Online generation<a target="_blank" rel="ugc" href="https://tensor.art/models/1023242308545421378">https://tensor.art/models/952032833109521879/2127-Z-Image-Asian-Utopian-Turbo</a><hr /><h2 id="v4.6-turbo">v4.6 - Turbo</h2>額外加入 600 張不同亞洲臉孔，增加表情的變化，並降低部分臉孔過於光滑的材質表現問題。An additional 600 different Asian faces have been added to increase the variety of expressions and reduce the material performance problem of some faces being too smooth.さらに 600 種類のアジア人の顔が追加され、表現の多様性が増し、一部の顔が滑らかすぎるというマテリアル パフォーマンスの問題が軽減されました。표현의 다양성을 높이고 일부 얼굴이 너무 매끄러워지는 재질 성능 문제를 줄이기 위해 600개의 다양한 아시아인 얼굴이 추가되었습니다.<hr /><h2 id="v4.0-turbo">v4.0 - Turbo</h2>加入 2000 張自行拍攝景色照片，200 張細緻臉孔照片，調整了體態、光澤表現。Added 2000 self-taken landscape photos and 200 detailed facial photos, adjusting body shape and gloss.自撮り風景写真2000枚と詳細な顔写真200枚を追加し、体型と光沢を調整しました。직접 촬영한 풍경 사진 2000장과 얼굴 디테일 사진 200장을 추가하고, 체형과 광택을 조정했습니다.<hr /><h2 id="v3.6-baseturbo-ffv">v3.6 - Base/Turbo - FFV</h2>使用 v2.0-FF 加入全新的訓練集，調整了光澤、臉孔、體態。We used v2.0-FF to add a brand new training set and adjusted the gloss, face, and body shape.バージョン2.0-FFを使用して、全く新しいトレーニングセットを追加し、光沢、顔、体型を調整しました。저희는 v2.0-FF를 사용하여 완전히 새로운 학습 데이터 세트를 추가하고 광택, 얼굴 및 몸매를 조정했습니다.<hr /><h2 id="v3.0-turbo-ffdpo">v3.0 - Turbo - FFDPO</h2>使用 v2.0 fp32 版本，並重新融合 Lokr 與 <a target="_blank" rel="ugc" href="https://huggingface.co/F16/z-image-turbo-flow-dpo">FDPO Lora</a>。轉換為 bf16 與 Q8_0 gguf。Use version v2.0 fp32 and merge Lokr and <a target="_blank" rel="ugc" href="https://huggingface.co/F16/z-image-turbo-flow-dpo">FDPO Lora</a>. Convert to bf16 and Q8_0 gguf.バージョンv2.0 fp32を使用し、Lokrと <a target="_blank" rel="ugc" href="https://huggingface.co/F16/z-image-turbo-flow-dpo">FDPO Lora </a>を再マージします。bf16とQ8_0 ggufに変換します。버전 v2.0 fp32를 사용하고 Lokr 및 <a target="_blank" rel="ugc" href="https://huggingface.co/F16/z-image-turbo-flow-dpo">FDPO Lora </a>를 다시 병합합니다. bf16 및 Q8_0 gguf로 변환합니다.<hr /><h2 id="v2.0-ff-v2.0-base">v2.0 - FF / v2.0 - Base</h2>FF = 全片幅，全精度使用 2.4MP 高解析度資料集訓練 fp32, dim 64 Lora，並重新調整 v1.5 - ZSVD 融合配重，重新製作為 fp32 全精度版本，轉換為 bf16, GGUF。 FF = Full Frame, Full PrecisionThe fp32, dim 64 Lora dataset was trained using a 2.4MP high-resolution dataset, and the v1.5 - ZSVD fusion weights were readjusted to create a new fp32 full-precision version, which was then converted to bf16 and GGUF.FF = フルフレーム、フル精度fp32、dim 64のLoraデータセットは、2.4MPの高解像度データセットを用いて学習され、v1.5 - ZSVD融合重みが再調整されて新しいfp32フル精度バージョンが作成され、その後bf16とGGUFに変換されました。FF = 풀 프레임, 풀 정확도fp32, dim 64 의 Lora 데이터 세트는 2.4MP의 고해상도 데이터 세트를 사용하여 학습되었으며 v1.5 - ZSVD 융합 가중치가 재조정되어 새로운 fp32 풀 정확도 버전이 생성된 다음 bf16 및 GGUF 로 변환되었습니다.