Initial K3 snapshot: 0.5B KDA/MLA/MoE train path

Standalone tree split from LLMRL/projects/kda. Includes Triton dt_bias
backward fix, train_k3 --preset 0.5b, SFT, Docker runtime, and tests.

kda/layers/rmsnorm.py

@@ -0,0 +1,17 @@
+"""RMSNorm used by attention, FFN, and the final LM stem."""
+from __future__ import annotations
+
+import torch
+from torch import nn
+
+
+class RMSNorm(nn.Module):
+    def __init__(self, dim: int, eps: float = 1e-6):
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(dim))
+        self.eps = eps
+
+    def forward(self, x: torch.Tensor):
+        dtype = x.dtype
+        x = x.float()
+        return (x * torch.rsqrt(x.square().mean(-1, keepdim=True) + self.eps)).to(dtype) * self.weight