Caffe源碼-幾種優化算法

SGD簡介

caffe中的SGDSolver類中實現了帶動量的梯度降低法,其原理以下,\(lr\)爲學習率,\(m\)爲動量參數。html

  1. 計算新的動量:history_data = local_rate * param_diff + momentum * history_data
    \(\nu_{t+1}=lr*\nabla_{\theta_{t}}+m*\nu_{t}\)
  2. 計算更新時使用的梯度:param_diff = history_data
    \(\Delta\theta_{t+1}=\nu_{t+1}\)
  3. 應用更新:param_data = param_data - param_diff
    \(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

步驟1和步驟2均在SGDSolver類的ComputeUpdateValue()函數中實現,步驟3對每一個優化方法來講都是相同的,代碼可參考以前的博客:Caffe源碼-SGDSolver類git

NAG(Nesterov Accelerated Gradient)簡介

NAG算法在NesterovSolver類中實現,NAG與SGD相比惟一區別在於梯度的計算上。如上,SGD使用的梯度是參數\(\theta_{t}\)在當前位置的梯度\(\nabla_{\theta_{t}}\),而NAG中使用的是當前參數\(\theta_{t}\)在施加了動量以後的位置的梯度\(\nabla_{(\theta_{t}-m*\nu_{t})}\),其原理爲:github

  1. 應用臨時更新:\(\tilde{\theta}_{t+1}=\theta_{t}-m*\nu_{t}\)
  2. 計算該位置的梯度:\(\nabla_{\tilde{\theta}_{t+1}}\)
  3. 計算新的動量:\(\nu_{t+1}=lr*\nabla_{\tilde{\theta}_{t+1}}+m*\nu_{t}\)
  4. 獲得更新時使用的梯度:\(\Delta\theta_{t+1}=\nu_{t+1}\)
  5. 應用更新:\(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

網絡上有一張常見的圖用於表示SGD和NAG的過程。
算法

對於SGD算法,藍色向量\(p_{1}\)爲當前參數\(\theta_{t}\)在該位置的梯度\(lr*\nabla_{\theta_{t}}\),藍色向量\(p_{2}\)爲動量\(m*\nu_{t}\),而\(p_{1}+p_{2}\)即爲參數一次的更新量\(\Delta\theta_{t+1}\)
對於NAG算法,\(O_{1}\)爲參數\(\theta_{t}\)的初始位置,棕色向量\(p_{3}=p_{2}\),先計算運用動量後的參數\(\tilde{\theta}_{t+1}\)的位置\(O_{2}\),而後計算該位置梯度\(\nabla_{\tilde{\theta}_{t+1}}\),即爲圖中的紅色向量\(p_{4}\),而\(p_{5}=p_{3}+p_{4}\)即爲參數一次的更新量\(\Delta\theta_{t+1}=\nu_{t+1}\)。以後仿照該步驟計算下一次迭代的動量\(m*\nu_{t+1}\)(棕色向量\(p_{6}\))和梯度\(\nabla_{\tilde{\theta}_{t+2}}\)(紅色向量\(p_{7}\)),獲得更新量\(p_{8}\)網絡

NAG算法的原理仍是很好理解的,可是實現起來卻有一個很是難理解的地方,即如何計算參數臨時更新位置的梯度\(\nabla_{\tilde{\theta}_{t+1}}\)?神經網絡這種複雜的系統中想要根據當前位置的梯度\(\nabla_{\theta_{t}}\)來估算另外一位置的梯度\(\nabla_{\tilde{\theta}_{t+1}}\)幾乎是不可能的。網絡上關於該算法的實現細節很是少,不過結合caffe代碼和其餘的開源代碼等,能夠判斷出,NAG算法每次迭代時保存的參數是臨時參數\(\tilde{\theta}_{t+1}\)(位置\(O_{2}\)),而非初始\(O_{1}\)位置處的參數\(\theta_{t}\),這樣每次反向傳播計算出的梯度實際上就是紅色向量\(p_{4}\)。而後每次更新時,會根據動量\(p_{3}\)先將參數從位置\(O_{2}\)退回\(O_{1}\),而後計算獲得一次迭代的更新量\(p_{5}\),使參數更新\(\theta_{t+1}\)(位置\(O_{3}\)),並保存下一次迭代時須要使用的臨時參數\(\tilde{\theta}_{t+2}\)(位置\(O_{4}\))。app

