fix bugs

paddlepalm/interface.py

@@ -154,7 +154,11 @@ class task_paradigm(object):
 
         raise NotImplementedError()
 
-    def build(self, inputs):
+    @property
+    def epoch_inputs_attrs(self):
+        return {}
+
+    def build(self, inputs, scope_name=""):
         """建立task_layer的计算图。将符合inputs_attrs描述的来自各个对象集的静态图Variables映射成符合outputs_attr描述的静态图Variable输出。
         Args:
             inputs: dict类型。字典中包含inputs_attrs中的对象名到计算图Variable的映射，inputs中至少会包含inputs_attr中定义的对象
@@ -168,6 +172,6 @@ class task_paradigm(object):
         """每个训练或推理step后针对当前batch的task_layer的runtime计算结果进行相关后处理。注意，rt_outputs除了包含build方法，还自动包含了loss的计算结果。"""
         pass
         
-    def post_postprocess(self, global_buffer):
+    def epoch_postprocess(self, post_inputs):
         pass
 

paddlepalm/mtl_controller.py

@@ -182,21 +182,17 @@ def _fit_attr(conf, fit_attr, strict=False):
 
 class Controller(object):
 
-    def __init__(self, config=None, task_dir='.', for_train=True):
+    def __init__(self, config, task_dir='.', for_train=True):
         """
         Args:
             config: (str|dict) 字符串类型时，给出yaml格式的config配置文件路径；
         """
 
         self._for_train = for_train
-        # default mtl_conf
-        # if config is None and config_path is None:
-        #     raise ValueError('For config and config_path, at least one of them should be set.')
+        assert isinstance(config, str) or isinstance(config, dict), "a config dict or config file path is required to create a Controller."
 
         if isinstance(config, str):
             mtl_conf = _parse_yaml(config, support_cmd_line=True)
-            # if config is not None:
-            #     mtl_conf = _merge_conf(config, mtl_conf)
         else:
             mtl_conf = config
                 
@@ -518,6 +514,11 @@ class Controller(object):
 
     def _init_pred(self, instance, infer_model_path):
         inst = instance
+        if 'pred_output_path' not in inst.config:
+            inst.config['pred_output_path'] = os.path.join(inst.config.get('save_path', '.'), inst.name)
+
+        if not os.path.exists(inst.config['pred_output_path']):
+            os.makedirs(inst.config['pred_output_path'])
 
         pred_backbone = self.Backbone(self.bb_conf, phase='pred')
         pred_parad = inst.Paradigm(inst.config, phase='pred', backbone_config=self.bb_conf)
@@ -563,7 +564,12 @@ class Controller(object):
         finish = []
         for inst in instances:
             if inst.is_target:
-                finish.append(False)
+                if inst.expected_train_steps > 0:
+                    finish.append(False)
+                else:
+                    finish.append(True)
+                    print(inst.name+': train finished!')
+                    inst.save()
         
         def train_finish():
             for inst in instances:
@@ -641,9 +647,11 @@ class Controller(object):
         pred_prog = self._init_pred(instance, inference_model_dir)
                 
         inst = instance
+        print(inst.name+": loading data...")
         inst.reader['pred'].load_data()
         fetch_names, fetch_vars = inst.pred_fetch_list
 
+        print('predicting...')
         mapper = {k:v for k,v in inst.pred_input}
         buf = []
         for feed in inst.reader['pred'].iterator():
@@ -653,12 +661,13 @@ class Controller(object):
             rt_outputs = self.exe.run(pred_prog, feed, fetch_vars)
             rt_outputs = {k:v for k,v in zip(fetch_names, rt_outputs)}
             inst.postprocess(rt_outputs, phase='pred')
-        reader_outputs = inst.reader['pred'].get_epoch_outputs()
+        if inst.task_layer['pred'].epoch_inputs_attrs:
+            reader_outputs = inst.reader['pred'].get_epoch_outputs()
+        else:
+            reader_outputs = None
         inst.epoch_postprocess({'reader':reader_outputs}, phase='pred')
 
