{
  "architectures": [
    "CDMMoELM"
  ],
  "model_type": "cdm_moe_poincare",
  "d_model": 768,
  "n_layers": 7,
  "n_heads": 12,
  "K": 8,
  "num_experts": 4,
  "d_ff": 3072,
  "max_seq": 512,
  "vocab_size": 248320,
  "geometry": "poincare",
  "curvature": 1.4427,
  "learn_curvature": true,
  "use_telepathy": true,
  "beta_tele": 0.5,
  "best_val_bpb": 0.9605,
  "best_step": 55500,
  "hepta_cap": true,
  "training_steps": 100000,
  "teacher": "Qwen/Qwen3.8-27B",
  "dataset_composition": "50% FineWeb-Edu + 50% SmolTalk (Hybrid Reasoning & Instruct)"
}