“Chao Xing 2015-12-14”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
(以“Last Week: ---- Solution: * None. Plan to do: 1. Broadcast last week mnist results to Chinese images. 2. Give a report after read some papers. 3. Tes...”为内容创建页面)
 
 
(相同用户的3个中间修订版本未显示)
第3行: 第3行:
 
Solution:
 
Solution:
 
* None.
 
* None.
 
 
Plan to do:
 
Plan to do:
 
 
     1. Broadcast last week mnist results to Chinese images.
 
     1. Broadcast last week mnist results to Chinese images.
 
 
     2. Give a report after read some papers.
 
     2. Give a report after read some papers.
 
 
     3. Test some word sense vector models.
 
     3. Test some word sense vector models.
  
第15行: 第11行:
 
----
 
----
 
What have done:
 
What have done:
 
 
     1. Test MNIST images given a one-hot vector during training and test processing.
 
     1. Test MNIST images given a one-hot vector during training and test processing.
 
 
     2. Coding Fine tune & balance frequency code.
 
     2. Coding Fine tune & balance frequency code.
 
 
     3. Code two image test program.
 
     3. Code two image test program.
 
 
Plan to do:
 
Plan to do:
 
 
     1. Test same number multiple samples image.
 
     1. Test same number multiple samples image.
 
 
     2. Test different numbers given a single images.
 
     2. Test different numbers given a single images.
 
 
     3. Code balance word frequency program within word2vec.
 
     3. Code balance word frequency program within word2vec.
 
+
     4. Code other fine tune strategies.
     4. Code another fine tune functions.
+
 
+
 
+
    3. Test some word sense vector models.
+

2015年12月14日 (一) 00:58的最后版本

Last Week:


Solution:

  • None.

Plan to do:

   1. Broadcast last week mnist results to Chinese images.
   2. Give a report after read some papers.
   3. Test some word sense vector models.

This Week:


What have done:

   1. Test MNIST images given a one-hot vector during training and test processing.
   2. Coding Fine tune & balance frequency code.
   3. Code two image test program.

Plan to do:

   1. Test same number multiple samples image.
   2. Test different numbers given a single images.
   3. Code balance word frequency program within word2vec.
   4. Code other fine tune strategies.