 
-
-
 if __name__ == '__main__':
     assert len(sys.argv) == 2, "Usage: python mtl_controller.py <mtl_conf_path>"
     conf_path = sys.argv[1]

paddlepalm/reader/match.py

@@ -93,10 +93,6 @@ class Reader(reader):
         for batch in self._data_generator():
             yield list_to_dict(batch)
 
-    def get_epoch_outputs(self):
-        return {'examples': self._reader.get_examples(self._phase),
-                'features': self._reader.get_features(self._phase)}
-
     @property
     def num_examples(self):
         return self._reader.get_num_examples(phase=self._phase)

paddlepalm/task_paradigm/cls.py

@@ -14,8 +14,10 @@
 # limitations under the License.
 
 import paddle.fluid as fluid
-from paddlepalm.interface import task_paradigm
 from paddle.fluid import layers
+from paddlepalm.interface import task_paradigm
+import numpy as np
+import os
 
 class TaskParadigm(task_paradigm):
     '''
@@ -35,6 +37,8 @@ class TaskParadigm(task_paradigm):
             self._dropout_prob = config['dropout_prob']
         else:
             self._dropout_prob = backbone_config.get('hidden_dropout_prob', 0.0)
+        self._pred_output_path = config.get('pred_output_path', None)
+        self._preds = []
 
     @property
     def inputs_attrs(self):
@@ -78,3 +82,20 @@ class TaskParadigm(task_paradigm):
         else:
             return {"logits":logits}
 
+    def postprocess(self, rt_outputs):
+        if not self._is_training:
+            logits = rt_outputs['logits']
+            preds = np.argmax(logits, -1)
+            self._preds.extend(preds.tolist())
+
+    def epoch_postprocess(self, post_inputs):
+        # there is no post_inputs needed and not declared in epoch_inputs_attrs, hence no elements exist in post_inputs
+        if not self._is_training:
+            if self._pred_output_path is None:
+                raise ValueError('argument pred_output_path not found in config. Please add it into config dict/file.')
+            with open(os.path.join(self._pred_output_path, 'predictions.json'), 'w') as writer:
+                for p in self._preds:
+                    writer.write(str(p)+'\n')
+            print('Predictions saved at '+os.path.join(self._pred_output_path, 'predictions.json'))
+
+                

paddlepalm/task_paradigm/match.py

@@ -14,8 +14,10 @@
 # limitations under the License.
 
 import paddle.fluid as fluid
-from paddlepalm.interface import task_paradigm
 from paddle.fluid import layers
+from paddlepalm.interface import task_paradigm
+import numpy as np
+import os
 
 class TaskParadigm(task_paradigm):
     '''
@@ -35,6 +37,9 @@ class TaskParadigm(task_paradigm):
         else:
             self._dropout_prob = backbone_config.get('hidden_dropout_prob', 0.0)
 
+        self._pred_output_path = config.get('pred_output_path', None)
+        self._preds = []
+
     
     @property
     def inputs_attrs(self):
@@ -50,7 +55,7 @@ class TaskParadigm(task_paradigm):
         if self._is_training:
             return {"loss": [[1], 'float32']}
         else:
-            return {"logits": [[-1, 1], 'float32']}
+            return {"logits": [[-1, 2], 'float32']}
 
     def build(self, inputs, scope_name=""):
         if self._is_training:
@@ -81,3 +86,20 @@ class TaskParadigm(task_paradigm):
         else:
             return {'logits': logits}
 
+    def postprocess(self, rt_outputs):
+        if not self._is_training:
+            logits = rt_outputs['logits']
+            preds = np.argmax(logits, -1)
+            self._preds.extend(preds.tolist())
+
+    def epoch_postprocess(self, post_inputs):
+        # there is no post_inputs needed and not declared in epoch_inputs_attrs, hence no elements exist in post_inputs
+        if not self._is_training:
+            if self._pred_output_path is None:
+                raise ValueError('argument pred_output_path not found in config. Please add it into config dict/file.')
+            with open(os.path.join(self._pred_output_path, 'predictions.json'), 'w') as writer:
+                for p in self._preds:
+                    writer.write(str(p)+'\n')
+            print('Predictions saved at '+os.path.join(self._pred_output_path, 'predictions.json'))
+
+