Add ONNX Export (#2061)

* Add ONNX Export

* deal with comments

* deal with comments

* deal with comments

Author: wjj19950828

paddlenlp/trainer/trainer_base.py

@@ -1042,14 +1042,16 @@ def save_model(self, output_dir: Optional[str]=None):
     def export_model(self,
                      input_spec=None,
                      load_best_model=False,
-                     output_dir: Optional[str]=None):
-        """ Export paddle inference model.
+                     output_dir: Optional[str]=None,
+                     model_format: Optional[str]="paddle"):
+        """ Export paddle inference model or onnx model.
 
         Args:
             input_spec (paddle.static.InputSpec, optional): InputSpec describes the signature information of the model input, 
                 such as shape , dtype , name. Defaults to None.
             load_best_model (bool, optional): Load best model. Defaults to False.
             output_dir (Optional[str], optional): Output dir to save the exported model. Defaults to None.
+            model_format (Optional[str], optional): Export model format. There are two options: paddle or onnx, defaults to paddle.
         """
 
         if output_dir is None:
@@ -1079,14 +1081,26 @@ def export_model(self,
         model = unwrap_model(self.model)
         model.eval()
 
-        # Convert to static graph with specific input description
-        model = paddle.jit.to_static(model, input_spec=input_spec)
-
-        # Save in static graph model.
-        save_path = os.path.join(output_dir, "inference", "infer")
-        logger.info("Exporting inference model to %s" % save_path)
-        paddle.jit.save(model, save_path)
-        logger.info("Inference model exported.")
+        model_format = model_format.lower()
+        if model_format == "paddle":
+            # Convert to static graph with specific input description
+            model = paddle.jit.to_static(model, input_spec=input_spec)
+
+            # Save in static graph model.
+            save_path = os.path.join(output_dir, "inference", "infer")
+            logger.info("Exporting inference model to %s" % save_path)
+            paddle.jit.save(model, save_path)
+            logger.info("Inference model exported.")
+        elif model_format == "onnx":
+            # Export ONNX model.
+            save_path = os.path.join(output_dir, "onnx", "model")
+            logger.info("Exporting ONNX model to %s" % save_path)
+            paddle.onnx.export(model, save_path, input_spec=input_spec)
+            logger.info("ONNX model exported.")
+        else:
+            logger.info(
+                "This export format is not supported, please select paddle or onnx!"
+            )
 
     def _save_checkpoint(self, model, metrics=None):
         # assert unwrap_model(model) is self.model, "internal model should be a reference to self.model"

requirements.txt

@@ -8,3 +8,4 @@ datasets
 tqdm
 paddlefsl
 sentencepiece
+paddle2onnx