code-loader 1.0.178__tar.gz → 1.0.179.dev1__tar.gz

{code_loader-1.0.178 → code_loader-1.0.179.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: code-loader
-Version: 1.0.178
+Version: 1.0.179.dev1
 Summary: 
 Home-page: https://github.com/tensorleap/code-loader
 License: MIT

{code_loader-1.0.178 → code_loader-1.0.179.dev1}/code_loader/inner_leap_binder/leapbinder_decorators.py

@@ -1226,9 +1226,8 @@ def tensorleap_metadata(
         def _validate_result(result):
             supported_result_types = (type(None), int, str, bool, float, dict, np.floating,
                                       np.bool_, np.unsignedinteger, np.signedinteger, np.integer)
-            if isinstance(result, tuple):
-                validate_output_structure(result, func_name=user_function.__name__,
-                                          expected_type_name=supported_result_types)
+            validate_output_structure(result, func_name=user_function.__name__,
+                                      expected_type_name=supported_result_types)
             assert isinstance(result, supported_result_types), \
                 (f'{user_function.__name__}() validation failed: '
                  f'Unsupported return type. Got {type(result)}. should be any of {str(supported_result_types)}')
@@ -1293,8 +1292,20 @@ def tensorleap_custom_latent_space():
 
         def _validate_result(result):
             assert isinstance(result, np.ndarray), \
-                (f'tensorleap_custom_loss validation failed: '
+                (f'tensorleap_custom_latent_space validation failed: '
                  f'The return type should be a numpy array. Got {type(result)}.')
+            if result.ndim > 1:
+                warning_key = ("tensorleap_custom_latent_space_flatten", tuple(result.shape))
+                if warning_key not in _STORED_WARNING_KEYS:
+                    _STORED_WARNING_KEYS.add(warning_key)
+                    flat_dim = int(np.prod(result.shape))
+                    warnings.warn(
+                        f"tensorleap_custom_latent_space returned per-sample shape {tuple(result.shape)} "
+                        f"(ndim={result.ndim}). Tensorleap assumes per-sample shape (d, ...) and will "
+                        f"flatten to ({flat_dim},) before downstream visualization and clustering. "
+                        f"If you want a different aggregation (e.g. global average pooling), do it "
+                        f"inside your function."
+                    )
 
         def inner_without_validate(sample_id, preprocess_response):
             global _called_from_inside_tl_decorator

{code_loader-1.0.178 → code_loader-1.0.179.dev1}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "code-loader"
-version = "1.0.178"
+version = "1.0.179.dev1"
 description = ""
 authors = ["dorhar <doron.harnoy@tensorleap.ai>"]
 license = "MIT"