code-loader 1.0.182.dev0__tar.gz → 1.0.184.dev0__tar.gz

{code_loader-1.0.182.dev0 → code_loader-1.0.184.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: code-loader
-Version: 1.0.182.dev0
+Version: 1.0.184.dev0
 Summary: 
 Home-page: https://github.com/tensorleap/code-loader
 License: MIT

{code_loader-1.0.182.dev0 → code_loader-1.0.184.dev0}/code_loader/leaploader.py

@@ -44,6 +44,7 @@ class LeapLoader(LeapLoaderBase):
         super().__init__(code_path, code_entry_name)
 
         self._preprocess_result_cached = None
+        self._synthetic_lookup: Dict[str, Tuple[PreprocessResponse, Any]] = {}
 
         try:
             from code_loader.mixpanel_tracker import track_code_loader_loaded
@@ -189,6 +190,13 @@ class LeapLoader(LeapLoaderBase):
 
     def get_sample(self, state: DataStateEnum, sample_id: Union[int, str], instance_id: int = None) -> DatasetSample:
         self.exec_script()
+
+        if isinstance(sample_id, str) and sample_id in self._synthetic_lookup:
+            sim_preprocess, original_local_id = self._synthetic_lookup[sample_id]
+            return self._get_sample_from_preprocess(
+                sim_preprocess, original_local_id, synthetic_index=sample_id
+            )
+
         preprocess_result = self._preprocess_result()
         if state == DataStateEnum.unlabeled and sample_id not in preprocess_result[state].sample_ids:
             self._preprocess_result(update_unlabeled_preprocess=True)
@@ -211,6 +219,50 @@ class LeapLoader(LeapLoaderBase):
                                instance_masks=instance_mask)
         return sample
 
+    def _get_sample_from_preprocess(
+        self,
+        preprocess: "PreprocessResponse",
+        original_sample_id: Union[int, str],
+        synthetic_index: str,
+    ) -> DatasetSample:
+        inputs = {}
+        for handler in global_leap_binder.setup_container.inputs:
+            inputs[handler.name] = handler.function(original_sample_id, preprocess)
+
+        gt = {}
+        for handler in global_leap_binder.setup_container.ground_truths:
+            gt[handler.name] = handler.function(original_sample_id, preprocess)
+
+        metadata = {}
+        metadata_is_none = {}
+        for handler in global_leap_binder.setup_container.metadata:
+            handler_result = handler.function(original_sample_id, preprocess)
+            if isinstance(handler_result, dict):
+                for k, v in handler_result.items():
+                    key = "{}_{}".format(handler.name, k)
+                    metadata[key], metadata_is_none[key] = self._convert_metadata_to_correct_type(key, v)
+            else:
+                metadata[handler.name], metadata_is_none[handler.name] = self._convert_metadata_to_correct_type(
+                    handler.name, handler_result
+                )
+
+        custom_latent_space = None
+        if global_leap_binder.setup_container.custom_latent_space is not None:
+            custom_latent_space = global_leap_binder.setup_container.custom_latent_space.function(
+                original_sample_id, preprocess
+            )
+
+        return DatasetSample(
+            inputs=inputs,
+            gt=gt if gt else None,
+            metadata=metadata,
+            metadata_is_none=metadata_is_none,
+            index=synthetic_index,
+            state=DataStateEnum.additional,
+            custom_latent_space=custom_latent_space,
+            instance_masks=None,
+        )
+
     def check_dataset(self) -> DatasetIntegParseResult:
         test_payloads: List[DatasetTestResultPayload] = []
         setup_response = None
@@ -345,8 +397,8 @@ class LeapLoader(LeapLoaderBase):
             result_payloads.append(test_result)
         return result_payloads
 
