Set model params nn torch manual#

Now next part is more detailed, but it is the main point of the post. 'African-American','Asian','Hispanic','Native American','Other', Recid_prep = np.random.binomial(1,0.75,len(recid_prep)) # reference category is separated/unknown/widowedĭum_mar = pd.get_dummies(recid) 'juv_fel_count','YearsScreening']].copy()

# Prepping the Compas data and making train/test

So first, I load the libraries and then prep the recidivism data before I fit my predictive models. Here is the prepped CSV file to download to follow along. For an example use case I will just use a prior Compas recidivism data I have used for past examples on the blog (see ROC/Calibration plots, and Balancing False Positives). I recently wrote some example code to make this process somewhat more like the sklearn approach, where you instantiate an initial model object, and then use a mod.fit(X, y) function call to fit the pytorch model. Best practices are to both evaluate the loss in-sample and wait for it to flatten out, as well as evaluate out of sample. Loss = crit(y_pred,y) #y is tensor of outcomesĪnd this would use backpropogation to adjust our model parameters to minimize the loss function, here just the mean square error, over 20,000 iterations. Y_pred = mod(x) #x is tensor of independent vars Opt = (mod.parameters(), lr=1e-4)Ĭrit = torch.nn.MSELoss(reduction='mean') So typically something like this: # Example fitting a pytorch model

Set model params nn torch manual#

Out of the box when fitting pytorch models we typically run through a manual loop.