{
  "args": {
    "model_id": "Qwen/Qwen3-4B",
    "load_in_4bit": true,
    "torch_dtype": "bf16",
    "device_map": "auto",
    "use_lora": false,
    "lora_r": 8,
    "lora_alpha": 16,
    "lora_dropout": 0.05,
    "lora_target_modules": "all-linear",
    "gradient_checkpointing": true,
    "optimizer": "paged_adamw_8bit",
    "modulus": 97,
    "max_steps": 4,
    "train_min_len": 1,
    "train_max_len": 12,
    "curriculum_stages": "smoke:1:2:2",
    "answer_train_min_len": 1,
    "answer_train_max_len": 24,
    "eval_lengths": "2,4",
    "train_template_mode": "mixed",
    "paired_train": true,
    "paired_batches": true,
    "paired_template_modes": "standard,paraphrase",
    "pair_size": 2,
    "answer_train_template_mode": "mixed",
    "eval_template_modes": "standard,paraphrase",
    "paired_eval": true,
    "paired_eval_template_modes": "standard,paraphrase",
    "train_size": 4,
    "answer_train_size": 4,
    "eval_size": 4,
    "train_batch_size": 2,
    "eval_batch_size": 2,
    "bootstrap_steps": 0,
    "answer_steps": 0,
    "eval_every": 1,
    "stage_eval_every": 1,
    "lr": 0.0002,
    "answer_lr_scale": 0.1,
    "weight_decay": 0.0,
    "grad_clip": 1.0,
    "head_width": 64,
    "register_width": 64,
    "register_layers": 1,
    "register_heads": 4,
    "register_dropout": 0.05,
    "trace_loss_weight": 1.0,
    "init_trace_loss_weight": 4.0,
    "op_trace_loss_weight": 1.0,
    "arg_trace_loss_weight": 4.0,
    "executor_loss_weight": 1.0,
    "state_loss_weight": 0.25,
    "variants": "register_trace",
    "max_length": 256,
    "seed": 7,
    "train_seed": -1,
    "output_dir": "experiments/qwen_register_token_latent_compiler/runs/smoke_frozen_register_trace",
    "checkpoint_dir": "large_artifacts/qwen_register_token_latent_compiler/checkpoints/smoke_frozen_register_trace"
  },
  "metadata": {
    "python": "3.12.3",
    "platform": "Linux-6.8.0-62-generic-x86_64-with-glibc2.39",
    "torch": "2.8.0+cu128",
    "cuda_available": true,
    "transformers_loader": "AutoModelForCausalLM",
    "peft_installed": true,
    "use_lora": false,
    "lora_r": 8,
    "gpu_name": "NVIDIA RTX 6000 Ada Generation",
    "gpu_vram_gb": 47.383
  },
  "hidden_dim": 2560,
  "register_count": 9,
  "dataset": {
    "train_size": 4,
    "train_examples": 8,
    "curriculum_stages": [
      {
        "name": "smoke",
        "min_len": 1,
        "max_len": 2,
        "steps": 2
      }
    ],
    "answer_train_size": 4,
    "eval_size": 4,
    "eval_lengths": [
      2,
      4
    ],
    "train_template_mode": "mixed",
    "paired_train": true,
    "paired_template_modes": [
      "standard",
      "paraphrase"
    ],
    "answer_train_template_mode": "mixed",
    "eval_template_modes": [
      "standard",
      "paraphrase"
    ],
    "paired_eval": true,
    "paired_eval_template_modes": [
      "standard",
      "paraphrase"
    ]
  },
  "variants": {
    "register_trace": {
      "variant": "register_trace",
      "train_seconds": 4.414611101150513,
      "checkpoints": [
        "large_artifacts/qwen_register_token_latent_compiler/checkpoints/smoke_frozen_register_trace/register_trace/frozen_backbone.txt",
        "large_artifacts/qwen_register_token_latent_compiler/checkpoints/smoke_frozen_register_trace/register_trace/heads.pt"
      ],
      "train_log": [
        {
          "variant": "register_trace",
          "stage": "trace_smoke",
          "local_step": 1,
          "step": 1,
          "loss": 41.794063568115234,
          "trace_loss_active": true,
          "executor_loss_active": true,
          "state_loss_active": false,
          "lr_scale": 1.0,
          "executor_loss": 4.579510688781738,
