MaLoW-Qwen3.5-2B-MLP is an open-weight model from Zhuoming Liu, released under Apache License 2.0. It has 175M parameters. At 16-bit it needs about 0.4 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Inference checkpoint for Memory as Weights: Internalizing Long-Term History for Streaming Videos. Evaluation requires the MaLoW code release. The code has been prepared locally; public GitHub publication is pending. Intended repository: MaLoW (forthcoming).
Runs On
What it takes to serve MaLoW-Qwen3.5-2B-MLP (175M 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 | 0.3 GB | 0.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.2 GB | 0.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.1 GB | 0.1 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.
MaLoW-Qwen3.5-2B-MLP on every accelerator the SAVRN Index prices, at every precision
Model Card
By Zhuoming Liu, published under apache-2.0, revision b7b83d13a1d8.
Inference checkpoint for Memory as Weights: Internalizing Long-Term History for Streaming Videos. Evaluation requires the MaLoW code release. The code has been prepared locally; public GitHub publication is pending. Intended repository: MaLoW (forthcoming). Download the base model separately: Qwen/Qwen3.5-2B. This repository contains inference modules, not a standalone base model. Follow docs/vlaevaluation.md in the MaLoW code release for Simpler WidowX. This checkpoint additionally requires the Qwen3.5-2B-MLP baseline adapter as the frozen policy base. Both bundles are required. Reported reference success rates (not a fresh evaluation of this export): datasetstatistics.json supplies action…
Read Zhuoming Liu's full model card
MaLoW (Q3.5-2B-MLP)
Inference checkpoint for Memory as Weights: Internalizing Long-Term History for Streaming Videos.
Evaluation requires the MaLoW code release. The code has been prepared locally; public GitHub publication is pending. Intended repository: MaLoW (forthcoming).
Download the base model separately: Qwen/Qwen3.5-2B. This repository contains inference modules, not a standalone base model.
Follow docs/vla_evaluation.md in the MaLoW code release for Simpler WidowX.
This checkpoint additionally requires the Qwen3.5-2B-MLP baseline adapter as the frozen policy base. Both bundles are required.
Reported reference success rates (not a fresh evaluation of this export):
| Stack green block | Carrot on plate | Spoon on towel | Eggplant in basket | Overall |
|---|---|---|---|---|
| 0.2916 | 0.75 | 0.875 | 0.8333 | 0.687475 |
dataset_statistics.json supplies action normalization; use the oxe_bridge key for WidowX. The action-head tensor dictionary uses PyTorch's restricted weights_only=True loader.
Weights and inference metadata are distributed under Apache-2.0. See LICENSE and NOTICE; base models and benchmark assets retain their respective upstream terms.
Identity and Version
- Repository
- zhuomingliu/MaLoW-Qwen3.5-2B-MLP
- Publisher
- Zhuoming Liu
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 175M parameters
- Languages
- Not stated by the source
- Revision
- b7b83d13a1d886b62a7c945cda8c60af1c6517b6
- First published
- 2026-10-02
- Last updated
- 2026-10-02
Files and Weights
19 files, 726.3 MB in total. The weights are 3 files totalling 706.2 MB in pt, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| c2l_metadata.safetensors | Weights | 272 B | c30a94be5155 |
| model.safetensors | Weights | 699.8 MB | 3927b2d91636 |
| vla_trainable_params.pt | Weights | 6.3 MB | 9674d4f9ea21 |
| c2l_metadata.json | Configuration | 69.2 KB | — |
| dataset_statistics.json | Configuration | 2.9 KB | — |
| processor_config.json | Configuration | 1.2 KB | — |
| release_manifest.json | Configuration | 2.9 KB | — |
| vla_config.json | Configuration | 1.4 KB | — |
| LICENSE | Documentation | 11.3 KB | — |
| NOTICE | Documentation | 1.3 KB | — |
| README.md | Documentation | 1.4 KB | — |
| LICENSES/SimplerEnv-MIT.txt | Other | 1.1 KB | — |
| LICENSES/StreamingBench-MIT.txt | Other | 1.1 KB | — |
| LICENSES/doc-to-lora-MIT.txt | Other | 1.1 KB | — |
| SHA256SUMS | Other | 1.5 KB | — |
| chat_template.jinja | Other | 7.8 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 20.0 MB | 06b9509352d2 |
| tokenizer_config.json | Tokenizer | 1.2 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 706.2 MB
Released by Zhuoming Liu through its official repository on Hugging Face. Read the license.
Built From
- Derived from Qwen/Qwen3.5-2B
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 706.2 MB |
| 16-bit | 0.3 GB |
| 8-bit | 0.2 GB |
| 4-bit | 0.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About MaLoW-Qwen3.5-2B-MLP
How much GPU memory does MaLoW-Qwen3.5-2B-MLP need?
About 0.4 GB at 16-bit and 0.1 GB at 4-bit: the weights (175M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run MaLoW-Qwen3.5-2B-MLP 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 MaLoW-Qwen3.5-2B-MLP commercially?
Yes. MaLoW-Qwen3.5-2B-MLP 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.