from torch import nn
import torch.nn.functional as F

class FeedForward(nn.Module):
  def __init__(self, hidden_dim, d_ff):
    super(FeedForward, self).__init__()
    self.fc1 = nn.Linear(hidden_dim, d_ff)
    self.fc2 = nn.Linear(d_ff, hidden_dim)

  def forward(self, x):
    # x shape: (seqlen, batch, hiddendim)
    x = self.fc1(x)
    x = F.relu(x)
    x = self.fc2(x)
    return x