Skip to content

chore(adapters): bump granitelib-core pin for fixed requirement-check io.yaml #1706

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
planetf1 merged 2 commits into from
Oct 6, 2026
Merged

chore(adapters): bump granitelib-core pin for fixed requirement-check io.yaml #1706

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e52399d
chore(adapters): bump granitelib-core pin for fixed requirement-check…
planetf1 Oct 5, 2026
b291670
docs: sync stale granitelib-core SHA references to the new pin
planetf1 Oct 5, 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
2 changes: 1 addition & 1 deletion docs/docs/tutorials/08-adapter-schema-migrations.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ from mellea.backends.adapters.catalog import fetch_intrinsic_metadata

metadata = fetch_intrinsic_metadata("requirement-check")
print(metadata.repo_id, metadata.revision)
# ibm-granite/granitelib-core-r1.0 d0a2a96a4cd07e96f0fe7ca29a42bfe088299d43
# ibm-granite/granitelib-core-r1.0 c4b4fc6ec4fa05b6c8f02b25ac8822da085962b5
```

`LocalFileBinding.resolved_revision()` uses this pinned value whenever you
Expand Down
2 changes: 1 addition & 1 deletion mellea/backends/adapters/catalog.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -163,7 +163,7 @@ def effective_capability(self) -> str:
# `mellea/backends/adapters/adapter.py`) only proves the adapter switched on,
# not that its verdicts are trustworthy.
_RAG_SHA = "2f0b2c79c6731068625aca8045c2eb2e8912b353" # main @ 2026-05-26
_CORE_R1_SHA = "d0a2a96a4cd07e96f0fe7ca29a42bfe088299d43" # main @ 2026-05-26
_CORE_R1_SHA = "c4b4fc6ec4fa05b6c8f02b25ac8822da085962b5" # main @ 2026-10-05, fixes requirement-check aLoRA io.yaml invocation sequence (#1699)
_GUARDIAN_SHA = "773b254e98f993a605ec4b6259634906e0e64e8e" # main @ 2026-05-26


Expand Down
7 changes: 0 additions & 7 deletions test/formatters/granite/base/test_base_alora_activation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -307,13 +307,6 @@ class TestAloraDifferentialEndToEnd:
scores measured 0.047 vs 0.999 in the #1679 diagnostic eval.
"""

@pytest.mark.xfail(
strict=True,
raises=AssertionError,
reason="the published requirement-check aLoRA io.yaml instruction does "
"not tokenise to its declared invocation sequence, so the adapter "
"cannot activate until the adapter is republished",
)
def test_requirement_check_adapter_moves_score(self, alora_backend):
from mellea.stdlib.components import Message
from mellea.stdlib.components.intrinsic import core
Expand Down
2 changes: 1 addition & 1 deletion test/scripts/build_ollama_uncertainty_adapter.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@ fi
BASE_MODEL="${OLLAMA_BASE_MODEL:-granite4.1:3b}"
OUTPUT_MODEL="${MELLEA_OLLAMA_UNCERTAINTY_MODEL:-mellea-test/uncertainty-alora:latest}"
ADAPTER_REPO="ibm-granite/granitelib-core-r1.0"
ADAPTER_REVISION="d0a2a96a4cd07e96f0fe7ca29a42bfe088299d43"
ADAPTER_REVISION="c4b4fc6ec4fa05b6c8f02b25ac8822da085962b5"
ADAPTER_PATH="uncertainty/granite-4.1-3b/alora"
BASE_REPO="ibm-granite/granite-4.1-3b"
BASE_REVISION="c0650403e44e78ec0262dab1c90914c65b196c4e"
Expand Down
Loading