Solver可理解為模型舵手。因為它控制著模型的訓(xùn)練和測試階段，并規(guī)定了學(xué)習(xí)的方法和學(xué)習(xí)的參數(shù)。
The solver orchestrates （安排）model optimization by coordinating（協(xié)調(diào)） the network’s forward inference and backward gradients to form（產(chǎn)生） parameter updates that attempt to improve the loss.
The Solver optimizes a model by：
（1）calls network forward to compute the output and loss（前向）
（2）calls network backward to compute the gradients（反向）
（3）incorporates the gradients into parameter updates according to the solver method （權(quán)重更新）
（4）updates the solver state according to learning rate, history, and method（分離解耦）

The Caffe solvers mathod are:

Stochastic Gradient Descent (type: "SGD"),#隨機(jī)梯度下降 AdaDelta (type: "AdaDelta"), Adaptive Gradient (type: "AdaGrad"), Adam (type: "Adam"), Nesterov’s Accelerated Gradient (type: "Nesterov") and RMSprop (type: "RMSProp")

The solver 作用
1. 參數(shù)字典
2. 迭代優(yōu)化參數(shù)
3. 周期性的得到測試結(jié)果
4. snapshots the model and solver state throughout the optimization（？）

Snapshotting and Resuming
The weight snapshots export the learned model while the solver snapshots allow training to be resumed from a given point.

Solver::Snapshot()和Solver::SnapshotSolverState()保存指定迭代次數(shù)中的參數(shù)并停止，相當(dāng)于斷點(diǎn)，由Solver::Restore()和Solver::RestoreSolverState()繼續(xù)運(yùn)行。
參見博文http://blog.csdn.net/u013508213/article/details/53208434#solver

總結(jié)

以上是生活随笔為你收集整理的caffe教程笔记《Solver》的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。