import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("SanDiegoDude/Cosmos3-Super-Text2Image-nf4", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]Cosmos3-Super-Text2Image — NF4 4-bit Pre-Quantized Transformer
Pre-quantized NF4 (4-bit, double-quantized) version of NVIDIA's
nvidia/Cosmos3-Super-Text2Image
— the 64B Cosmos 3 checkpoint specialized for high-fidelity text-to-image —
created with bitsandbytes.
Only the large Cosmos3OmniTransformer is quantized; the VAE and tokenizers are
bundled unchanged at bf16, so the repo is self-contained and drop-in.
This makes the 64B model practical on a single GPU and loads in ~1–2 minutes with no runtime quantization pass (on-the-fly NF4 of the bf16 original takes ~13 minutes every load).
Key Details
| Property | Value |
|---|---|
| Repo size | 35 GB (vs ~130 GB bf16) |
| Quantized component | transformer — NF4 (vs ~128 GB bf16) |
| Quantization | NF4 (bitsandbytes), double quantization, bnb_4bit_compute_dtype=bfloat16 |
| Mode | text-to-image |
| Base params | 64B |
| VRAM (loaded) | ~37 GB |
| Source weights | nvidia/Cosmos3-Super-Text2Image (bf16) |
| Tested on | NVIDIA GB10 (DGX Spark) |
Usage
Requires a diffusers build with Cosmos 3 support (currently from source) plus
bitsandbytes. The NF4 config is embedded — do not pass a
quantization_config, and do not call .to(dtype) on a 4-bit model.
pip install "git+https://github.com/huggingface/diffusers.git" bitsandbytes accelerate
import torch
from diffusers import Cosmos3OmniPipeline
pipe = Cosmos3OmniPipeline.from_pretrained(
"SanDiegoDude/Cosmos3-Super-Text2Image-nf4",
torch_dtype=torch.bfloat16,
enable_safety_checker=False, # skips the optional cosmos_guardrail dependency
).to("cuda")
result = pipe("A weathered lighthouse on a cliff at golden hour, photoreal, 50mm.")
frames = result.video[0] # text-to-image returns a single frame
frames[0].save("out.png")
ComfyUI
A turnkey loader + T2I node is available in scg-Cosmos3. The loader auto-detects this pre-quantized layout and skips the re-quant pass.
Related Repos
- Original model (bf16, source): nvidia/Cosmos3-Super-Text2Image
- 16B omnimodal variant (NF4): SanDiegoDude/Cosmos3-Nano-nf4
License
Released under NVIDIA's OpenMDW 1.1 License, inherited from the base model. Quantization only changes the weight encoding.
- Downloads last month
- 19
Model tree for SanDiegoDude/Cosmos3-Super-Text2Image-nf4
Base model
nvidia/Cosmos3-Super-Text2Image