Merge pull request #57 from hpcaitech/feature/accumulate

update bert

examples/bert/bert.py

@@ -246,16 +246,19 @@ def __init__(self,
 
     def forward(self, hidden_states=None, input_ids=None, attention_mask=None, seq_lens=None):
 
-        batch_size = None
-        max_padding_size = None
-        if seq_lens is not None:
-            batch_size = input_ids.shape[0]
-            max_padding_size = input_ids.shape[1]
+        batch_size = input_ids.shape[0]
+        max_padding_size = input_ids.shape[1]
 
         if self.first:
             hidden_states = self.embed(input_ids=input_ids, position_ids=None, tokentype_ids=None, seq_lens=seq_lens,
                                        batch_size=batch_size, max_padding_size=max_padding_size)  # , seq_lens
 
+        if attention_mask is not None:
+            attention_mask = attention_mask.view(batch_size, -1)
+            attention_mask = attention_mask.unsqueeze(1).unsqueeze(2)
+            attention_mask = attention_mask.to(dtype=hidden_states.dtype)  # fp16 compatibility
+            attention_mask = (1.0 - attention_mask) * -10000.0
+
         for block in self.blocks:
             hidden_states = block(hidden_states=hidden_states, attention_mask=attention_mask, seq_lens=seq_lens,
                                   batch_size=batch_size, max_padding_size=max_padding_size)

examples/bert/bert_server.py

@@ -19,7 +19,7 @@ def run():
     batch_size = 32
 
     input_ids = torch.randint(1, 10, (batch_size, seq_len), dtype=torch.int64)
-    attention_mask = torch.randint(0, 1, (batch_size, 1, seq_len, seq_len), dtype=torch.int64)
+    attention_mask = torch.randint(0, 1, (batch_size, 1, seq_len), dtype=torch.int64)
     # seq_lens = torch.randint(1, 128, (batch_size, ), dtype=torch.int64) # generate seq_lens randomly
     hidden_states = None
     sample = dict(hidden_states=hidden_states, input_ids=input_ids, attention_mask=attention_mask)
@@ -36,8 +36,8 @@ def run():
     batch_size = 32
 
     input_ids = torch.randint(1, 10, (batch_size, seq_len), dtype=torch.int64)
-    attention_mask = torch.randint(0, 1, (batch_size, 1, seq_len, seq_len), dtype=torch.int64)
-    seq_lens = torch.randint(1, 128, (batch_size, ), dtype=torch.int64) # generate seq_lens randomly
+    attention_mask = torch.randint(0, 1, (batch_size, 1, seq_len), dtype=torch.int64)
+    seq_lens = torch.randint(1, 128, (batch_size, ), dtype=torch.int) # generate seq_lens randomly
     hidden_states = None
     sample = dict(hidden_states=hidden_states, input_ids=input_ids, attention_mask=attention_mask, seq_lens=seq_lens)
 

examples/hf_gpt2/hf_gpt2.py

@@ -377,7 +377,7 @@ def __init__(self,
             self.head = GPTLMHead1D(dim=dim, vocab_size=vocab_size,
                                     dtype=dtype)  # word_embeeding_weight=self.embed.word_embedding_weight not in the same process
 
-    def forward(self, hidden_states=None, input_ids=None, attention_mask=None):
+    def forward(self, hidden_states=None, input_ids=None, attention_mask=None, seq_lens=None):
         topk = 5 # TODO: add as a parameter
         if self.first:
             hidden_states = self.embed(input_ids)