File size: 2,334 Bytes
1444629 3fb1a9c 1444629 3fb1a9c 1444629 3fb1a9c 1444629 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 | ---
language:
- en
library_name: lerobot
pipeline_tag: robotics
tags:
- vision-language-action
- imitation-learning
- lerobot
inference: false
license: gemma
base_model: lerobot/pi05_base
---
# pi05_subtask base checkpoint
A derived copy of [`lerobot/pi05_base`](https://huggingface.co/lerobot/pi05_base) prepared for the
[`lerobot_policy_pi05_subtask`](https://github.com/) LeRobot plugin (π0.5 with System-1 subtask
generation, following [openpi_subtask](https://github.com/Ke-Wang1017/openpi_subtask)).
**The model weights are bit-identical to `lerobot/pi05_base`.** Only two metadata files differ:
- `config.json`: policy `type` is `pi05_subtask` instead of `pi05`.
- `policy_preprocessor.json`: pi05's `pi05_prepare_state_tokenizer_processor_step` +
`tokenizer_processor` steps are replaced by `pi05_subtask_context_tokenize_processor_step`,
which tokenizes the context segment of the two-segment prompt
```
context (no loss): "Task: {task}.; State: {256-bin ints}; Subtask: "
predicted (CE loss): "{subtask}.;\nAction: " + EOS
```
into `observation.language.context_tokens`/`.context_mask`. The predicted segment is model
territory: the policy appends it at training time (computing the CE region mask next to the
loss) and decodes it autoregressively at inference.
This exists because LeRobot instantiates processor pipelines from the checkpoint's
`policy_preprocessor.json` (by processor-step registry name), so fine-tuning the subtask policy
directly from `lerobot/pi05_base` would silently install plain pi05's tokenizer steps.
The same artifact is the base for the `framepick` policy type (`lerobot_policy_framepick`):
framepick has no processor steps of its own — frame selection, chunk re-expression and the
assignment prefix all live in the policy, configured from the policy config.
## Usage
Requires the `lerobot_policy_pi05_subtask` plugin installed (LeRobot auto-imports
`lerobot_policy_*` distributions). In a training config:
```json
"policy": {
"type": "pi05_subtask",
"pretrained_path": "dgrachev/pi05_subtask_base",
...
}
```
During training the subtask string is read from `complementary_data["subtask"]` when the dataset
provides one, and falls back to the task string otherwise.
Regenerate with `scripts/prepare_base_checkpoint.py` from the plugin repository.
|