{
  "mode": "hook",
  "metadata": {
    "created_unix": 1781910589.0051832,
    "created_utc": "2026-06-19T23:09:49Z",
    "python": "3.12.3 (main, Aug 14 2025, 17:47:21) [GCC 13.3.0]",
    "platform": "Linux-6.8.0-62-generic-x86_64-with-glibc2.39",
    "torch_version": "2.8.0+cu128",
    "transformers_version": "5.12.1",
    "bitsandbytes_version": "0.49.2",
    "cuda": {
      "available": true,
      "torch_cuda": "12.8",
      "device_count": 1,
      "devices": [
        {
          "index": 0,
          "name": "NVIDIA RTX 6000 Ada Generation",
          "total_memory_gb": 47.38275146484375,
          "major": 8,
          "minor": 9
        }
      ]
    },
    "loader_name": "AutoModelForMultimodalLM",
    "script": "/workspace/latent_qwen_fastweight_experiment.py",
    "script_sha256": "0367b3a1d51564c9f44ac1f7ff45d9dbe3f652336e2a78c2f722bcf2243153c5",
    "args": {
      "model_id": "Qwen/Qwen3.5-4B",
      "text_model_fallback": "Qwen/Qwen3-4B",
      "model_class": "auto",
      "mode": "hook",
      "load_in_4bit": true,
      "torch_dtype": "bf16",
      "device_map": "auto",
      "seed": 7,
      "verbose": false,
      "modulus": 97,
      "num_choices": 5,
      "min_steps": 3,
      "max_steps": 8,
      "max_length": 384,
      "hook_layer": -4,
      "runtime_dim": 256,
      "workspace_tokens": 8,
      "runtime_heads": 4,
      "num_bases": 12,
      "adapter_rank": 16,
      "mem_dim": 128,
      "dropout": 0.05,
      "init_scale": 0.05,
      "disable_fast_memory": false,
      "disable_dynamic_lowrank": false,
      "train_steps": 300,
      "batch_size": 4,
      "grad_accum": 1,
      "lr": 0.0002,
      "weight_decay": 0.01,
      "max_grad_norm": 1.0,
      "train_k": "0",
      "eval_k": "0,1,2,4,8",
      "aux_value_loss": 0.0,
      "gate_entropy_bonus": 0.0,
      "log_every": 10,
      "eval_every": 100,
      "eval_batches": 25,
      "eval_batch_size": 4,
      "candidate_batch_size": 20,
      "skip_initial_eval": true,
      "output_dir": "runs/control_qwen35_hook_traink0_seed7"
    }
  },
  "architecture": {
    "layer_path": "model.language_model.layers",
    "num_layers": 32,
    "hook_layer_index": 28,
    "hidden_dim": 2560,
    "runtime_dim": 256,
    "workspace_tokens": 8,
    "runtime_heads": 4,
    "num_bases": 12,
    "adapter_rank": 16,
    "mem_dim": 128,
    "disable_fast_memory": false,
    "disable_dynamic_lowrank": false,
    "aux_value_loss": 0.0
  },
  "train": [
    {
      "step": 1,
      "loss": 4.118315696716309,
      "delta_rms": 1.2338939995970577e-05,
      "scale": 0.05000000074505806,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 10,
      "loss": 3.410245418548584,
      "delta_rms": 0.013242789544165134,
      "scale": 0.05008026957511902,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 20,
      "loss": 1.6667759418487549,
      "delta_rms": 0.08235674351453781,
      "scale": 0.0501885861158371,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 30,
      "loss": 1.4925601482391357,
      "delta_rms": 0.18737785518169403,
      "scale": 0.05028362572193146,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 40,
      "loss": 1.5611211061477661,
      "delta_rms": 0.24239741265773773,
      "scale": 0.05031329393386841,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 50,
      "loss": 1.482926607131958,
      "delta_rms": 0.19795474410057068,
      "scale": 0.05031409487128258,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 60,
      "loss": 1.572636365890503,
      "delta_rms": 0.1750953048467636,
      "scale": 0.05031139776110649,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 70,
      "loss": 1.5793439149856567,
      "delta_rms": 0.17916129529476166,
      "scale": 0.0503123439848423,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 80,
      "loss": 1.5625226497650146,
      "delta_rms": 0.2056991159915924,
      "scale": 0.05031696334481239,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 90,
      "loss": 1.6950101852416992,
      "delta_rms": 0.22630104422569275,
      "scale": 0.050322722643613815,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 100,
      "loss": 1.5529403686523438,
      "delta_rms": 0.19247639179229736,
