Python Pytorch Tutorials # 2 Transfer Learning : Inference with ImageNet Models

preview_player
Показать описание
Pytorch is another deep learning framework, which I am finding to be more intuitive than the other popular framework Tensorflow.
As we continue with this mini series of transfer learning with Pytorch,
we will now learn about inference and imagenet Models.

Imagenet models are open-sourced models that were trained on million-image dataset on 1000 categories.
These models will form the basis of transfer learning, which as the concept name reads, the transfering of
a model to a custom dataset; our dataset.

Before we initiate transfer learning we need to first understand a little more about imagenet models as well as
inference.

Just created a facebook page:

Here is my reddit account for sharing links:

Here is my twitter account for programming:

Here is my github account:

Рекомендации по теме
Комментарии
Автор

one of the first videos i understood in my deep learning journey.

datasciencetoday
Автор

Nice vid, thank you! Just a small correction, in this example model(img) gives us the raw values. We can get the probabilities if we then apply the softmax.

dimitrisanastasiou
Автор

Thaks a lot for the tutorial, you rock!. Hope you are well. I will be waiting for the next video :)

xtian_neuralx
Автор

we need more such tutorials. please make more videos on transfer learning

TrendingWave
Автор

Great tutorial, eagerly looking forward to next video. Thanks a lot and keep it up

GoKotlinJava
Автор

When are we going to see the next part ? I really need it! :)

AM-patreon
Автор

Thanks for the video, I think sorrel is out for the horse image because of the field, which the horse is standing in is looking forward for the next video soon :)

youssefdiaa
Автор

Thanks. Do you know the exact issue when not setting eval() ? In what way is the behaviour different? Happy holidays

DanielWeikert
Автор

Hii great tutorials. Can you make video on how to use pretrained weights as encoder part of unet model for segmentation

rs
Автор

Thanks for this! Are you still planning on publishing the next video?

mordechailis
Автор

Great video! Thanks for simple explanation! When do you plan to make a next tutorial?

bartomiejjozwiak
Автор

For the labels you reference to a json file. How does the structure of the json file need to like like?

wellml