code-loader 1.0.77__tar.gz → 1.0.77.dev2__tar.gz

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: code-loader
-Version: 1.0.77
+Version: 1.0.77.dev2
 Summary: 
 Home-page: https://github.com/tensorleap/code-loader
 License: MIT

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/code_loader/contract/datasetclasses.py

@@ -4,7 +4,8 @@ import re
 import numpy as np
 import numpy.typing as npt
 
-from code_loader.contract.enums import DataStateType, DataStateEnum, LeapDataType, ConfusionMatrixValue, MetricDirection
+from code_loader.contract.enums import DataStateType, DataStateEnum, LeapDataType, ConfusionMatrixValue, \
+    MetricDirection, DatasetMetadataType
 from code_loader.contract.visualizer_classes import LeapImage, LeapText, LeapGraph, LeapHorizontalBar, \
     LeapTextMask, LeapImageMask, LeapImageWithBBox, LeapImageWithHeatmap
 
@@ -196,6 +197,7 @@ class GroundTruthHandler(DatasetBaseHandler):
 class MetadataHandler:
     name: str
     function: MetadataSectionCallableInterface
+    metadata_type: Optional[Union[DatasetMetadataType, Dict[str, DatasetMetadataType]]] = None
 
 
 @dataclass
@@ -232,6 +234,6 @@ class DatasetIntegrationSetup:
 class DatasetSample:
     inputs: Dict[str, npt.NDArray[np.float32]]
     gt: Optional[Dict[str, npt.NDArray[np.float32]]]
-    metadata: Dict[str, Union[str, int, bool, float]]
+    metadata: Dict[str, Union[Optional[str], int, bool, Optional[float]]]
     index: Union[int, str]
     state: DataStateEnum

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/code_loader/inner_leap_binder/leapbinder.py

@@ -11,7 +11,7 @@ from code_loader.contract.datasetclasses import SectionCallableInterface, InputH
     CustomCallableInterfaceMultiArgs, ConfusionMatrixCallableInterfaceMultiArgs, LeapData, \
     CustomMultipleReturnCallableInterfaceMultiArgs, DatasetBaseHandler, custom_latent_space_attribute, \
     RawInputsForHeatmap, VisualizerHandlerData, MetricHandlerData, CustomLossHandlerData, SamplePreprocessResponse
-from code_loader.contract.enums import LeapDataType, DataStateEnum, DataStateType, MetricDirection
+from code_loader.contract.enums import LeapDataType, DataStateEnum, DataStateType, MetricDirection, DatasetMetadataType
 from code_loader.contract.responsedataclasses import DatasetTestResultPayload
 from code_loader.contract.visualizer_classes import map_leap_data_type_to_visualizer_class
 from code_loader.default_losses import loss_name_to_function
@@ -333,7 +333,8 @@ class LeapBinder:
 
         self._encoder_names.append(name)
 
-    def set_metadata(self, function: MetadataSectionCallableInterface, name: str) -> None:
+    def set_metadata(self, function: MetadataSectionCallableInterface, name: str,
+                     metadata_type: Optional[Union[DatasetMetadataType, Dict[str, DatasetMetadataType]]] = None) -> None:
         """
         Set the metadata handler function. This function is used for measuring and analyzing external variable values per sample, which is recommended for analysis within the Tensorleap platform.
 
@@ -368,7 +369,7 @@ class LeapBinder:
             leap_binder.set_metadata(metadata_handler_index, name='metadata_index')
             leap_binder.set_metadata(metadata_handler_image_mean, name='metadata_image_mean')
         """
-        self.setup_container.metadata.append(MetadataHandler(name, function))
+        self.setup_container.metadata.append(MetadataHandler(name, function, metadata_type))
 
