I wonder if this pose can be generated with just the prompt using the Pony checkpoint, and I just don't know about it...?

This LoRA is designed to render seated figures in what is called the "agura" pose in Japanese. In this version, it supports views from the "front," "side," and "above."

(While this pose is said to be achievable using the prompt "indian style, sitting," at least within the scope of my attempts, I was unable to generate it with the prompt alone.)

It performs best when the character is barefoot. When wearing boots or similar footwear, there is a tendency for the ankles to appear fused, so you may need to try multiple times.

Please note that even if you specify a direction, when used in combination with character LoRAs, it tends to be influenced by the direction those are more likely to generate (e.g., specifying "from above" but the image is drawn "from the front").

Trigger words:

agura, sitting, from front, 
agura, sitting, from side, 
agura, sitting, from above, 

Optional:

If specifying the direction (from XXX) alone does not yield the desired results, replacing "agura" with the dedicated prompts below may improve the outcome. The term "agura" has been trained comprehensively, while the following prompts are trained using data specific to those directions.

aguraFromFront, sitting, from front, 
aguraFromSide, sitting, from side, 
aguraFromAbove, sitting, from above, 

Negative prompts:

yokozuwari, wariza, seiza, fused ankles,  fused legs, fused footwear,

It may not always be necessary, but sometimes these elements are drawn, so they are included just as a precaution.

Strength:

0.9 - 1.0

私が知らないだけでPonyでもプロンプトで出せるんですかね…？

このLoRAはあぐらをかくポーズを描画します。このバージョンでは「前から」「横から」「上から」に対応しています。

靴を履いていない状態で最も高いパフォーマンスを発揮します。ブーツなどは足首が結合されたものが出やすいため、何度か挑戦する必要があるかもしれません。

なお方向指定してもキャラクターLoRAなどと併用する場合、そちらの描画されやすい方向に引っ張られる傾向があるようです。（from aboveを指定してもfrom frontで描画されるなど）

Trigger words:

agura, sitting, from front, 
agura, sitting, from side, 
agura, sitting, from above, 

Optional:

方向指定だけでは出にくい場合、aguraを外して以下の専用のプロンプトに置き換えると改善する可能性があります。aguraは包括的に学習されたワードで、以下のものはその方向のデータのみを学習しています。

aguraFromFront, sitting, from front, 
aguraFromSide, sitting, from side, 
aguraFromAbove, sitting, from above, 

Negative prompts:

yokozuwari, wariza, seiza, fused ankles,  fused legs, fused footwear,

無くてもいいですが、たまにこの辺が描画されたりしますのでお守り的な意味で。

Strength:

0.9 - 1.0