Compatible with Langchain version 0.3.0 (#724)

* Compatible with Langchain version 0.3.0

* Change clova default batch 8 to 5

* Change UpstageLayoutAnalysis to DocumentParse

* upgrade langchain-core version 0.3.0 at requirements.txt

* set limit batch size at clova.py

autorag/__init__.py

@@ -15,7 +15,7 @@
 from llama_index.llms.ollama import Ollama
 from llama_index.llms.openai import OpenAI
 from llama_index.llms.openai_like import OpenAILike
-from langchain.embeddings import OpenAIEmbeddings
+from langchain_openai.embeddings import OpenAIEmbeddings
 from rich.logging import RichHandler
 
 version_path = os.path.join(os.path.dirname(os.path.realpath(__file__)), "VERSION")

autorag/data/__init__.py

@@ -15,7 +15,7 @@
 	DirectoryLoader,
 )
 from langchain_unstructured import UnstructuredLoader
-from langchain_upstage import UpstageLayoutAnalysisLoader
+from langchain_upstage import UpstageDocumentParseLoader
 
 from llama_index.core.node_parser import (
 	TokenTextSplitter,
@@ -56,7 +56,7 @@
 	# 6. All files
 	"directory": DirectoryLoader,
 	"unstructured": UnstructuredLoader,
-	"upstagelayoutanalysis": UpstageLayoutAnalysisLoader,
+	"upstagedocumentparse": UpstageDocumentParseLoader,
 }
 
 chunk_modules = {

autorag/data/parse/clova.py

@@ -16,7 +16,7 @@ def clova_ocr(
 	data_path_list: List[str],
 	url: Optional[str] = None,
 	api_key: Optional[str] = None,
-	batch: int = 8,
+	batch: int = 5,
 	table_detection: bool = False,
 ) -> Tuple[List[str], List[str], List[int]]:
 	"""
@@ -46,6 +46,8 @@ def clova_ocr(
 			"Please set the API key for Clova OCR in the environment variable CLOVA_API_KEY "
 			"or directly set it on the config YAML file."
 		)
+	if batch > 5:
+		raise ValueError("The batch size should be less than or equal to 5.")
 
 	image_data_lst = list(
 		map(lambda data_path: pdf_to_images(data_path), data_path_list)

requirements.txt

@@ -49,7 +49,7 @@ streamlit
 gradio
 
 ### Langchain ###
-langchain-core>=0.1.6
+langchain-core>=0.3.0
 langchain_unstructured
 langchain-upstage