      "scale": 0.050315894186496735,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 110,
      "loss": 1.418046474456787,
      "delta_rms": 0.21077851951122284,
      "scale": 0.05031900480389595,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 120,
      "loss": 1.559584379196167,
      "delta_rms": 0.22480829060077667,
      "scale": 0.050322599709033966,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 130,
      "loss": 1.671646237373352,
      "delta_rms": 0.1827739030122757,
      "scale": 0.0503157414495945,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 140,
      "loss": 1.4890546798706055,
      "delta_rms": 0.15373314917087555,
      "scale": 0.05030447617173195,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 150,
      "loss": 1.3672163486480713,
      "delta_rms": 0.1527094542980194,
      "scale": 0.05029599741101265,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 160,
      "loss": 1.528685450553894,
      "delta_rms": 0.15731757879257202,
      "scale": 0.050298526883125305,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 170,
      "loss": 1.6485496759414673,
      "delta_rms": 0.18916106224060059,
      "scale": 0.05031439661979675,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 180,
      "loss": 1.4408936500549316,
      "delta_rms": 0.1848999410867691,
      "scale": 0.05031449347734451,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 190,
      "loss": 1.6117140054702759,
      "delta_rms": 0.1651257574558258,
      "scale": 0.05030502751469612,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 200,
      "loss": 1.6543309688568115,
      "delta_rms": 0.11990927159786224,
      "scale": 0.05028168112039566,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 210,
      "loss": 1.5425611734390259,
      "delta_rms": 0.14221517741680145,
      "scale": 0.05028907582163811,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 220,
      "loss": 1.5269031524658203,
      "delta_rms": 0.16764028370380402,
      "scale": 0.05030532553792,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 230,
      "loss": 1.7374258041381836,
      "delta_rms": 0.214109405875206,
      "scale": 0.05031983181834221,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 240,
      "loss": 1.6125596761703491,
      "delta_rms": 0.21669496595859528,
      "scale": 0.050324782729148865,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 250,
      "loss": 1.3991734981536865,
      "delta_rms": 0.1887584626674652,
      "scale": 0.05032076686620712,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 260,
      "loss": 1.205233097076416,
      "delta_rms": 0.21762928366661072,
      "scale": 0.05033363029360771,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 270,
      "loss": 1.7257670164108276,
      "delta_rms": 0.20904573798179626,
      "scale": 0.05032789334654808,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 280,
      "loss": 1.587977409362793,
      "delta_rms": 0.16587971150875092,
      "scale": 0.05030671879649162,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 290,
      "loss": 1.6300983428955078,
      "delta_rms": 0.16120421886444092,
      "scale": 0.050307437777519226,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    },
    {
      "step": 300,
      "loss": 1.666135549545288,
      "delta_rms": 0.14778535068035126,
      "scale": 0.050301693379879,
      "gate_entropy": 0.0,
      "mem_rms": 0.0
    }
  ],
  "eval": [
    {
      "step": 100,
      "val": {
        "0": 0.25,
        "1": 0.25,
        "2": 0.24,
        "4": 0.22,
        "8": 0.22
      },
      "hard": {
        "0": 0.2,
        "1": 0.2,
        "2": 0.18,
        "4": 0.2,
        "8": 0.23
      }
    },
    {
      "step": 200,
      "val": {
        "0": 0.19,
        "1": 0.19,
        "2": 0.17,
        "4": 0.16,
        "8": 0.21
      },
      "hard": {
        "0": 0.26,
        "1": 0.25,
        "2": 0.28,
        "4": 0.3,
        "8": 0.28
      }
    },
    {
      "step": 300,
      "val": {
        "0": 0.22,
        "1": 0.22,
        "2": 0.2,
        "4": 0.17,
        "8": 0.2
      },
      "hard": {
        "0": 0.19,
        "1": 0.17,
        "2": 0.19,
        "4": 0.16,
        "8": 0.2
      }
    }
  ]
}