{
  "args": {
    "model_id": "Qwen/Qwen3.5-4B",
    "text_model_fallback": "Qwen/Qwen3-4B",
    "model_class": "auto",
    "load_in_4bit": true,
    "torch_dtype": "bf16",
    "device_map": "auto",
    "modulus": 97,
    "max_steps": 24,
    "train_min_len": 1,
    "train_max_len": 12,
    "answer_train_min_len": 1,
    "answer_train_max_len": 12,
    "eval_lengths": "4,8,12,24",
    "train_template_mode": "mixed",
    "answer_train_template_mode": "mixed",
    "eval_template_modes": "standard,paraphrase",
    "train_size": 1024,
    "answer_train_size": 256,
    "eval_size": 128,
    "feature_batch_size": 8,
    "train_batch_size": 128,
    "eval_batch_size": 128,
    "bootstrap_steps": 1200,
    "answer_steps": 0,
    "eval_every": 600,
    "stage_eval_every": 600,
    "lr": 0.001,
    "answer_lr_scale": 0.1,
    "weight_decay": 0.01,
    "grad_clip": 1.0,
    "head_width": 512,
    "trace_loss_weight": 1.0,
    "init_trace_loss_weight": 4.0,
    "op_trace_loss_weight": 1.0,
    "arg_trace_loss_weight": 1.0,
    "attention_loss_weight": 1.0,
    "executor_loss_weight": 1.0,
    "variants": "compiler_trace",
    "max_length": 512,
    "feature_seq_len": 320,
    "seed": 7,
    "output_dir": "experiments/qwen_span_free_compiler/runs/pilot_qwen35_attention_len12_mixed",
    "checkpoint_dir": "",
    "verbose": false
  },
  "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": "AutoModelForMultimodalLM",
    "peft_installed": true,
    "gpu_name": "NVIDIA RTX 6000 Ada Generation",
    "gpu_vram_gb": 47.383
  },
  "hidden_dim": 2560,
  "dataset": {
    "bootstrap_train_size": 1024,
    "answer_train_size": 256,
    "eval_size": 128,
    "bootstrap_train_lengths": [
      1,
      12
    ],
    "answer_train_lengths": [
      1,
      12
    ],
    "eval_lengths": [
      4,
      8,
      12,
      24
    ],
    "train_template_mode": "mixed",
    "answer_train_template_mode": "mixed",
    "eval_template_modes": [
      "standard",
      "paraphrase"
    ],
    "feature_seq_len": 320
  },
  "variants": {
    "compiler_trace": {
      "variant": "compiler_trace",
      "train_seconds": 141.2418177127838,
      "checkpoints": [
        "large_artifacts/qwen_span_free_compiler/checkpoints/pilot_qwen35_attention_len12_mixed/compiler_trace/checkpoint_final.pt"
      ],
      "train_log": [
        {
          "stage": "trace_bootstrap",
          "local_step": 1,
          "step": 1,
          "loss": 42.631309509277344,
          "trace_loss_active": true,
          "lr_scale": 1.0,
          "executor_loss": 4.580158710479736,
          "init_loss": 4.613909721374512,
          "op_loss": 1.1049611568450928,
          "arg_loss": 4.616775035858154,
          "init_attn_loss": 4.459136962890625,
          "op_attn_loss": 4.712254047393799,
          "arg_attn_loss": 4.70238733291626,
          "standard_len4_direct_accuracy": NaN,
          "standard_len4_direct_target_mass": NaN,
          "standard_len4_executor_accuracy": 0.0078125,
          "standard_len4_executor_target_mass": 0.010309100151062012,
          "standard_len4_init_accuracy": 0.0,
          "standard_len4_op_accuracy": 0.314453125,
          "standard_len4_arg_accuracy": 0.05078125,
          "standard_len4_program_exact": 0.0,
          "standard_len8_direct_accuracy": NaN,
          "standard_len8_direct_target_mass": NaN,
          "standard_len8_executor_accuracy": 0.0234375,
          "standard_len8_executor_target_mass": 0.010309698060154915,
          "standard_len8_init_accuracy": 0.0,
          "standard_len8_op_accuracy": 0.333984375,
          "standard_len8_arg_accuracy": 0.052734375,
          "standard_len8_program_exact": 0.0,
          "standard_len12_direct_accuracy": NaN,
          "standard_len12_direct_target_mass": NaN,
          "standard_len12_executor_accuracy": 0.0078125,
          "standard_len12_executor_target_mass": 0.01030859537422657,
          "standard_len12_init_accuracy": 0.0078125,
          "standard_len12_op_accuracy": 0.3268229166666667,
          "standard_len12_arg_accuracy": 0.050130208333333336,
          "standard_len12_program_exact": 0.0,
          "standard_len24_direct_accuracy": NaN,
          "standard_len24_direct_target_mass": NaN,
