Write text to file with disabling buffer in Python3

In Python2 era, we could use these code to write the file without buffer:

But in Python3 we can only write binary file by disabling buffer:

The only way to write text file without buffering is:

Adding ‘flush()’ everywhere is a terrible experience for a programmer who need to migrate his code from Python2 to Python3. I really want to know: what’s in Python3’s developers mind ?

A successful rescue for a remote server

After installed CUDA-9.2 on a remote server, I found that the system can’t load nvidia.ko (kernel module) with dmesg:

The reason is the current kernel running on my system has turned on the CONFIG_CC_STACKPROTECTOR compiler option. Therefore I change the default entry of grub2 and reboot the server, for entering a new kernel without this option.
But unfortunately, the server never start up again. All my code and data (includes my colleague’s code and data) are on this server, so we get a little nervous then.

Since the server is in a remote datacenter, we can’t just plugin in a keyboard and a screen to debug. Thus I use the out-of-bound system to reboot this server to diskless-mode. After entering this mode, I mount the disk for ‘/boot’ directory:

and manually change the ‘/boot/grub2/grubenv’ like this (the ‘save_entry’ is 2 before):

Then reboot the server again. This time, the server started up smoothly now. All our code and data is untainted.

Some tips about python this week

List of lists in python
Created a list of lists by using multiply symbol:

It’s weird that adding one item to first list have side-effect on second list! Seems ‘* 2’ makes two reference for one list.
How to avoid this then? The answer is using normal syntax:

A difference between python2 and python3
There are many trivial differences between python3 and python3, and I luckily found one this week:

In python3, the values() of a dictionary is not type of ‘list’ but ‘dict_values’. To be compatible to python2, we need to add

for old python2 codes.

Compare implementation of tf.AdamOptimizer to its paper

When I reviewed the implementation of Adam optimizer in tensorflow yesterday, I noticed that it’s code is different from the formulas that I saw in Adam’s paper. In tensorflow’s formulas for Adam are:




But the algorithm in the paper is:




Then quickly I found these words in the document of tf.AdamOptimizer:

Note that since AdamOptimizer uses the formulation just before Section 2.1 of the Kingma and Ba paper rather than the formulation in Algorithm 1, the “epsilon” referred to here is “epsilon hat” in the paper.

And this time I did find the ‘Algo 2’ in the paper:



But how does ‘Algo 1’ tranform to ‘Algo 2’? Let me try to deduce them from ‘Algo 1’:


\theta_t \gets \theta_{t-1} - \frac{\alpha \cdot \hat{m_t}}{(\sqrt{\hat{v_t}} + \epsilon)}
\implies   \theta_t \gets \theta_{t-1} - \alpha \cdot \frac{m_t}{1 - \beta_1^t} \cdot \frac{1}{(\sqrt{\hat{v_t}} + \epsilon)} \quad \text{       (put } \hat{m_t} \text{ in) }
\implies   \theta_t \gets \theta_{t-1} - \alpha \cdot \frac{m_t}{1 - \beta_1^t} \cdot \frac{\sqrt{1-\beta_2^t}}{\sqrt{v_t}} \quad \text{       (put } \hat{v_t} \text{ in and ignore } \epsilon \text{) }
\implies   \theta_t \gets \theta_{t-1} - \alpha_t \cdot \frac{m_t}{\sqrt{v_t} + \hat{\epsilon}} \quad \text{add new } \hat{\epsilon} \text { to avoid zero-divide}

The bug about using hooks and MirroredStrategy in tf.estimator.Estimator

When I was using MirroedStrategy in my tf.estimator.Estimator:

and add hooks for training:

The tensorflow report errors:

Without finding any answers on google, I have to look into the code of ‘estimator.py’ in tensorflow. Fortunately, the code defect is obvious:

class Estimator havn’t any private argument named ‘_distribution’ but only have ‘_train_distribution’ and ‘_eval_distribution’. So the fix is just change ‘self._distribution.unwrap(per_device_hook)[0]’ to ‘self._train_distribution.unwrap(per_device_hook)[0]’.

I had submitted a request pull for tensorflow to fix this bug in branch 1.11

Some lessons from Kaggle’s competition

About two months ago, I joined the competition of ‘RSNA Pneumonia Detection’ in Kaggle. It’s ended yesterday, but I still have many experiences and lessons to be rethinking.

1. Augmentation is extremely crucial. After using tf.image.sample_distorted_bounding_box() in my program, the mAP(mean Average Precision) of evaluating dataset thrived to a perfect number. Then I realised that I should have used radical augmentation method in the first place. Otherwise, for machine learning job such as image detection and image classification, the number of samples is only about tens of thousands which is quite small for extracting dense features. Thus we need use more powerful augmentation strategy or tools (albumentation may be a good choice).

