admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-07-23 17:56:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

tests(annotations): fix stale canned-VLM markers + action_record style assertion

Browse Source 
The annotation tests had never actually run in CI (collection failed on
the missing 'datasets' extra); now that they do, three stale assertions
surfaced against the evolved pipeline:

  * test_module1_plan_memory_subtask_smoke: the memory canned-responder
    marker 'Update the memory' no longer appears in module_1_memory.txt
    (now 'compressed semantic memory'), so the stub returned no memory
    row and the {subtask,plan,memory} subset check failed. Marker
    updated to match the current prompt.
  * test_module2_mid_episode_emits_paired_interjection_and_speech: the
    interjection marker 'Write ONE interjection' is now 'Write ONE
    compact interjection' in module_2_interjection.txt, so 0 interjections
    were emitted. Marker updated.
  * tests/datasets/test_language.py::test_style_registry_routes_columns:
    PERSISTENT_STYLES gained 'action_record' in this PR; add it to the
    expected set.

These are test/prompt-marker syncs — no production behavior change.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Pepijn
2026-06-03 16:21:17 +02:00
parent 273a8fc335
commit a18d969753
2 changed files with 6 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/annotations/test_modules.py
+5 -7
View File
@@ -88,7 +88,7 @@ def test_module1_plan_memory_subtask_smoke(fixture_dataset_root: Path, tmp_path:
{"text": "place the sponge into the sink", "start": 0.8, "end": 1.1}, {"text": "place the sponge into the sink", "start": 0.8, "end": 1.1},
] ]
}, },
"Update the memory": {"memory": "wiped the counter once"}, "compressed semantic memory": {"memory": "wiped the counter once"},
}, },
) )
module = PlanSubtasksMemoryModule(vlm=vlm, config=PlanConfig()) module = PlanSubtasksMemoryModule(vlm=vlm, config=PlanConfig())
@@ -151,12 +151,10 @@ def test_module2_mid_episode_emits_paired_interjection_and_speech(
{ {
"acknowledgement the robot": {"text": "OK."}, "acknowledgement the robot": {"text": "OK."},
# Marker matches the distinctive line of # Marker matches the distinctive line of
# ``module_2_interjection.txt``. The old marker # ``module_2_interjection.txt`` ("Write ONE compact
# ("ONE realistic interruption") came from a previous prompt # interjection ..."). Keep this in sync with that prompt's
# version that asked for counterfactual interjections; the # wording — the canned responder matches on substring.
# current design anchors on subtask boundaries instead, so "Write ONE compact interjection": {
# the prompt and its marker changed.
"Write ONE interjection": {
"interjection": "now wipe the counter please", "interjection": "now wipe the counter please",
"speech": "On it.", "speech": "On it.",
}, },
tests/datasets/test_language.py
+1 -1
View File
@@ -64,7 +64,7 @@ def test_validate_feature_language_warns_only_on_non_empty_value(caplog):
def test_style_registry_routes_columns(): def test_style_registry_routes_columns():
assert {"subtask", "plan", "memory", "motion", "task_aug"} == PERSISTENT_STYLES assert {"subtask", "plan", "memory", "motion", "task_aug", "action_record"} == PERSISTENT_STYLES
assert {"interjection", "vqa", "trace"} == EVENT_ONLY_STYLES assert {"interjection", "vqa", "trace"} == EVENT_ONLY_STYLES
assert PERSISTENT_STYLES | EVENT_ONLY_STYLES <= STYLE_REGISTRY assert PERSISTENT_STYLES | EVENT_ONLY_STYLES <= STYLE_REGISTRY