - lr: 1.2500e-05\n", "Epoch 39/40\n", "3750/3750 [==============================] - 54s 14ms/step - loss: 0.0198 - accuracy: 0.9988 - top3_acc: 0.6406 - val_loss: 0.0258 - val_accuracy: 0.9974 - val_top3_acc: 0.7015 - lr: 1.2500e-05\n", "Epoch 40/40\n", "3750/3750 [==============================] - 53s 14ms/step - loss: 0.0195 - accuracy: 0.9988 - top3_acc: 0.6420 - val_loss: 0.0192 - val_accuracy: 0.9984 - val_top3_acc: 0.6848 - lr: 1.2500e-05\n" ] } ], "source": [ "history=model.fit(train_ds,validation_data=val_ds,epochs=40,callbacks=callbacks)" ] }, { "cell_type": "code", "execution_count": 19, "id": "77da217f-1dd4-4b7f-99fc-905915bbcb3e", "metadata": {}, "outputs": [], "source": [ "model=load_model(\"best_rice_cnn.keras\")" ] }, { "cell_type": "code", "execution_count": 20, "id": "47ff4197-39ae-4a65-b212-b119d4d508da", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "938/938 [==============================] - 5s 5ms/step - loss: 0.0179 - accuracy: 0.9994 - top3_acc: 0.5762\n", "\n", "Test-Resultate:\n", "loss: 0.0179\n", "accuracy: 0.9994\n", "top3_acc: 0.5762\n" ] } ], "source": [ "results = model.evaluate(val_ds)\n", "\n", "print(\"\\nTest-Resultate:\")\n", "for name, value in zip(model.metrics_names, results):\n", " print(f\"{name}: {value:.4f}\")" ] }, { "cell_type": "code", "execution_count": 32, "id": "eb333290-fa0c-426e-8305-1a8e115b6fed", "metadata": {}, "outputs": [], "source": [ "import pickle\n", "\n", "with open(\"last_rice_cnn_history.pkl\", \"wb\") as f:\n", " pickle.dump(history.history, f)" ] }, { "cell_type": "markdown", "id": "327fc242-9eed-43ea-9241-516a63097794", "metadata": {}, "source": [ "## 4. Results\n", "\n", "The trained model achieved very high performance on the validation dataset:\n", "\n", "- **Validation Accuracy:** **99.94%**\n", "- **Validation Loss:** **0.0179**\n", "\n", "The **confusion matrix** shows that almost all predictions are correct. \n", "The **classification report** shows **precision, recall and F1-score of 1.00 for all five classes**.\n", "\n", "Only **9 images were misclassified** out of 75,000 total images. The errors mainly occurred between visually similar rice types.\n", "\n", "The training and validation curves show that the model learned very well and did **not strongly overfit**.\n" ] }, { "cell_type": "code", "execution_count": 21, "id": "3a74c2ce-b09a-42ce-9728-594816d3f36f", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "epochs = range(len(history.history[\"accuracy\"]))\n", "\n", "plt.figure(figsize=(12, 5))\n", "\n", "# Accuracy\n", "plt.subplot(1, 2, 1)\n", "plt.plot(epochs, history.history[\"accuracy\"], label=\"Train Accuracy\")\n", "plt.plot(epochs, history.history[\"val_accuracy\"], label=\"Val Accuracy\")\n", "plt.title(\"Training vs Validation Accuracy\")\n", "plt.xlabel(\"Epoch\")\n", "plt.ylabel(\"Accuracy\")\n", "plt.legend()\n", "\n", "# Loss \n", "plt.subplot(1, 2, 2)\n", "plt.plot(epochs, history.history[\"loss\"], label=\"Train Loss\")\n", "plt.plot(epochs, history.history[\"val_loss\"], label=\"Val Loss\")\n", "plt.title(\"Training vs Validation Loss\")\n", "plt.xlabel(\"Epoch\")\n", "plt.ylabel(\"Loss\")\n", "plt.legend()\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 24, "id": "1e4e6adb-6f48-4ccb-8684-89befa744093", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "Classification Report:\n", "\n", " precision recall f1-score support\n", "\n", " Arborio 1.00 1.00 1.00 3022\n", " Basmati 1.00 1.00 1.00 3005\n", " Ipsala 1.00 1.00 1.00 3007\n", " Jasmine 1.00 1.00 1.00 3015\n", " Karacadag 1.00 1.00 1.00 2951\n", "\n", " accuracy 1.00 15000\n", " macro avg 1.00 1.00 1.00 15000\n", "weighted avg 1.00 1.00 1.00 15000\n", "\n" ] } ], "source": [ "y_true = []\n", "y_pred = []\n", "\n", "for images, labels in val_ds:\n", " preds = model.predict(images, verbose=0)\n", " y_true.extend(labels.numpy())\n", " y_pred.extend(np.argmax(preds, axis=1))\n", "\n", "y_true = np.array(y_true)\n", "y_pred = np.array(y_pred)\n", "\n", "# absoulut Confusion matrix\n", "cm = confusion_matrix(y_true, y_pred)\n", "\n", "plt.figure(figsize=(6, 5))\n", "plt.imshow(cm, interpolation=\"nearest\")\n", "plt.title(\"Confusion Matrix\")\n", "plt.colorbar()\n", "\n", "ticks = np.arange(len(class_names))\n", "plt.xticks(ticks, class_names, rotation=45)\n", "plt.yticks(ticks, class_names)\n", "\n", "plt.xlabel(\"Predicted\")\n", "plt.ylabel(\"True\")\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "# Normalized Confusion matrix\n", "cm_norm = cm.astype(\"float\") / cm.sum(axis=1, keepdims=True)\n", "\n", "plt.figure(figsize=(6, 5))\n", "plt.imshow(cm_norm, interpolation=\"nearest\")\n", "plt.title(\"Confusion Matrix (Normalized)\")\n", "plt.colorbar()\n", "\n", "plt.xticks(ticks, class_names, rotation=45)\n", "plt.yticks(ticks, class_names)\n", "\n", "plt.xlabel(\"Predicted\")\n", "plt.ylabel(\"True\")\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "# Classification report\n", "print(\"Classification Report:\\n\")\n", "print(classification_report(y_true, y_pred, target_names=class_names))" ] }, { "cell_type": "code", "execution_count": 25, "id": "ca6b8669-0f4b-40aa-b50e-42979b1f6d61", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Finded Fail Classification: 9\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "2025-12-05 20:24:14.334198: I tensorflow/core/framework/local_rendezvous.cc:404] Local rendezvous is aborting with status: OUT_OF_RANGE: End of sequence\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "wrong_images = []\n", "wrong_true = []\n", "wrong_pred = []\n", "\n", "# Failing Classification\n", "for images, labels in val_ds:\n", " preds = model.predict(images, verbose=0)\n", " pred_labels = np.argmax(preds, axis=1)\n", "\n", " for img, true_lbl, pred_lbl in zip(images, labels, pred_labels):\n", " if true_lbl != pred_lbl:\n", " wrong_images.append(img.numpy().astype(\"uint8\"))\n", " wrong_true.append(int(true_lbl.numpy()))\n", " wrong_pred.append(int(pred_lbl))\n", " if len(wrong_images) >= 12: \n", " break\n", " if len(wrong_images) >= 12:\n", " break\n", "\n", "print(f\"Finded Fail Classification: {len(wrong_images)}\")\n", "\n", "# Plotten\n", "plt.figure(figsize=(12, 8))\n", "for i in range(len(wrong_images)):\n", " ax = plt.subplot(3, 4, i + 1)\n", " plt.imshow(wrong_images[i])\n", " t = class_names[wrong_true[i]]\n", " p = class_names[wrong_pred[i]]\n", " plt.title(f\"True: {t}\\nPred: {p}\")\n", " plt.axis(\"off\")\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 29, "id": "883aba92-bad9-42ac-ba0e-f27467e69062", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Last Conv: conv2d_9\n" ] } ], "source": [ "def get_last_conv_layer_name(model):\n", " for layer in reversed(model.layers):\n", " if isinstance(layer, Conv2D):\n", " return layer.name\n", " raise ValueError(\"Not find Conv2D.\")\n", "\n", "last_conv_layer_name = get_last_conv_layer_name(model)\n", "print(\"Last Conv:\", last_conv_layer_name)" ] }, { "cell_type": "code", "execution_count": 30, "id": "66516bfd-301b-45bd-b4be-a526132c7c71", "metadata": {}, "outputs": [], "source": [ "def make_gradcam_heatmap(img_array, model, last_conv_layer_name, pred_index=None):\n", " grad_model = tf.keras.models.Model(\n", " [model.inputs],\n", " [model.get_layer(last_conv_layer_name).output, model.output]\n", " )\n", "\n", " with tf.GradientTape() as tape:\n", " conv_outputs, predictions = grad_model(img_array)\n", " if pred_index is None:\n", " pred_index = tf.argmax(predictions[0])\n", " class_channel = predictions[:, pred_index]\n", "\n", " #Gradients\n", " grads = tape.gradient(class_channel, conv_outputs)\n", " pooled_grads = tf.reduce_mean(grads, axis=(0, 1, 2))\n", "\n", " conv_outputs = conv_outputs[0]\n", "\n", " heatmap = conv_outputs @ pooled_grads[..., tf.newaxis]\n", " heatmap = tf.squeeze(heatmap)\n", "\n", " heatmap = tf.maximum(heatmap, 0) / tf.reduce_max(heatmap)\n", " return heatmap.numpy()" ] }, { "cell_type": "code", "execution_count": 31, "id": "e0e8abec-2500-407d-94e2-010809043d11", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1764966609.670484 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.671005 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.671472 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.671828 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.672197 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.672484 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.672839 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.673190 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.673410 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.673541 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.673754 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.673899 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.676969 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.677333 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.677430 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.677559 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.677743 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678093 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678210 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678324 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678435 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678549 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.678730 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.679074 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.679202 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.679338 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.679465 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.679681 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.680280 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.692383 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.692692 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.692787 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.692918 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693035 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693163 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693272 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693392 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693498 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693748 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693855 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.693975 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.694122 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.694247 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.694379 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.694576 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.724689 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.727942 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.731942 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.732163 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.732313 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.732597 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.732749 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.732899 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.733057 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.733222 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.733432 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.733739 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.733893 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.734071 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.734413 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.734626 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.734955 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.735269 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.798933 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.814363 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.819601 