Autoodtwarzanie
Uruchom lokalnie
LTXV_2-DeforuMorph_v1_000004500.safetensors
588 MB · SafeTensor
Pobieranie obsługuje Civitai. Niektórzy twórcy wymagają najpierw zalogowania się na Civitai.
AD
Zainstaluj / Uruchom na Graydient
Wymaga członkostwa
<deforum-evolution-ltx2>
O tym modelu
LTX2 · LoRA
by s4f3ty_marc - LTX-2_Deforum_Evolution_v1
Notatki twórcy
LTX-2_Deforum_Evolution_v1
Model description
LTX-2 Deforum Evolution LoRA This LoRA was fine-tuned on a curated, high-quality dataset of my personal Deforum animations generated with Flux. It includes 45 × 10-second clips and 45 × 5-second clips, all at 768×768 resolution and 24 fps, created using the "Journey-Through-Flux" style pipeline. Each training video incorporated its own original audio track. While the audio conditioning shows some interesting emergent behaviors, it often converges toward similar soundscapes rather than faithfully following unique inputs. (I'm planning a follow-up version trained without audio to focus purely on visual evolution.) The LoRA was optimised for img2vid workflows in AI Toolkit . It doesn't adhere extremely rigidly to the input image—tending instead toward creative interpretation—but it does influence motion and style direction noticeably when used with a strong starting frame.
Recommended trigger & usage:
Use the trigger phrase: [deforumorph] Best results come from prompts emphasising trippy, evolving, morphing visuals, psychedelic transformations, fluid organic shifts, surreal mutations, or endless morphing sequences.
Example prompt starters:
"[deforumorph] A digital figure composed of glowing white wireframe lines pulses rhythmically across its form, transforming subtly within the dark void. The hooded silhouette flickers with neon outlines that shift rapidly, creating fluid geometric patterns along the contours of the head and shoulders. As the sequence progresses, the eyes glow brighter, their pupils dilating and contracting in sync with internal rhythms, revealing faint inner lights beneath the surface. The camera slowly rotates clockwise, pulling back to reveal more of the surrounding space—a grid-like structure extending outward in concentric waves behind the figure’s back. Ambient hums echo through the environment; soft electronic tones blend with distant mechanical whirring, suggesting advanced technology operating under dim lighting. The air vibrates gently, producing subtle echoes of shifting data streams. The entire scene breathes forward with calm tension, each pulse resonant against silence."
Weight recommendation:
Start at 1.0 - 2.0 and adjust based on how strongly you want the Deforum-style evolution to dominate. I found a setting of 1.5 worked well.
This LoRA excels at generating hypnotic, ever-changing animations reminiscent of classic Deforum journeys but infused with Flux-level detail and coherence. Great for abstract art, music visuals, experimental video loops, or trippy background content.
SHA256
6386DB0C5DBEB2C307923D35BE328FE20AF470262F85FA997A437BBEDE605BC9
ID modelu ModelTrends.ai
#21145
Zgłoś treść
Podobne modele
Modele AI do generowania obrazów, wideo i tekstu
deforum evolution ltx2 to LoRA z rodziny Ltx2 wymieniony na ModelTrends.ai, katalogu tylko do odczytu otwartych modeli AI do obrazów, wideo i tekstu. Porównaj go z innymi modelami Ltx2, sprawdź wynik popularności, aby zobaczyć trend, i otwórz na PirateDiffusion lub BitVector, aby wypróbować.
Przeglądaj według rodziny modeli
Modele Text / LLM
411
Modele Anima
509
Modele Chroma
14
Modele Flux
1,071
Modele Flux 2 / Klein
1,237
Modele MiniMax H3
331
Modele Hunyuan
197
Modele Ideogram
20
Modele Krea2
674
Modele Qwen
17
Modele Qwen2
36
Modele Ltx2
150
Modele Stable Diffusion 1.5
6,465
Modele Stable Diffusion XL 1.0
15,432
Modele Zimage
804
Modele Wan
919

Model
Trends
.ai
© ModelTrends.ai
|
Wyprodukowano w Japonii
|
© 2026