          "standard_len24_executor_accuracy": 0.0078125,
          "standard_len24_executor_target_mass": 0.010308679193258286,
          "standard_len24_init_accuracy": 0.0,
          "standard_len24_op_accuracy": 0.337890625,
          "standard_len24_arg_accuracy": 0.043294270833333336,
          "standard_len24_program_exact": 0.0,
          "paraphrase_len4_direct_accuracy": NaN,
          "paraphrase_len4_direct_target_mass": NaN,
          "paraphrase_len4_executor_accuracy": 0.0,
          "paraphrase_len4_executor_target_mass": 0.010309863835573196,
          "paraphrase_len4_init_accuracy": 0.0,
          "paraphrase_len4_op_accuracy": 0.384765625,
          "paraphrase_len4_arg_accuracy": 0.044921875,
          "paraphrase_len4_program_exact": 0.0,
          "paraphrase_len8_direct_accuracy": NaN,
          "paraphrase_len8_direct_target_mass": NaN,
          "paraphrase_len8_executor_accuracy": 0.015625,
          "paraphrase_len8_executor_target_mass": 0.01030894834548235,
          "paraphrase_len8_init_accuracy": 0.0,
          "paraphrase_len8_op_accuracy": 0.3251953125,
          "paraphrase_len8_arg_accuracy": 0.056640625,
          "paraphrase_len8_program_exact": 0.0,
          "paraphrase_len12_direct_accuracy": NaN,
          "paraphrase_len12_direct_target_mass": NaN,
          "paraphrase_len12_executor_accuracy": 0.015625,
          "paraphrase_len12_executor_target_mass": 0.010308641940355301,
          "paraphrase_len12_init_accuracy": 0.0078125,
          "paraphrase_len12_op_accuracy": 0.3151041666666667,
          "paraphrase_len12_arg_accuracy": 0.055338541666666664,
          "paraphrase_len12_program_exact": 0.0,
          "paraphrase_len24_direct_accuracy": NaN,
          "paraphrase_len24_direct_target_mass": NaN,
          "paraphrase_len24_executor_accuracy": 0.015625,
          "paraphrase_len24_executor_target_mass": 0.01030859537422657,
          "paraphrase_len24_init_accuracy": 0.0,
          "paraphrase_len24_op_accuracy": 0.3271484375,
          "paraphrase_len24_arg_accuracy": 0.046875,
          "paraphrase_len24_program_exact": 0.0
        },
        {
          "stage": "trace_bootstrap",
          "local_step": 600,
          "step": 600,
          "loss": 5.092926979064941,
          "trace_loss_active": true,
          "lr_scale": 1.0,
          "executor_loss": 2.3384735584259033,
          "init_loss": 0.00016102036170195788,
          "op_loss": 0.4779512584209442,
          "arg_loss": 0.6962012052536011,
          "init_attn_loss": 5.820388105348684e-05,
          "op_attn_loss": 0.8070896863937378,
          "arg_attn_loss": 0.7725088000297546,
          "standard_len4_direct_accuracy": NaN,
          "standard_len4_direct_target_mass": NaN,
          "standard_len4_executor_accuracy": 0.3828125,
          "standard_len4_executor_target_mass": 0.357939213514328,
          "standard_len4_init_accuracy": 1.0,
          "standard_len4_op_accuracy": 0.90625,
          "standard_len4_arg_accuracy": 0.912109375,
          "standard_len4_program_exact": 0.3828125,
          "standard_len8_direct_accuracy": NaN,
          "standard_len8_direct_target_mass": NaN,
          "standard_len8_executor_accuracy": 0.0078125,
          "standard_len8_executor_target_mass": 0.008953405544161797,
          "standard_len8_init_accuracy": 1.0,
          "standard_len8_op_accuracy": 0.68359375,
          "standard_len8_arg_accuracy": 0.630859375,
          "standard_len8_program_exact": 0.0,
          "standard_len12_direct_accuracy": NaN,
          "standard_len12_direct_target_mass": NaN,
          "standard_len12_executor_accuracy": 0.0078125,
          "standard_len12_executor_target_mass": 0.010435199365019798,
          "standard_len12_init_accuracy": 1.0,
          "standard_len12_op_accuracy": 0.5638020833333334,
          "standard_len12_arg_accuracy": 0.462890625,
          "standard_len12_program_exact": 0.0,
          "standard_len24_direct_accuracy": NaN,
          "standard_len24_direct_target_mass": NaN,
          "standard_len24_executor_accuracy": 0.0078125,
          "standard_len24_executor_target_mass": 0.010310395620763302,
          "standard_len24_init_accuracy": 1.0,
          "standard_len24_op_accuracy": 0.4479166666666667,