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.820085 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.820488 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.820923 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.821065 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.821306 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.821654 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.821860 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.822170 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.822315 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.822475 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.822812 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.823158 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.823488 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.823738 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.824134 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.837046 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848118 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848390 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848513 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848662 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848796 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.848943 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.849204 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.849313 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.849555 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.849908 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.850146 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.850341 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.863259 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.866499 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.866741 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.866873 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867000 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867189 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867542 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867693 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867817 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.867939 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868046 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868151 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868259 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868373 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868469 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868726 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.868905 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.869011 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.872231 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.872564 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.872736 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.872994 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.873085 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.873292 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.873405 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.873626 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.873877 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.874152 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.874293 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.874788 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.875506 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.878888 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.879356 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.879633 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880026 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880343 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880453 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880578 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880744 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.880903 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881053 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881238 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881359 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881464 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881609 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881760 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.881939 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.882131 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890102 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890334 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890421 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890537 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890631 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890791 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.890920 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.891010 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.891129 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.891277 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.891418 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.891904 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.892198 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.894364 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.894622 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.894733 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.894844 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.894994 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.895288 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.895533 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.895624 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.895736 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.895861 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896077 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896181 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896299 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896453 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896746 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.896989 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.897183 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.902404 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.902678 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.902763 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903029 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903275 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903407 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903547 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903650 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903767 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.903935 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.904085 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.904478 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.906597 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.906887 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.906993 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907080 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907217 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907289 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907369 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907500 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907574 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907688 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907805 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.907923 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.908071 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.908269 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.911107 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.911529 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913263 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913534 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913622 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913747 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913868 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.913973 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.914157 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.914403 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.914585 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.914792 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.914926 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.915917 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917076 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917340 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917455 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917520 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917675 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917781 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.917969 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.918206 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.918366 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.918627 3691 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.955605 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.955913 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.955987 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956106 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956229 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956305 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956370 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956473 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956567 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956738 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.956919 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.957485 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959323 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959461 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959532 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959621 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959686 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959764 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959860 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.959953 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960039 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960189 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960299 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960470 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960570 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.960877 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.961213 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.961522 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.964836 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.965233 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.965371 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.965499 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.965665 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.965843 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.966060 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.966258 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.966478 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.966785 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.967023 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.967317 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.967583 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.967917 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.968327 