fix pad_1

dataset.py

@@ -188,6 +188,7 @@ class Data_Encoder_mol(data.Dataset):
         token_type_ids = np.concatenate((np.zeros((len(d) + 2), dtype=np.int), np.ones((len(p) + 1), dtype=np.int)))
         if len(input_seq) > self.max_len:
             input_seq = input_seq[:self.max_len-1] + [self.sep_id]
+            token_type_ids = token_type_ids[:self.max_len]
         else:
             token_type_ids = np.pad(token_type_ids, (0, self.max_len - len(input_seq)), 'constant', constant_values=0)
         input, input_mask = seq2emb_encoder(input_seq, self.max_len, self.vocab)

test.py

 from transformers import BertTokenizer
 from modeling_bert import BertForMaskedLM
 import torch
+import numpy as np
 
-tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
-# model = BertForMaskedLM.from_pretrained('bert-base-uncased')
-seq_a = "The capital of France is [MASK]."
-seq_b = "The capital of France is Paris."
-choice0 = "It is eaten with a fork and a knife."
-choice1 = "It is eaten while held in the hand."
-
-
-inputs = tokenizer([[seq_a, seq_b], [choice0, choice1]], padding=True)
-labels = tokenizer("The capital of France is Paris.", return_tensors="pt") #["input_ids"]
-
+a = np.ones((5))
 
+b = a[:4]
 print('----------------')
\ No newline at end of file