Hugging Face's logo

TrNi
/
efficient-cube3d

Text-to-3D
torchao
3d-generation
quantized
int4
rtn
Model card Files
xet
 Community
TrNi commited on
Commit
098bfc4
·
verified ·
1 Parent(s): 81133a1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -23,9 +23,9 @@ Presenting the **first INT4 quantized version** of [Cube3D v0.5](https://hugging
23
 
24
  | | BF16 + Engine | BF16 + EngineFast | **INT4 + EngineFast** |
25
  |---|:-:|:-:|:-:|
26
- | ⏱️ Latency | 90.9 s | 15.0 s | **14.2 s** |
27
  | 🎮 Peak VRAM | 21.7 GB | 25.4 GB | **14.3 GB (44%↓)** ✨ |
28
  | 📦 Setup time | 19.4 s | 206.9 s | **25.1 s (88%↓)** |
 
29
 
30
  > 💡 The 44% VRAM reduction means this model now fits on a single 16 GB GPU (e.g. NVIDIA L4, A10 etc.), bringing high-quality text-to-3D generation to individual researchers and end-user hardware.
31
 
 
23
 
24
  | | BF16 + Engine | BF16 + EngineFast | **INT4 + EngineFast** |
25
  |---|:-:|:-:|:-:|
 
26
  | 🎮 Peak VRAM | 21.7 GB | 25.4 GB | **14.3 GB (44%↓)** ✨ |
27
  | 📦 Setup time | 19.4 s | 206.9 s | **25.1 s (88%↓)** |
28
+ | ⏱️ Latency | 90.9 s | 15.0 s | **14.2 s** |
29
 
30
  > 💡 The 44% VRAM reduction means this model now fits on a single 16 GB GPU (e.g. NVIDIA L4, A10 etc.), bringing high-quality text-to-3D generation to individual researchers and end-user hardware.
31