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.969306 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.972218 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.972767 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.972969 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.973202 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.973467 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.973689 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.973935 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.974214 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.974473 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.974991 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.975404 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.975849 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.976315 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.976920 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.977866 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.979763 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.984614 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.984833 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.985020 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.985298 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.985509 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.985758 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.986008 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.986302 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.986553 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.987000 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.987433 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.987865 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.988329 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.988723 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.989500 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1764966609.990510 3825 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# One Picture + Label from val_ds\n", "for images, labels in val_ds.take(1):\n", " img = images[0] # Tensor\n", " true_label = int(labels[0].numpy())\n", " break\n", "\n", "# Batch-Dimension\n", "img_batch = tf.expand_dims(img, axis=0)\n", "\n", "# Heatmap\n", "heatmap = make_gradcam_heatmap(img_batch, model, last_conv_layer_name)\n", "heatmap_resized = tf.image.resize(\n", " heatmap[..., tf.newaxis],\n", " (img_size[0], img_size[1])\n", ").numpy().squeeze()\n", "\n", "# Prediction for this picture\n", "preds = model.predict(img_batch, verbose=0)\n", "pred_idx = int(tf.argmax(preds[0]).numpy())\n", "pred_class = class_names[pred_idx]\n", "true_class = class_names[true_label]\n", "\n", "# Plot, Original, Heatmap, Overlay\n", "plt.figure(figsize=(12, 4))\n", "\n", "plt.subplot(1, 3, 1)\n", "plt.title(f\"Original\\nTrue: {true_class}\")\n", "plt.imshow(img.numpy().astype(\"uint8\"))\n", "plt.axis(\"off\")\n", "\n", "plt.subplot(1, 3, 2)\n", "plt.title(\"Grad-CAM Heatmap\")\n", "plt.imshow(heatmap_resized, cmap=\"jet\")\n", "plt.axis(\"off\")\n", "\n", "plt.subplot(1, 3, 3)\n", "plt.title(f\"Overlay\\nPred: {pred_class}\")\n", "plt.imshow(img.numpy().astype(\"uint8\"))\n", "plt.imshow(heatmap_resized, cmap=\"jet\", alpha=0.4)\n", "plt.axis(\"off\")\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "3551605d-5d7e-467d-829f-ffaee604fea0", "metadata": {}, "source": [ "## 5. Reflection\n", "\n", "This project shows that a well-designed CNN from scratch can achieve **excellent performance** on a clean and balanced image dataset.\n", "\n", "Data augmentation, batch normalization, dropout and L2 regularization helped to make the training **stable and robust**.\n", "\n", "However, the dataset is very clean and simple, with a black background. In real-world environments, the performance might be lower.\n", "\n", "In future work, I plan to:\n", "\n", "- Test the model on real photos\n", "- Apply stronger augmentations\n", "- Compare this CNN with a **transfer learning model** such as ResNet50" ] }, { "cell_type": "markdown", "id": "38c782be-fb4c-492b-a465-c93b52025b8d", "metadata": {}, "source": [ "# Resnet50 Model" ] }, { "cell_type": "markdown", "id": "6c84f4e3-4a9e-43c7-9170-4c9d97317629", "metadata": {}, "source": [ "## 1. Introduction\n", "\n", "The goal of this project is to classify different rice types using a deep learning model with **transfer learning**. \n", "For this task, a pre-trained **ResNet50** model was used and adapted for a custom rice dataset.\n", "\n", "The problem is a **multi-class image classification task** with five different rice classes. \n", "Transfer learning was chosen because it allows faster training and better performance with limited data." ] }, { "cell_type": "code", "execution_count": 5, "id": "da451e42-e0dd-46a3-90fa-3bfca789d8d3", "metadata": {}, "outputs": [], "source": [ "from tensorflow.keras.applications import ResNet50\n", "from tensorflow.keras.applications.resnet50 import preprocess_input" ] }, { "cell_type": "code", "execution_count": 6, "id": "479e105a-4e27-4fd5-8425-cd847f014e79", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset\n" ] }, { "data": { "text/plain": [ "[PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Ipsala'),\n", " PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Karacadag'),\n", " PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Arborio'),\n", " PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Basmati'),\n", " PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Jasmine'),\n", " PosixPath('/root/.cache/kagglehub/datasets/muratkokludataset/rice-image-dataset/versions/1/Rice_Image_Dataset/Rice_Citation_Request.txt')]" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "print(data_dir)\n", "list(data_dir.iterdir())" ] }, { "cell_type": "code", "execution_count": 7, "id": "59454369-55db-49ff-b4ef-d53b9834b50c", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Found 60000 images belonging to 5 classes.\n", "Found 15000 images belonging to 5 classes.\n" ] } ], "source": [ "img_size = (224, 224)\n", "batch_size = 16\n", "\n", "train_datagen = ImageDataGenerator(\n", " preprocessing_function=preprocess_input,\n", " validation_split=0.2,\n", " rotation_range=25,\n", " width_shift_range=0.15,\n", " height_shift_range=0.15,\n", " zoom_range=0.15,\n", " shear_range=0.15,\n", " horizontal_flip=True,\n", " fill_mode='nearest'\n", ")\n", "\n", "val_datagen = ImageDataGenerator(\n", " preprocessing_function=preprocess_input,\n", " validation_split=0.2\n", ")\n", "\n", "train_data = train_datagen.flow_from_directory(\n", " data_dir,\n", " target_size=img_size,\n", " batch_size=batch_size,\n", " subset='training',\n", " class_mode='sparse',\n", " shuffle=True\n", ")\n", "\n", "val_data = val_datagen.flow_from_directory(\n", " data_dir,\n", " target_size=img_size,\n", " batch_size=batch_size,\n", " subset='validation',\n", " class_mode='sparse',\n", " shuffle=False\n", ")" ] }, { "cell_type": "markdown", "id": "8076592e-520e-42f3-9881-d2bbe29364c6", "metadata": {}, "source": [ "## 2. Analysis\n", "\n", "Before training, the dataset was analyzed and prepared using image generators.\n", "\n", "Main points from the analysis:\n", "\n", "- The dataset contains **five rice types**\n", "- All images have similar background and colors\n", "- Some rice types look **very similar in shape and texture**\n", "- This makes classification difficult even for humans\n", "\n", "To improve generalization, **data augmentation** was used." ] }, { "cell_type": "code", "execution_count": 8, "id": "6311a191-03d3-4dc3-95a1-cec28e8f2515", "metadata": {}, "outputs": [], "source": [ "callbacks = [\n", " EarlyStopping(\n", " monitor='val_loss',\n", " patience=3,\n", " restore_best_weights=True\n", " ),\n", "\n", " ModelCheckpoint(\n", " filepath='best_rice_resnet50.keras',\n", " monitor='val_loss',\n", " save_best_only=True\n", " ),\n", "\n", " ReduceLROnPlateau(\n", " monitor='val_loss',\n", " factor=0.5, \n", " patience=2, \n", " min_lr=1e-6,\n", " verbose=1\n", " )\n", "]" ] }, { "cell_type": "code", "execution_count": 9, "id": "fed7007d-8627-41aa-bd4a-75798ee2df3a", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n", "I0000 00:00:1765021877.543076 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021878.165643 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021878.165705 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021878.169666 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021878.169720 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021878.169731 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021879.484446 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "I0000 00:00:1765021879.484548 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "2025-12-06 11:51:19.484563: I tensorflow/core/common_runtime/gpu/gpu_device.cc:2112] Could not identify NUMA node of platform GPU id 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n", "I0000 00:00:1765021879.484619 208 cuda_executor.cc:1001] could not open file to read NUMA node: /sys/bus/pci/devices/0000:02:00.0/numa_node\n", "Your kernel may have been built without NUMA support.\n", "2025-12-06 11:51:19.485454: I tensorflow/core/common_runtime/gpu/gpu_device.cc:2021] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 10861 MB memory: -> device: 0, name: NVIDIA GeForce RTX 5080 Laptop GPU, pci bus id: 0000:02:00.0, compute capability: 12.0\n" ] } ], "source": [ "base_model = ResNet50(\n", " weights='imagenet',\n", " include_top=False,\n", " input_shape=(224, 224, 3)\n", ")\n", "\n", "base_model.trainable = False\n", "\n", "model = Sequential()\n", "model.add(base_model)\n", "model.add(GlobalAveragePooling2D())\n", "model.add(Dense(256, activation='relu'))\n", "model.add(Dropout(0.5))\n", "model.add(Dense(9, activation='softmax'))\n", "\n", "model.compile(\n", " optimizer=Adam(learning_rate=1e-3),\n", " loss='sparse_categorical_crossentropy',\n", " metrics=['accuracy', tf.keras.metrics.SparseTopKCategoricalAccuracy(k=3, name='top3_acc')]\n", ")" ] }, { "cell_type": "markdown", "id": "2c3903cf-ef7e-4c56-8920-7991458e5dcd", "metadata": {}, "source": [ "## 3. Methods\n", "\n", "### Data Loading & Preprocessing\n", "\n", "The images were loaded using `ImageDataGenerator` with the following settings:\n", "\n", "- ResNet50 preprocessing (`preprocess_input`)\n", "- Rotation\n", "- Width shift\n", "- Height shift\n", "- Zoom\n", "- Shear\n", "- Horizontal flip\n", "- Validation split: **20%**\n", "- Class mode: **sparse**\n", "\n", "---\n", "\n", "### Model Architecture\n", "\n", "- Base model: **ResNet50 (ImageNet weights)**\n", "- `include_top = False`\n", "- All base model layers were **frozen**\n", "\n", "Custom classifier:\n", "\n", "- Global Average Pooling\n", "- Dense layer with **256 units (ReLU)**\n", "- Dropout **0.5**\n", "- Output layer with **Softmax**\n", "\n", "---\n", "\n", "### Training Setup\n", "\n", "- Optimizer: **Adam**\n", "- Learning rate: **0.001**\n", "- Loss function: **Sparse Categorical Crossentropy**\n", "- Metrics:\n", " - Accuracy\n", " - Top-3 Accuracy\n", "\n", "---\n", "\n", "### Callbacks Used\n", "\n", "- **EarlyStopping**\n", "- **ModelCheckpoint**\n", "- **ReduceLROnPlateau**\n", "\n", "These callbacks helped to:\n", "- Stop training early\n", "- Save the best model\n", "- Reduce the learning rate automatically" ] }, { "cell_type": "code", "execution_count": 10, "id": "a05e9fda-dc2a-4dc8-835d-91a648cab6fc", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/15\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "2025-12-06 11:51:30.190494: I external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:531] Loaded cuDNN version 90701\n", "W0000 00:00:1765021890.648833 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.683775 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.690729 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.723755 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.727039 