loaded backbone from experiments/runs/p4_tiny_backbone_seed01_20261029232626/checkpoint missing: 35
readout_scale set to backbone residual rms = 5.251
/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/training/train_editor.py:55: UserWarning: The given NumPy array is not writable, and PyTorch does not support non-writable tensors. This means writing to this tensor will result in undefined behavior. You may want to copy the array to protect its data or make it writable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at /Users/runner/work/pytorch/pytorch/torch/csrc/utils/tensor_numpy.cpp:214.)
  return torch.from_numpy(acts).to(device), torch.from_numpy(np.ascontiguousarray(b)).to(device)
step 0 loss 0.2574 lr 1.00e-05 gn 0.02 (1s)
step 50 loss 0.2963 lr 5.10e-04 gn 0.02 (35s)
step 100 loss 0.2928 lr 1.00e-03 gn 0.03 (68s)
step 150 loss 0.2752 lr 9.99e-04 gn 0.04 (101s)
step 200 loss 0.2912 lr 9.97e-04 gn 0.03 (134s)
step 250 loss 0.2685 lr 9.94e-04 gn 0.02 (168s)
step 300 loss 0.2521 lr 9.89e-04 gn 0.02 (200s)
step 350 loss 0.3760 lr 9.84e-04 gn 0.10 (234s)
step 400 loss 0.2934 lr 9.76e-04 gn 0.03 (266s)
step 450 loss 0.2019 lr 9.68e-04 gn 0.06 (299s)
  val loss by budget: {0: 0.4144, 1: 0.4139, 2: 0.4133, 4: 0.4131, 8: 0.4149}
step 500 loss 0.3621 lr 9.58e-04 gn 0.03 (327s)
step 550 loss 0.2949 lr 9.48e-04 gn 0.04 (349s)
step 600 loss 0.3686 lr 9.36e-04 gn 0.07 (372s)
step 650 loss 0.3048 lr 9.22e-04 gn 0.03 (407s)
step 700 loss 0.3155 lr 9.08e-04 gn 0.04 (440s)
step 750 loss 0.2576 lr 8.93e-04 gn 0.02 (478s)
step 800 loss 0.3423 lr 8.77e-04 gn 0.04 (527s)
step 850 loss 0.3550 lr 8.59e-04 gn 0.07 (579s)
step 900 loss 0.3444 lr 8.41e-04 gn 0.06 (627s)
step 950 loss 0.3196 lr 8.22e-04 gn 0.05 (671s)
  val loss by budget: {0: 0.4144, 1: 0.4129, 2: 0.4124, 4: 0.4123, 8: 0.4146}
step 1000 loss 0.3007 lr 8.03e-04 gn 0.06 (751s)
step 1050 loss 0.2752 lr 7.82e-04 gn 0.10 (820s)
step 1100 loss 0.2404 lr 7.61e-04 gn 0.08 (889s)
step 1150 loss 0.3453 lr 7.39e-04 gn 0.05 (948s)
step 1200 loss 0.3205 lr 7.17e-04 gn 0.08 (987s)
step 1250 loss 0.3766 lr 6.94e-04 gn 0.11 (1018s)
step 1300 loss 0.2507 lr 6.70e-04 gn 0.03 (1053s)
step 1350 loss 0.2750 lr 6.47e-04 gn 0.09 (1087s)
step 1400 loss 0.2685 lr 6.23e-04 gn 0.09 (1124s)
step 1450 loss 0.3713 lr 5.99e-04 gn 0.09 (1158s)
  val loss by budget: {0: 0.4144, 1: 0.412, 2: 0.412, 4: 0.4141, 8: 0.4231}
step 1500 loss 0.2672 lr 5.74e-04 gn 0.11 (1197s)
step 1550 loss 0.3123 lr 5.50e-04 gn 0.10 (1231s)
step 1600 loss 0.2820 lr 5.26e-04 gn 0.09 (1259s)
step 1650 loss 0.3476 lr 5.01e-04 gn 0.06 (1289s)
step 1700 loss 0.3768 lr 4.77e-04 gn 0.09 (1318s)
step 1750 loss 0.2676 lr 4.53e-04 gn 0.05 (1347s)
step 1800 loss 0.2471 lr 4.30e-04 gn 0.05 (1382s)
step 1850 loss 0.2802 lr 4.06e-04 gn 0.10 (1418s)
step 1900 loss 0.2870 lr 3.83e-04 gn 0.07 (1452s)
step 1950 loss 0.2158 lr 3.61e-04 gn 0.09 (1488s)