     def set_custom_layer(self, custom_layer: Type[Any], name: str, inspect_layer: bool = False,
                          kernel_index: Optional[int] = None, use_custom_latent_space: bool = False) -> None:
@@ -465,23 +466,55 @@ class LeapBinder:
     @staticmethod
     def check_handler(
             preprocess_response: PreprocessResponse, test_result: List[DatasetTestResultPayload],
-            dataset_base_handler: Union[DatasetBaseHandler, MetadataHandler]) -> List[DatasetTestResultPayload]:
+            dataset_base_handler: Union[DatasetBaseHandler, MetadataHandler], state: DataStateEnum) -> List[DatasetTestResultPayload]:
         assert preprocess_response.sample_ids is not None
         raw_result = dataset_base_handler.function(preprocess_response.sample_ids[0], preprocess_response)
         handler_type = 'metadata' if isinstance(dataset_base_handler, MetadataHandler) else None
-        if isinstance(dataset_base_handler, MetadataHandler) and isinstance(raw_result, dict):
-            metadata_test_result_payloads = [
-                DatasetTestResultPayload(f'{dataset_base_handler.name}_{single_metadata_name}')
-                for single_metadata_name, single_metadata_result in raw_result.items()
-            ]
-            for i, (single_metadata_name, single_metadata_result) in enumerate(raw_result.items()):
-                metadata_test_result = metadata_test_result_payloads[i]
-
-                result_shape = get_shape(single_metadata_result)
-                metadata_test_result.shape = result_shape
-                metadata_test_result.raw_result = single_metadata_result
-                metadata_test_result.handler_type = handler_type
-            test_result = metadata_test_result_payloads
+        if isinstance(dataset_base_handler, MetadataHandler):
+            if isinstance(raw_result, dict):
+                metadata_test_result_payloads = [
+                    DatasetTestResultPayload(f'{dataset_base_handler.name}_{single_metadata_name}')
+                    for single_metadata_name, single_metadata_result in raw_result.items()
+                ]
+                for i, (single_metadata_name, single_metadata_result) in enumerate(raw_result.items()):
+                    metadata_test_result = metadata_test_result_payloads[i]
+
+                    metadata_type = None
+                    if single_metadata_result is None:
+                        if state != DataStateEnum.training and test_result[i].name == f'{dataset_base_handler.name}_{single_metadata_name}':
+                            metadata_test_result_payloads[i] = test_result[i]
+                            continue
+
+                        if dataset_base_handler.metadata_type is None:
+                            raise Exception(f"Metadata {single_metadata_name} is None and no metadata type is provided")
+                        elif isinstance(dataset_base_handler.metadata_type, dict):
+                            if single_metadata_name not in dataset_base_handler.metadata_type:
+                                raise Exception(f"Metadata {single_metadata_name} is None and no metadata type is provided")
+                            metadata_type = dataset_base_handler.metadata_type[single_metadata_name]
+                        else:
+                            raise Exception(f"Metadata {single_metadata_name} is None and no metadata type is provided")
+
+                    result_shape = get_shape(single_metadata_result)
+                    metadata_test_result.shape = result_shape
+                    metadata_test_result.raw_result = (
+                        single_metadata_result) if single_metadata_result is not None else metadata_type
+                    metadata_test_result.handler_type = handler_type
+                test_result = metadata_test_result_payloads
+            else:
+                if raw_result is None:
+                    if state != DataStateEnum.training:
+                        return test_result
+
+                    if dataset_base_handler.metadata_type is None:
+                        raise Exception(f"Metadata {dataset_base_handler.name} is None and no metadata type is provided")
+                    elif isinstance(dataset_base_handler.metadata_type, dict):
+                        raise Exception(f"Metadata {dataset_base_handler.name} is None and no metadata type is provided")
+                    metadata_type = dataset_base_handler.metadata_type
+
+                result_shape = get_shape(raw_result)
+                test_result[0].shape = result_shape
+                test_result[0].raw_result = raw_result if raw_result is not None else metadata_type
+                test_result[0].handler_type = handler_type
         else:
             result_shape = get_shape(raw_result)
             test_result[0].shape = result_shape

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/code_loader/inner_leap_binder/leapbinder_decorators.py

@@ -9,7 +9,7 @@ from code_loader.contract.datasetclasses import CustomCallableInterfaceMultiArgs
     CustomMultipleReturnCallableInterfaceMultiArgs, ConfusionMatrixCallableInterfaceMultiArgs, CustomCallableInterface, \
     VisualizerCallableInterface, MetadataSectionCallableInterface, PreprocessResponse, SectionCallableInterface, \
     ConfusionMatrixElement, SamplePreprocessResponse
-from code_loader.contract.enums import MetricDirection, LeapDataType
+from code_loader.contract.enums import MetricDirection, LeapDataType, DatasetMetadataType
 from code_loader import leap_binder
 from code_loader.contract.visualizer_classes import LeapImage, LeapImageMask, LeapTextMask, LeapText, LeapGraph, \
     LeapHorizontalBar, LeapImageWithBBox, LeapImageWithHeatmap
@@ -160,14 +160,15 @@ def tensorleap_custom_visualizer(name: str, visualizer_type: LeapDataType,
     return decorating_function
 