nesterov_solver.cpp源碼

//根據當前迭代次數對應的學習率rate,計算網絡中第param_id個可學習參數在更新時使用的梯度
template <typename Dtype>
void NesterovSolver<Dtype>::ComputeUpdateValue(int param_id, Dtype rate) {
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //網絡中的全部可學習參數
  const vector<float>& net_params_lr = this->net_->params_lr();   //網絡中每一個參數對應的學習率係數
  Dtype momentum = this->param_.momentum();                       //求解器設置的動量
  Dtype local_rate = rate * net_params_lr[param_id];              //獲得當前參數對應的學習率
  switch (Caffe::mode()) {
  case Caffe::CPU: {    //CPU模式
    // save history momentum for stepping back
    caffe_copy(net_params[param_id]->count(), this->history_[param_id]->cpu_data(),
        this->update_[param_id]->mutable_cpu_data());   //將歷史數據history_拷貝至update_中,update_data = history_data

    // update history     //history_data = local_rate * net_params_diff + momentum * history_data
    caffe_cpu_axpby(net_params[param_id]->count(), local_rate, net_params[param_id]->cpu_diff(), momentum,
        this->history_[param_id]->mutable_cpu_data());

    // compute update: step back then over step   //update_data = (1 + momentum) * history_data + (-momentum) * update_data
    caffe_cpu_axpby(net_params[param_id]->count(), Dtype(1) + momentum,
        this->history_[param_id]->cpu_data(), -momentum,
        this->update_[param_id]->mutable_cpu_data());

    // copy   //net_params_diff = update_data
    caffe_copy(net_params[param_id]->count(), this->update_[param_id]->cpu_data(),
        net_params[param_id]->mutable_cpu_diff());
    break;
  }
  case Caffe::GPU: {
#ifndef CPU_ONLY
    // gpu的操做同理
    // h_temp = history_data
    // history_data = momentum * h_temp + local_rate * net_params_diff
    // net_params_diff = (1+momentum) * history_data - momentum * h_temp
    nesterov_update_gpu(net_params[param_id]->count(), net_params[param_id]->mutable_gpu_diff(),
        this->history_[param_id]->mutable_gpu_data(), momentum, local_rate);
#else
    NO_GPU;
#endif
    break;
  }
  default:
    LOG(FATAL) << "Unknown caffe mode: " << Caffe::mode();
  }
}

對應上述的說明,代碼中的各步操做爲:ide

  1. 當前迭代的動量\(\nu_{t}\)update_data = history_data
  2. net_params_diff臨時位置的參數的梯度\(\nabla_{\tilde{\theta}_{t+1}}\),計算新的動量:history_data = local_rate * net_params_diff + momentum * history_data
    \(\nu_{t+1}=lr*\nabla_{\tilde{\theta}_{t+1}}+m*\nu_{t}\)
  3. 計算下一次迭代的臨時參數相對於當前臨時參數的更新量\(\Delta\tilde{\theta}_{t+2}\)update_data = (1 + momentum) * history_data + (-momentum) * update_data
    \(\Delta\tilde{\theta}_{t+2}=(1+m)*\nu_{t+1}-m*\nu_{t}\)
    注意,當前臨時參數在位置\(O_{2}\),須要減去向量\(p_{3}\)\(p_{3}=m*\nu_{t}\)),再加上向量\(p_{5}\)\(p_{6}\)\(p_{5}=\nu_{t+1},p_{6}=m*\nu_{t+1}\))才能獲得新的臨時位置\(O_{4}\)
  4. 保存參數更新量:net_params_diff = update_data
  5. 應用更新:\(\tilde{\theta}_{t+2}=\tilde{\theta}_{t+1}-\Delta\tilde{\theta}_{t+2}\)