          "init_loss": 4.385224342346191,
          "op_loss": 1.1968920230865479,
          "arg_loss": 4.6191911697387695,
          "standard_len2_direct_accuracy": NaN,
          "standard_len2_direct_target_mass": NaN,
          "standard_len2_executor_accuracy": 0.0,
          "standard_len2_executor_target_mass": 0.010254076682031155,
          "standard_len2_init_accuracy": 0.0,
          "standard_len2_op_accuracy": 0.5,
          "standard_len2_arg_accuracy": 0.0,
          "standard_len2_program_exact": 0.0,
          "standard_len2_state_accuracy": 0.0,
          "standard_len2_state_all_exact": 0.0,
          "standard_len2_state_prefix_fraction": 0.0,
          "standard_len4_direct_accuracy": NaN,
          "standard_len4_direct_target_mass": NaN,
          "standard_len4_executor_accuracy": 0.0,
          "standard_len4_executor_target_mass": 0.010244335513561964,
          "standard_len4_init_accuracy": 0.0,
          "standard_len4_op_accuracy": 0.375,
          "standard_len4_arg_accuracy": 0.0,
          "standard_len4_program_exact": 0.0,
          "standard_len4_state_accuracy": 0.0,
          "standard_len4_state_all_exact": 0.0,
          "standard_len4_state_prefix_fraction": 0.0,
          "paraphrase_len2_direct_accuracy": NaN,
          "paraphrase_len2_direct_target_mass": NaN,
          "paraphrase_len2_executor_accuracy": 0.0,
          "paraphrase_len2_executor_target_mass": 0.010254761669784784,
          "paraphrase_len2_init_accuracy": 0.0,
          "paraphrase_len2_op_accuracy": 0.375,
          "paraphrase_len2_arg_accuracy": 0.0,
          "paraphrase_len2_program_exact": 0.0,
          "paraphrase_len2_state_accuracy": 0.0,
          "paraphrase_len2_state_all_exact": 0.0,
          "paraphrase_len2_state_prefix_fraction": 0.0,
          "paraphrase_len4_direct_accuracy": NaN,
          "paraphrase_len4_direct_target_mass": NaN,
          "paraphrase_len4_executor_accuracy": 0.0,
          "paraphrase_len4_executor_target_mass": 0.010254837572574615,
          "paraphrase_len4_init_accuracy": 0.0,
          "paraphrase_len4_op_accuracy": 0.4375,
          "paraphrase_len4_arg_accuracy": 0.0,
          "paraphrase_len4_program_exact": 0.0,
          "paraphrase_len4_state_accuracy": 0.0,
          "paraphrase_len4_state_all_exact": 0.0,
          "paraphrase_len4_state_prefix_fraction": 0.0,
          "paired_len2_direct_accuracy": NaN,
          "paired_len2_direct_target_mass": NaN,
          "paired_len2_executor_accuracy": 0.0,
          "paired_len2_executor_target_mass": 0.010259798727929592,
          "paired_len2_init_accuracy": 0.25,
          "paired_len2_op_accuracy": 0.5,
          "paired_len2_arg_accuracy": 0.0,
          "paired_len2_program_exact": 0.0,
          "paired_len2_state_accuracy": 0.0,
          "paired_len2_state_all_exact": 0.0,
          "paired_len2_state_prefix_fraction": 0.0,
          "paired_len2_pair_true_answer_consistency": 1.0,
          "paired_len2_executor_pair_answer_consistency": 0.0,
          "paired_len2_executor_pair_both_correct": 0.0,
          "paired_len2_compiler_pair_program_consistency": 0.0,
          "paired_len2_compiler_pair_state_consistency": 0.0,
          "paired_len4_direct_accuracy": NaN,
          "paired_len4_direct_target_mass": NaN,
          "paired_len4_executor_accuracy": 0.0,
          "paired_len4_executor_target_mass": 0.010252463864162564,
          "paired_len4_init_accuracy": 0.0,
          "paired_len4_op_accuracy": 0.46875,
          "paired_len4_arg_accuracy": 0.0,
          "paired_len4_program_exact": 0.0,
          "paired_len4_state_accuracy": 0.0,
