def _update()

in higher/optim.py [0:0]

• 25 lines of code
• 8 McCabe index (conditional complexity)

    def _update(self, grouped_grads: _GroupedGradsType, **kwargs) -> None:
        zipped = zip(self.param_groups, grouped_grads)
        for group_idx, (group, grads) in enumerate(zipped):
            weight_decay = group['weight_decay']
            momentum = group['momentum']
            dampening = group['dampening']
            nesterov = group['nesterov']

            for p_idx, (p, g) in enumerate(zip(group['params'], grads)):
                if g is None:
                    continue

                if weight_decay != 0:
                    g = _add(g, weight_decay, p)
                if momentum != 0:
                    param_state = self.state[group_idx][p_idx]
                    if 'momentum_buffer' not in param_state:
                        buf = param_state['momentum_buffer'] = g
                    else:
                        buf = param_state['momentum_buffer']
                        buf = _add(buf.mul(momentum), 1 - dampening, g)
                        param_state['momentum_buffer'] = buf
                    if nesterov:
                        g = _add(g, momentum, buf)
                    else:
                        g = buf

                group['params'][p_idx] = _add(p, -group['lr'], g)