[s2s] s/alpha_loss_encoder/alpha_encoder_loss/ (#7298)

fix to match `distillation.py:        self.alpha_encoder_loss`

examples/seq2seq/test_seq2seq_examples.py

@@ -86,7 +86,7 @@ CHEAP_ARGS = {
     "n_val": -1,
     "n_test": -1,
     "student_encoder_layers": 1,
-    "alpha_loss_encoder": 0.0,
+    "alpha_encoder_loss": 0.0,
     "freeze_encoder": False,
     "auto_scale_batch_size": False,
 }