lib/torch/optim/adamax.rb in torch-rb-0.3.7 vs lib/torch/optim/adamax.rb in torch-rb-0.4.0

@@ -44,10 +44,10 @@
             if group[:weight_decay] != 0
               grad = grad.add(group[:weight_decay], p.data)
             end
 
             # Update biased first moment estimate.
-            exp_avg.mul!(beta1).add!(1 - beta1, grad)
+            exp_avg.mul!(beta1).add!(grad, alpha: 1 - beta1)
             # Update the exponentially weighted infinity norm.
             norm_buf = Torch.cat([
                 exp_inf.mul!(beta2).unsqueeze(0),
                 grad.abs.add!(eps).unsqueeze!(0)
             ], 0)