[Perf]: CFG parallel abstraction (#851)

Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>

vllm_omni/diffusion/data.py

@@ -55,6 +55,7 @@ class DiffusionParallelConfig:
         assert self.ulysses_degree > 0, "Ulysses degree must be > 0"
         assert self.ring_degree > 0, "Ring degree must be > 0"
         assert self.cfg_parallel_size > 0, "CFG parallel size must be > 0"
+        assert self.cfg_parallel_size in [1, 2], f"CFG parallel size must be 1 or 2, but got {self.cfg_parallel_size}"
         assert self.sequence_parallel_size == self.ulysses_degree * self.ring_degree, (
             "Sequence parallel size must be equal to the product of ulysses degree and ring degree,"
             f" but got {self.sequence_parallel_size} != {self.ulysses_degree} * {self.ring_degree}"