Improve model card: Add Radial Attention paper, project, code links and update metadata
Browse filesThis PR improves the model card for `vrgamedevgirl/Wan14BT2VFusioniX` by:
- Linking to the associated research paper: [Radial Attention: $O(n\log n)$ Sparse Attention with Energy Decay for Long Video Generation](https://huggingface.co/papers/2506.19852).
- Adding links to the Radial Attention project page (https://hanlab.mit.edu/projects/radial-attention) and its GitHub repository (https://github.com/mit-han-lab/radial-attention).
- Updating metadata: explicitly setting `pipeline_tag: text-to-video`, adding `library_name: diffusers`, and correcting `license` to `cc-by-nc-sa-4.0` to reflect the non-commercial usage restrictions mentioned in the model card.
README.md
CHANGED
|
@@ -1,46 +1,64 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
tags:
|
| 3 |
- text-to-video
|
| 4 |
- diffusion
|
| 5 |
- merged-model
|
| 6 |
- video-generation
|
| 7 |
- wan2.1
|
| 8 |
-
|
| 9 |
widget:
|
| 10 |
-
- text:
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
output:
|
| 13 |
url: videos/Video_00063.mp4
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
|
|
|
|
|
|
|
|
|
| 17 |
output:
|
| 18 |
url: videos/AnimateDiff_00001.mp4
|
| 19 |
-
|
| 20 |
-
-
|
| 21 |
-
|
|
|
|
|
|
|
| 22 |
output:
|
| 23 |
url: videos/FusionX_00012.mp4
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
|
|
|
|
|
|
|
|
|
| 27 |
output:
|
| 28 |
url: videos/FusionX_00005.mp4
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
|
|
|
|
|
|
|
|
|
| 32 |
output:
|
| 33 |
url: videos/Video_00069.mp4
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
base_model:
|
| 38 |
-
- Wan-AI/Wan2.1-T2V-14B
|
| 39 |
-
license: apache-2.0
|
| 40 |
---
|
| 41 |
|
| 42 |
# 🌀 Wan2.1_14B_FusionX
|
| 43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
**High-Performance Merged Text-to-Video Model**
|
| 45 |
Built on WAN 2.1 and fused with research-grade components for cinematic motion, detail, and speed — optimized for ComfyUI and rapid iteration in as few as 6 steps.
|
| 46 |
|
|
@@ -256,4 +274,4 @@ For commercial use or monetization, please consult a legal advisor and verify al
|
|
| 256 |
|
| 257 |
And thanks to the open-source community!
|
| 258 |
|
| 259 |
-
---
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model:
|
| 3 |
+
- Wan-AI/Wan2.1-T2V-14B
|
| 4 |
+
license: cc-by-nc-sa-4.0
|
| 5 |
+
pipeline_tag: text-to-video
|
| 6 |
+
library_name: diffusers
|
| 7 |
tags:
|
| 8 |
- text-to-video
|
| 9 |
- diffusion
|
| 10 |
- merged-model
|
| 11 |
- video-generation
|
| 12 |
- wan2.1
|
|
|
|
| 13 |
widget:
|
| 14 |
+
- text: 'Prompt: A gritty close-up of an elven princess kneeling in a rocky ravine,
|
| 15 |
+
calming a wounded, desert dragon. Its scales are cracked, dry, She wears a crimson
|
| 16 |
+
sash over bone-colored armor, her auburn hair half-tied back. The camera dollies
|
| 17 |
+
in rapidly as she reaches for its eye ridge. Lighting comes from golden sunlight
|
| 18 |
+
reflecting off surrounding rock, casting a warm, earthy hue with no artificial
|
| 19 |
+
glow.'
|
| 20 |
output:
|
| 21 |
url: videos/Video_00063.mp4
|
| 22 |
+
- text: 'Prompt: Tight close-up of her smiling lips and sparkling eyes, catching golden
|
| 23 |
+
hour sunlight. She wears a white sundress with floral prints and a wide-brimmed
|
| 24 |
+
straw hat. Camera pulls back in a dolly motion, revealing her twirling under a
|
| 25 |
+
cherry blossom tree. Petals flutter in the air, casting playful shadows. Soft
|
| 26 |
+
lens flares enhance the euphoric, dreamlike vibe. (Before vs After — Left: Wan2.1
|
| 27 |
+
| Right: Merged model Wan14BT2V_MasterModel)'
|
| 28 |
output:
|
| 29 |
url: videos/AnimateDiff_00001.mp4
|
| 30 |
+
- text: 'Prompt: A gritty close-up of a dwarven beastmaster’s face, his grey beard
|
| 31 |
+
braided tightly, brows furrowed as he looks just off-camera. The camera dollies
|
| 32 |
+
out over his shoulder, revealing a perched gryphon watching him from a boulder,
|
| 33 |
+
its feathers rustling slightly in the breeze. The moment holds stillness and mutual
|
| 34 |
+
trust. Lighting is early daylight, clean and sharp with strong environmental clarity.'
|
| 35 |
output:
|
| 36 |
url: videos/FusionX_00012.mp4
|
| 37 |
+
- text: 'Prompt: A gritty close-up of a jungle tracker crouching low, face flushed
|
| 38 |
+
with focus as she watches a perched macaw a few feet ahead. Her cheek twitches
|
| 39 |
+
as she shifts forward, beads of sweat visible on her brow. The camera slowly dollies
|
| 40 |
+
in from below her line of sight, capturing the moment her eyes widen in fascination.
|
| 41 |
+
Lighting is rich and directional from above, creating a warm glow over her face
|
| 42 |
+
with minimal shadows.'
|
| 43 |
output:
|
| 44 |
url: videos/FusionX_00005.mp4
|
| 45 |
+
- text: 'Prompt: A gritty close-up of a battle-worn ranger kneeling in a scorched
|
| 46 |
+
clearing, calming a wounded gryphon whose wing is torn and bloodied. Its feathers
|
| 47 |
+
are dusky bronze with streaks of ash-gray. She wears soot-covered hunter green
|
| 48 |
+
armor, her blonde hair pulled into a loose braid. The camera dollies in as her
|
| 49 |
+
hand brushes the creature''s sharp beak. Lighting comes from late afternoon sun
|
| 50 |
+
filtering through smoke, casting a burnt-orange haze across the frame.'
|
| 51 |
output:
|
| 52 |
url: videos/Video_00069.mp4
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
---
|
| 54 |
|
| 55 |
# 🌀 Wan2.1_14B_FusionX
|
| 56 |
|
| 57 |
+
This model, Wan2.1_14B_FusionX, incorporates advancements from the research on [Radial Attention: $O(n\log n)$ Sparse Attention with Energy Decay for Long Video Generation](https://huggingface.co/papers/2506.19852).
|
| 58 |
+
|
| 59 |
+
Project Page: https://hanlab.mit.edu/projects/radial-attention
|
| 60 |
+
Code: https://github.com/mit-han-lab/radial-attention
|
| 61 |
+
|
| 62 |
**High-Performance Merged Text-to-Video Model**
|
| 63 |
Built on WAN 2.1 and fused with research-grade components for cinematic motion, detail, and speed — optimized for ComfyUI and rapid iteration in as few as 6 steps.
|
| 64 |
|
|
|
|
| 274 |
|
| 275 |
And thanks to the open-source community!
|
| 276 |
|
| 277 |
+
---
|