---
title: SenseNova-U1.5-8B-MoT-pose-8bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit"
md_url: "https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit.md"
repository: mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit
last_updated: 2026-09-03
license: apache-2.0
pipeline_tag: image-to-image
tasks:
  - image-to-image
model_type:
  - neo_chat
architectures:
  - NEOChatModel
base_model:
  - sensenova/SenseNova-U1.5-8B-MoT
base_model_relation: finetune
parameters: 5.9B
tensor_type:
  - BF16
  - U32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 19
stars: 0
tags:
  - mlx
  - image-to-image
  - image-editing
  - pose
  - lora-merged
  - apple-silicon
---

# SenseNova-U1.5-8B-MoT-pose-8bit

> SenseNova-U1.5-8B-MoT-pose-8bit - mlx-community 在 ModelScope 开源的模型。SenseNova-U1.5-8B-MoT-pose-8bit

mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit 是 ModelScope 魔搭社区上的 5.9B 参数image-to-image模型，采用 apache-2.0 许可，基于 sensenova/SenseNova-U1.5-8B-MoT 构建。

- **Repository**: mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit
- **License**: apache-2.0
- **Tasks**: image-to-image
- **Parameters**: 5.9B
- **Base model**: sensenova/SenseNova-U1.5-8B-MoT
- **Tags**: mlx, image-to-image, image-editing, pose, lora-merged, apple-silicon
- **Downloads**: 19
- **Stars**: 0
- **Last updated**: 2026-09-03

Source: https://www.modelscope.cn/models/mlx-community/SenseNova-U1.5-8B-MoT-pose-8bit

---

# SenseNova-U1.5-8B-MoT-pose-8bit

A **pose-transfer** tier of [SenseNova-U1.5-8B-MoT](https://huggingface.co/sensenova/SenseNova-U1.5-8B-MoT)
for [sensenova-u1-swift](https://github.com/xocialize/sensenova-u1-swift) on Apple
silicon: the RefControl pose adapter merged into the base checkpoint, then quantized to 8-bit (group 64) on the two transformer streams.

## What it does

Restages the person in a reference photo into the pose of an OpenPose-style
skeleton, keeping their identity, clothing and scene. It is a **two-reference
edit**, and the slot order matters:

```
Image-1: <image>
Image-2: <image>
apply pose from image 1 with reference from image 2
```

Image 1 is the skeleton, image 2 the identity/appearance frame. Swap them and the
model reproduces the appearance frame's own pose and ignores the skeleton.

```bash
sensenova-cli --weights SenseNova-U1.5-8B-MoT-pose-8bit \
  --edit-image skeleton.png --edit-image identity.jpg \
  --prompt $'Image-1: <image>\nImage-2: <image>\napply pose from image 1 with reference from image 2' \
  --width 768 --height 768 --steps 28 --cfg 4.0 --out posed.npy
```

From Swift, through the MLXEngine wrapper:

```swift
let package = SenseNovaU1Package(configuration: .init(variant: .pose8))
try await package.load()
let response = try await package.run(
    IEditRequest(
        images: [skeleton, identity],            // 1 = pose, 2 = appearance
        prompt: SenseNovaVariant.posePrompt,
        width: 768, height: 768, steps: 28, seed: 42))
```

## Performance (M5 Max)

| | |
|---|---|
| 768² two-reference edit, 28 steps, cfg 4 | **20.2 s** (0.72 s/step) |
| peak memory | 22.5 GB |
| resident after load | 19.0 GB |
| artifact load | 2.6 s |

## Provenance

- **Base:** `sensenova/SenseNova-U1.5-8B-MoT` @ `07d76f61` (Apache-2.0). Hub
  `main` has since moved to `19bc874e`, a bf16 re-shard of the same weights; this
  artifact was converted from the pinned revision.
- **Adapter:** RefControl pose LoRA v1 @ 3000 steps — rank 32 / alpha 32,
  `neo_hf_lora` layout, 294 gen-stream (`*_mot_gen`) projections, trained on
  (skeleton, reference, target) triples. Merged in fp32 as `W += (alpha/rank)·BA`
  and cast back; all 294 targets matched.
- **Conversion:** `sensenova-cli --weights <base> --lora <adapter> --quant 8 --convert <out>`.
- The merge is exactly as wide as the adapter claims: against the plain
  8-bit artifact, 882 tensors differ — 294 targets × {{weight, scales, biases}} —
  and nothing else. A zero-B copy of the adapter through the same path reproduces
  the plain artifact tensor for tensor.

Training data: Wikimedia Commons (CC-BY / CC-BY-SA) and Pexels footage. The source
frames are not redistributed; these are merged weights.

## Tiers

`pose-8bit` reproduces the `pose-bf16` renders on all five valid held-out fixtures,
so it is the recommended tier. `pose-bf16` exists for reference-precision work.

This is a **base-checkpoint merge, not a distillation** — text-to-image, think mode
and visual question answering all still run on these weights, though the generation
stream is pose-biased. For plain generation use the
[base tiers](https://huggingface.co/mlx-community/SenseNova-U1.5-8B-MoT-8bit).

## Licence

Apache-2.0, following the base checkpoint. Port code is MIT.