forked from huggingface/diffusers
-
Notifications
You must be signed in to change notification settings - Fork 10
Home
Seongsu Park edited this page Jul 22, 2023
·
36 revisions
- (ํน์ ์ฉ์ด๋ฅผ ๊ฐ๋ฆฌํค๋) ๋๋ฌธ์๋ก ์์ํ๋ ๊ธ์๋ค
- Hugging Face
- Diffusers
- text-to-image
- sliced attention
- memory-efficient attention
- tracing
- forward pass
- just-in-time
- Channels Last (๋ฉ๋ชจ๋ฆฌ ํ์)
- diffusion
- unconditional, conditional
- toolbox
- shape
- optimizer
- textual-inversion
- LoRA
- PyTorch
- ToMe, Token Merging
- Python
- snapshot
- push: ํธ์
- step: ์คํ
- convolution: ์ปจ๋ณผ๋ฃจ์
- section: ์น์
- fine tuning: ํ์ธํ๋
- licence: ๋ผ์ด์ ์ค
- load: ๋ถ๋ฌ์ค๊ธฐ
- inference: ์ถ๋ก
- inference time: ์ถ๋ก ์๊ฐ
- custom: ์ปค์คํ
- code snippet: ์์ ์ฝ๋...
- submodule: ์๋ธ๋ชจ๋
- branch: ๋ธ๋์น
- directory: ๋๋ ํ ๋ฆฌ
- checkpoint: ์ฒดํฌํฌ์ธํธ
- resume: ์ฌ๊ฐ
- gradient: ๊ทธ๋๋์ธํธ
- model: ๋ชจ๋ธ
- system: ์์คํ
- lession: ๋จ์
- reproducibility: ์ฌํ์ฑ
- prompt: ํ๋กฌํํธ
- pre-train: ์ฌ์ ํ์ต
- instance: ์ธ์คํด์ค
- inpainting: ์ธํ์ธํ
- denoise: ๋๋ ธ์ด์ฆ
- repository: ๋ฆฌํฌ์งํ ๋ฆฌ
- ์ฐจ์์์(dimensions ordering)
- ๋ค์ค ์ด๋ฏธ์ง ๋ฐฐ์น(multi-image batches)
- ํผํฉ ์ ๋ฐ๋(mixed precision) // ํผํฉ ์ ๋ฐ๋๋ผ๋ ํํ์ด ์๋ ์ค ๋ชฐ๋๋ค์(ํ๊ท)
- ๋ฐ์ ๋ฐ๋(half precision)
- attention: ์ดํ ์
- prior-preserving(์ฌ์ ๋ณด์กด)
- batch(๋ฐฐ์น)
- ์ข ์์ฑ(dependency)