!!!!!!very important!!!!!!
The LoRA adaptation for the Minimax H3 model has been successfully developed and validated. Evaluation of the image-to-video (i2v) outputs demonstrates consistently high visual fidelity, coherent motion, and strong adherence to the intended conditioning. Text-to-video (t2v) results are serviceable, though they exhibit comparatively lower consistency and detail retention under identical prompting conditions.
In addition, existing LoRA modules may be employed as effective substitutes. These can be further refined through targeted fine-tuning to incorporate explicit anatomical features, specifically the inclusion of female and male genitalia, thereby expanding the model’s capacity for precise adult content generation.
as Helper loras:
First, you need both the high-noise and low-noise versions. If you only use the low-noise version of one, for example, nothing will work. You also need the erect_penis_80-lora to ensure it looks correct. otherwise, you’ll just end up with something worm-like, which won’t be useful for her.
weight:
High noise: 0.8-1.0
Low Noise: 0.8-1.0
Erect_Penis_80-Lora (wan2.1 old version.): 1.0 Works!!
Description
FAQ
Comments (1)
Hi. trying to train an i2v ltx 2.3 lora but getting awful result and I think this come maybe from my WF, may I ask you ur i2v ltx2.3 WF ?
Details
Files
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
Mirrors
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
Futanari transformation (Futanarization) - LTX-2.3 - FutaTF.safetensors
Futanari transformation (Futanarization) - LTX-2.3 - FutaTF.safetensors
Futanari transformation (Futanarization) - LTX-2.3 - FutaTF.safetensors
LTX-2_3_Futanari_TF_lora_v3_1.safetensors
Futanari transformation (Futanarization) - LTX-2.3 - FutaTF.safetensors
Futanari transformation (Futanarization) - LTX-2.3 - FutaTF.safetensors