AdaGrad簡介

AdaGrad算法經過縮放每一個參數反比於其全部梯度歷史平方值總和的平方跟,可以使得具備較大梯度的參數可以快速降低,使具備小偏導的參數可以緩慢降低。
其原理以下,初始累積變量\(r=0\)\(\delta\)爲較小常數,防止除法除數太小而不穩定。函數

  1. 累加平方梯度(\(\odot\)爲逐元素點乘):\(r_{t+1}=r_{t}+\nabla_{\theta_{t}}\odot\nabla_{\theta_{t}}\)
  2. 計算梯度的更新量:\(\Delta\theta_{t+1}=\frac{lr}{\delta+\sqrt{r_{t+1}}}\odot\nabla_{\theta_{t}}\)
  3. 應用更新:\(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

adagrad_solver.cpp源碼

template <typename Dtype>
void AdaGradSolver<Dtype>::ComputeUpdateValue(int param_id, Dtype rate) {
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();
  const vector<float>& net_params_lr = this->net_->params_lr();
  Dtype delta = this->param_.delta();
  Dtype local_rate = rate * net_params_lr[param_id];
  switch (Caffe::mode()) {
  case Caffe::CPU: {
    // compute square of gradient in update
    caffe_powx(net_params[param_id]->count(),
        net_params[param_id]->cpu_diff(), Dtype(2),
        this->update_[param_id]->mutable_cpu_data());   //update_data = net_params ^ 2

    // update history
    caffe_add(net_params[param_id]->count(),
        this->update_[param_id]->cpu_data(),
        this->history_[param_id]->cpu_data(),
        this->history_[param_id]->mutable_cpu_data());  //history_data = update_data + history_data

    // prepare update
    caffe_powx(net_params[param_id]->count(), this->history_[param_id]->cpu_data(), Dtype(0.5),
        this->update_[param_id]->mutable_cpu_data());   //update_data = history_data ^ 0.5

    caffe_add_scalar(net_params[param_id]->count(),
        delta, this->update_[param_id]->mutable_cpu_data());  //update_data += delta

    caffe_div(net_params[param_id]->count(),
              net_params[param_id]->cpu_diff(),
              this->update_[param_id]->cpu_data(),
              this->update_[param_id]->mutable_cpu_data());   //update_data = net_params_diff / update_data

    // scale and copy
    caffe_cpu_axpby(net_params[param_id]->count(), local_rate,
        this->update_[param_id]->cpu_data(), Dtype(0),
        net_params[param_id]->mutable_cpu_diff());    //net_params_diff = local_rate * update_data + 0 * net_params_diff
    break;
  }
  case Caffe::GPU: {    //gpu操做同理
#ifndef CPU_ONLY
    // gi = net_params_diff;
    // hi = history_data = history_data + gi*gi;
    // net_params_diff = local_rate * gi / (sqrt(hi) + delta);
    adagrad_update_gpu(net_params[param_id]->count(),
        net_params[param_id]->mutable_gpu_diff(),
        this->history_[param_id]->mutable_gpu_data(), delta, local_rate);
#else
    NO_GPU;
#endif
    break;
  }
  default:
    LOG(FATAL) << "Unknown caffe mode: " << Caffe::mode();
  }
}

AdaGrad/RMSProp/AdaDelta/Adam算法的caffe代碼很容易找到對應的公式,再也不詳細介紹。學習

RMSProp簡介

RMSProp算法在AdaGrad基礎上增長一個衰減係數\(\rho\),以便將很早以前的歷史梯度數據丟棄。
其原理以下,初始累積變量\(r=0\)\(\delta\)一樣爲較小常數。優化

