Update run_classifier.py

fixed code

fine-tuning/run_classifier.py

@@ -39,7 +39,7 @@ def forward(self, src, tgt, seg, soft_tgt=None):
         """
         # Embedding.
         emb = self.embedding(src, seg)
-        # Encoder. output输出是[bz,seq_len,768]，[cls]的embedding应取[bz,1,768]
+        # Encoder.
         output = self.encoder(emb, seg)
         temp_output = output
         # Target.
@@ -157,7 +157,7 @@ def read_dataset(args, path):
                 seg = [1] * len(src_a) + [2] * len(src_b)
 
             if len(src) > args.seq_length:
-                src = src[: args.seq_length] # 截取文本的前512
+                src = src[: args.seq_length]
                 seg = seg[: args.seq_length]
             while len(src) < args.seq_length:
                 src.append(0)
@@ -285,7 +285,7 @@ def main():
     model = model.to(args.device)
 
     # Training phase.
-    trainset = read_dataset(args, args.train_path) # trainset是一个list,包含[src,tgt,seg]，src是样本内容，tgt是标签，seg是用于标记不同样本的符号（当前样本是1，则补0和其他样本为0）
+    trainset = read_dataset(args, args.train_path)
     random.shuffle(trainset)
     instances_num = len(trainset)
     batch_size = args.batch_size