ebm4subjects 0.5.5__tar.gz → 0.5.6__tar.gz

{ebm4subjects-0.5.5 → ebm4subjects-0.5.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ebm4subjects
-Version: 0.5.5
+Version: 0.5.6
 Summary: Embedding Based Matching for Automated Subject Indexing
 Author: Deutsche Nationalbibliothek
 Maintainer-email: Clemens Rietdorf <c.rietdorf@dnb.de>, Maximilian Kähler <m.kaehler@dnb.de>

{ebm4subjects-0.5.5 → ebm4subjects-0.5.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ebm4subjects"
-version = "0.5.5"
+version = "0.5.6"
 description = "Embedding Based Matching for Automated Subject Indexing"
 authors = [
     {name = "Deutsche Nationalbibliothek"},

{ebm4subjects-0.5.5 → ebm4subjects-0.5.6}/src/ebm4subjects/ebm_model.py

@@ -44,7 +44,7 @@ class EbmModel:
         use_altLabels: bool = True,
         hnsw_index_params: dict | str | None = None,
         embedding_model_name: str | None = None,
-        embedding_model_deployment: str = "offline-inference",
+        embedding_model_deployment: str = "mock",
         embedding_model_args: dict | str | None = None,
         encode_args_vocab: dict | str | None = None,
         encode_args_documents: dict | str | None = None,
@@ -101,7 +101,7 @@ class EbmModel:
 
         # Parameters for embedding generator
         self.generator = None
-        self.embedding_model_deployment = embedding_model_deployment
+        self.embedding_model_deployment = embedding_model_deployment.lower()
         self.embedding_model_name = embedding_model_name
         self.embedding_dimensions = int(embedding_dimensions)
         if isinstance(embedding_model_args, str) or not embedding_model_args:
@@ -182,7 +182,7 @@ class EbmModel:
         """
         if self.generator is None:
             if self.embedding_model_deployment == "in-process":
-                self.logger.info("initializing offline-inference embedding generator")
+                self.logger.info("initializing in-process embedding generator")
                 self.generator = EmbeddingGeneratorInProcess(
                     model_name=self.embedding_model_name,
                     embedding_dimensions=self.embedding_dimensions,
@@ -192,7 +192,7 @@ class EbmModel:
             elif self.embedding_model_deployment == "mock":
                 self.logger.info("initializing mock embedding generator")
                 self.generator = EmbeddingGeneratorMock(self.embedding_dimensions)
-            elif self.embedding_model_deployment == "HuggingFaceTEI":
+            elif self.embedding_model_deployment == "huggingfacetei":
                 self.logger.info("initializing API embedding generator")
                 self.generator = EmbeddingGeneratorHuggingFaceTEI(
                     model_name=self.embedding_model_name,
@@ -200,7 +200,7 @@ class EbmModel:
                     logger=self.logger,
                     **self.embedding_model_args,
                 )
-            elif self.embedding_model_deployment == "OpenAI":
+            elif self.embedding_model_deployment == "openai":
                 self.logger.info("initializing API embedding generator")
                 self.generator = EmbeddingGeneratorOpenAI(
                     model_name=self.embedding_model_name,
@@ -209,8 +209,7 @@ class EbmModel:
                     **self.embedding_model_args,
                 )
             else:
-                self.logger.error("unsupportet API for embedding generator")
-                raise NotImplementedError
+                raise NotImplementedError("Unsupportet API for embedding generator")
 
     def init_logger(
         self,
@@ -670,7 +669,7 @@ class EbmModel:
             )
             self.logger.info("training successful finished")
         except xgb.core.XGBoostError:
-            self.logger.critical(
+            self.logger.warn(
                 "XGBoost can't train with candidates equal to gold standard "
                 "or candidates with no match to gold standard at all - "
                 "Check if your training data and gold standard are correct"

{ebm4subjects-0.5.5 → ebm4subjects-0.5.6}/src/ebm4subjects/embedding_generator.py

@@ -80,17 +80,7 @@ class EmbeddingGeneratorHuggingFaceTEI(EmbeddingGenerator):
             self.logger.debug(
                 "API call successful. Everything seems to be working fine."
             )
-        elif response.status_code == 404:
-            self.logger.error(
-                "API not found under given adress! Please check the corresponding parameter!"
-            )
-            raise RuntimeError(
-                "API not found under given adress! Please check the corresponding parameter!"
-            )
         else:
-            self.logger.error(
-                "Request to API not possible! Please check the corresponding parameters!"
-            )
             raise RuntimeError(
                 "Request to API not possible! Please check the corresponding parameters!"
             )
@@ -188,29 +178,12 @@ class EmbeddingGeneratorOpenAI(EmbeddingGenerator):
         """
         Tests if the API is working with the given parameters
         """
-        try:
-            _ = self.client.embeddings.create(
-                input="This is a test request!",
-                model=self.model_name,
-                encoding_format="float",
-            )
-            self.logger.debug(
-                "API call successful. Everything seems to be working fine."
-            )
-        except NotFoundError:
-            self.logger.error(
-                "API not found under given adress! Please check the corresponding parameter!"
-            )
-            raise RuntimeError(
-                "API not found under given adress! Please check the corresponding parameter!"
-            )
-        except BadRequestError:
-            self.logger.error(
-                "Request to API not possible! Please check the corresponding parameters!"
-            )
-            raise RuntimeError(
-                "Request to API not possible! Please check the corresponding parameters!"
-            )
+        _ = self.client.embeddings.create(
+            input="This is a test request!",
+            model=self.model_name,
+            encoding_format="float",
+        )
+        self.logger.debug("API call successful. Everything seems to be working fine.")
 
     def generate_embeddings(self, texts: list[str], **kwargs) -> np.ndarray:
         """