Skip to content

A community-supported supercharged document management system: scan, index and archive all your documents

Fix: Use a persistent, writeable location for hugging face models (#12771)

Trenton Hda0f25b546d1
diff --git a/src/paperless_ai/embedding.py b/src/paperless_ai/embedding.py
index 7fb594a97..e03ddae72 100644
--- a/src/paperless_ai/embedding.py
+++ b/src/paperless_ai/embedding.py
@@ -39,6 +39,7 @@ def get_embedding_model() -> "BaseEmbedding":
return HuggingFaceEmbedding(
model_name=config.llm_embedding_model
or "sentence-transformers/all-MiniLM-L6-v2",
+ cache_folder=str(settings.DATA_DIR / "hf_cache"),
)
case LLMEmbeddingBackend.OLLAMA:
from llama_index.embeddings.ollama import OllamaEmbedding
diff --git a/src/paperless_ai/tests/test_embedding.py b/src/paperless_ai/tests/test_embedding.py
index 3fb5c39ce..6ee1994b8 100644
--- a/src/paperless_ai/tests/test_embedding.py
+++ b/src/paperless_ai/tests/test_embedding.py
@@ -3,6 +3,7 @@ from unittest.mock import MagicMock
from unittest.mock import patch
import pytest
+from django.conf import settings
from documents.models import Document
from paperless.models import LLMEmbeddingBackend
@@ -116,6 +117,7 @@ def test_get_embedding_model_huggingface(mock_ai_config):
model = get_embedding_model()
MockHuggingFaceEmbedding.assert_called_once_with(
model_name="sentence-transformers/all-MiniLM-L6-v2",
+ cache_folder=str(settings.DATA_DIR / "hf_cache"),
)
assert model == MockHuggingFaceEmbedding.return_value