Skip to content

fix(ci): use the cpu variant of xllamacpp in llm2 #214

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
marcelklehr merged 2 commits into from
Jul 20, 2026
Merged

fix(ci): use the cpu variant of xllamacpp in llm2 #214

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
bf48b03
fix(ci): use the cpu variant of xllamacpp in llm2
kyteinsky Jul 20, 2026
c342164
fix(ci): test for get_file_tree tool call and count to 50 instead of 500
kyteinsky Jul 20, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 14 additions & 4 deletions .github/workflows/integration_test.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -152,6 +152,10 @@ jobs:
pipx install poetry
mkdir -p ../llm2-persistent_storage
poetry install
# pyproject.toml pins xllamacpp to the cu128 wheel, which needs libcuda.so.1 at import
# time. The CI runner has no NVIDIA driver, so we swap in the PyPI CPU build for tests.
# see https://github.com/nextcloud/llm2/blob/d1c5604d75d008359937fcb16f1d82e1f421a87c/.github/workflows/integration_test.yml#L133-L135
poetry run pip install --force-reinstall --no-deps xllamacpp

- name: Cache llm2 models
uses: actions/cache/restore@v5
Expand Down Expand Up @@ -225,7 +229,7 @@ jobs:
MCP_URL="http://localhost:8080/index.php/apps/app_api/proxy/context_agent/mcp/"
CREDS=$(OC_PASS=alice ./occ user:add-app-password --password-from-env alice | tail -n 1)

sleep 300
sleep 60

ATTEMPT=0
until [ "$ATTEMPT" -ge 20 ]; do
Expand Down Expand Up @@ -253,7 +257,7 @@ jobs:
[ "$STATUS" = "200" ]
echo "$BODY"
echo "$BODY" | grep -q 'list_calendars'

- name: Cache llm2 models
uses: actions/cache/save@v5
env:
Expand All @@ -262,6 +266,11 @@ jobs:
path: llm2-persistent_storage/
key: ${{ steps.cache-llm2-models-restore.outputs.cache-primary-key }}

- name: Create a sample file in alice's directory
run: |
echo "hello nextcloud" > data/alice/files/my_test_file.txt
./occ files:scan alice

- name: Run task
env:
CREDS: "alice:alice"
Expand All @@ -283,7 +292,8 @@ jobs:
[ "$TASK_STATUS" == '"STATUS_SUCCESSFUL"' ]
echo $TASK | jq '.ocs.data.task.output.output'
echo $TASK | jq '.ocs.data.task.output.sources'
echo $TASK | jq '.ocs.data.task.output.sources' | grep -q 'get_folder_tree'
echo $TASK | jq '.ocs.data.task.output.sources' | grep -qE 'get_folder_tree|get_file_tree'
echo $TASK | jq '.ocs.data.task.output.output' | grep -q 'my_test_file'

- name: Run streaming task
if: ${{ matrix.server-versions == 'master' }}
Expand All @@ -292,7 +302,7 @@ jobs:
run: |
TASK=$(curl -X POST -u "$CREDS" -H "oCS-APIRequest: true" -H "Content-type: application/json" \
"http://localhost:8080/ocs/v2.php/taskprocessing/schedule?format=json" \
--data-raw '{"input": {"input": "Count to 500", "confirmation":1, "conversation_token": ""},"type":"core:contextagent:interaction", "appId": "test", "customId": "", "preferStreaming": true}')
--data-raw '{"input": {"input": "Count to 50", "confirmation":1, "conversation_token": ""},"type":"core:contextagent:interaction", "appId": "test", "customId": "", "preferStreaming": true}')
echo "$TASK"
TASK_ID=$(echo "$TASK" | jq '.ocs.data.task.id')

Expand Down
Loading