{
  "args": {
    "model_id": "Qwen/Qwen3-4B",
    "load_in_4bit": true,
    "torch_dtype": "bf16",
    "device_map": "auto",
    "use_lora": true,
    "lora_r": 4,
    "lora_alpha": 8,
    "lora_dropout": 0.05,
    "lora_target_modules": "all-linear",
    "gradient_checkpointing": true,
    "optimizer": "paged_adamw_8bit",
    "modulus": 97,
    "max_steps": 6,
    "beam_count": 2,
    "train_min_len": 1,
    "train_max_len": 12,
    "curriculum_stages": "smoke:1:3:2",
    "eval_lengths": "3",
    "train_template_mode": "mixed",
    "paired_train": true,
    "paired_batches": true,
    "paired_template_modes": "standard,paraphrase",
    "pair_size": 2,
    "eval_template_modes": "standard,paraphrase",
    "paired_eval": true,
    "paired_eval_template_modes": "standard,paraphrase",
    "train_size": 8,
    "eval_size": 8,
    "train_batch_size": 2,
    "eval_batch_size": 2,
    "eval_every": 2,
    "stage_eval_every": 2,
    "lr": 0.0002,
    "weight_decay": 0.0,
    "grad_clip": 1.0,
    "register_width": 128,
    "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": 1.0,
    "selector_loss_weight": 0.2,
    "diversity_loss_weight": 0.03,
    "pair_set_consistency_loss_weight": 0.05,
    "set_loss_tau": 0.5,
    "pair_set_tau": 0.5,
    "max_length": 1024,
    "seed": 11,
    "output_dir": "experiments/qwen_latent_beam_program_compiler/runs/smoke_beam2",
    "checkpoint_dir": "large_artifacts/qwen_latent_beam_program_compiler/checkpoints/smoke_beam2"
  },
  "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": true,
    "lora_r": 4,
    "gpu_name": "NVIDIA RTX 6000 Ada Generation",
    "gpu_vram_gb": 47.383
  },
  "hidden_dim": 2560,
  "beam_count": 2,
  "register_count": 13,
  "dataset": {
    "train_size": 8,
    "curriculum_stages": [
      {
        "name": "smoke",
        "min_len": 1,
        "max_len": 3,
        "steps": 2
      }
    ],
    "eval_size": 8,
    "eval_lengths": [
      3
    ],
    "train_template_mode": "mixed",
    "paired_train": true,
    "paired_template_modes": [
      "standard",
      "paraphrase"
    ],
    "eval_template_modes": [
      "standard",
      "paraphrase"
    ],
    "paired_eval": true,
    "paired_eval_template_modes": [
      "standard",
      "paraphrase"
    ]
  },
  "train_seconds": 7.8675854206085205,
  "checkpoints": [
    "large_artifacts/qwen_latent_beam_program_compiler/checkpoints/smoke_beam2/adapter",
    "large_artifacts/qwen_latent_beam_program_compiler/checkpoints/smoke_beam2/heads.pt"
  ],
  "train_log": [
    {
      "stage": "smoke",
      "local_step": 1,
      "step": 1,
      "loss": 45.92642593383789,
      "trace_set_loss": 36.98747253417969,
      "executor_loss": 4.583456516265869,
      "state_set_loss": 4.236883163452148,
      "selector_loss": 0.6781326532363892,
      "diversity_loss": 0.010346652939915657,
      "pair_set_loss": -0.3465491533279419,
      "selector_matches_best_trace": 0.5,
      "best_soft_answer_mass": 0.010224202647805214,
      "mean_soft_answer_mass": 0.010219511575996876,
      "best_target_prob_beam_fraction": 1.0,
      "standard_len3_n": 8.0,
      "standard_len3_beam_count": 2.0,
      "standard_len3_beam0_accuracy": 0.0,
      "standard_len3_selected_accuracy": 0.0,
      "standard_len3_prior_accuracy": 0.0,
      "standard_len3_oracle_accuracy": 0.0,
      "standard_len3_beam0_program_exact": 0.0,
      "standard_len3_selected_program_exact": 0.0,
      "standard_len3_prior_program_exact": 0.0,