2. SGD is good for generalisation. Previously I used Adam to acquire outstanding training accuracy. But soon after, I found it is useless since evaluating accuracy is poor. For the samples are too few, I can’t use my evaluating dataset (10% cut from original data) to correctly evaluate the score on competition leaderboard. Without choice, I have to use only SGD to train my model in last stage.

3. Use more visual monitor tools. At first my model have high training accuracy and low evaluating accuracy, but after I added too many regularisation methods (such as dropout, weight decay) both training accuracy and evaluating accuracy thrinked to a too low value. The key for regularistaion of DNN is “Keep the fitting capability of training, and then try to rise evaluating accuracy”. So if I could monitor both training and evaluating accuracy at realtime, I would not trapped in dilemma.

4. Thinking more, experimenting less. Spend more time to understand and check the source code, the mechanism of model, instead of only adjusting hyper-parameters in vain.

There are still some questions I can’t answer at present:

1. Why even Resnet-50 can’t raise my training mAP up to 0.5 ?

2. Why perfect mAP value in my evaluating dataset can’t gain good score in this competition’s leaderboard ?

Will go on to discover them.

How could it possible to assign an integer to string?

The snippet below could be compiled and run:

The result is:

I noticed that the corresponding value of key ‘banana’ is empty. The reason is I assign an integer directly to key ‘banana’ by mistake. But how could c++ compiler allow me to do this? Why doesn’t it report a compiling error?
To reveal the truth, I write another snippet:

This code could also be compiled correctly!
Then I change my code to:

This time, the compiler complained that

Seems the std::string ‘s constructor and assignment operator have totally different behavier.
After checking the document, I found the reason: std::string has assignment operator for ‘char’ !(ref)

Thus we should be much more carefully when assign number to std::string.

Move semantics in C++11

      No Comments on Move semantics in C++11

After studying an example for Move Semantics of C++11, I write a more complete code snippet:

Pay attention to last two lines in ‘move constructor’:

Since ‘move constructor’ will not set initial value for m_size and m_data of ‘v4’, the m_size and m_data of ‘v3’ will be uninitial after swaps. Adding the two lines of code means to initialize m_size and m_data of ‘v3’.

How Tensorflow set device for each Operation ?

In Tensorflow, we only need to use snippet below to assign a device to a Operation:

How dose it implement? Let’s take a look.

There is a mechanism called ‘context manager’ in Python. For example, we can use it to add a wrapper for a few codes:

The result of running this script is:

Function ‘tag()’ works like a decorator. It will do something before and after those codes laying under its ‘context’.

Tensorflow uses the same principle.

This will call class Graph’s function ‘device()’. Its implementation:

The key line is ‘self._add_device_to_stack()’. Context of ‘device’ will add device name into stack of python, and when developer create an Operation it will fetch device name from stack and set it to this Operation.
Let’s check the code routine of creating Operation:

‘self._device_function_stack.peek_objs’ is where it peek the device name from stack.

Some tips about using google’s TPU (Cont.)

Sometimes I get this error from TPUEstimator:

And after stop and restart TPU in console of GCP, the error disappeared. TPU doesn’t allow users to use it directly like GPU. You can’t see the device in VM looks like ‘/dev/tpu’ or something like this. Google provides TPU as RPC service, so you can only run DNN training through this service. I think this RPC service is not stable enough so sometimes it can’t work and lead to the error ‘Deadline Exceeded’.

When I get this type of error from TPU:

The only solution is to create a new TPU instance and delete the old one in GCP console. Seems Google need to improve the robust of their TPU RPC service.

Running 10000 steps and get ‘loss’ for every turn:

It’s quite strange that the ‘loss’ can’t go low enough. I still need to do more experiments.

Previously, I run MobileNet_v2 in a machine with Geforce GTX 960 and it could process 100 samples per second. And by using 8 TPUs of Version 2, it can process about 500 samples per second. Firstly, I am so disappointed about the performance-boosting of TPUv2, for it only has about 1.4 TFLOPS for each. But then I noticed that may be the bottleneck is not the performance of TPU, since IO is usually the limit for training speed. Besides, my model is MobileNet_v2, which is too simple and light so it can’t excavate all the capability of TPU.
Therefore I set ‘depth_multiplier=4’ for MobileNet_v2. Under this model, GTX 960 could process 21 samples per second, and TPUv2-8 could process 275 samples per second. This time, I can estimate each TPUv2 has about 4 TFLOPS. I know this metric seems too low from Google’s official 45 TFLOPS. But considering the possible bottlenecks of storage IO and network bandwidth, it becomes understandable. And also, there is another possibility: Google’s 45 TFLOPS means the half-precision operation performance 🙂

Google has just release Tensorflow 1.11 for TPU clusters. At first, I think I can use hooks in TPUEstimatorSpec now, but after adding

it reports

Certainly, the TPU is much harder to use and debug than GPU/CPU.