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.730747 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.735248 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.840456 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.845374 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.901563 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.912741 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021890.929762 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.195489 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.201779 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.202472 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.209212 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.211174 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.214633 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.216684 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.218407 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.219434 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.253181 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.255726 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.257685 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.258927 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.260937 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.262447 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.264089 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.332984 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021891.353473 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1765021892.149337 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.155074 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.155787 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.158502 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.160449 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.161275 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.162700 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.164757 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.165526 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.166978 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.169692 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.171682 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.173324 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.175197 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.176821 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.178631 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.183880 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.204279 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.230502 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.289365 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.295889 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.296962 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.297987 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.298763 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.299599 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.300394 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.301162 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.302177 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.303005 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.304080 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.305439 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.306854 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.308476 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.311827 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.312326 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.312837 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.313576 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.314437 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.315205 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.315797 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.316358 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.316957 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.317559 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.318089 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.318853 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.319918 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.320510 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.321155 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.323394 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.323866 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.324399 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.324924 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.325401 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.326097 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.326695 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.327385 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.328197 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.329226 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.330390 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.331540 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.332885 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.334383 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.335939 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.337473 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.338915 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.342578 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.343389 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.344062 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.344834 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.345241 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.345795 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.346252 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.346975 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.347842 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.349054 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.350188 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.351171 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.352182 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.353650 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.355246 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.356848 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.358101 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.362734 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.364393 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.365913 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.367442 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.368616 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.369846 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.371144 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.371810 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.373032 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.374099 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.375280 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.377397 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.379514 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.385615 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.387323 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.390395 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.391184 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.392160 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.393470 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.394724 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.395369 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.395823 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.397705 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.398946 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.399286 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.399546 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.400468 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.401387 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.401730 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.402684 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.405654 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.406156 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.407126 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.408147 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.409088 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.410514 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.411788 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.412778 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.413569 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.414658 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.415681 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.417029 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.417915 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.418892 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.420720 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.422167 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.426295 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.432702 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.433530 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.434510 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.436055 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.437170 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.438265 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.439269 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.444693 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.446074 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.447335 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.448685 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.450274 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.451694 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.452547 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.454195 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.456018 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.460031 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.463900 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.465745 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.466425 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.467096 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.468152 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.469324 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.470139 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.470813 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.471518 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.472316 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.473246 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.474025 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.475730 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.476878 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.478630 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.481521 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.482136 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.482541 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.482988 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.483444 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.484003 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.484398 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.485102 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.485553 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.485872 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.486492 