huggingface / huggingface/diffusers
PeRFlow: Piecewise Rectified Flow as Universal Plug-and-Play Accelerator
- 主要言語
- Python
- スター
- 34.5k
- フォーク
- 7.3k
- 平均マージ
- 3日 3時間
- マージ済み PR(30日)
- 91
説明
### Model/Pipeline/Scheduler description
A competitor of LCM-Lora, with higher generation quality and consistency regardless number of steps
Project page: https://piecewise-rectified-flow.github.io/
Github: https://github.com/magic-research/piecewise-rectified-flow/tree/main
### Open source status
- [X] The model implementation is available.
- [X] The model weights are available (Only relevant if addition is not a scheduler).
### Provide useful links for the implementation
Maybe we can add the new scheduler to Diffusers
コントリビューションガイド
調査の方向性
Start with the PeRFlow project page and linked model implementation, then compare its scheduler requirements with the existing Diffusers scheduler conventions. The issue names no Diffusers files or tests; done would require a reviewed scheduler addition with validation against the available implementation and weights.
索引モデルが issue の本文から書いたものです。
評価
- 技術スタック
- python, pytorch
- 領域
- machine-learning
- issue の種類
- 機能追加
- 難易度
- 5/5
- 見積もり時間
- 1週間以上
- 活発さ
- 停滞
- 明瞭さ
- 説明が足りない
- 初心者へのやさしさ
- 25/100