1batch 0/300 r_pre=0 mean gain 0.00397 frac>0 0.4942/Users/simon-pierreboucher/Desktop/Recherche-UQO/ia/prisme-slm/training/train_editor.py:57: 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.)3 return torch.from_numpy(acts).to(device), torch.from_numpy(np.ascontiguousarray(b)).to(device)4batch 10/300 r_pre=3 mean gain 0.00043 frac>0 0.3335batch 20/300 r_pre=3 mean gain -0.00053 frac>0 0.2846batch 30/300 r_pre=2 mean gain -0.00279 frac>0 0.3857batch 40/300 r_pre=3 mean gain -0.00303 frac>0 0.2718batch 50/300 r_pre=3 mean gain -0.00278 frac>0 0.3049batch 60/300 r_pre=0 mean gain 0.01724 frac>0 0.51010batch 70/300 r_pre=0 mean gain 0.01201 frac>0 0.53311batch 80/300 r_pre=2 mean gain 0.00534 frac>0 0.41212batch 90/300 r_pre=2 mean gain -0.00454 frac>0 0.32813batch 100/300 r_pre=3 mean gain -0.00403 frac>0 0.24514batch 110/300 r_pre=2 mean gain -0.00116 frac>0 0.52915batch 120/300 r_pre=0 mean gain 0.00268 frac>0 0.51416batch 130/300 r_pre=3 mean gain 0.00058 frac>0 0.29917batch 140/300 r_pre=2 mean gain -0.00481 frac>0 0.34618batch 150/300 r_pre=3 mean gain 0.00099 frac>0 0.30919batch 160/300 r_pre=2 mean gain 0.00548 frac>0 0.44320batch 170/300 r_pre=3 mean gain 0.00025 frac>0 0.29021batch 180/300 r_pre=0 mean gain 0.01394 frac>0 0.50922batch 190/300 r_pre=3 mean gain -0.00209 frac>0 0.27823batch 200/300 r_pre=1 mean gain 0.00116 frac>0 0.44324batch 210/300 r_pre=1 mean gain -0.00147 frac>0 0.41925batch 220/300 r_pre=2 mean gain 0.00312 frac>0 0.40726batch 230/300 r_pre=0 mean gain -0.00251 frac>0 0.47727batch 240/300 r_pre=1 mean gain -0.00129 frac>0 0.40028batch 250/300 r_pre=3 mean gain -0.00223 frac>0 0.28229batch 260/300 r_pre=3 mean gain -0.00351 frac>0 0.27530batch 270/300 r_pre=1 mean gain 0.00268 frac>0 0.47031batch 280/300 r_pre=1 mean gain 0.00504 frac>0 0.51132batch 290/300 r_pre=2 mean gain -0.00309 frac>0 0.36933train {'file': 'experiments/runs/labels_v1ctrl_seed01_20261030004018/labels_train.pt', 'n_states': 49965, 'n_labels': 199860, 'mean_gain': 0.0013899323530495167, 'frac_positive': 0.40608927607536316, 'frac_harm': 0.38673070073127747, 'seconds': 61.8, 'n_audit_batches': 75}34batch 0/40 r_pre=3 mean gain 0.00410 frac>0 0.27335batch 10/40 r_pre=3 mean gain -0.00087 frac>0 0.30336batch 20/40 r_pre=1 mean gain -0.00083 frac>0 0.50237batch 30/40 r_pre=2 mean gain 0.00212 frac>0 0.44638calib {'file': 'experiments/runs/labels_v1ctrl_seed01_20261030004018/labels_calib.pt', 'n_states': 6729, 'n_labels': 26916, 'mean_gain': -0.00023525116557721049, 'frac_positive': 0.40730419754981995, 'frac_harm': 0.38687026500701904, 'seconds': 10.2, 'n_audit_batches': 10}39run labels_v1ctrl_seed01_2026103000401840