nvidia
/

Cosmos3-Super-Image2Video-4Step

@@ -522,7 +522,7 @@ sglang serve \
   --num-gpus 4
 ```
-Support for this checkpoint is added in [sgl-project/sglang#31590](https://github.com/sgl-project/sglang/pull/31590) and currently requires the SGLang main branch. Switch to a stable SGLang release once it is included there.
 The 64B transformer is roughly 128 GB in BF16. Sequence and classifier-free-guidance parallelism divide activations rather than weights, so every GPU still holds a full copy of the model. On devices with less memory, such as H100-class GPUs, shard the weights across GPUs with `--use-fsdp-inference`, or stream them layer by layer with `--dit-layerwise-offload`:

   --num-gpus 4
 ```
+Cosmos 3 support in SGLang Diffusion currently requires the SGLang main branch. Switch to a stable SGLang release once Cosmos 3 support is included there.
 The 64B transformer is roughly 128 GB in BF16. Sequence and classifier-free-guidance parallelism divide activations rather than weights, so every GPU still holds a full copy of the model. On devices with less memory, such as H100-class GPUs, shard the weights across GPUs with `--use-fsdp-inference`, or stream them layer by layer with `--dit-layerwise-offload`: