---
title: expimodel
canonical_url: "https://www.modelscope.cn/models/ApacheOne/expimodel"
md_url: "https://www.modelscope.cn/models/ApacheOne/expimodel.md"
repository: ApacheOne/expimodel
last_updated: 2026-05-11
parameters: 19.7B
tensor_type:
  - U8
  - F8_E4M3
  - BF16
  - F32
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 6
stars: 0
tags:
  - hidream
  - qwen3-vl
  - modelopt
  - nvfp4
  - experimental
---

# expimodel

> expimodel - ApacheOne 在 ModelScope 开源的模型。HiDream-O1-Image-Dev ModelOpt Upload

- **Repository**: ApacheOne/expimodel
- **Parameters**: 19.7B
- **Tags**: hidream, qwen3-vl, modelopt, nvfp4, experimental
- **Downloads**: 6
- **Stars**: 0
- **Last updated**: 2026-05-11

Source: https://www.modelscope.cn/models/ApacheOne/expimodel

---

# HiDream-O1-Image-Dev ModelOpt Upload

This repo was uploaded from Colab.

Included folders:

- `modelopt-v4/` — ModelOpt NVFP4 conversion output folder.
- `merged-load-v3/` — temporary merged-load HF folder used for loading/conversion.

Notes:

- This is experimental.
- Check whether `modelopt-v4/hf_quant_config.json` exists.
- If missing, the ModelOpt export did not create a clean unified HF quantized checkpoint.