3D_GAMECG_slider_z_v1.safetensors

20 MB · SafeTensor

Mitgliedschaft erforderlich

<3d-game-cg-zit>

About ZIT version Actually, you can completely work without trigger words. However, I would recommend using a lower weight + trigger word. Simply put, LoRa has a wide range of effects, and too high a weight will have too many negative consequences. It's more natural to use a trigger word with a low weight. ZIT almost always works this way. Then this time, the trigger word is "ultra-high resolution game rendering ray-traced image" in both Chinese and English. Chinese actually works better. But one annoying problem with ZIT is that it doesn't recognize many game celebrities, but Tifa is always reliable, so she's the example. The rest mainly involves the scene. A little greasy? That's right, that's the kind of greasy smell we're looking for—real yet unreal. 其實不用觸發詞也完全可以 但是我會比較推薦用較小的權重+觸發詞 簡單來說,LORA作用的層面較廣泛,權重太高會影響到太多,不如用觸發詞加上低權重來的自然ZIT幾乎都如此 然後這次觸發詞",超高解析度遊戲渲染光追圖片"同時中英文,中文其實效果比較好 但ZIT一個討厭的問題是很多遊戲名人他不認,但是蒂法老師永遠可以信賴,所以範例是她 剩下主要是場景 有點油膩?那就對了,要的就是那種真實又不真實的油膩味

SHA256

81EE2E95E5CF6D9DD6B388F1FA706DDC952E488328E4A17C2AE60B25E5E077B4

ModelTrends.ai-Modell-ID

#23615

Sampler

Euler

Schritte

9

Guidance

1

Seed

1980842822

Größe

832x1216

Verwendetes Gewicht

2

Sampler

Euler a

Schritte

8

Guidance

2

Seed

-1

Größe

1024x1024

Clip skip

2