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.
2026-08-25 14:43:17 +08:00
commit 584f7e9e73
140 changed files with 21592 additions and 0 deletions
@@ -0,0 +1,14 @@
+"""KDA operators, composable layers, and CausalLM."""
+
+from .models.causal_lm import CausalLM
+from .models.config import KDAConfig
+from .models.k3_config import K3Config
+from .ops import chunk_kda
+
+__all__ = [
+    "CausalLM",
+    "KDAConfig",
+    "K3Config",
+    "chunk_kda",
+]
+__version__ = "0.0.1"