mirror of
https://github.com/huggingface/lerobot.git
synced 2026-05-15 16:49:55 +00:00
Pepijn
a635a32290
Adds the steerable annotation pipeline (`lerobot-annotate`) that populates the `language_persistent` and `language_events` columns introduced in PR 1 directly into `data/chunk-*/file-*.parquet`. No flavor namespace, no sidecar tree. Modules produced: - Module 1 (plan_subtasks_memory): Pi0.7-style subtasks, plan (init + refresh on interjection), MEM-style memory at subtask boundaries. - Module 2 (interjections_and_speech): t=0 speech-only acknowledgement, mid-episode paired interjection + speech tool-call atom. - Module 3 (general_vqa): bbox/keypoint/count/attribute/spatial pairs at configurable cadence with one-retry JSON validation. Writer enforces: per-episode persistent identity, exact-frame event timestamps, column routing per `column_for_style`, dataset-level `tools` column with the `say` schema, drops legacy `subtask_index`. Validator runs against staged JSONL artifacts before the writer rewrites parquet. Adds `lerobot-annotate` console script, `annotations` extra (datatrove + optional vllm), `make annotation-e2e` opt-in smoke target, and `docs/source/annotation_pipeline.mdx`. Branched from PR 1 (`feat/language-columns`). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
59 lines
2.0 KiB
Python
59 lines
2.0 KiB
Python
#!/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.
|
|
"""Helpers shared across annotation-pipeline tests."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
from typing import Any
|
|
|
|
from lerobot.annotations.steerable_pipeline.vlm_client import StubVlmClient
|
|
|
|
|
|
def make_canned_responder(
|
|
responses_by_marker: dict[str, Any],
|
|
default: Any = None,
|
|
) -> StubVlmClient:
|
|
"""Return a stub that picks a response by inspecting the user prompt.
|
|
|
|
For each call the responder examines the last user-message text and
|
|
returns the response keyed by the first marker substring it contains.
|
|
Falls back to ``default`` if no marker matches.
|
|
"""
|
|
|
|
def responder(messages: list[dict[str, Any]]) -> Any:
|
|
last_user_text = ""
|
|
for message in messages:
|
|
if message.get("role") != "user":
|
|
continue
|
|
content = message.get("content")
|
|
if isinstance(content, str):
|
|
last_user_text = content
|
|
elif isinstance(content, list):
|
|
for block in content:
|
|
if isinstance(block, dict) and block.get("type") == "text":
|
|
last_user_text = block.get("text", "")
|
|
for marker, response in responses_by_marker.items():
|
|
if marker in last_user_text:
|
|
return response
|
|
return default
|
|
|
|
return StubVlmClient(responder=responder)
|
|
|
|
|
|
def encode_vqa_answer(payload: dict[str, Any]) -> str:
|
|
return json.dumps(payload, sort_keys=True)
|