-    def run_simulation(self, sim_name, params=None, n_samples=1, seed=0):
-        # type: (str, Optional[Dict[str, Any]], int, int) -> Dict[str, npt.NDArray[np.float32]]
+    def run_simulation(self, sim_name, params=None, n_samples=1, seed=0, sample_ids=None):
+        # type: (str, Optional[Dict[str, Any]], int, int, Optional[List[str]]) -> Dict[str, Any]
         self.exec_script()
         sim = next(
             (s for s in global_leap_binder.setup_container.simulations if s.name == sim_name),
@@ -364,11 +416,48 @@ class LeapLoader(LeapLoaderBase):
             _simulation_context["active"] = False
         sim_preprocess.state = DataStateType.additional
         sim_preprocess.tl_generated = True
+        original_sample_ids = list(sim_preprocess.sample_ids)
         per_encoder = {handler.name: [] for handler in global_leap_binder.setup_container.inputs}
-        for sample_id in sim_preprocess.sample_ids:
+        for sample_id in original_sample_ids:
             for handler in global_leap_binder.setup_container.inputs:
                 per_encoder[handler.name].append(handler.function(sample_id, sim_preprocess))
-        return {name: np.stack(arrays) for name, arrays in per_encoder.items()}
+        encoded = {name: np.stack(arrays) for name, arrays in per_encoder.items()}
+        if sample_ids is not None:
+            if len(sample_ids) != len(original_sample_ids):
+                raise ValueError(
+                    "sample_ids length ({}) does not match simulation output length ({})".format(
+                        len(sample_ids), len(original_sample_ids)
+                    )
+                )
+            for sid in sample_ids:
+                if not isinstance(sid, str):
+                    raise TypeError(
+                        "All sample_ids must be of type str. Got: {}".format(type(sid))
+                    )
+            for synth_id, original_local_id in zip(sample_ids, original_sample_ids):
+                self._synthetic_lookup[synth_id] = (sim_preprocess, original_local_id)
+            self._extend_additional_preprocess(list(sample_ids))
+            returned_sample_ids = list(sample_ids)
+        else:
+            returned_sample_ids = original_sample_ids
+        return {"encoded": encoded, "sample_ids": returned_sample_ids}
+
+    def _extend_additional_preprocess(self, new_sample_ids: List[str]) -> None:
+        if self._preprocess_result_cached is None:
+            self._preprocess_result()
+        additional = self._preprocess_result_cached.get(DataStateEnum.additional)
+        if additional is None:
+            placeholder = PreprocessResponse(
+                sample_ids=list(new_sample_ids),
+                data={},
+                state=DataStateType.additional,
+                sample_id_type=str,
+            )
+            placeholder.tl_generated = True
+            self._preprocess_result_cached[DataStateEnum.additional] = placeholder
+        else:
+            additional.sample_ids = list(additional.sample_ids) + list(new_sample_ids)
+            additional.length = len(additional.sample_ids)
 
     @staticmethod
     def _get_all_dataset_base_handlers() -> List[Union[DatasetBaseHandler, MetadataHandler]]:
@@ -583,7 +672,13 @@ class LeapLoader(LeapLoaderBase):
                               state: DataStateEnum, sample_id: Union[int, str]) -> Dict[str, npt.NDArray[np.float32]]:
         result_agg = {}
         preprocess_result = self._preprocess_result()
-        preprocess_state = preprocess_result[state]
+        if (state == DataStateEnum.additional and isinstance(sample_id, str)
+                and sample_id in self._synthetic_lookup):
+            sim_preprocess, original_local_id = self._synthetic_lookup[sample_id]
+            preprocess_state = sim_preprocess
+            sample_id = original_local_id
+        else:
+            preprocess_state = preprocess_result[state]
         for handler in handlers:
             handler_result = handler.function(sample_id, preprocess_state)
             handler_name = handler.name
@@ -597,7 +692,13 @@ class LeapLoader(LeapLoaderBase):
         if instance_id is None:
             return None
         preprocess_result = self._preprocess_result()
-        preprocess_state = preprocess_result[state]
+        if (state == DataStateEnum.additional and isinstance(sample_id, str)
+                and sample_id in self._synthetic_lookup):
+            sim_preprocess, original_local_id = self._synthetic_lookup[sample_id]
+            preprocess_state = sim_preprocess
+            sample_id = original_local_id
+        else:
+            preprocess_state = preprocess_result[state]
         result_agg = {}
         for handler in global_leap_binder.setup_container.instance_masks:
             handler_result = handler.function(sample_id, preprocess_state, instance_id)
@@ -657,7 +758,13 @@ class LeapLoader(LeapLoaderBase):
         result_agg = {}
         is_none = {}
         preprocess_result = self._preprocess_result()
-        preprocess_state = preprocess_result[state]
+        if (state == DataStateEnum.additional and isinstance(sample_id, str)
+                and sample_id in self._synthetic_lookup):
+            sim_preprocess, original_local_id = self._synthetic_lookup[sample_id]
+            preprocess_state = sim_preprocess
+            sample_id = original_local_id
+        else:
+            preprocess_state = preprocess_result[state]
         for handler in global_leap_binder.setup_container.metadata:
             if requested_metadata_names:
                 if not is_metadata_name_starts_with_handler_name(handler):

{code_loader-1.0.182.dev0 → code_loader-1.0.184.dev0}/code_loader/leaploaderbase.py

@@ -154,8 +154,8 @@ class LeapLoaderBase:
         pass
 
     @abstractmethod
-    def run_simulation(self, sim_name, params=None, n_samples=1, seed=0):
-        # type: (str, Optional[Dict[str, Any]], int, int) -> Dict[str, Any]
+    def run_simulation(self, sim_name, params=None, n_samples=1, seed=0, sample_ids=None):
+        # type: (str, Optional[Dict[str, Any]], int, int, Optional[List[str]]) -> Dict[str, Any]
         pass
 
     def is_custom_latent_space(self) -> bool:

{code_loader-1.0.182.dev0 → code_loader-1.0.184.dev0}/pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name = "code-loader"
 
-version = "1.0.182.dev0"
+version = "1.0.184.dev0"
 description = ""
 authors = ["dorhar <doron.harnoy@tensorleap.ai>"]
 license = "MIT"