      "standard_len3_oracle_program_exact": 0.0,
      "standard_len3_selected_state_prefix_fraction": 0.0,
      "standard_len3_oracle_state_prefix_fraction": 0.0,
      "standard_len3_avg_distinct_answers": 1.0,
      "standard_len3_selector_entropy": 0.6923007071018219,
      "standard_len3_selected_oracle_gap_recovered": NaN,
      "paraphrase_len3_n": 8.0,
      "paraphrase_len3_beam_count": 2.0,
      "paraphrase_len3_beam0_accuracy": 0.0,
      "paraphrase_len3_selected_accuracy": 0.0,
      "paraphrase_len3_prior_accuracy": 0.0,
      "paraphrase_len3_oracle_accuracy": 0.0,
      "paraphrase_len3_beam0_program_exact": 0.0,
      "paraphrase_len3_selected_program_exact": 0.0,
      "paraphrase_len3_prior_program_exact": 0.0,
      "paraphrase_len3_oracle_program_exact": 0.0,
      "paraphrase_len3_selected_state_prefix_fraction": 0.0,
      "paraphrase_len3_oracle_state_prefix_fraction": 0.0,
      "paraphrase_len3_avg_distinct_answers": 1.0,
      "paraphrase_len3_selector_entropy": 0.6925987154245377,
      "paraphrase_len3_selected_oracle_gap_recovered": NaN,
      "paired_len3_n": 16.0,
      "paired_len3_beam_count": 2.0,
      "paired_len3_beam0_accuracy": 0.0,
      "paired_len3_selected_accuracy": 0.0,
      "paired_len3_prior_accuracy": 0.0,
      "paired_len3_oracle_accuracy": 0.0,
      "paired_len3_beam0_program_exact": 0.0,
      "paired_len3_selected_program_exact": 0.0,
      "paired_len3_prior_program_exact": 0.0,
      "paired_len3_oracle_program_exact": 0.0,
      "paired_len3_selected_state_prefix_fraction": 0.041666666666666664,
      "paired_len3_oracle_state_prefix_fraction": 0.041666666666666664,
      "paired_len3_avg_distinct_answers": 1.0,
      "paired_len3_selector_entropy": 0.692365288734436,
      "paired_len3_selected_oracle_gap_recovered": NaN,
      "paired_len3_selected_pair_answer_consistency": 1.0,
      "paired_len3_selected_pair_both_correct": 0.0,
      "paired_len3_oracle_pair_both_correct": 0.0,
      "paired_len3_selected_pair_state_consistency": 1.0,
      "paired_len3_oracle_pair_state_consistency": 1.0
    },
    {
      "stage": "smoke",
      "local_step": 2,
      "step": 2,
      "loss": 45.880069732666016,
      "trace_set_loss": 36.941749572753906,
      "executor_loss": 4.583071708679199,
      "state_set_loss": 4.234369277954102,
      "selector_loss": 0.6894411444664001,
      "diversity_loss": 0.010379346087574959,
      "pair_set_loss": -0.3465644121170044,
      "selector_matches_best_trace": 0.5,
      "best_soft_answer_mass": 0.010227276012301445,
      "mean_soft_answer_mass": 0.010223442688584328,
      "best_target_prob_beam_fraction": 0.5,
      "standard_len3_n": 8.0,
      "standard_len3_beam_count": 2.0,
      "standard_len3_beam0_accuracy": 0.0,
      "standard_len3_selected_accuracy": 0.0,
      "standard_len3_prior_accuracy": 0.0,
      "standard_len3_oracle_accuracy": 0.0,
      "standard_len3_beam0_program_exact": 0.0,
      "standard_len3_selected_program_exact": 0.0,
      "standard_len3_prior_program_exact": 0.0,
      "standard_len3_oracle_program_exact": 0.0,
      "standard_len3_selected_state_prefix_fraction": 0.0,
      "standard_len3_oracle_state_prefix_fraction": 0.0,
      "standard_len3_avg_distinct_answers": 1.0,
      "standard_len3_selector_entropy": 0.6924464404582977,
      "standard_len3_selected_oracle_gap_recovered": NaN,
      "paraphrase_len3_n": 8.0,
      "paraphrase_len3_beam_count": 2.0,
      "paraphrase_len3_beam0_accuracy": 0.0,
