update test codes so that it generates correct structure of input data

tests/test_permutation.py

@@ -48,7 +48,7 @@ class TestPermutation:
         """
         self.test_data_dir = os.path.join(os.getcwd(),"tests/test_data")
         self.label_ids = ['label_1','label_1','label_2','label_2','label_2']
-        self.asym_id = [1]*9+[2]*9+[3]*13+[4]*13 + [5]*13
+        self.asym_id = [0]*9+[1]*9+[2]*13+[3]*13 + [4]*13
 
     def affine_vector_to_4x4(self,affine):
         r = Rigid.from_tensor_7(affine)
@@ -111,6 +111,7 @@ class TestPermutation:
         batch["seq_mask"] = torch.randint(low=0, high=2, size=(n_res,)).float()
         batch.update(data_transforms.make_atom14_masks(batch))
         batch["no_recycling_iters"] = torch.tensor(2.)
+        batch["seq_length"] = torch.from_numpy(np.array([n_res] * n_res))
 
         if consts.is_multimer:
             #
@@ -120,7 +121,7 @@ class TestPermutation:
             asym_id = self.asym_id
             batch["asym_id"] = torch.tensor(asym_id,dtype=torch.float64)
             # batch["entity_id"] = torch.randint(0, 1, size=(n_res,))
-            batch['entity_id'] = torch.tensor([1]*18+[2]*39,dtype=torch.float64)
+            batch['entity_id'] = torch.tensor([0]*18+[1]*39,dtype=torch.float64)
             batch["sym_id"] = torch.tensor(asym_id,dtype=torch.float64)
             # batch["num_sym"] = torch.tensor([1]*18+[2]*13,dtype=torch.int64) # currently there are just 2 chains
             batch["extra_deletion_matrix"] = torch.randint(0, 2, size=(n_extra_seq, n_res))