Mobilebert QAT (#55)

* Remove duplicate quantization of vocabulary.

* Remove duplicate quantization of vocabulary.

Author: anmarques

src/transformers/models/mobilebert/modeling_mobilebert.py

@@ -152,6 +152,23 @@ def forward(self, input_tensor):
 
 NORM2FN = {"layer_norm": nn.LayerNorm, "no_norm": NoNorm}
 
+class QATEmbeddingTransformation(nn.Module):
+    def __init__(self, embedded_input_size, hidden_size):
+        super().__init__()
+
+        # Behaves like normal Linear module unless a SparseML QuantizationModifier
+        # is initialized.
+        # When initialized, does not quantize inputs.
+        # Only weights are quantized (inputs come quantized from embeddings)
+        self.linear = nn.Linear(embedded_input_size, hidden_size)
+        self.wrap_qat = True
+        self.qat_wrapper_kwargs = {
+            "num_inputs": 0,
+            "num_outputs": 1,
+        }
+
+    def forward(self, x: torch.Tensor):
+        return self.linear(x)
 
 class MobileBertEmbeddings(nn.Module):
     """Construct the embeddings from word, position and token_type embeddings."""
@@ -168,7 +185,7 @@ def __init__(self, config):
 
         embed_dim_multiplier = 3 if self.trigram_input else 1
         embedded_input_size = self.embedding_size * embed_dim_multiplier
-        self.embedding_transformation = nn.Linear(embedded_input_size, config.hidden_size)
+        self.embedding_transformation = QATEmbeddingTransformation(embedded_input_size, config.hidden_size)
 
         self.LayerNorm = NORM2FN[config.normalization_type](config.hidden_size)
         self.dropout = nn.Dropout(config.hidden_dropout_prob)