finish

hyunwoongko · Dec 18, 2019 · ad2b684 · ad2b684
1 parent c19810c
commit ad2b684
Show file tree

Hide file tree

Showing 5 changed files with 7 additions and 19 deletions.
diff --git a/.idea/workspace.xml b/.idea/workspace.xml
diff --git a/__pycache__/conf.cpython-36.pyc b/__pycache__/conf.cpython-36.pyc
diff --git a/__pycache__/train.cpython-36.pyc b/__pycache__/train.cpython-36.pyc
diff --git a/conf.py b/conf.py
@@ -9,11 +9,9 @@
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
 # model parameter setting
-batch_size = 32
+batch_size = 128
 max_len = 512
 d_model = 512
-sinusoid = True
-d_k = d_v = d_model * 4
 n_layers = 6
 n_heads = 8
 ffn_hidden = 256

diff --git a/train.py b/train.py
@@ -63,7 +63,7 @@ def train(model, iterator, optimizer, criterion, clip):
  torch.nn.utils.clip_grad_norm_(model.parameters(), clip)
  optimizer.step()
  epoch_loss += loss.item()
- print('step :', i, ' , loss :', loss.item())
+ print('step :', round((i/len(iterator)) * 100, 2), '% , loss :', loss.item())
 
  return epoch_loss / len(iterator)
 
@@ -97,8 +97,8 @@ def run(total_epoch, best_loss):
 
  if valid_loss < best_loss:
  best_loss = valid_loss
- torch.save(model.state_dict(), 'model.pt')
+ torch.save(model.state_dict(), 'model-{0}.pt'.format(valid_loss))
 
- print(f'Epoch: {epoch + 1:02} | Time: {epoch_mins}m {epoch_secs}s')
+ print(f'Epoch: {step + 1} | Time: {epoch_mins}m {epoch_secs}s')
  print(f'\tTrain Loss: {train_loss:.3f} | Train PPL: {math.exp(train_loss):7.3f}')
  print(f'\t Val. Loss: {valid_loss:.3f} | Val. PPL: {math.exp(valid_loss):7.3f}')