          "paired_len4_state_all_exact": 0.0,
          "paired_len4_state_prefix_fraction": 0.0,
          "paired_len4_pair_true_answer_consistency": 1.0,
          "paired_len4_executor_pair_answer_consistency": 0.5,
          "paired_len4_executor_pair_both_correct": 0.0,
          "paired_len4_compiler_pair_program_consistency": 0.5,
          "paired_len4_compiler_pair_state_consistency": 0.5
        },
        {
          "variant": "register_trace",
          "stage": "trace_smoke",
          "local_step": 2,
          "step": 2,
          "loss": 42.714210510253906,
          "trace_loss_active": true,
          "executor_loss_active": true,
          "state_loss_active": false,
          "lr_scale": 1.0,
          "executor_loss": 4.575713157653809,
          "init_loss": 4.655272483825684,
          "op_loss": 1.178600788116455,
          "arg_loss": 4.5847015380859375,
          "standard_len2_direct_accuracy": NaN,
          "standard_len2_direct_target_mass": NaN,
          "standard_len2_executor_accuracy": 0.0,
          "standard_len2_executor_target_mass": 0.010258573107421398,
          "standard_len2_init_accuracy": 0.0,
          "standard_len2_op_accuracy": 0.25,
          "standard_len2_arg_accuracy": 0.0,
          "standard_len2_program_exact": 0.0,
          "standard_len2_state_accuracy": 0.0,
          "standard_len2_state_all_exact": 0.0,
          "standard_len2_state_prefix_fraction": 0.0,
          "standard_len4_direct_accuracy": NaN,
          "standard_len4_direct_target_mass": NaN,
          "standard_len4_executor_accuracy": 0.0,
          "standard_len4_executor_target_mass": 0.010251559782773256,
          "standard_len4_init_accuracy": 0.0,
          "standard_len4_op_accuracy": 0.1875,
          "standard_len4_arg_accuracy": 0.0,
          "standard_len4_program_exact": 0.0,
          "standard_len4_state_accuracy": 0.0,
          "standard_len4_state_all_exact": 0.0,
          "standard_len4_state_prefix_fraction": 0.0,
          "paraphrase_len2_direct_accuracy": NaN,
          "paraphrase_len2_direct_target_mass": NaN,
          "paraphrase_len2_executor_accuracy": 0.0,
          "paraphrase_len2_executor_target_mass": 0.01025943923741579,
          "paraphrase_len2_init_accuracy": 0.0,
          "paraphrase_len2_op_accuracy": 0.375,
          "paraphrase_len2_arg_accuracy": 0.0,
          "paraphrase_len2_program_exact": 0.0,
          "paraphrase_len2_state_accuracy": 0.0,
          "paraphrase_len2_state_all_exact": 0.0,
          "paraphrase_len2_state_prefix_fraction": 0.0,
          "paraphrase_len4_direct_accuracy": NaN,
          "paraphrase_len4_direct_target_mass": NaN,
          "paraphrase_len4_executor_accuracy": 0.0,
          "paraphrase_len4_executor_target_mass": 0.010260708630084991,
          "paraphrase_len4_init_accuracy": 0.0,
          "paraphrase_len4_op_accuracy": 0.1875,
          "paraphrase_len4_arg_accuracy": 0.0,
          "paraphrase_len4_program_exact": 0.0,
          "paraphrase_len4_state_accuracy": 0.0,
          "paraphrase_len4_state_all_exact": 0.0,
          "paraphrase_len4_state_prefix_fraction": 0.0,
          "paired_len2_direct_accuracy": NaN,
          "paired_len2_direct_target_mass": NaN,
          "paired_len2_executor_accuracy": 0.0,
          "paired_len2_executor_target_mass": 0.01026474661193788,
          "paired_len2_init_accuracy": 0.25,
          "paired_len2_op_accuracy": 0.25,
          "paired_len2_arg_accuracy": 0.0,
          "paired_len2_program_exact": 0.0,
          "paired_len2_state_accuracy": 0.0,
          "paired_len2_state_all_exact": 0.0,
          "paired_len2_state_prefix_fraction": 0.0,