          "standard_len24_arg_accuracy": 0.21647135416666666,
          "standard_len24_program_exact": 0.0,
          "paraphrase_len4_direct_accuracy": NaN,
          "paraphrase_len4_direct_target_mass": NaN,
          "paraphrase_len4_executor_accuracy": 0.3046875,
          "paraphrase_len4_executor_target_mass": 0.2748168110847473,
          "paraphrase_len4_init_accuracy": 1.0,
          "paraphrase_len4_op_accuracy": 0.900390625,
          "paraphrase_len4_arg_accuracy": 0.85546875,
          "paraphrase_len4_program_exact": 0.3046875,
          "paraphrase_len8_direct_accuracy": NaN,
          "paraphrase_len8_direct_target_mass": NaN,
          "paraphrase_len8_executor_accuracy": 0.0234375,
          "paraphrase_len8_executor_target_mass": 0.010595961473882198,
          "paraphrase_len8_init_accuracy": 1.0,
          "paraphrase_len8_op_accuracy": 0.7392578125,
          "paraphrase_len8_arg_accuracy": 0.6318359375,
          "paraphrase_len8_program_exact": 0.0,
          "paraphrase_len12_direct_accuracy": NaN,
          "paraphrase_len12_direct_target_mass": NaN,
          "paraphrase_len12_executor_accuracy": 0.0078125,
          "paraphrase_len12_executor_target_mass": 0.010192983783781528,
          "paraphrase_len12_init_accuracy": 1.0,
          "paraphrase_len12_op_accuracy": 0.625,
          "paraphrase_len12_arg_accuracy": 0.4869791666666667,
          "paraphrase_len12_program_exact": 0.0,
          "paraphrase_len24_direct_accuracy": NaN,
          "paraphrase_len24_direct_target_mass": NaN,
          "paraphrase_len24_executor_accuracy": 0.015625,
          "paraphrase_len24_executor_target_mass": 0.010308265686035156,
          "paraphrase_len24_init_accuracy": 1.0,
          "paraphrase_len24_op_accuracy": 0.4645182291666667,
          "paraphrase_len24_arg_accuracy": 0.234375,
          "paraphrase_len24_program_exact": 0.0
        },
        {
          "stage": "trace_bootstrap",
          "local_step": 1200,
          "step": 1200,
          "loss": 4.709189414978027,
          "trace_loss_active": true,
          "lr_scale": 1.0,
          "executor_loss": 2.136873722076416,
          "init_loss": 5.316223541740328e-05,
          "op_loss": 0.4826711416244507,
          "arg_loss": 0.6029970645904541,
          "init_attn_loss": 6.0216560086701065e-05,
          "op_attn_loss": 0.7621296048164368,
          "arg_attn_loss": 0.7242451906204224,
          "standard_len4_direct_accuracy": NaN,
          "standard_len4_direct_target_mass": NaN,
          "standard_len4_executor_accuracy": 0.375,
          "standard_len4_executor_target_mass": 0.35270047187805176,
          "standard_len4_init_accuracy": 1.0,
          "standard_len4_op_accuracy": 0.890625,
          "standard_len4_arg_accuracy": 0.91015625,
          "standard_len4_program_exact": 0.375,
          "standard_len8_direct_accuracy": NaN,
          "standard_len8_direct_target_mass": NaN,
          "standard_len8_executor_accuracy": 0.0078125,
          "standard_len8_executor_target_mass": 0.012810004875063896,
          "standard_len8_init_accuracy": 1.0,
          "standard_len8_op_accuracy": 0.693359375,
          "standard_len8_arg_accuracy": 0.640625,
          "standard_len8_program_exact": 0.0,
          "standard_len12_direct_accuracy": NaN,
          "standard_len12_direct_target_mass": NaN,
          "standard_len12_executor_accuracy": 0.0,
          "standard_len12_executor_target_mass": 0.01013963297009468,
          "standard_len12_init_accuracy": 1.0,
          "standard_len12_op_accuracy": 0.5826822916666666,
          "standard_len12_arg_accuracy": 0.4752604166666667,
          "standard_len12_program_exact": 0.0,
          "standard_len24_direct_accuracy": NaN,
          "standard_len24_direct_target_mass": NaN,
          "standard_len24_executor_accuracy": 0.03125,
          "standard_len24_executor_target_mass": 0.010332968086004257,
          "standard_len24_init_accuracy": 1.0,
          "standard_len24_op_accuracy": 0.4501953125,
          "standard_len24_arg_accuracy": 0.23079427083333334,
          "standard_len24_program_exact": 0.0,
          "paraphrase_len4_direct_accuracy": NaN,
          "paraphrase_len4_direct_target_mass": NaN,
