Text-to-Image
Diffusers
Safetensors
English
Chinese
Russian
QwenImage21Pipeline
image-editing
qwen-image
orbitquant
w4a4
4-bit precision
quantized
turbo
few-step
8-bit precision
Instructions to use WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4", 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
Download benchmark/vae_test.json from WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4: direct link, hf CLI and curl.
- Browser
- Download file 1.4 kB
-
https://huggingface.co/WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4/resolve/main/benchmark/vae_test.json
- Command line
-
hf download hf://WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4/benchmark/vae_test.json
-
curl -L -o vae_test.json https://huggingface.co/WaveCut/Turbo-Image-2.1-OrbitQuant-W4A4/resolve/main/benchmark/vae_test.json
1.4 kB
| { | |
| "1024_fp32_tiled": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 1.0318022351711988, | |
| "peak_gib": 14.72630786895752 | |
| }, | |
| "1024_fp16_full": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 0.22755242511630058, | |
| "peak_gib": 19.82466459274292, | |
| "psnr_vs_fp32_tiled": 38.28855840355152 | |
| }, | |
| "1024_fp16_tiled": { | |
| "nonfinite": 668, | |
| "absmax": 1.0, | |
| "s": 0.7701441468670964, | |
| "peak_gib": 13.70357084274292, | |
| "psnr_vs_fp32_tiled": 48.24812055156028 | |
| }, | |
| "1024_bf16_tiled": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 0.7385302293114364, | |
| "peak_gib": 13.70405912399292, | |
| "psnr_vs_fp32_tiled": 54.60409748195562 | |
| }, | |
| "1024_bf16_full": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 0.22161237802356482, | |
| "peak_gib": 19.84077787399292, | |
| "psnr_vs_fp32_tiled": 38.21441072672072 | |
| }, | |
| "2048_fp32_tiled": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 3.208422244992107, | |
| "peak_gib": 14.82591724395752 | |
| }, | |
| "2048_fp16_full": { | |
| "oom": true | |
| }, | |
| "2048_fp16_tiled": { | |
| "nonfinite": 3572, | |
| "absmax": 1.0, | |
| "s": 2.404657072853297, | |
| "peak_gib": 13.95285701751709, | |
| "psnr_vs_fp32_tiled": 47.489242123907964 | |
| }, | |
| "2048_bf16_tiled": { | |
| "nonfinite": 0, | |
| "absmax": 1.0, | |
| "s": 2.35006901063025, | |
| "peak_gib": 14.01742935180664, | |
| "psnr_vs_fp32_tiled": 54.04782997831014 | |
| }, | |
| "2048_bf16_full": { | |
| "oom": true | |
| } | |
| } | |