anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

[Doc] validate Cosmos3-Super online FP8 (#4584)

Browse Source 
Signed-off-by: Kwokho Tsui <kh_tsui@163.com>
Co-authored-by: Kwokho Tsui <kh_tsui@163.com>
This commit is contained in:
Kwokho TsuiandKwokho Tsui authored and GitHub committed 2026-06-22 02:24:04 +00:00
1 parent 36bd8029f3
commit 6bdd85b384
3 files changed
+6 -3
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

recipes/cosmos3/Cosmos3-Super.md
+4 -1
View File
@@ -57,7 +57,10 @@ vllm serve nvidia/Cosmos3-Super \
Guardrails are on by default (gated `nvidia/Cosmos-1.0-Guardrail` — `pip install
cosmos-guardrail`, accept the license, set `HF_TOKEN`); add `--no-guardrails` to
disable. `--enable-layerwise-offload` reduces VRAM on smaller GPUs.
disable. `--enable-layerwise-offload` reduces VRAM on smaller GPUs;
`--quantization fp8` (online, no calibration) cuts peak VRAM for 720p video
generation from ~83 GB to ~55 GB per GPU (2-GPU) with BF16-level quality (T2V
composition can shift at the same seed).
#### Verification