  1. 累加平方梯度:\(r_{t+1}=\rho*r_{t}+(1-\rho)*\nabla_{\theta_{t}}\odot\nabla_{\theta_{t}}\)
  2. 計算梯度的更新量:\(\Delta\theta_{t+1}=\frac{lr}{\delta+\sqrt{r_{t+1}}}\odot\nabla_{\theta_{t}}\)
  3. 應用更新:\(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

rmsprop_solver.cpp源碼

template <typename Dtype>
void RMSPropSolver<Dtype>::ComputeUpdateValue(int param_id, Dtype rate) {
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //全部可學習參數
  const vector<float>& net_params_lr = this->net_->params_lr();     //參數對應的學習率係數
  // get the learning rate
  Dtype delta = this->param_.delta();                   //常數delta
  Dtype rms_decay = this->param_.rms_decay();           //衰減速率
  Dtype local_rate = rate * net_params_lr[param_id];    //參數對應的學習率

  switch (Caffe::mode()) {
  case Caffe::CPU:
    // compute square of gradient in update
    caffe_powx(net_params[param_id]->count(), net_params[param_id]->cpu_diff(), Dtype(2),
        this->update_[param_id]->mutable_cpu_data());   //update_data = net_params_diff ^ 2

    // update history       //history_data = (1-rms_decay) * update_data + rms_decay * history_data
    caffe_cpu_axpby(net_params[param_id] -> count(), Dtype(1-rms_decay), this->update_[param_id]->cpu_data(),
        rms_decay, this->history_[param_id]-> mutable_cpu_data());

    // prepare update
    caffe_powx(net_params[param_id]->count(), this->history_[param_id]->cpu_data(), Dtype(0.5),
        this->update_[param_id]->mutable_cpu_data());   //update_data = history_data ^ 0.5

    caffe_add_scalar(net_params[param_id]->count(),
        delta, this->update_[param_id]->mutable_cpu_data());    //update_data += delta

    //update_data = net_params_diff / update_data
    caffe_div(net_params[param_id]->count(), net_params[param_id]->cpu_diff(),
        this->update_[param_id]->cpu_data(), this->update_[param_id]->mutable_cpu_data());

    // scale and copy
    caffe_cpu_axpby(net_params[param_id]->count(), local_rate,
        this->update_[param_id]->cpu_data(), Dtype(0),
        net_params[param_id]->mutable_cpu_diff());  //net_params_diff = local_rate * update_data + 0 * net_params_diff
    break;
  case Caffe::GPU:
#ifndef CPU_ONLY
    // g = net_params_diff
    // h = history_data
    // gi = g[i];
    // hi = h[i] = rms_decay*h[i] + (1-rms_decay)*gi*gi;
    // g[i] = local_rate * g[i] / (sqrt(hi) + delta);
    rmsprop_update_gpu(net_params[param_id]->count(),
        net_params[param_id]->mutable_gpu_diff(),
        this->history_[param_id]->mutable_gpu_data(),
        rms_decay, delta, local_rate);
#else
    NO_GPU;
#endif
    break;
  default:
    LOG(FATAL) << "Unknown caffe mode: " << Caffe::mode();
  }
}

AdaDelta簡介

AdaDelta也像RMSProp算法同樣在AdaGrad基礎上增長一個衰減係數\(\rho\),而且還額外維護一個狀態量\(x\)
其原理以下,初始累積變量\(x=0, r=0\)\(\delta\)一樣爲較小常數。

  1. 累加平方梯度:\(r_{t+1}=\rho*r_{t}+(1-\rho)*\nabla_{\theta_{t}}\odot\nabla_{\theta_{t}}\)
  2. 計算不帶學習率的梯度的更新量:\(\Delta\tilde{\theta}_{t+1}=\sqrt{\frac{x_{t}+\delta}{r_{t+1}+\delta}}\odot\nabla_{\theta_{t}}\)
  3. 更新狀態量:\(x_{t+1}=\rho*x_{t}+(1-\rho)*\Delta\tilde{\theta}_{t+1}\odot\Delta\tilde{\theta}_{t+1}\)
  4. 計算帶學習率的梯度的更新量:\(\Delta\theta_{t+1}=lr*\Delta\tilde{\theta}_{t+1}\)
    參考 4中的說明不一樣,caffe代碼中仍然有使用學習率\(lr\)
  5. 應用更新:\(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

adadelta_solver.cpp源碼

template <typename Dtype>
void AdaDeltaSolver<Dtype>::AdaDeltaPreSolve() {        //AdaDeltaSolver類在構造時會調用該函數
  // Add the extra history entries for AdaDelta after those from SGDSolver::PreSolve
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //當前網絡中的全部可學習參數
  for (int i = 0; i < net_params.size(); ++i) {
    const vector<int>& shape = net_params[i]->shape();  //第i個可學習參數的形狀
    //在SGDSolver<Dtype>::PreSolve中history_已經存入一個與參數blob相同形狀的空blob,此處再存入一個
    this->history_.push_back(shared_ptr<Blob<Dtype> >(new Blob<Dtype>(shape)));
  }
}

