act-pick-and-stack is an open-weight model for robotics from Hayato Nakamura, released under Apache License 2.0. It has 52M parameters. At 16-bit it needs about 0.1 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Parameters52M
Context—
Weights—
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—
Runs On
What it takes to serve act-pick-and-stack (52M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 22, 2026.
act-pick-and-stack on every accelerator the SAVRN Index prices, at every precision
Model Card
Identity and Version
- Repository
- Tron-Hayato/act-pick-and-stack
- Publisher
- Hayato Nakamura
- Task
- Robotics
- Modality
- Control
- Library
- lerobot
- Parameters
- 52M parameters
- Languages
- act
- Revision
- 12de1cb83c59821e76ce91f036dc39a767e054a2
- First published
- 2026-10-19
- Last updated
- 2026-10-19
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
Download from Hayato Nakamura
Released by Hayato Nakamura through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:2304.13705
- Trained on (disclosed) Tron-Hayato/record-pick-and-stack
Memory Requirements
| Precision | Weights in memory |
|---|---|
| 16-bit | 0.1 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About act-pick-and-stack
How much GPU memory does act-pick-and-stack need?
About 0.1 GB at 16-bit and 0 GB at 4-bit: the weights (52M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run act-pick-and-stack on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use act-pick-and-stack commercially?
Yes. act-pick-and-stack is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
Similar Models
NAJY 11단계 ACT 체크포인트 (1호기 DGX1 학습). 분석·채점용 업로드 — 실기 배포 후보 확정과는 별개. 받은 뒤: sha256sum model.safetensors 가 위와 같은지. 채점 도구는 같은 폴더의 multimanifest.json 에서 플래그를 읽는다 (--checkpoint; pretrainedmodel 구조 그대로이므로 stageconddiag.py 등에 바로 쓴다). 배경: trossen-ai-simulation docs/mobilebaseinvestigation.md §94.
Action Chunking with Transformers (ACT) is an imitation-learning method that predicts short action chunks instead of single steps. It learns from teleoperated data and often achieves high success rates. This policy has been trained and pushed to the Hub using LeRobot. Learn how to train and run it in the LeRobot act guide, or browse the full documentation. The policy consumes these observation features and produces these action features. Inputs Outputs New to LeRobot? These guides cover the full workflow: - Install LeRobot — set up the lerobot package. - Hardware setup — assemble, wire, and calibrate your robot and cameras. - Record data & train a policy — the end-to-end imitation-learning…
LeRobot 0.3.3 ACT policy for the RoboSynChallenge simulation task handlebasket (CobotMagic dual-arm robot, three 480x640 RGB cameras, 14-D joint state/action). - Initialised from RoboSynChallenge/ACTsimmixeroperating weights (normalisation statistics recomputed from the task dataset). - Fine-tuned for 14,000 steps on RoboSynChallenge/cobotmagicSimhandlebasket (1000 episodes), batch size 16, lr 1e-5, AMP; default ACT architecture (ResNet-18, dim 512, 4 enc / 1 dec layers, VAE, chunk size 50). - Local check (random setting, seeds 0 and 1, 10 episodes each): 11/20 successes, vs. 5/20 for the released RoboSynChallenge/ACTsimhandlebasket under the same protocol. Evaluate with the…
NAJY 11단계 ACT 체크포인트 (1호기 DGX1 학습). 분석·채점용 업로드 — 실기 배포 후보 확정과는 별개. 받은 뒤: sha256sum model.safetensors 가 위와 같은지. 채점 도구는 같은 폴더의 multimanifest.json 에서 플래그를 읽는다 (--checkpoint; pretrainedmodel 구조 그대로이므로 stageconddiag.py 등에 바로 쓴다). 배경: trossen-ai-simulation docs/mobilebaseinvestigation.md §94.