Merge pull request #17 from lukovnikov/master

activation function in BERTIntermediate

modeling.py

@@ -25,6 +25,7 @@ import six
 import torch
 import torch.nn as nn
 from torch.nn import CrossEntropyLoss
+from six import string_types
 
 def gelu(x):
     """Implementation of the gelu activation function.
@@ -34,6 +35,13 @@ def gelu(x):
     return x * 0.5 * (1.0 + torch.erf(x / math.sqrt(2.0)))
 
 
+def swish(x):
+    return x * torch.sigmoid(x)
+
+
+ACT2FN = {"gelu": gelu, "relu": torch.nn.functional.relu, "swish": swish}
+
+
 class BertConfig(object):
     """Configuration class to store the configuration of a `BertModel`.
     """
@@ -60,7 +68,7 @@ class BertConfig(object):
             intermediate_size: The size of the "intermediate" (i.e., feed-forward)
                 layer in the Transformer encoder.
             hidden_act: The non-linear activation function (function or string) in the
-                encoder and pooler.
+                encoder and pooler. If string, "gelu", "relu" and "swish" are supported.
             hidden_dropout_prob: The dropout probabilitiy for all fully connected
                 layers in the embeddings, encoder, and pooler.
             attention_probs_dropout_prob: The dropout ratio for the attention
@@ -237,7 +245,8 @@ class BERTIntermediate(nn.Module):
     def __init__(self, config):
         super(BERTIntermediate, self).__init__()
         self.dense = nn.Linear(config.hidden_size, config.intermediate_size)
-        self.intermediate_act_fn = gelu
+        self.intermediate_act_fn = ACT2FN[config.hidden_act] \
+            if isinstance(config.hidden_act, string_types) else config.hidden_act
 
     def forward(self, hidden_states):
         hidden_states = self.dense(hidden_states)