verl/models/llama/megatron/layers/parallel_mlp.py

• ParallelLlamaMLP · class · L31-L74 — class ParallelLlamaMLP(nn.Module)
• init · method · L33-L69 — def init(self, config, megatron_config: ModelParallelConfig = None) -> None
• forward · method · L71-L74 — def forward(self, x)