      "paraphrase_len3_selected_accuracy": 0.0,
      "paraphrase_len3_prior_accuracy": 0.0,
      "paraphrase_len3_oracle_accuracy": 0.0,
      "paraphrase_len3_beam0_program_exact": 0.0,
      "paraphrase_len3_selected_program_exact": 0.0,
      "paraphrase_len3_prior_program_exact": 0.0,
      "paraphrase_len3_oracle_program_exact": 0.0,
      "paraphrase_len3_selected_state_prefix_fraction": 0.0,
      "paraphrase_len3_oracle_state_prefix_fraction": 0.0,
      "paraphrase_len3_avg_distinct_answers": 1.0,
      "paraphrase_len3_selector_entropy": 0.6925888359546661,
      "paraphrase_len3_selected_oracle_gap_recovered": NaN,
      "paired_len3_n": 16.0,
      "paired_len3_beam_count": 2.0,
      "paired_len3_beam0_accuracy": 0.0,
      "paired_len3_selected_accuracy": 0.0,
      "paired_len3_prior_accuracy": 0.0,
      "paired_len3_oracle_accuracy": 0.0,
      "paired_len3_beam0_program_exact": 0.0,
      "paired_len3_selected_program_exact": 0.0,
      "paired_len3_prior_program_exact": 0.0,
      "paired_len3_oracle_program_exact": 0.0,
      "paired_len3_selected_state_prefix_fraction": 0.0,
      "paired_len3_oracle_state_prefix_fraction": 0.0,
      "paired_len3_avg_distinct_answers": 1.0625,
      "paired_len3_selector_entropy": 0.6924744918942451,
      "paired_len3_selected_oracle_gap_recovered": NaN,
      "paired_len3_selected_pair_answer_consistency": 1.0,
      "paired_len3_selected_pair_both_correct": 0.0,
      "paired_len3_oracle_pair_both_correct": 0.0,
      "paired_len3_selected_pair_state_consistency": 1.0,
      "paired_len3_oracle_pair_state_consistency": 1.0
    }
  ],
  "final_metrics": {
    "standard_len3": {
      "n": 8.0,
      "beam_count": 2.0,
      "beam0_accuracy": 0.0,
      "selected_accuracy": 0.0,
      "prior_accuracy": 0.0,
      "oracle_accuracy": 0.0,
      "beam0_program_exact": 0.0,
      "selected_program_exact": 0.0,
      "prior_program_exact": 0.0,
      "oracle_program_exact": 0.0,
      "selected_state_prefix_fraction": 0.0,
      "oracle_state_prefix_fraction": 0.0,
      "avg_distinct_answers": 1.0,
      "selector_entropy": 0.6924464404582977,
      "selected_oracle_gap_recovered": NaN
    },
    "paraphrase_len3": {
      "n": 8.0,
      "beam_count": 2.0,
      "beam0_accuracy": 0.0,
      "selected_accuracy": 0.0,
      "prior_accuracy": 0.0,
      "oracle_accuracy": 0.0,
      "beam0_program_exact": 0.0,
      "selected_program_exact": 0.0,
      "prior_program_exact": 0.0,
      "oracle_program_exact": 0.0,
      "selected_state_prefix_fraction": 0.0,
      "oracle_state_prefix_fraction": 0.0,
      "avg_distinct_answers": 1.0,
      "selector_entropy": 0.6925888359546661,
      "selected_oracle_gap_recovered": NaN
    },
    "paired_len3": {
      "n": 16.0,
      "beam_count": 2.0,
      "beam0_accuracy": 0.0,
      "selected_accuracy": 0.0,
      "prior_accuracy": 0.0,
      "oracle_accuracy": 0.0,
      "beam0_program_exact": 0.0,
      "selected_program_exact": 0.0,
      "prior_program_exact": 0.0,
      "oracle_program_exact": 0.0,
      "selected_state_prefix_fraction": 0.0,
      "oracle_state_prefix_fraction": 0.0,
      "avg_distinct_answers": 1.0625,
      "selector_entropy": 0.6924744918942451,
      "selected_oracle_gap_recovered": NaN,
      "selected_pair_answer_consistency": 1.0,
      "selected_pair_both_correct": 0.0,
      "oracle_pair_both_correct": 0.0,
      "selected_pair_state_consistency": 1.0,
      "oracle_pair_state_consistency": 1.0
    }
  }
}