 
-def tensorleap_metadata(name: str):
+def tensorleap_metadata(
+        name: str, metadata_type: Optional[Union[DatasetMetadataType, Dict[str, DatasetMetadataType]]] = None):
     def decorating_function(user_function: MetadataSectionCallableInterface):
         for metadata_handler in leap_binder.setup_container.metadata:
             if metadata_handler.name == name:
                 raise Exception(f'Metadata with name {name} already exists. '
                                 f'Please choose another')
 
-        leap_binder.set_metadata(user_function, name)
+        leap_binder.set_metadata(user_function, name, metadata_type)
 
         def _validate_input_args(sample_id: Union[int, str], preprocess_response: PreprocessResponse):
             assert isinstance(sample_id, (int, str)), \
@@ -182,7 +183,7 @@ def tensorleap_metadata(name: str):
                  f'{preprocess_response.sample_id_type}. Got {type(sample_id)}.')
 
         def _validate_result(result):
-            supported_result_types = (int, str, bool, float, dict, np.floating,
+            supported_result_types = (type(None), int, str, bool, float, dict, np.floating,
                                       np.bool_, np.unsignedinteger, np.signedinteger, np.integer)
             assert isinstance(result, supported_result_types), \
                 (f'tensorleap_metadata validation failed: '

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/code_loader/leaploader.py

@@ -46,7 +46,7 @@ class LeapLoader(LeapLoaderBase):
 
     def evaluate_module(self) -> None:
         def append_path_recursively(full_path: str) -> None:
-            if self.code_path not in full_path or full_path == '/':
+            if '/' not in full_path or full_path == '/':
                 return
 
             parent_path = str(Path(full_path).parent)
@@ -210,7 +210,7 @@ class LeapLoader(LeapLoaderBase):
                 state_name = state.name
                 try:
                     test_result = global_leap_binder.check_handler(
-                        preprocess_response, test_result, dataset_base_handler)
+                        preprocess_response, test_result, dataset_base_handler, state)
                 except Exception as e:
                     line_number, file_name, stacktrace = get_root_exception_file_and_line_number()
                     test_result[0].display[
@@ -339,6 +339,11 @@ class LeapLoader(LeapLoaderBase):
                 continue
             if hasattr(handler_test_payload.raw_result, 'tolist'):
                 handler_test_payload.raw_result = handler_test_payload.raw_result.tolist()
+            if isinstance(handler_test_payload.raw_result, DatasetMetadataType):
+                dataset_metadata_type = handler_test_payload.raw_result
+                metadata_instances.append(DatasetMetadataInstance(name=handler_test_payload.name,
+                                                                  type=dataset_metadata_type))
+                continue
             metadata_type = type(handler_test_payload.raw_result)
             if metadata_type == int or isinstance(handler_test_payload.raw_result,
                                                   (np.unsignedinteger, np.signedinteger)):
@@ -447,21 +452,22 @@ class LeapLoader(LeapLoaderBase):
             DatasetMetadataType.int: int
         }
         metadata_type_to_default_value = {
-            DatasetMetadataType.float: -1,
-            DatasetMetadataType.string: "",
-            DatasetMetadataType.boolean: False,
-            DatasetMetadataType.int: -1
+            DatasetMetadataType.float: np.nan,
+            DatasetMetadataType.string: 'None',
+            DatasetMetadataType.boolean: np.nan,
+            DatasetMetadataType.int: np.nan
         }
 
         try:
+            if value is None:
+                raise ValueError()
             converted_value = metadata_type_to_python_type[metadata_name_to_type[metadata_name]](value)
         except ValueError:
             converted_value = metadata_type_to_default_value[metadata_name_to_type[metadata_name]]
 
         return converted_value
 
-    def _get_metadata(self, state: DataStateEnum, sample_id: Union[int, str]) -> Dict[
-        str, Union[str, int, bool, float]]:
+    def _get_metadata(self, state: DataStateEnum, sample_id: Union[int, str]) -> Dict[str, Union[str, int, bool, float]]:
         result_agg = {}
         preprocess_result = self._preprocess_result()
         preprocess_state = preprocess_result[state]

{code_loader-1.0.77 → code_loader-1.0.77.dev2}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "code-loader"
-version = "1.0.77"
+version = "1.0.77.dev2"
 description = ""
 authors = ["dorhar <doron.harnoy@tensorleap.ai>"]
 license = "MIT"