You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a question regarding training using snntorch.
I am starting to work on a task where I will be needing to train a very small SNN in a rate-based task. I am starting with a simple test experiment to understand snntorch training etc using just one single LIF neuron like this:
I am attempting to pass in a frequency of anywhere between 100Hz and 1kHz as input to the neuron, and would like to train the single weight using snntorch to have the LIF neuron spike at a pre-defined frequency, like 30Hz for example. The training loop will eventually lower the weight until the exact LIF output spike count is 30 using the mse_count_loss(). I use the built-in spikegen function to produce the frequency input very easily.
Any advice on how to implement this simply using the snntorch libraries? I assume the mse_count_loss() requires labelled data?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I have a question regarding training using snntorch.
I am starting to work on a task where I will be needing to train a very small SNN in a rate-based task. I am starting with a simple test experiment to understand snntorch training etc using just one single LIF neuron like this:
I am attempting to pass in a frequency of anywhere between 100Hz and 1kHz as input to the neuron, and would like to train the single weight using snntorch to have the LIF neuron spike at a pre-defined frequency, like 30Hz for example. The training loop will eventually lower the weight until the exact LIF output spike count is 30 using the
mse_count_loss()
. I use the built-in spikegen function to produce the frequency input very easily.Any advice on how to implement this simply using the snntorch libraries? I assume the mse_count_loss() requires labelled data?
Code so far is shown below:
Beta Was this translation helpful? Give feedback.
All reactions