Regenerate uv.lock, revert hacky Dockerfile stuff

Co-Authored-By: Trenton H <797416+stumpylog@users.noreply.github.com>

Dockerfile

@@ -196,7 +196,11 @@ RUN set -eux \
     && apt-get install --yes --quiet --no-install-recommends ${BUILD_PACKAGES} \
   && echo "Installing Python requirements" \
     && uv export --quiet --no-dev --all-extras --format requirements-txt --output-file requirements.txt \
-    && uv pip install --no-cache --system --no-python-downloads --python-preference system --requirements requirements.txt \
+    && uv pip install --no-cache --system --no-python-downloads --python-preference system \
+      --index https://pypi.org/simple \
+      --index https://download.pytorch.org/whl/cpu \
+      --index-strategy unsafe-best-match \
+      --requirements requirements.txt \
   && echo "Installing NLTK data" \
     && python3 -W ignore::RuntimeWarning -m nltk.downloader -d "/usr/share/nltk_data" snowball_data \
     && python3 -W ignore::RuntimeWarning -m nltk.downloader -d "/usr/share/nltk_data" stopwords \

pyproject.toml

@@ -28,7 +28,7 @@ dependencies = [
   #          Only patch versions are guaranteed to not introduce breaking changes.
   "django~=5.2.5",
   "django-allauth[mfa,socialaccount]~=65.12.1",
-  "django-auditlog~=3.4.1",
+  "django-auditlog~=3.3.0",
   "django-cachalot~=2.8.0",
   "django-celery-results~=2.6.0",
   "django-compression-middleware~=0.5.0",
@@ -47,7 +47,7 @@ dependencies = [
   "faiss-cpu>=1.10",
   "filelock~=3.20.0",
   "flower~=2.0.1",
-  "gotenberg-client~=0.13.1",
+  "gotenberg-client~=0.12.0",
   "httpx-oauth~=0.16",
   "imap-tools~=1.11.0",
   "inotifyrecursive~=0.3",
@@ -60,7 +60,7 @@ dependencies = [
   "llama-index-llms-openai>=0.3.38",
   "llama-index-vector-stores-faiss>=0.3",
   "nltk~=3.9.1",
-  "ocrmypdf~=16.13.0",
+  "ocrmypdf~=16.12.0",
   "openai>=1.76",
   "pathvalidate~=3.3.1",
   "pdf2image~=1.17.0",
@@ -77,6 +77,7 @@ dependencies = [
   "sentence-transformers>=4.1",
   "setproctitle~=1.3.4",
   "tika-client~=0.10.0",
+  "torch~=2.7.0",
   "tqdm~=4.67.1",
   "watchdog~=6.0",
   "whitenoise~=6.9",
@@ -91,7 +92,7 @@ optional-dependencies.postgres = [
   "psycopg[c,pool]==3.2.12",
   # Direct dependency for proper resolution of the pre-built wheels
   "psycopg-c==3.2.12",
-  "psycopg-pool==3.3",
+  "psycopg-pool==3.2.7",
 ]
 optional-dependencies.webserver = [
   "granian[uvloop]~=2.5.1",
@@ -126,7 +127,7 @@ testing = [
 ]
 
 lint = [
-  "pre-commit~=4.5.1",
+  "pre-commit~=4.4.0",
   "pre-commit-uv~=4.2.0",
   "ruff~=0.14.0",
 ]
@@ -169,6 +170,15 @@ zxing-cpp = [
   { url = "https://github.com/paperless-ngx/builder/releases/download/zxing-2.3.0/zxing_cpp-2.3.0-cp312-cp312-linux_aarch64.whl", marker = "sys_platform == 'linux' and platform_machine == 'aarch64' and python_version == '3.12'" },
 ]
 
+torch = [
+  { index = "pytorch-cpu" },
+]
+
+[[tool.uv.index]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+explicit = true
+
 [tool.ruff]
 target-version = "py310"
 line-length = 88

src/paperless_ai/tests/test_chat.py

@@ -11,12 +11,14 @@ from paperless_ai.chat import stream_chat_with_documents
 @pytest.fixture(autouse=True)
 def patch_embed_model():
     from llama_index.core import settings as llama_settings
-    from llama_index.core.embeddings.utils import MockEmbedding
 
-    mock_embed_model = MockEmbedding(embed_dim=8)
-    llama_settings.Settings.embed_model = mock_embed_model
+    mock_embed_model = MagicMock()
+    mock_embed_model._get_text_embedding_batch.return_value = [
+        [0.1] * 1536,
+    ]  # 1 vector per input
+    llama_settings.Settings._embed_model = mock_embed_model
     yield
-    llama_settings.Settings.embed_model = None
+    llama_settings.Settings._embed_model = None
 
 
 @pytest.fixture(autouse=True)