LTX2.3_Fr0z3nElsa_V1_000007000.safetensors

643 MB · SafeTensor · bf16

Requires membership

<frozen-elsa-ltx23>

Full Video & Voice Training! The full LTX2.3 LoRA for Elsa (Frozen) is finally here! This LoRA was trained using ai-toolkit against a large set of short sample video clips for both video & audio! So that means that you can simply refer to your Elsa character by the trigger word Fr0z3nElsa naturally for entire videos, including speaking! Start with a closeup of Fr0z3nElsa, she is wearing an a t-shirt and jeans, she is walking outside in the snowy mountains, talking to the camera she says "So ... Anna told me something interesting the other day ... I guess she turned out to be allergic to ice ... hehe ... who knew?" The training was intentionally done for the most flexibility with the characters, so sometimes they might come out looking more like ... real humans haha. That might be what you want! And that's the point! The character LoRA should allow you to specify and override just about anything. If you want to better ensure the original animation style, a simple prefix is all you need, and Just prefix or specify something like this: P1x4r 3D animated style. I just ended up having this as a static prefix for all of my prompts, they all worked out pretty well for fidelity. The full example combining these would be as simple as: P1x4r 3D animated style. Start with a closeup of Fr0z3nElsa, she is wearing an a t-shirt and jeans, she is walking outside in the snowy mountains, talking to the camera she says "So ... Anna told me something interesting the other day ... I guess she turned out to be allergic to ice ... hehe ... who knew?" Most of my testing was done at standard 1.0 strength, for best fidelity would expect 0.8 - 1.0, but play around and see what works best for you! These have been tested with and actually work pretty well, but as usual, the best character fidelity will be from using the full LTX2.3-DEV checkpoint. This is my first serious attempt at a real LoRA, please share your generations here (Add Post Below) and provide any feedback you may have <3 This LoRA is created as part of a fan project for research purposes only and is not intended for commercial use. It is based on the Movies "Frozen" which is protected by copyright. Users utilize the model at their own risk. Users are obligated to comply with copyright laws and applicable regulations. The model has been developed for non-commercial purposes, and it is not my intention to infringe on any copyright. I assume no responsibility for any damages or legal consequences arising from the use of the model

SHA256

3DC0054A60F6D96A03A7DF5636DF39ECF06BAD796334DA676EA908580FAABC1F

ModelTrends.ai Model ID

#25351

Steps

8

Guidance

1

Size

1280x704

Sampler

Euler a

Steps

8

Guidance

1

Seed

655950580950797

Size

704x1280

Steps

8

Guidance

1

Size

1280x704

Sampler

Euler a

Steps

8

Guidance

1

Seed

150349528944234

Size

704x1280

Sampler

Euler a

Steps

8

Guidance

1

Seed

428350559526476

Size

704x1280

Seed

220229463590448

Size

576x1024