  val loss by budget: {0: 0.4144, 1: 0.4119, 2: 0.4129, 4: 0.4139, 8: 0.4241}
step 2000 loss 0.2418 lr 3.39e-04 gn 0.09 (1524s)
step 2050 loss 0.2704 lr 3.18e-04 gn 0.12 (1553s)
step 2100 loss 0.2859 lr 2.97e-04 gn 0.15 (1583s)
step 2150 loss 0.2684 lr 2.78e-04 gn 0.06 (1611s)
step 2200 loss 0.2357 lr 2.59e-04 gn 0.11 (1640s)
step 2250 loss 0.2331 lr 2.41e-04 gn 0.06 (1674s)
step 2300 loss 0.2750 lr 2.23e-04 gn 0.08 (1709s)
step 2350 loss 0.2091 lr 2.07e-04 gn 0.10 (1742s)
step 2400 loss 0.1803 lr 1.92e-04 gn 0.08 (1774s)
step 2450 loss 0.3071 lr 1.78e-04 gn 0.09 (1804s)
  val loss by budget: {0: 0.4144, 1: 0.41, 2: 0.41, 4: 0.4111, 8: 0.4188}
step 2500 loss 0.3278 lr 1.64e-04 gn 0.13 (1842s)
step 2550 loss 0.2849 lr 1.52e-04 gn 0.09 (1877s)
step 2600 loss 0.2684 lr 1.42e-04 gn 0.12 (1911s)
step 2650 loss 0.2542 lr 1.32e-04 gn 0.16 (1946s)
step 2700 loss 0.1905 lr 1.24e-04 gn 0.08 (1978s)
step 2750 loss 0.2685 lr 1.16e-04 gn 0.13 (2007s)
step 2800 loss 0.2449 lr 1.11e-04 gn 0.10 (2034s)
step 2850 loss 0.2741 lr 1.06e-04 gn 0.17 (2059s)
step 2900 loss 0.2866 lr 1.03e-04 gn 0.13 (2084s)
step 2950 loss 0.2046 lr 1.01e-04 gn 0.07 (2104s)
step 2999 loss 0.3118 lr 1.00e-04 gn 0.08 (2120s)
  val loss by budget: {0: 0.4144, 1: 0.4093, 2: 0.4092, 4: 0.411, 8: 0.4192}
  exact match dev_iid budget 0: 0.4375 (n=400) by family {'fsm': 0.640625, 'bind': 0.38372093023255816, 'arith': 0.22784810126582278, 'excl': 0.27906976744186046, 'sched': 0.6941176470588235}
Traceback (most recent call last):
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/training/train_editor.py", line 186, in <module>
    metrics_file=str(run.dir / "metrics.jsonl"))
          ^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/training/train_editor.py", line 173, in main
    for name, vs in val_sets.items():
                 ^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/.venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 158, in no_grad
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/data/synthetic_dataset.py", line 105, in exact_match
    out = generate(model, prompt, max_new, max_budget=max_budget, cost=cost, temperature=0.0, eos_id=tok.eot_id,
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/.venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 158, in no_grad
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/prisme/generation.py", line 115, in generate
    logits, _, _ = step_logits(model, prompt, cache, max_budget, cost, fa, lambda_cost, kappa_risk)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/.venv/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 158, in no_grad
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/prisme/generation.py", line 43, in step_logits
    if cfg.ctrl_use_entropy:
       ^^^^^^^^^^^^^^^^^^^^
AttributeError: 'PrismeConfig' object has no attribute 'ctrl_use_entropy'