cs231n 学习笔记（5）- 神经网络-3：Learning and Evaluation

目录 1. Gradient checks 2. Sanity Checks - Before learning 3. Babysitting the learning process 3.1 Loss function 3.2 Train/Val accuracy 3.3 Ratio of weights:updates 权重的更新幅度 3.4 Activation / Gradients di

>>阅读原文<<