Commit History
-
Keep the SGLang support reference out of the model card
Verified signature95a3835 -
Link the SGLang support pull request
Verified signaturef8c9a1e -
Add SGLang usage section to the Cosmos3-Super-Image2Video-4Step model card
Verified signatureAdds an SGLang Diffusion usage section to the model card, mirroring the structure of the `nvidia/Cosmos3-Super-Image2Video` card. Contents: - Install block and serve command. - A note that the 64B transformer is ~128 GB in BF16 and that sequence and classifier-free-guidance parallelism divide activations rather than weights, so on GPUs with less memory the weights need `--use-fsdp-inference` (sharding) or `--dit-layerwise-offload` (layer streaming). - The fixed 4-step distilled schedule: SGLang reads it from `scheduler/scheduler_config.json`, so `num_inference_steps` and `flow_shift` are ignored and `guidance_scale` is forced to `1.0`. - An example request, verified end-to-end on 4xH100. Opened as a draft until Cosmos 3 support is merged into SGLang.45cfd14
-
Refine model card overview (#21)
- Refine I2V 4-step model card overview (98a91d4daae4d23d2dfee1116db8ecff55420fa1) Co-authored-by: Qianli <qianlim@users.noreply.huggingface.co>fda1636
-
Initial release
Co-authored-by: qianlim <qianlim@users.noreply.huggingface.co> Co-authored-by: rickznv <rickznv@users.noreply.huggingface.co> Co-authored-by: fsergiy <fsergiy@users.noreply.huggingface.co> Co-authored-by: yaoxuhf <yaoxuhf@users.noreply.huggingface.co> Co-authored-by: yzhautouskay <yzhautouskay@users.noreply.huggingface.co>226cebf