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.486970 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.487399 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.487878 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.488808 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.491591 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.492694 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.493315 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.493825 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.494380 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.494995 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.495566 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.496472 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.497473 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.498280 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.499839 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.500980 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.502064 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.503516 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.504984 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.506461 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.511355 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.527829 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.528514 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.529052 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.529928 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.531046 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.531521 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.532158 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.532744 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.533908 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.534995 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.537359 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.538554 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.539723 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.541369 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.542736 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.544431 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021892.549291 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n", "I0000 00:00:1765021893.071456 398 service.cc:146] XLA service 0x773ca8391d90 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n", "I0000 00:00:1765021893.071511 398 service.cc:154] StreamExecutor device (0): NVIDIA GeForce RTX 5080 Laptop GPU, Compute Capability 12.0\n", "2025-12-06 11:51:33.098493: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:268] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "I0000 00:00:1765021893.259627 398 device_compiler.h:188] Compiled cluster using XLA! This line is logged at most once for the lifetime of the process.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 1/3750 [..............................] - ETA: 8:02:49 - loss: 3.4520 - accuracy: 0.1250 - top3_acc: 0.3750" ] }, { "name": "stderr", "output_type": "stream", "text": [ "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "'+ptx85' is not a recognized feature for this target (ignoring feature)\n", "W0000 00:00:1765021893.491408 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.492810 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.494635 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 2/3750 [..............................] - ETA: 13:00 - loss: 4.7873 - accuracy: 0.1562 - top3_acc: 0.4062 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021893.496793 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.498752 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.500626 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.503308 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.505860 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.507963 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.510388 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.513028 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.516715 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.527976 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.536885 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.539689 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.541531 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.542939 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.545228 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.546956 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.548282 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.549809 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.551435 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.553345 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.554582 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.557167 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.558909 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.561106 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.565160 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.565903 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.567155 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.568245 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.569954 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.571688 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.573291 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.574482 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.575097 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.576470 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.577472 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.578902 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.580380 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.581699 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.582766 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.587091 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.590619 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.592305 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.593710 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.595270 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.596680 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.597870 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.599344 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.600992 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.602183 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.604279 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.605521 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.607855 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.608974 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.610942 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.614704 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.615581 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.616870 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.617549 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.618022 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.618867 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.620303 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.621358 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.622034 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.622918 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.623368 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.624689 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.631093 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.632471 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.633309 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.638262 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.639954 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.641453 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.642746 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.644342 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.645083 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.646510 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.647829 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.649400 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.650666 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.651863 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.653677 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.655416 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.656421 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.658144 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.660802 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.661785 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.662151 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.662847 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.663886 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.664617 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.665554 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.666866 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.668074 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.668860 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.669782 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.670483 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.671783 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.672499 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.673739 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 3/3750 [..............................] - ETA: 10:53 - loss: 3.7761 - accuracy: 0.2500 - top3_acc: 0.5208" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021893.702268 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.703767 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.705415 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.707236 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.708599 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.710624 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.712934 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.714864 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.716703 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.718518 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.720572 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.723279 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.733045 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.743514 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.745591 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.746601 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.747764 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.749118 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.750587 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.751973 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.753399 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.755332 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.756965 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.758753 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.760863 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.763109 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.764546 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.768246 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.768913 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.769624 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.770528 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.771888 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.772246 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.773421 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.773910 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.775194 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.776299 