Instructions to use ProCreations/Image-2.1-Calibrated-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use ProCreations/Image-2.1-Calibrated-NVFP4 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("ProCreations/Image-2.1-Calibrated-NVFP4", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Release calibrated Image2.1 NVFP4 transformer with dynamic scaling and BF16 rank correction, native SM120 runtime, quality evidence and real-time demo
1961af5 verified Download reports/quantization_summary.json from ProCreations/Image-2.1-Calibrated-NVFP4: direct link, hf CLI and curl.
- Browser
- Download file 715 Bytes
-
https://huggingface.co/ProCreations/Image-2.1-Calibrated-NVFP4/resolve/main/reports/quantization_summary.json
- Command line
-
hf download hf://ProCreations/Image-2.1-Calibrated-NVFP4/reports/quantization_summary.json
-
curl -L -o quantization_summary.json https://huggingface.co/ProCreations/Image-2.1-Calibrated-NVFP4/resolve/main/reports/quantization_summary.json
715 Bytes
| { | |
| "modules": 224, | |
| "transformer_bytes": 4870924032, | |
| "mean_diagnostic_nrmse": 0.057958053558000495, | |
| "max_diagnostic_nrmse": 0.1279429793357849, | |
| "naive_mean_diagnostic_nrmse": 0.07584479178019267, | |
| "protocol": "64 original BF16 trajectories;384 fit and384 disjoint diagnostic rows per layer, stratified across prompts and timesteps. Five smoothing exponents, two weight-scale methods. Rank128 randomized SVD (seed1234,q160,niter3); optional activation-weighted15-point block-scale MSE search. Native W4A4 kernels with dynamic activation amax used for all calibrated selection/diagnostic output errors. Plain reference uses original static range. No held-out image evaluation prompts used for calibration." | |
| } |