          "paired_len2_pair_true_answer_consistency": 1.0,
          "paired_len2_executor_pair_answer_consistency": 1.0,
          "paired_len2_executor_pair_both_correct": 0.0,
          "paired_len2_compiler_pair_program_consistency": 1.0,
          "paired_len2_compiler_pair_state_consistency": 1.0,
          "paired_len4_direct_accuracy": NaN,
          "paired_len4_direct_target_mass": NaN,
          "paired_len4_executor_accuracy": 0.0,
          "paired_len4_executor_target_mass": 0.010259605944156647,
          "paired_len4_init_accuracy": 0.0,
          "paired_len4_op_accuracy": 0.25,
          "paired_len4_arg_accuracy": 0.03125,
          "paired_len4_program_exact": 0.0,
          "paired_len4_state_accuracy": 0.0,
          "paired_len4_state_all_exact": 0.0,
          "paired_len4_state_prefix_fraction": 0.0,
          "paired_len4_pair_true_answer_consistency": 1.0,
          "paired_len4_executor_pair_answer_consistency": 0.5,
          "paired_len4_executor_pair_both_correct": 0.0,
          "paired_len4_compiler_pair_program_consistency": 0.5,
          "paired_len4_compiler_pair_state_consistency": 0.5
        }
      ],
      "final_metrics": {
        "standard_len2": {
          "variant": "register_trace",
          "n": 4.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.010258573107421398,
          "init_accuracy": 0.0,
          "op_accuracy": 0.25,
          "arg_accuracy": 0.0,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0
        },
        "standard_len4": {
          "variant": "register_trace",
          "n": 4.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.010251559782773256,
          "init_accuracy": 0.0,
          "op_accuracy": 0.1875,
          "arg_accuracy": 0.0,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0
        },
        "paraphrase_len2": {
          "variant": "register_trace",
          "n": 4.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.01025943923741579,
          "init_accuracy": 0.0,
          "op_accuracy": 0.375,
          "arg_accuracy": 0.0,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0
        },
        "paraphrase_len4": {
          "variant": "register_trace",
          "n": 4.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.010260708630084991,
          "init_accuracy": 0.0,
          "op_accuracy": 0.1875,
          "arg_accuracy": 0.0,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0
        },
        "paired_len2": {
          "variant": "register_trace",
          "n": 8.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.01026474661193788,
          "init_accuracy": 0.25,
          "op_accuracy": 0.25,
          "arg_accuracy": 0.0,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0,
          "pair_true_answer_consistency": 1.0,
          "executor_pair_answer_consistency": 1.0,
          "executor_pair_both_correct": 0.0,
          "compiler_pair_program_consistency": 1.0,
          "compiler_pair_state_consistency": 1.0
        },
        "paired_len4": {
          "variant": "register_trace",
          "n": 8.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.010259605944156647,
          "init_accuracy": 0.0,
          "op_accuracy": 0.25,
          "arg_accuracy": 0.03125,
          "program_exact": 0.0,
          "state_accuracy": 0.0,
          "state_all_exact": 0.0,
          "state_prefix_fraction": 0.0,
          "pair_true_answer_consistency": 1.0,
          "executor_pair_answer_consistency": 0.5,
          "executor_pair_both_correct": 0.0,
          "compiler_pair_program_consistency": 0.5,
          "compiler_pair_state_consistency": 0.5
        }
      }
    }
  }
}