42 points by quantum_coder 1 year ago flag hide 13 comments
deep_learning_enthusiast 1 year ago next
This is really groundbreaking! A new era of neural networks with less computational demand.
ghost_in_the_shell 1 year ago next
Indeed, the deep learning field has been waiting for a solution like this for quite a while.
coding_chimp 1 year ago next
It will be interesting to see how this affects the field of edge intelligence.
sentient_being 1 year ago prev next
How can this method be integrated into existing models? Are there any limitations?
neural_scientist 1 year ago next
The paper discusses various strategies for incorporating this into established models.
ai_novice 1 year ago next
Can this be used for pruning or quantization of the weights or architectures?
algorithm_warrior 1 year ago next
I have applied this technique to my YOLOv2 project, and it reduced my model size significantly.
pytorch_genius 1 year ago next
Make the code available, please? Would be nice to test this out myself.
ml_architect 1 year ago next
It's interesting to think about how the optimization landscape is affected by this. Anyone have thoughts to share?
tensor_queen 1 year ago prev next
I think we can expect many Github projects implementing this method now.
keras_prodigy 1 year ago next
I've created a simple implementation on Github you can all check out.
networks_freak 1 year ago prev next
The authors state it can be implemented for both pruning and quantization seamlessly.
edge_master 1 year ago prev next
That's promising! I'm looking forward to the improvements in edge device computing.