#ifndef CPU_ONLY
template <typename Dtype>
void adadelta_update_gpu(int N, Dtype* g, Dtype* h, Dtype* h2, Dtype momentum,
    Dtype delta, Dtype local_rate);
#endif

template <typename Dtype>
void AdaDeltaSolver<Dtype>::ComputeUpdateValue(int param_id, Dtype rate) {
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //網絡中的全部可學習參數
  const vector<float>& net_params_lr = this->net_->params_lr();     //每一個參數對應的學習率係數
  Dtype delta = this->param_.delta();                   //AdaDelta方法中的一個參數
  Dtype momentum = this->param_.momentum();             //動量係數
  Dtype local_rate = rate * net_params_lr[param_id];    //獲得當前參數對應的學習率
  size_t update_history_offset = net_params.size();     //網絡的參數個數
  //history_在AdaDeltaPreSolve()中又存入了一次與全部參數形狀相同的空blob,下面將
  //history_[param_id]表示成 history_former, history_[update_history_offset + param_id]表示成 history_latter
  switch (Caffe::mode()) {
  case Caffe::CPU: {
    // compute square of gradient in update
    caffe_powx(net_params[param_id]->count(), net_params[param_id]->cpu_diff(), Dtype(2),
        this->update_[param_id]->mutable_cpu_data());       //update_data = net_params_diff ^ 2

    // update history of gradients    //history_former_data = (1 - momentum) * update_data + momentum * history_former_data
    caffe_cpu_axpby(net_params[param_id]->count(), Dtype(1) - momentum, this->update_[param_id]->cpu_data(),
         momentum, this->history_[param_id]->mutable_cpu_data());

    // add delta to history to guard against dividing by zero later
    caffe_set(net_params[param_id]->count(), delta,
        this->temp_[param_id]->mutable_cpu_data());     //temp_中每一個元素都置爲delta, temp_data = delta

    caffe_add(net_params[param_id]->count(),
        this->temp_[param_id]->cpu_data(),
        this->history_[update_history_offset + param_id]->cpu_data(),
        this->update_[param_id]->mutable_cpu_data());   //update_data = temp_data + history_latter_data

    caffe_add(net_params[param_id]->count(),
        this->temp_[param_id]->cpu_data(),
        this->history_[param_id]->cpu_data(),
        this->temp_[param_id]->mutable_cpu_data()); //temp_data = temp_data + history_former_data

    // divide history of updates by history of gradients
    caffe_div(net_params[param_id]->count(),
        this->update_[param_id]->cpu_data(),
        this->temp_[param_id]->cpu_data(),
        this->update_[param_id]->mutable_cpu_data());   //update_data = update_data / temp_data

    // jointly compute the RMS of both for update and gradient history
    caffe_powx(net_params[param_id]->count(),
        this->update_[param_id]->cpu_data(), Dtype(0.5),
        this->update_[param_id]->mutable_cpu_data());   //update_data = update_data ^ 0.5

    // compute the update
    caffe_mul(net_params[param_id]->count(),
        net_params[param_id]->cpu_diff(),
        this->update_[param_id]->cpu_data(),
        net_params[param_id]->mutable_cpu_diff());  //net_params_diff = net_params_diff * update_data

    // compute square of update
    caffe_powx(net_params[param_id]->count(),
        net_params[param_id]->cpu_diff(), Dtype(2),
        this->update_[param_id]->mutable_cpu_data());   //update_data = net_params_diff ^ 2

    // update history of updates    //history_latter_data = (1 - momentum) * update_data + momentum * history_latter_data
    caffe_cpu_axpby(net_params[param_id]->count(), Dtype(1) - momentum,
        this->update_[param_id]->cpu_data(), momentum,
        this->history_[update_history_offset + param_id]->mutable_cpu_data());