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.777478 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.779017 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.780189 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.781258 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.782143 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.787890 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.795892 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.797424 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.798336 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.799389 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.800102 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.801526 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.802239 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.803739 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.804875 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.811111 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.812386 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.813936 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.815665 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.817074 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.843449 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.845166 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.852011 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.854219 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.856689 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.859145 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.861312 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.863506 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.865243 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.867196 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.869841 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.872349 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.882035 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.892220 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.893981 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.895556 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.897167 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.898432 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.899682 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.900858 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.902473 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.904149 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.905348 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 5/3750 [..............................] - ETA: 9:11 - loss: 3.2476 - accuracy: 0.2875 - top3_acc: 0.6000 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021893.906745 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.908984 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.910374 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.911778 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.915423 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.916815 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.917667 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.918657 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.919664 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.920447 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.920775 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.921777 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.922261 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.923577 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.924467 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.925579 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.926825 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.927659 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.929121 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.934394 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.942627 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.943639 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.943883 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.945297 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.945542 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.947220 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.947734 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.948646 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.949166 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.949971 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.950622 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.951407 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.952903 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.953942 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.980662 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.982320 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.983824 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.986044 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.987615 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.989092 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.990944 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.992604 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.993876 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.995513 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021893.997785 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.000777 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.010659 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.020836 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.021726 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.022632 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.024843 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.025513 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.026381 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.026861 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.027477 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.028494 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.029360 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.030235 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.031020 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.032039 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.033139 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.037812 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.046902 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.047881 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.048792 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.049866 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.050143 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.050423 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.051814 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.052396 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.052839 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.053914 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.055215 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.056091 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.056554 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.057424 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.082554 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.083443 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.084321 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.086604 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.088837 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.090609 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.092838 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.095138 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.096455 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.098813 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.101113 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.104009 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 8/3750 [..............................] - ETA: 7:21 - loss: 2.7950 - accuracy: 0.3359 - top3_acc: 0.6953" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021894.113558 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.124436 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.124888 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.126138 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.127211 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.128483 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.129507 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.130731 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.131673 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.133199 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.134310 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.135503 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.136382 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.137177 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.137928 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.142394 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.152578 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.153202 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.154223 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.154909 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.155159 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.156181 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.156571 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.157236 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.158001 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.158847 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.159540 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.160205 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.160860 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.161615 400 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.205474 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.221130 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.222105 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.223182 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.223938 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.224853 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.225729 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.226285 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.226837 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.227355 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.228268 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.228867 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.229597 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.230710 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.231151 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.261990 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.279473 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.279950 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.281103 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.282115 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.282969 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.283827 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.285735 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.286572 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.287301 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.288364 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.289309 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.291039 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.292550 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.293398 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 