Skip to content

Commit

Permalink
Merge pull request #105 from llm-jp/f/stable-diffusions
Browse files Browse the repository at this point in the history 
add Stable Diffusion by TohokuNLP and Stability AI
  • Loading branch information
@kaisugi
kaisugi authored Dec 3, 2023
2 parents 5ee4c1d + 5362ef8 commit 87a53af
Show file tree
Hide file tree
Showing 3 changed files with 9 additions and 3 deletions.
4 changes: 3 additions & 1 deletion README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -220,7 +220,9 @@
|:---|:---:|:---:|:---:|:---:|:---:|
| [日本語CLIP](https://rinna.co.jp/news/2022/05/20220512.html) | CLIP <br>(画像エンコーダは google/vit-base-patch16-224 で重みが初期化された ViT-B/16、<br>テキストエンコーダは rinna RoBERTa で重みが初期化された RoBERTa(base)) | CC12M のキャプションを日本語に翻訳したもの | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-clip-vit-b-16) |
| [日本語CLOOB](https://rinna.co.jp/news/2022/05/20220512.html) | CLOOB <br>(画像エンコーダは google/vit-base-patch16-224 で重みが初期化された ViT-B/16、<br>テキストエンコーダは rinna RoBERTa で重みが初期化された RoBERTa(base)) | CC12M のキャプションを日本語に翻訳したもの | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-cloob-vit-b-16) |
| [日本語 Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion (最初にテキストエンコーダのみ日本語キャプション付き画像を用いて追加学習を行い、次にテキストエンコーダと生成モデルのパラメータを同時に更新する追加学習を行う) | LAION-5B データセットのうちキャプションが日本語のもの（画像約 1 億枚）| rinna | [CreativeML OpenRAIL-M License](https://huggingface.co/spaces/CompVis/stable-diffusion-license) | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |
| [Japanese Stable Diffusion XL](https://ja.stability.ai/blog/japanese-stable-diffusion-xl) | Stable Diffusion | 不明 | Stability AI | STABILITY AI JAPANESE STABLE DIFFUSION XL COMMUNITY LICENSE | [◯](https://huggingface.co/stabilityai/japanese-stable-diffusion-xl) |
| [東北大Stable Diffusion](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0) | Stable Diffusion | WMT2023 Shared Task の日英対訳コーパス、laion2B-multi のキャプション約 1,300 万件 | 東北大<br>自然言語処理研究グループ | CreativeML OpenRAIL-M License | ◯ ([base](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0), [refiner](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-refiner-1.0)) |
| [rinna Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion | LAION-5B データセットのうちキャプションが日本語のもの（画像約 1 億枚）| rinna | CreativeML OpenRAIL-M License | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |

<a id="benchmark-suites"></a>
## 日本語LLM評価ベンチマーク/データセットまとめ
Expand Down
4 changes: 3 additions & 1 deletion README_en.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -220,7 +220,9 @@ Please point out any errors on the [issues page](https://github.com/llm-jp/aweso
|:---|:---:|:---:|:---:|:---:|:---:|
| [JapaneseCLIP](https://rinna.co.jp/news/2022/05/20220512.html) | CLIP <br>(Image encoding with google/vit-base-patch16-224 initialized ViT-B/16 model, <br>text encoding with rinna RoBERTa initialized RoBERTa(base) model) | CC12M translated to Japanese | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-clip-vit-b-16) |
| [JapaneseCLOOB](https://rinna.co.jp/news/2022/05/20220512.html) | CLOOB <br>(Image encoding with google/vit-base-patch16-224 initialized ViT-B/16 model, <br>text encoding with rinna RoBERTa initialized RoBERTa(base) model) | CC12M translated to Japanese | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-cloob-vit-b-16) |
| [Japanese Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion (First training the text encoder on Japanese image-caption pairs, then training both the encoder and the diffusion model) | LAION-5B Japanese Subset (100M images) | rinna | [CreativeML OpenRAIL-M License](https://huggingface.co/spaces/CompVis/stable-diffusion-license) | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |
| [Japanese Stable Diffusion XL](https://ja.stability.ai/blog/japanese-stable-diffusion-xl) | Stable Diffusion | Unknown | Stability AI | STABILITY AI JAPANESE STABLE DIFFUSION XL COMMUNITY LICENSE | [◯](https://huggingface.co/stabilityai/japanese-stable-diffusion-xl) |
| [TohokuUniversity Stable Diffusion](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0) | Stable Diffusion | WMT2023 Shared Task English-Japanese parallel corpus, about 13 million captions from laion2B-multi | Tohoku University NLP Group | CreativeML OpenRAIL-M License | ◯ ([base](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0), [refiner](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-refiner-1.0)) |
| [rinna Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion | LAION-5B Japanese Subset (100M images) | rinna | CreativeML OpenRAIL-M License | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |

<a id="benchmark-suites"></a>
## Evaluation Benchmarks for Japanese LLMs
Expand Down
4 changes: 3 additions & 1 deletion README_fr.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -220,7 +220,9 @@ N'hésitez pas à signaler les erreurs sur la page [issues](https://github.com/l
|:---|:---:|:---:|:---:|:---:|:---:|
| [JapaneseCLIP](https://rinna.co.jp/news/2022/05/20220512.html) | CLIP <br>(Encodage d'image avec google/vit-base-patch16-224 initialisé par modèle ViT-B/16, <br>Encodage textuelle avec rinna RoBERTa initialisé RoBERTa(base) model) | CC12M traduit en japonais | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-clip-vit-b-16) |
| [JapaneseCLOOB](https://rinna.co.jp/news/2022/05/20220512.html) | CLOOB <br>(Image encoding with google/vit-base-patch16-224 initialized ViT-B/16 model, <br>Encodage textuelle avec rinna RoBERTa initialisé RoBERTa(base) model) | CC12M traduit en japonais | rinna | Apache 2.0 | [◯](https://huggingface.co/rinna/japanese-cloob-vit-b-16) |
| [Japanese Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion (Premier apprentissage de l'encodeur textuelle en japonais avec des pairs image-caption, puis apprentissage double de l'encodeur et du modèle de diffision) | LAION-5B Japanese Subset (100M images) | rinna | [CreativeML OpenRAIL-M License](https://huggingface.co/spaces/CompVis/stable-diffusion-license) | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |
| [Japanese Stable Diffusion XL](https://ja.stability.ai/blog/japanese-stable-diffusion-xl) | Stable Diffusion | Inconnu | Stability AI | STABILITY AI JAPANESE STABLE DIFFUSION XL COMMUNITY LICENSE | [◯](https://huggingface.co/stabilityai/japanese-stable-diffusion-xl) |
| [TohokuUniversity Stable Diffusion](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0) | Stable Diffusion | Corpus parallèle anglais-japonais de la tâche partagée WMT2023, environ 13 millions de légendes de laion2B-multi | Université de Tohoku - Groupe TAL | CreativeML OpenRAIL-M License | ◯ ([base](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-base-1.0), [refiner](https://huggingface.co/cl-tohoku/stable-diffusion-xl-jp-refiner-1.0)) |
| [rinna Stable Diffusion](https://rinna.co.jp/news/2022/09/20220909.html) | Stable Diffusion | LAION-5B Japanese Subset (100M images) | rinna | CreativeML OpenRAIL-M License | [◯](https://huggingface.co/rinna/japanese-stable-diffusion) |

<a id="benchmark-suites"></a>
## Standard d'évaluation pour les LLM en japonais
Expand Down

0 comments on commit 87a53af

Please sign in to comment.