    // apply learning rate
    caffe_cpu_scale(net_params[param_id]->count(), local_rate,
        net_params[param_id]->cpu_diff(),
        net_params[param_id]->mutable_cpu_diff());  //net_params_diff = local_rate * net_params_diff
    break;
  }
  case Caffe::GPU: {
#ifndef CPU_ONLY
    // g = net_params_diff;
    // h = history_former_data;
    // h2 = history_latter_data;
    // gi = g[i];
    // hi = h[i] = momentum * h[i] + (1-momentum) * gi * gi;
    // gi = gi * sqrt((h2[i] + delta) / (hi + delta));
    // h2[i] = momentum * h2[i] + (1-momentum) * gi * gi;
    // g[i] = local_rate * gi;
    adadelta_update_gpu(net_params[param_id]->count(),
        net_params[param_id]->mutable_gpu_diff(),
        this->history_[param_id]->mutable_gpu_data(),
        this->history_[update_history_offset + param_id]->mutable_gpu_data(),
        momentum, delta, local_rate);
#else
    NO_GPU;
#endif
    break;
  }
  default:
    LOG(FATAL) << "Unknown caffe mode: " << Caffe::mode();
  }
}

Adam簡介

Adam算法包含兩個衰減參數\(\rho_{1}\)\(\rho_{2}\),通常\(\rho_{1}=0.9, \rho_{2}=0.999\)。還包含一階矩和二階矩變量\(s, r\),時間步\(t\)
初始時\(s=0, r=0, t=0\)\(\delta\)一樣爲較小常數。

  1. 更新一階矩:\(s_{t+1}=\rho_{1}*s_{t}+(1-\rho_{1})*\nabla_{\theta_{t}}\)
  2. 更新二階矩:\(r_{t+1}=\rho_{2}*r_{t}+(1-\rho_{2})*\nabla_{\theta_{t}}\odot\nabla_{\theta_{t}}\)
  3. 修正一階矩的誤差:\(\tilde{s}_{t+1}=\frac{s_{t+1}}{1-\rho_{1}^{t+1}}\)
  4. 修正二階矩的誤差:\(\tilde{r}_{t+1}=\frac{r_{t+1}}{1-\rho_{2}^{t+1}}\)
  5. 計算梯度的更新量:\(\Delta\theta_{t+1}=lr*\frac{\tilde{s}_{t+1}}{\sqrt{\tilde{r}_{t+1}}+\delta}\)

  6. 應用更新:\(\theta_{t+1}=\theta_{t}-\Delta\theta_{t+1}\)

adam_solver.cpp源碼

template <typename Dtype>
void AdamSolver<Dtype>::AdamPreSolve() {
  // Add the extra history entries for Adam after those from SGDSolver::PreSolve
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //全部可學習參數
  for (int i = 0; i < net_params.size(); ++i) {
    const vector<int>& shape = net_params[i]->shape();  //第i個可學習參數對應的形狀
    this->history_.push_back(shared_ptr<Blob<Dtype> >(new Blob<Dtype>(shape))); //history_再存入一個與參數大小相同的空blob
  }
}

#ifndef CPU_ONLY
template <typename Dtype>
void adam_update_gpu(int N, Dtype* g, Dtype* m, Dtype* v, Dtype beta1,
    Dtype beta2, Dtype eps_hat, Dtype corrected_local_rate);
#endif

template <typename Dtype>
void AdamSolver<Dtype>::ComputeUpdateValue(int param_id, Dtype rate) {
  const vector<Blob<Dtype>*>& net_params = this->net_->learnable_params();  //全部可學習參數
  const vector<float>& net_params_lr = this->net_->params_lr(); //參數的學習率係數
  Dtype local_rate = rate * net_params_lr[param_id];            //當前參數的學習率
  const Dtype beta1 = this->param_.momentum();    //兩個動量係數
  const Dtype beta2 = this->param_.momentum2();

