fix bug in predictor tail_process_fn

funasr/models/predictor/cif.py

@@ -208,7 +208,8 @@ class CifPredictorV2(nn.Module):
             mask_2 = torch.cat([ones_t, mask], dim=1)
             mask = mask_2 - mask_1
             tail_threshold = mask * tail_threshold
-            alphas = torch.cat([alphas, tail_threshold], dim=1)
+            alphas = torch.cat([alphas, zeros_t], dim=1)
+            alphas = torch.add(alphas, tail_threshold)
         else:
             tail_threshold = torch.tensor([tail_threshold], dtype=alphas.dtype).to(alphas.device)
             tail_threshold = torch.reshape(tail_threshold, (1, 1))
@@ -654,4 +655,4 @@ class CifPredictorV3(nn.Module):
 
         predictor_alignments = index_div_bool_zeros_count_tile_out
         predictor_alignments_length = predictor_alignments.sum(-1).type(encoder_sequence_length.dtype)
-        return predictor_alignments.detach(), predictor_alignments_length.detach()
+        return predictor_alignments.detach(), predictor_alignments_length.detach()