<hr /><h2 id="v1.5-zsvd">v1.5 - ZSVD</h2>額外使用 1000 張高畫質訓練 dim 128 Lora 並 SVD 融合 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora v3.78</a> 與原版 Z Image Turbo bf16。An additional 1000 high-quality images were used to train dim 128 Lora and SVD fused with <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora v3.78</a> and the original Z Image Turbo bf16.追加の 1000 枚の高品質画像を使用して、 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora v3.78</a> とオリジナルの Z Image Turbo bf16 を融合した dim 128 Lora と SVD をトレーニングしました。추가로 1000개의 고품질 이미지를 사용하여 dim 128 Lora 와 SVD를 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora v3.78</a> 및 기존 Z Image Turbo bf16과 융합하여 학습시켰습니다.<hr /><h2 id="v1.0-turbo">v1.0 - Turbo</h2>這是使用了 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora</a> v3.78 與原版 Z Image Turbo bf16，另外加入了兩組 10k 資料集（我擁有所有使用權，其實，都是我自行拍攝的風景照），使用 <code>gradient_accumulation: 4</code> 額外訓練的 Lora 來調配權重，並根據 AI-Toolkit 的訓練結果中，將 DiT 根據 30 個不同的 Blocks 各別調整融合成的模型。 如果可以的話，請幫我返圖。This model was created using <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora</a> v3.78 and the original Z Image Turbo bf16, with two additional 10k datasets (I own all rights to these datasets; they are actually landscape photos I took myself). It uses an additional Lora dataset trained with <code>gradient_accumulation: 4</code> to adjust the weights, and the DiT algorithm is adjusted and fused based on the training results from the AI-Toolkit, tailored to 30 different blocks.Feel free to upload the images you generated.このモデルは、 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora</a> v3.68 とオリジナルの Z Image Turbo bf16 を使用し、さらに2つの10kデータセット（すべて私が著作権を所有しており、実際に私が撮影した風景写真です）を追加しました。重み調整のため <code>gradient_accumulation: 4</code> で学習した Lora モデルを追加し、DiT アルゴリズムは AI-Toolkit の学習結果に基づいて調整・融合され、30 個の異なるブロックに合わせて調整されています。可能であれば、生成した画像をアップロードしてください。이 모델은 <a target="_blank" rel="ugc" href="https://civarchive.com/models/2185778/z-image-turbo-asian-mix-lora">[Z Image Turbo] Asian Mix Lora</a> v3.78 과 원래 Z Image Turbo bf16 을 사용했으며, 두 개의 10k 데이터 세트 (모두 내가 저작권을 소유하고 실제로 내가 찍은 풍경 사진입니다) 을 추가했습니다. 가중치 조정을 위해, <code>gradient_accumulation: 4</code> 에서 학습한 Lora 모델을 추가하고, DiT 알고리즘은 AI-Toolkit 의 학습 결과에 근거해 조정·융합되어, 30 개의 다른 블록에 맞추어 조정되고 있습니다.가능하면 생성된 이미지를 업로드하세요.

2127 - Z Image Asian Utopian - Turbo/Base BF16

2127ZImageAsianUtopian_v20FF_full_bf16.safetensors

2127ZImageAsianUtopian_v20FF_full_fp8.safetensors

2127ZImageAsianUtopian_v20FF.gguf

v2.0 - FF

v4.6 - Turbo

v4.0 - Turbo

v3.6 - Turbo - FFV

v3.6 - Base - FFV

v3.0 - Turbo - FFDPO

v2.0 - Base

v1.5 - ZSVD

v1.0 - Turbo

Files with SHA256: 16a84568a7de9cbbf490841e0b602f94881ee89f9d88b20322db3173bf1da4fb