10/3750 [..............................] - ETA: 7:14 - loss: 2.4756 - accuracy: 0.3812 - top3_acc: 0.7563" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021894.360112 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.384577 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.385193 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.386321 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.387702 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.388150 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.389015 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.390100 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.391412 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.393369 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.394473 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.395036 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.396312 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.396782 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.397968 398 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.475934 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.498064 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.499009 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.499878 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.500719 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.501259 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.501862 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.503116 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.503841 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.504582 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.506025 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.507069 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.508220 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.508940 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.509942 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " 12/3750 [..............................] - ETA: 7:00 - loss: 2.2332 - accuracy: 0.4062 - top3_acc: 0.7865" ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765021894.582361 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.605246 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.606209 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.607769 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.608870 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.609811 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.611181 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.611900 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.612958 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.613507 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.614478 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.616624 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.617641 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.618794 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765021894.620359 397 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - ETA: 0s - loss: 0.1433 - accuracy: 0.9510 - top3_acc: 0.9985 " ] }, { "name": "stderr", "output_type": "stream", "text": [ "W0000 00:00:1765022276.212023 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.215066 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.216188 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.217207 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.218169 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.219241 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.221126 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.222629 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.224024 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.225407 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.226580 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.228668 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.235811 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.236747 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.237659 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.238590 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.239412 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.240175 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.241285 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.242234 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.242992 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.244225 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.245340 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.246474 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.247357 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.248610 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.249707 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.250789 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.252020 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.269322 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.273606 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.276283 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.276985 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.277892 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.278831 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.279856 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.280579 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.281464 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.282676 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.283524 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.284342 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.285499 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.286744 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.287792 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.288867 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.291149 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.291500 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.292064 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.292539 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.293010 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.293478 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.294079 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.294654 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.295090 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.295629 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.296138 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.296747 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.297288 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.297792 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.298356 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.298876 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.305332 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.306232 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.307185 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.308020 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.308507 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.309085 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.309920 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.310563 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.311381 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.312560 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.313726 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.314927 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.315739 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.316854 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.317703 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.318509 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.319788 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.321183 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.341395 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.342169 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.342895 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.343621 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.344319 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.344965 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.345584 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.346239 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.347039 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.347852 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.348626 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.349445 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.350424 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.351267 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.352447 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.354933 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.355751 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.356489 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.356940 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.357408 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.358234 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.358758 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.359401 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.360108 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.360823 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.361573 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.362325 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.362781 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.363184 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.363766 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.364131 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.369833 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.370847 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.371431 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.372064 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.372965 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.374030 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.375411 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.376442 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.377430 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.378551 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.380048 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.381170 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.382557 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.383404 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.384531 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.385457 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.389522 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.405724 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.410466 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.413931 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.415072 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.416286 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.417218 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.418323 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.419450 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.420452 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.421433 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.422170 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.423214 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.424868 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.425811 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.427014 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.428412 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.431761 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.432562 