Instructions to use MochunniaN1/One-to-All-14b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use MochunniaN1/One-to-All-14b with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image, export_to_video # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("MochunniaN1/One-to-All-14b", dtype=torch.bfloat16, device_map="cuda") pipe.to("cuda") prompt = "A man with short gray hair plays a red electric guitar." image = load_image( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png" ) output = pipe(image=image, prompt=prompt).frames[0] export_to_video(output, "output.mp4") - Notebooks
- Google Colab
- Kaggle
VRAM requirements?
#3
by agronholm - opened
The README does not mention how much VRAM the model requires to run. Knowing if this can be run on consumer cards would be helpful!
The 1.3B and 14B models run on (81 × 576 × 1024, 30 steps, bfloat16, H100) use roughly:
• 1.3B: ~29 GB
• 14B: ~65 GB
These numbers are from tests without any memory‑saving optimizations. I don’t have consumer GPUs to verify, but with proper optimization (quantization or activation checkpointing) the 1.3B model might run on consumer cards.