-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
78574ae
commit 5f1f96b
Showing
15 changed files
with
208 additions
and
0 deletions.
There are no files selected for viewing
Binary file not shown.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,39 @@ | ||
## 1. Overview | ||
|
||
* 培养看文档的能力(当自己的pytorch版本和讲解的不一样的时候) | ||
* 推理 & 预测Prediction | ||
* ILSVRC竞赛 (AlexNet->VGG->GoogleNet->ResNet) | ||
* Back Propagation | ||
* 学会构造模型的套路,然后去构造自己的模型 | ||
* Popular learning frameworks | ||
* Theano / TensorFlow / Caffe / PyTorch | ||
|
||
## 2. Linear Model | ||
|
||
* 三步走:1.准备数据集 2.选择模型 3.训练模型 4.inferring 进行推理 | ||
* 模型可能过拟合,所以我们希望模型有很好的泛化功能 | ||
* y = x * w 一开始随机产生一个w,然后评估 (y_hat-y)^2来修正w | ||
* visdom | ||
|
||
|
||
|
||
明天先 读英语 + 把线性模型那个作业跑出来 + 完成仿生优化算法的作业 + 看论文 + 看新的视频 + 完成大数据作业+完成自己导师作业 + leetcode12道题 | ||
|
||
|
||
|
||
把线性模型那个作业跑出来 + 看论文 + 看新的视频 + 完成大数据作业 + 完成自己导师作业 + leetcode10道题 | ||
|
||
|
||
|
||
今日计划: | ||
|
||
leetcode每日一题 + 动态规划2道题 + 读英语(单词阅读听力口语)+ 大数据作业写完 + 线性模型跑出来 + 看两个pytorch视频 + 看论文(那一篇文章要看完!!!) | ||
|
||
|
||
|
||
读英语(单词阅读听力) + 线性模型跑出来 + 看两个pytorch视频 + 看论文(那一篇文章要看完!!!) | ||
|
||
|
||
|
||
|
||
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,147 @@ | ||
## 四则运算 | ||
|
||
### 加法 减法 运算 | ||
|
||
c = torch.add(a, b) -> tensor(3) | ||
|
||
c = a + b | ||
|
||
a.add(b) a.add_(b)(这里会对a的值进行修改,变成a+b的和) | ||
|
||
a .sub(b) a.sub_(b) | ||
|
||
### 乘法 除法 运算(对应元素相乘 哈达玛积 element wise) | ||
|
||
c = a * b c = torch.mul(a,b) 这个是点乘 也就是对应元素相乘 不是矩阵的那种乘法 | ||
|
||
a.mul(b) a.mul_(b) a.div(b) a.div__(b) | ||
|
||
### 矩阵的乘法 | ||
|
||
#### 二维矩阵 torch.mm(a,b) torch.matmul(a,b) a @ b | ||
|
||
a.matmul(b) a.mm(b) `注意a为 m * n 的矩阵,b就要为 n* p的矩阵` | ||
|
||
#### 高维矩阵 | ||
|
||
![matrix_matmul](/Users/yannie/Desktop/Pytorch/1.png) | ||
|
||
### 幂运算 | ||
|
||
torch.pow(a,2) a.pow(2) a**2 a.pow_(2) | ||
|
||
torch.exp(a) 表示e^n a.exp_() | ||
|
||
### 开方运算 | ||
|
||
a.sqrt() a.sqrt_() | ||
|
||
### 对数运算 | ||
|
||
torch.log2(a) torch.log10(a) torch.log(a) -> 默认底数为e torch.log_(a) | ||
|
||
#### PyTorch中的in-place操作 | ||
|
||
* “就地操作”,即不允许使用临时变量 | ||
|
||
* add_ sub_ mul_ | ||
|
||
#### PyTorch 中的广播机制 | ||
|
||
* 张量参数可以自动拓展为相同大小 | ||
|
||
* 广播机制要满足右对齐 | ||
|
||
`torch.rand(2,1,1) + torch.rand(3)` 这里3前面会加上两个1变成 113 是否满足对齐要看当前两个元素是否相等或者是否有一个值为1 | ||
|
||
|
||
|
||
|
||
|
||
### Tensor的取整/余运算 | ||
|
||
![method](/Users/yannie/Desktop/Pytorch/2.png) | ||
|
||
#### Tensor的比较运算(其返回的也是一个tensor,由false和true构成 即mask张量) | ||
|
||
![](/Users/yannie/Desktop/Pytorch/3.png) | ||
|
||
![](/Users/yannie/Desktop/Pytorch/4.png) | ||
|
||
![](/Users/yannie/Desktop/Pytorch/5.png) | ||
|
||
### Tensor的三角函数和其他的数学函数 | ||
|
||
![](/Users/yannie/Desktop/Pytorch/6.png) | ||
|
||
![](/Users/yannie/Desktop/Pytorch/7.png) | ||
|
||
* sigmoid也叫做激活函数 由于符号函数没办法求导,所以有了sigmoid函数 | ||
* 通过sigmoid函数 我们的数据会被影射到[0,1]的区间上去 | ||
|
||
![](/Users/yannie/Desktop/Pytorch/8.png) | ||
|
||
![](/Users/yannie/Desktop/Pytorch/9.png) | ||
|
||
#### 该看3.15了 | ||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,21 @@ | ||
|
||
|
||
|
||
|
||
class PageRankVertex : public Vertex<Double, Void, Double> | ||
public void Compute(MessageIterator msgs) { | ||
if (superstep() >= 1) { | ||
double sum = 0; | ||
for (Message msg : msgs) { | ||
sum += msg.getValue(); | ||
} | ||
this.value = 0.15 / getNumOfVertices() + 0.85 * sum; | ||
} | ||
| ||
if (superstep() < 30) { | ||
int outSize = getOutEdgerIterator().getSize(); | ||
sendMessageToAllNeighbors(this.value / n); | ||
} else { | ||
voteToHalt(); | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1 @@ | ||
希望自己2月份之前能把这个课程刷完吧 |