  // we create aliases for convenience
  size_t update_history_offset = net_params.size();     //history_的大小爲2 * update_history_offset
  Blob<Dtype>* val_m = this->history_[param_id].get();
  Blob<Dtype>* val_v = this->history_[param_id + update_history_offset].get();
  Blob<Dtype>* val_t = this->temp_[param_id].get();

  const int t = this->iter_ + 1;                //步數
  const Dtype correction = std::sqrt(Dtype(1) - pow(beta2, t)) /
      (Dtype(1.) - pow(beta1, t));              //correction = sqrt(1 - beta2 ^ t) / (1 - beta1 ^ t)
  const int N = net_params[param_id]->count();  //參數的元素個數
  const Dtype eps_hat = this->param_.delta();   //微小值

  switch (Caffe::mode()) {
    case Caffe::CPU: {
    // update m <- \beta_1 m_{t-1} + (1-\beta_1)g_t
    caffe_cpu_axpby(N, Dtype(1)-beta1, net_params[param_id]->cpu_diff(), beta1,
        val_m->mutable_cpu_data()); //val_m = (1 - beta1) * net_params_diff + beta1 * val_m

    // update v <- \beta_2 m_{t-1} + (1-\beta_2)g_t^2
    caffe_mul(N, net_params[param_id]->cpu_diff(), net_params[param_id]->cpu_diff(),
        val_t->mutable_cpu_data()); //val_t = net_params_diff * net_params_diff
    caffe_cpu_axpby(N, Dtype(1)-beta2, val_t->cpu_data(), beta2,
        val_v->mutable_cpu_data()); //val_v = (1 - beta2) * val_t + beta2 * val_v

    // set update
    caffe_powx(N, val_v->cpu_data(), Dtype(0.5),
        val_t->mutable_cpu_data()); //val_t = val_v ^ 0.5
    caffe_add_scalar(N, eps_hat, val_t->mutable_cpu_data());  //val_t += eps_hat
    caffe_div(N, val_m->cpu_data(), val_t->cpu_data(),
        val_t->mutable_cpu_data()); //val_t = val_m / val_t

    caffe_cpu_scale(N, local_rate*correction, val_t->cpu_data(),
        net_params[param_id]->mutable_cpu_diff());  //net_params_diff = local_rate*correction * val_t
    break;
  }
  case Caffe::GPU: {
#ifndef CPU_ONLY
    // g = net_params_diff
    // m = val_m
    // v = val_v
    // gi = g[i];
    // mi = m[i] = m[i]*beta1 + gi*(1-beta1);
    // vi = v[i] = v[i]*beta2 + gi*gi*(1-beta2);
    // g[i] = local_rate * correction * mi / (sqrt(vi) + eps_hat);
    adam_update_gpu(N, net_params[param_id]->mutable_gpu_diff(),
        val_m->mutable_gpu_data(), val_v->mutable_gpu_data(), beta1, beta2,
        eps_hat, local_rate*correction);
#else
    NO_GPU;
#endif
    break;
  }
  default:
    LOG(FATAL) << "Unknown caffe mode: " << Caffe::mode();
  }
}

小結

  1. 不少地方的動量的符號與本文不用,是用\(\nu_{t+1}=-lr*\nabla_{\theta_{t}}+m*\nu_{t}\),而後\(\theta_{t+1}=\theta_{t}+\nu_{t+1}\),其實原理是一致的,本文只是爲了保持與caffe的代碼一致。

參考

  1. https://stats.stackexchange.com/questions/179915/whats-the-difference-between-momentum-based-gradient-descent-and-nesterovs-acc
  2. https://jlmelville.github.io/mize/nesterov.html
  3. https://zhuanlan.zhihu.com/p/22810533
  4. https://zh.d2l.ai/chapter_optimization/adadelta.html
  5. 《Deep Learning》-- Ian Goodfellow and Yoshua Bengio and Aaron Courville

Caffe的源碼筆者是第一次閱讀,一邊閱讀一邊記錄,對代碼的理解和分析可能會存在錯誤或遺漏,但願各位讀者批評指正,謝謝支持!

相關文章
相關標籤/搜索