          "paraphrase_len4_executor_accuracy": 0.328125,
          "paraphrase_len4_executor_target_mass": 0.30892354249954224,
          "paraphrase_len4_init_accuracy": 1.0,
          "paraphrase_len4_op_accuracy": 0.908203125,
          "paraphrase_len4_arg_accuracy": 0.869140625,
          "paraphrase_len4_program_exact": 0.3203125,
          "paraphrase_len8_direct_accuracy": NaN,
          "paraphrase_len8_direct_target_mass": NaN,
          "paraphrase_len8_executor_accuracy": 0.0234375,
          "paraphrase_len8_executor_target_mass": 0.01357295736670494,
          "paraphrase_len8_init_accuracy": 1.0,
          "paraphrase_len8_op_accuracy": 0.7431640625,
          "paraphrase_len8_arg_accuracy": 0.6337890625,
          "paraphrase_len8_program_exact": 0.0,
          "paraphrase_len12_direct_accuracy": NaN,
          "paraphrase_len12_direct_target_mass": NaN,
          "paraphrase_len12_executor_accuracy": 0.0,
          "paraphrase_len12_executor_target_mass": 0.009285049512982368,
          "paraphrase_len12_init_accuracy": 1.0,
          "paraphrase_len12_op_accuracy": 0.6438802083333334,
          "paraphrase_len12_arg_accuracy": 0.4895833333333333,
          "paraphrase_len12_program_exact": 0.0,
          "paraphrase_len24_direct_accuracy": NaN,
          "paraphrase_len24_direct_target_mass": NaN,
          "paraphrase_len24_executor_accuracy": 0.0078125,
          "paraphrase_len24_executor_target_mass": 0.010305731557309628,
          "paraphrase_len24_init_accuracy": 1.0,
          "paraphrase_len24_op_accuracy": 0.4762369791666667,
          "paraphrase_len24_arg_accuracy": 0.22688802083333334,
          "paraphrase_len24_program_exact": 0.0
        }
      ],
      "final_metrics": {
        "standard_len4": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.375,
          "executor_target_mass": 0.35270047187805176,
          "init_accuracy": 1.0,
          "op_accuracy": 0.890625,
          "arg_accuracy": 0.91015625,
          "program_exact": 0.375
        },
        "standard_len8": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0078125,
          "executor_target_mass": 0.012810004875063896,
          "init_accuracy": 1.0,
          "op_accuracy": 0.693359375,
          "arg_accuracy": 0.640625,
          "program_exact": 0.0
        },
        "standard_len12": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.01013963297009468,
          "init_accuracy": 1.0,
          "op_accuracy": 0.5826822916666666,
          "arg_accuracy": 0.4752604166666667,
          "program_exact": 0.0
        },
        "standard_len24": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.03125,
          "executor_target_mass": 0.010332968086004257,
          "init_accuracy": 1.0,
          "op_accuracy": 0.4501953125,
          "arg_accuracy": 0.23079427083333334,
          "program_exact": 0.0
        },
        "paraphrase_len4": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.328125,
          "executor_target_mass": 0.30892354249954224,
          "init_accuracy": 1.0,
          "op_accuracy": 0.908203125,
          "arg_accuracy": 0.869140625,
          "program_exact": 0.3203125
        },
        "paraphrase_len8": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0234375,
          "executor_target_mass": 0.01357295736670494,
          "init_accuracy": 1.0,
          "op_accuracy": 0.7431640625,
          "arg_accuracy": 0.6337890625,
          "program_exact": 0.0
        },
        "paraphrase_len12": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0,
          "executor_target_mass": 0.009285049512982368,
          "init_accuracy": 1.0,
          "op_accuracy": 0.6438802083333334,
          "arg_accuracy": 0.4895833333333333,
          "program_exact": 0.0
        },
        "paraphrase_len24": {
          "variant": "compiler_trace",
          "n": 128.0,
          "direct_accuracy": NaN,
          "direct_target_mass": NaN,
          "executor_accuracy": 0.0078125,
          "executor_target_mass": 0.010305731557309628,
          "init_accuracy": 1.0,
          "op_accuracy": 0.4762369791666667,
          "arg_accuracy": 0.22688802083333334,
          "program_exact": 0.0
        }
      }
    }
  }
}