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.433354 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.434088 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.434607 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.435131 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.435910 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.436680 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.437121 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.437698 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.438255 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.438852 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.439641 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.440138 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.440623 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.441167 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.448250 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.449071 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.449467 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.450029 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.450606 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.451533 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.452423 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.453366 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.454403 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.455387 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.456516 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.457695 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.459309 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.460758 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.462213 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.463889 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.477659 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n", "W0000 00:00:1765022276.494341 401 gpu_timer.cc:114] Skipping the delay kernel, measurement accuracy will be reduced\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "3750/3750 [==============================] - 392s 103ms/step - loss: 0.1433 - accuracy: 0.9510 - top3_acc: 0.9985 - val_loss: 0.0531 - val_accuracy: 0.9829 - val_top3_acc: 0.9999 - lr: 0.0010\n", "Epoch 2/15\n", "3750/3750 [==============================] - 337s 90ms/step - loss: 0.0945 - accuracy: 0.9679 - top3_acc: 0.9996 - val_loss: 0.0408 - val_accuracy: 0.9868 - val_top3_acc: 1.0000 - lr: 0.0010\n", "Epoch 3/15\n", "3750/3750 [==============================] - 347s 93ms/step - loss: 0.0812 - accuracy: 0.9731 - top3_acc: 0.9997 - val_loss: 0.0373 - val_accuracy: 0.9878 - val_top3_acc: 0.9999 - lr: 0.0010\n", "Epoch 4/15\n", "3750/3750 [==============================] - 340s 91ms/step - loss: 0.0763 - accuracy: 0.9760 - top3_acc: 0.9997 - val_loss: 0.0296 - val_accuracy: 0.9909 - val_top3_acc: 1.0000 - lr: 0.0010\n", "Epoch 5/15\n", "3750/3750 [==============================] - 351s 93ms/step - loss: 0.0702 - accuracy: 0.9773 - top3_acc: 0.9998 - val_loss: 0.0394 - val_accuracy: 0.9872 - val_top3_acc: 0.9999 - lr: 0.0010\n", "Epoch 6/15\n", "3750/3750 [==============================] - 368s 98ms/step - loss: 0.0663 - accuracy: 0.9790 - top3_acc: 0.9998 - val_loss: 0.0257 - val_accuracy: 0.9919 - val_top3_acc: 0.9999 - lr: 0.0010\n", "Epoch 7/15\n", "3750/3750 [==============================] - 838s 223ms/step - loss: 0.0650 - accuracy: 0.9798 - top3_acc: 0.9998 - val_loss: 0.0250 - val_accuracy: 0.9929 - val_top3_acc: 1.0000 - lr: 0.0010\n", "Epoch 8/15\n", "3750/3750 [==============================] - 874s 233ms/step - loss: 0.0623 - accuracy: 0.9810 - top3_acc: 0.9998 - val_loss: 0.0273 - val_accuracy: 0.9913 - val_top3_acc: 1.0000 - lr: 0.0010\n", "Epoch 9/15\n", "3750/3750 [==============================] - ETA: 0s - loss: 0.0577 - accuracy: 0.9819 - top3_acc: 0.9998 \n", "Epoch 9: ReduceLROnPlateau reducing learning rate to 0.0005000000237487257.\n", "3750/3750 [==============================] - 35400s 9s/step - loss: 0.0577 - accuracy: 0.9819 - top3_acc: 0.9998 - val_loss: 0.0334 - val_accuracy: 0.9906 - val_top3_acc: 0.9999 - lr: 0.0010\n", "Epoch 10/15\n", "3750/3750 [==============================] - 215s 57ms/step - loss: 0.0448 - accuracy: 0.9863 - top3_acc: 0.9999 - val_loss: 0.0259 - val_accuracy: 0.9921 - val_top3_acc: 1.0000 - lr: 5.0000e-04\n" ] } ], "source": [ "history = model.fit(train_data,validation_data=val_data,epochs=15,callbacks=callbacks)" ] }, { "cell_type": "code", "execution_count": 11, "id": "1cb27ae8-3052-48e4-a1c8-3bbafdc0da4d", "metadata": {}, "outputs": [], "source": [ "with open(\"rice_resnet50_history.pkl\", \"wb\") as f:\n", " pickle.dump(history.history, f)" ] }, { "cell_type": "code", "execution_count": 12, "id": "392af239-e8bf-4d69-9b12-8c247b2020b3", "metadata": {}, "outputs": [], "source": [ "best_model=tf.keras.models.load_model(\"best_rice_resnet50.keras\")" ] }, { "cell_type": "code", "execution_count": 27, "id": "8b6d7218-ac1b-423e-a67f-dfe390fb1879", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "938/938 [==============================] - 18s 19ms/step - loss: 0.0250 - accuracy: 0.9929 - top3_acc: 1.0000\n", "\n", "Test-Resultate:\n", "loss: 0.0250\n", "accuracy: 0.9929\n", "top3_acc: 1.0000\n" ] } ], "source": [ "results = model.evaluate(val_data)\n", "\n", "print(\"\\nTest-Resultate:\")\n", "for name, value in zip(model.metrics_names, results):\n", " print(f\"{name}: {value:.4f}\")" ] }, { "cell_type": "code", "execution_count": 14, "id": "84fb20e1-4477-42d1-86ec-6b3c695e94a9", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "938/938 [==============================] - 17s 17ms/step\n", " precision recall f1-score support\n", "\n", " Arborio 0.99 0.99 0.99 3000\n", " Basmati 1.00 0.99 0.99 3000\n", " Ipsala 1.00 1.00 1.00 3000\n", " Jasmine 0.99 0.99 0.99 3000\n", " Karacadag 0.99 0.99 0.99 3000\n", "\n", " accuracy 0.99 15000\n", " macro avg 0.99 0.99 0.99 15000\n", "weighted avg 0.99 0.99 0.99 15000\n", "\n" ] } ], "source": [ "y_true = val_data.classes\n", "y_pred_probs = best_model.predict(val_data)\n", "y_pred = np.argmax(y_pred_probs, axis=1)\n", "\n", "print(classification_report(\n", " y_true, \n", " y_pred, \n", " target_names=val_data.class_indices.keys()\n", "))" ] }, { "cell_type": "code", "execution_count": 15, "id": "038211cb-8172-401c-8c0f-9e7d12e35896", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Confusion Matrix\n", "cm=confusion_matrix(y_true, y_pred)\n", "\n", "plt.figure(figsize=(10, 8))\n", "sns.heatmap(\n", " cm,\n", " annot=True,\n", " fmt=\"d\",\n", " cmap=\"Blues\",\n", " xticklabels=list(val_data.class_indices.keys()),\n", " yticklabels=list(val_data.class_indices.keys())\n", ")\n", "\n", "plt.title(\"Confusion Matrix – Fish Classification (ResNet50)\")\n", "plt.xlabel(\"Predicted Label\")\n", "plt.ylabel(\"True Label\")\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 16, "id": "aa65d5f0-8c37-42c0-bedd-883aa37ea542", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Accuracy & Loss Plot\n", "\n", "plt.figure(figsize=(14, 5))\n", "\n", "# Accuracy\n", "plt.subplot(1, 2, 1)\n", "plt.plot(history.history['accuracy'], label='Train Accuracy')\n", "plt.plot(history.history['val_accuracy'], label='Validation Accuracy')\n", "plt.title('ResNet50 – Training & Validation Accuracy')\n", "plt.xlabel('Epoch')\n", "plt.ylabel('Accuracy')\n", "plt.legend()\n", "plt.grid(True)\n", "\n", "# Loss\n", "plt.subplot(1, 2, 2)\n", "plt.plot(history.history['loss'], label='Train Loss')\n", "plt.plot(history.history['val_loss'], label='Validation Loss')\n", "plt.title('ResNet50 – Training & Validation Loss')\n", "plt.xlabel('Epoch')\n", "plt.ylabel('Loss')\n", "plt.legend()\n", "plt.grid(True)\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 20, "id": "75858436-b01a-4668-a1e7-24ed27fe7ab8", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['Arborio', 'Basmati', 'Ipsala', 'Jasmine', 'Karacadag']\n" ] } ], "source": [ "class_names = [k for k, v in sorted(train_data.class_indices.items(), key=lambda x: x[1])]\n", "print(class_names)" ] }, { "cell_type": "code", "execution_count": 21, "id": "77c689d1-e09d-473c-b5ec-3cf22b062827", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Finded Fail Classification: 12\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "wrong_images = []\n", "wrong_true = []\n", "wrong_pred = []\n", "\n", "# Failing Classification\n", "for images, labels in val_data:\n", " preds = model.predict(images, verbose=0)\n", " pred_labels = np.argmax(preds, axis=1)\n", "\n", " for img, true_lbl, pred_lbl in zip(images, labels, pred_labels):\n", " if true_lbl != pred_lbl:\n", " wrong_images.append(img.astype(\"uint8\"))\n", " wrong_true.append(int(true_lbl))\n", " wrong_pred.append(int(pred_lbl))\n", " if len(wrong_images) >= 12: \n", " break\n", " if len(wrong_images) >= 12:\n", " break\n", "\n", "print(f\"Finded Fail Classification: {len(wrong_images)}\")\n", "\n", "# Plotten\n", "plt.figure(figsize=(12, 8))\n", "for i in range(len(wrong_images)):\n", " ax = plt.subplot(3, 4, i + 1)\n", " plt.imshow(wrong_images[i])\n", " t = class_names[wrong_true[i]]\n", " p = class_names[wrong_pred[i]]\n", " plt.title(f\"True: {t}\\nPred: {p}\")\n", " plt.axis(\"off\")\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "7fd64e67-c9c6-4a4c-9073-33a2557dffbe", "metadata": {}, "source": [ "## 4. Results\n", "\n", "### Final Model Performance\n", "\n", "After training, the best model was evaluated on the validation set.\n", "\n", "**Final Results:**\n", "\n", "- **Loss:** 0.0250 \n", "- **Accuracy:** 0.9929 → **99.29%**\n", "- **Top-3 Accuracy:** 1.0000 → **100%**\n", "\n", "This means that the model predicts the correct class almost perfectly.\n", "\n", "---\n", "\n", "### Classification Report\n", "\n", "All five rice classes achieved:\n", "\n", "- Precision ≈ **0.99 – 1.00**\n", "- Recall ≈ **0.99 – 1.00**\n", "- F1-score ≈ **0.99**\n", "\n", "This shows a **very balanced and stable performance** across all classes.\n", "\n", "---\n", "\n", "### Confusion Matrix\n", "\n", "- Most images are classified correctly.\n", "- The few errors appear mainly in the **Arborio** class.\n", "- Arborio is sometimes confused with:\n", " - Karacadag\n", " - Jasmine\n", " - Ipsala\n", "\n", "These mistakes happen because the rice grains look **very similar**.\n", "\n", "---\n", "\n", "### Misclassified Images\n", "\n", "12 misclassified images were visualized.\n", "\n", "Observations:\n", "\n", "- All misclassified samples belong to **Arborio**\n", "- Predictions were mainly:\n", " - Karacadag\n", " - Jasmine\n", " - Ipsala\n", "- The grains have:\n", " - Very similar shape\n", " - Similar texture\n", " - Similar color\n", "\n", "These errors are **normal and understandable**.\n", "\n", "---\n", "\n", "### Training Curves\n", "\n", "- Training accuracy increased steadily\n", "- Validation accuracy stayed above **99%**\n", "- Loss values decreased smoothly\n", "- No strong overfitting was observed" ] }, { "cell_type": "markdown", "id": "5f97fe33-5b5d-4192-91fc-5e66a6d8539b", "metadata": {}, "source": [ "## 5. Reflection\n", "\n", "This project showed that **transfer learning with ResNet50 works extremely well** for rice image classification.\n", "\n", "### What worked well:\n", "\n", "- Pretrained ResNet50 features\n", "- Data augmentation\n", "- Early stopping\n", "- Learning rate reduction\n", "\n", "### What can be improved:\n", "\n", "- Fine-tuning the last ResNet layers\n", "- Using more training data\n", "- Better background normalization\n", "\n", "---\n", "\n", "## Final Conclusion\n", "\n", "The ResNet50 transfer learning model achieved **excellent performance** with:\n", "\n", "- **99.29% accuracy**\n", "- **100% Top-3 accuracy**\n", "- Very few classification errors\n", "- Strong generalization\n", "\n", "The model is **stable, reliable, and ready for deployment**." ] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.3" } }, "nbformat": 4, "nbformat_minor": 5 }