Support string input for ofa export (PaddlePaddle#964)

* support string input for ofa export

Author: LiuChiachi

paddleslim/nas/ofa/ofa.py

@@ -79,6 +79,21 @@
 DistillConfig.__new__.__defaults__ = (None, ) * len(DistillConfig._fields)
 
 
+def to_tensor(string_values, name="text"):
+    """
+    Create the tensor that the value holds the list of string.
+    NOTICE: The value will be holded in the cpu place.
+
+    Parameters:
+        string_values(list[string]): The value will be setted to the tensor.
+        name(string): The name of the tensor.
+    """
+    tensor = paddle.Tensor(core.VarDesc.VarType.STRING, [], name,
+                           core.VarDesc.VarType.STRINGS, False)
+    tensor.value().set_string_list(string_values)
+    return tensor
+
+
 class OFABase(Layer):
     def __init__(self, model):
         super(OFABase, self).__init__()
@@ -531,6 +546,8 @@ def build_input(input_size, dtypes):
                     dtype = dtypes[0]
                 else:
                     dtype = dtypes
+                if dtype == core.VarDesc.VarType.STRINGS:
+                    return to_tensor([""])
                 return paddle.cast(paddle.rand(list(input_size)), dtype)
             if isinstance(input_size, dict):
                 inputs = {}

paddleslim/quant/quanter.py

@@ -307,6 +307,7 @@ def quant_post_static(
         quantize_model_path,
         batch_generator=None,
         sample_generator=None,
+        data_loader=None,
         model_filename=None,
         params_filename=None,
         save_model_filename='__model__',
@@ -346,6 +347,9 @@ def quant_post_static(
                 can be set. Beisdes, batch_generator supports lod tensor.
         sample_generator(Python Generator): The sample generator provides 
             calibrate data for DataLoader, and it only returns a sample every time.
+        data_loader(Python Generator, Paddle.io.DataLoader, optional): The
+            Generator or Dataloader provides calibrate data, and it could
+            return a batch every time.
         model_filename(str, optional): The name of model file. If parameters 
             are saved in separate files, set it as 'None'. Default: 'None'.
         params_filename(str, optional): The name of params file.
@@ -399,6 +403,7 @@ def quant_post_static(
         executor=executor,
         sample_generator=sample_generator,
         batch_generator=batch_generator,
+        data_loader=data_loader,
         model_dir=model_dir,
         model_filename=model_filename,
         params_filename=params_filename,