tensorboard

# TensorBoard：机器学习可视化工具包 ## 何时使用此技能 当您需要执行以下操作时，请使用 TensorBoard： - **可视化训练指标**，如随时间变化的损失（loss）和准确率（accuracy） - **调试模型**，通过直方图和分布图 - **比较实验**，跨多个运行周期 - **可视化模型图**和架构 - **投影嵌入（Embeddings）**到低维空间（t-SNE, PCA） - **跟踪超参数**实验 - **分析性能**并识别瓶颈 - **可视化图像和文本**在训练过程中 **用户量**：2000万+ 下载/年 | **GitHub 星标**：2.7万+ | **许可证**：Apache 2.0 ## 安装 bash # 安装 TensorBoard pip install tensorboard # PyTorch 集成 pip install torch torchvision tensorboard # TensorFlow 集成 (已包含 TensorBoard) pip install tensorflow # 启动 TensorBoard tensorboard --logdir=runs # 访问地址 http://localhost:6006 ## 快速入门 ### PyTorch python from torch.utils.tensorboard import SummaryWriter # 创建 writer writer = SummaryWriter('runs/experiment_1') # 训练循环 for epoch in range(10): train_loss = train_epoch() val_acc = validate() # 记录指标 writer.add_scalar('Loss/train', train_loss, epoch) writer.add_scalar('Accuracy/val', val_acc, epoch) # 关闭 writer writer.close() # 启动: tensorboard --logdir=runs ### TensorFlow/Keras python import tensorflow as tf # 创建回调 tensorboard_callback = tf.keras.callbacks.TensorBoard( log_dir='logs/fit', histogram_freq=1 ) # 训练模型 model.fit( x_train, y_train, epochs=10, validation_data=(x_val, y_val), callbacks=[tensorboard_callback] ) # 启动: tensorboard --logdir=logs ## 核心概念 ### 1. SummaryWriter (PyTorch) python from torch.utils.tensorboard import SummaryWriter # 默认目录: runs/CURRENT_DATETIME writer = SummaryWriter() # 自定义目录 writer = SummaryWriter('runs/experiment_1') # 自定义注释 (附加到默认目录) writer = SummaryWriter(comment='baseline') # 记录数据 writer.add_scalar('Loss/train', 0.5, step=0) writer.add_scalar('Loss/train', 0.3, step=1) # 刷新并关闭 writer.flush() writer.close() ### 2. 记录标量 (Scalars) python # PyTorch from torch.utils.tensorboard import SummaryWriter writer = SummaryWriter() for epoch in range(100): train_loss = train() val_loss = validate() # 记录单个指标 writer.add_scalar('Loss/train', train_loss, epoch) writer.add_scalar('Loss/v