init

vw-document-ai-indexer/env.yaml

@@ -0,0 +1,51 @@
+config: config.yaml
+njobs: 1
+
+
+search_service_name: https://<resource name>.search.windows.net
+search_admin_key:
+
+embedding_model_endpoint: https://<resource name>.openai.azure.com/openai/deployments/text-embedding-3-small/embeddings?api-version=2024-12-01-preview
+embedding_model_key:
+VECTOR_DIMENSION: 1536
+
+extract_method: di+vision-llm
+
+# extract_method=vision-llm
+
+form_rec_resource: https://<resource name>.cognitiveservices.azure.cn/
+form_rec_key:
+
+# Perform OCR at a higher resolution to handle documents with fine print
+di-hiRes: true
+# Enable the detection of mathematical expressions in the document.
+di-Formulas: true
+di_allow_features_ext: pdf;jpep;jpg;png;bmp;tiff;heif
+
+
+
+# 图片理解
+figure_caption:
+  include_di_content: false
+  description_gen_max_images: 0
+  model_endpoint: null
+  model_key: null
+  model: null                # azure  留空
+  azure_deployment: gpt-4o   # azure 部署名称,其他平台模型留空
+  api_version: 2024-08-01-preview  # azure api版本,其他平台留空
+
+
+FLAG_AOAI: "V3"
+#FLAG_EMBEDDING_MODEL: "qwen3-embedding-8b"
+FLAG_EMBEDDING_MODEL: "AOAI"
+
+
+FIGURE_BLOB_ACCOUNT_URL: https://blob sas url
+
+DI_BLOB_ACCOUNT_URL: https://blob sas url
+
+DB_URI: postgresql+psycopg2://user:passwords@localhost:5433/document_indexer
+
+header_fix: true
+
+