How to speed up training through parallelization across GPUs or CPUs in Machine Learning
$30-250 USD
已关闭
已发布超过 3 年前
$30-250 USD
货到付款
We are trying to reduce total training time for one of our models.
Currently CPU and GPU training is about the same: about 1 minute per epoch. When using the GPU, about half of that is just loading data onto the GPU, and another half is backpropagation.
We just tried an experiment to see how parallelizable the model is -- first only on the cpu, and then on distributed cpus with multiple processes. There's a parameter called num_threads which controls cpu level parallelization, which gives us a decent gain when going to two threads, but not much beyond that. This suggests that the model isn't very parallelizable, which also implies that distributing across multiple GPUs wouldn't be much of an improvement either.
Our typical flow for a customer is now:
• User clicks on an ad in Facebook, goes to our website
• User browses our website for a bit
• User goes directly to the app store, downloads the app
• Inside the app, the user makes an account and orders
Currently, FB is able to track this cross-platform movement, I believe via IDFA. Unfortunately, we're soon going to be left without many users' IDFAs with the upcoming Apple changes. We'd like to be able to continue reporting a conversion event along with the specific user who triggered it, so Facebook can continue optimizing our ad campaign for users who are similar to the ones who converted.
1. Even though it is supposed to be used only for web events, can we use the conversion API for in-app events? Even if we don't have IDFA?
2. If we use the conversion API, will Facebook optimize the ad campaign with that data point? Do we need to include more than just email address and name for this to work?
3. Why does the Conversion API say it requires us to configure the Facebook pixel in addition? Can we report conversions through the API without the FB pixel?
4. If the conversion API doesn't work for us, could we instead use the offline events API to report conversions for specific users, even if we don't have the user's IDFA?
Basically, I'm looking for some live troubleshooting for ideas to speed up training at this point. We're using Python 3 + PyTorch.
Hi,
I hope you are doing fine.
I have almost 10 years of experience in machine learning algorithms. I can implement various types of artificial intelligence algorithms including yours with Matlab, Python and etc. I have PhD from Tohoku University and have several journal publications on the subjects. You can see portfolio for my previous projects.
I read about your project and am interested in working with you. Please send me a message so that we can discuss more.
Best regards.
Hi i am an Senior Python ML , Pytorch Expert.
As 6+ years experiences in these field. I can give good quality work.
I have read the guidelines of your work.I believe that i can provide you the best quality works you are anticipating from this platfrom give me a chance to show you the best i can do at your service.
........................................................................................................... lets talk............. i am a deep learning professional .............
Thanks,
Pranay
Hi, i have a good experience of working with tensorflow and pytorch project in production over the GPUs. plz contact me back to discuss more about this project. I'll work till quality and satisfied results. Thanks.
Hi,
I can help you to do any kind of machine learning, image processing, Deep learning and MATLAB task. I will also show you demo of my previous work.
Although I am new on this platform but I'm professional machine learning expert with mainly expertise in python, MATLAB, deep learning algorithms, image processing and graphic designing having more than 5 years’ experience.
Below are few of my previous projects title:
Recognition of Urdu ligatures in Videos Frames
Glioblastoma tumor detection from MRI
Lukemia Cancer classification using DCNN.
Emotion Detection using Voice
Helmet Detection
And still there is a long list of portfolio items, which can be shared if requested.
I'm available to start immediately if you prefer.
Thanks.
Hello,
I have taken a brief note of your project description. I’m an expert developer. My preference is to satisfy all of your venture’s prerequisites. Let’s come over the chat to discuss the project module, so we can start this project instantly.
Please check my certifications to estimate my skills.
Regards,