Hy guys, I have stuck with my project about Deep learning. I have a plan to make bearing fault prediction using Transformers.
But I don't have idea how to use a transformer Specifically in time series data, like bearing fault.
I am very grateful if you guys can help me.
I try to copying code and test in my project. But it didn't work.
Related
I am going to build a chat bot from scratch with rasa.The biggest difficulty now is how to automate production training data.Training data includes nlu.md and stories.md .
I have tried rasa-nlu-trainer and Chatito，But there are still a lot of manual operations，If there are tens of thousands of corpora in the future.How to mark the data to make the data meet the data format of nlu.md and stories.md
Is there an automated tool or program to do this? Thanks a lot!
Well, if you're doing anything ML related, your data is the most important thing that you'll need for the model to learn from. And because we want the model to learn from that data, we create the data and then train the model with it. What you're asking for is for something to somehow create the data for it. It's precisely because there doesn't exist anything like that that we create datasets to train the AI on, by ourselves, so that the model learns form it. So, if you automate the data creation process, what do you expect the model to learn?
So, you can't create the data automatically because if that were possible, we would already have had Artificial General Intelligence (AGI) by now.
But if your goal is to just format the data then you can just write a script for that.
I am using couchdb and I need to stress test the db to make sure about performance limitation.
I have never done any sort of performance testing. So I decided to write a piece of code to do so:
import couchdb
couch= couchdb.Server('xxxxxxx')
db= couch["performance_test"]
doc={
"test": "test",
"db":"cocuh"
}
db.save(doc)
So this is just a beginning and I need to add a lot to it to have it working. For example in terms of benchmarks, measure ... I am not sure what to do. Before keep going and complete the code I want to make sure I am in the right path. Is it the best way to do so? or should I use a specific tool for it? any insight is appreciated
First you need to figure out what you're actually measuring -- concurrent read rates, write throughput etc.
There are lots of load testing frameworks around, for example Locust. Using an existing load testing tool gives you a lot of functionality you otherwise would ghave to implement yourself. Locust makes nice graphs etc.
There are also plenty of tools and libraries that can help you generate random test data, such as faker in Python.
Hello and thanks for checking out my question,
I am working on a project analysing film and visualizing the data I got from it. I'm quite new at programming and only have some basic experience in java and javascript.
For my project I want to store the db levels of a movie in a csv file, to later work with the data in processing. I couldn't find anything that wasn't too complex for me to comprehend for Mac (OSX.)
Help would be much appreciated!
Thank you.
You're going to have to break your problem down into smaller steps.
Step 1: Generating the CSV file.
There are probably a million different ways to do this, and that can be pretty confusing. But break this down into smaller sub-steps and then take those steps one at a time. Can you get a movie playing in Processing? There is a Video library that does just that. Then can you get the volume level every X seconds? You might start with a separate sketch that just prints something to the console every X seconds. For getting the volume, you might try out the Minim library. If that doesn't work, Google is your friend, and remember to keep breaking your problem down into smaller steps!
Step 2: Loading the CSV file.
Now that you have the CSV file, you have to load it into Proccessing. There are several functions in the reference that might come in handy. Again, start with an example program that just prints the values to the console. Get that working perfectly before moving on.
Step 3: Visualizing the data.
Now that you have the data in your Processing code, you can start thinking about how you want to visualize the data. Maybe a line chart that just shows the volume over time just to start with.
If you get stuck on a specific step, then try to break it down into smaller sub-steps. Create an example program that just tests one of those smaller sub-steps (also known as an MCVE), and you'll be able to ask a more specific code-oriented question. Good luck, sounds like an interesting project!
I wanted to know which one of these would be more time-efficient when implementing new features/whatever in a library:
Write it little by little, perfecting each stage
Write the entire feature, and debug the whole thing at once.
I just wanted some thoughts on this. Thanks :)
It is generally advised to do incremental coding ( your first option ) so that your are sure that at least some part of your feature is working before you move on.
Debugging everything at once will be a big headache as you will not know which part of the code the error is in.
In the long run , the first way will definitely be faster than the second one.
I have a question about best practice on how to tackle a new project, any project. When starting a new project how do you go about tackling the project, do you split it into sections, start writing code, draw up flow diagrams.
I'm asking this question because I'm looking for advice on how I can start new projects so I can get going on them quicker. I can have it planned, designed and starting coding with everything worked out.
Any advice?
Thanks
Stephen
It all really depends. Is the project for controlling a space shuttle with 200+ people working on it, or is it a hobby project with 1 person.
I'm guessing this is a small project. In that case, do whatever works for you. Write a list of things you think are required. If there are parts you know you need to learn more about or research, get reading the web, try some stuff out with prototype code to see whether it works or not. Don't turn prototype code into real code though, start again with production code and make sure you get all the appropriate error handling etc in.
When you think you've got a good feel for what's needed, get coding. If you hit a point where you think it's not working, go back to the design and rethink it and sketch some more diagrams, and then go back to the code again.
It is extremely doubtful that you can work everything out in your plan and that's how things will actually work out. So, there's little point in trying to plan too far ahead because you'll be wasting time. Just plan out far enough ahead to keep yourself focused on working on the right things and so that you've given yourself a reasonable chance that the code you're working on will fit the big picture and solve the problem you're trying to solve.
Start by writing a simple functional spec, a few paragraphs from the user's perspective: what they see, how they perform actions, what they expect to happen if they click widget X. This will glue the logic together in your head, and on paper.
From there you can work on the technical spec, which details the gritty things like database structure, special controls and components you need, SDK's if any, and all other developer-type details that you need to implement.