feat: language annotation pipeline (#3471)

Steerable annotation pipeline (lerobot-annotate) that populates the language_persistent and language_events columns introduced in PR 1 (#3467) directly into data/chunk-*/file-*.parquet.

This is PR 2 of the three-PR plan:

PR 1 (Add extensive language support #3467): schema + DSL + rendering, base of this PR
PR 2 (this PR): annotation pipeline writing into PR 1's columns
PR 3: model with language prediction and runtime
A VLM (Qwen-VL family, served on vLLM) watches each episode's video and emits grounded language annotations: subtasks, plans, memory, task rephrasings, interjections + speech, and per-camera VQA. The pipeline is built for production annotation at scale — single-camera grounding, embedded-frame inputs, a describe-then-segment grounding flow, and a deterministic full-episode coverage guarantee — informed by Scale's dense-captioning findings (representation > sampling, rules > reasoning, model capacity is the biggest lever, two-pass systems compound errors)

tests/annotations/conftest.py

@@ -0,0 +1,58 @@
+#!/usr/bin/env python
+
+# Copyright 2026 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Shared fixtures for annotation-pipeline tests.
+
+The on-disk dataset builder lives with the other dataset factories in
+``tests/fixtures/dataset_factories.py`` (:func:`build_annotation_dataset`);
+these fixtures only wire it into pytest.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+import pytest
+
+# ``build_annotation_dataset`` pulls in ``lerobot.datasets`` (HF ``datasets``
+# + ``pandas``, only in the ``dataset`` extra), so it's imported lazily inside
+# each fixture — this conftest stays importable without that extra. The test
+# modules ``pytest.importorskip("datasets")`` so they skip rather than error.
+
+
+@pytest.fixture
+def fixture_dataset_root(tmp_path: Path) -> Path:
+    """A tiny dataset with two episodes, 12 frames each at 10 fps."""
+    from tests.fixtures.dataset_factories import build_annotation_dataset
+
+    return build_annotation_dataset(
+        tmp_path / "ds",
+        episode_specs=[
+            (0, 12, "Could you tidy the kitchen please?"),
+            (1, 12, "Please clean up the kitchen"),
+        ],
+        fps=10,
+    )
+
+
+@pytest.fixture
+def single_episode_root(tmp_path: Path) -> Path:
+    from tests.fixtures.dataset_factories import build_annotation_dataset
+
+    return build_annotation_dataset(
+        tmp_path / "ds_one",
+        episode_specs=[(0, 30, "Pour water from the bottle into the cup.")],
+        fps=10,
+    )