admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-07-24 18:26:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs(annotate): cleaner architecture diagram layout

Browse Source 
Top-down flow (read episodes → 3 modules fan out → validator → writer →
parquet) with aligned boxes, instead of the cramped bordered version.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Pepijn
2026-06-04 11:59:31 +02:00
parent 7bec991cdf
commit 2af2402a0c
1 changed files with 26 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/source/annotation_pipeline.mdx
+26 -15
View File
@@ -14,21 +14,32 @@ trained on.
## How it fits together ## How it fits together
```text ```text
your dataset lerobot-annotate your dataset lerobot-annotate
(LeRobot v3.1) ┌──────────────────────────────────┐ (LeRobot v3.1)
│ │ read episodes │ │
└─────────────▶│ │ │ ▼
│ ▼ │ ┌─────────────────────────────────────────────────────┐
one shared │ ┌──────┐ ┌─────────────┐ ┌─────┐ │ each module writes │ read episodes │
Qwen-VL server ────▶│ │ plan │ │interjections│ │ vqa │ │ raw JSONL into └──────────────────────────┬──────────────────────────┘
(vLLM, OpenAI API) │ └──┬───┘ └──────┬──────┘ └──┬──┘ │ .annotate_staging/ │
│ └────────────┼───────────┘ │ ┌────────────────────┼────────────────────┐
│ ▼ │ ▼ ▼ ▼
│ validator │ checks everything ┌──────────┐ ┌───────────────┐ ┌──────────┐ one shared Qwen-VL
│ │ │ │ plan │ │ interjections │ │ vqa │ ◀── server (vLLM, OpenAI
│ ▼ │ └────┬─────┘ └───────┬───────┘ └────┬─────┘ API) drives all three
│ writer ──────────────┼─▶ data/chunk-*/file-*.parquet └────────────────────┼─────────────────────┘
└──────────────────────────────────┘ (+ meta/info.json tools) │ each module stages raw JSONL
▼ into .annotate_staging/
┌─────────────────┐
│ validator │ ◀── checks everything
└────────┬────────┘
▼
┌─────────────────┐
│ writer │
└────────┬────────┘
▼
data/chunk-*/file-*.parquet
(+ meta/info.json tools)
``` ```
Three modules (`plan`, `interjections`, `vqa`) all talk to **one** shared Three modules (`plan`, `interjections`, `vqa`) all talk to **one** shared