Instructions to use kandinskylab/Kandinsky-6.0-VSR-5s-Diffusers with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use kandinskylab/Kandinsky-6.0-VSR-5s-Diffusers with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("kandinskylab/Kandinsky-6.0-VSR-5s-Diffusers", 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
Commit History
-
Restore the flow-matching transformer: 64-wide output head and scheduler_scale 5.0
Verified signatureReplaces transformer/ uploaded on 2026-10-03 (commit f7888d7), which held the distilled checkpoint's weights and config (out_visual_dim 640, scheduler_scale 3.5). The weights are the 2026-10-01 checkpoint re-keyed to the current module names; the output layer is unchanged (64 rows). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>6e487aa
-
Re-key transformer state_dict for the FeedForward/TimestepEmbedding/attn/dec_block/private_/ModuleList/OutputHead module renames
Verified signaturef7888d7
-
Upload folder using huggingface_hub
Verified signature0530c6b