Update README.md

README.md

@@ -40,7 +40,7 @@ This model is jointly finetuned with [DMD](https://arxiv.org/pdf/2405.14867) and
 ### Training Infrastructure
 
 Training was conducted on **8 nodes with 64 H200 GPUs** in total, using a `global batch size = 64`.  
-We enable `gradient checkpointing`, set `HSDP_shard_dim = 8`, and use `learning rate = 1e-5`.  
+We enable `gradient checkpointing`, set `HSDP_shard_dim = 8`, 'sequence_parallel_size = 4', and use `learning rate = 1e-5`.  
 We set **VSA attention sparsity** to 0.9, and training runs for **3000 steps (~52 hours)**   
 The detailed training example script is available [here](https://github.com/hao-ai-lab/FastVideo/blob/main/examples/distill/Wan-Syn-480P/distill_dmd_VSA_t2v_14B_480P.slurm).