RinKana/trm-mamba2-sudoku

--- Starting Training on cuda --- Training on device: cuda --- Epoch 1 completed. Average Training Loss: 2.1852 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 12.33% Current Learning Rate: 1.000000e-04 --- Epoch 2 completed. Average Training Loss: 1.9631 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 15.23% Current Learning Rate: 1.000000e-04 --- Epoch 3 completed. Average Training Loss: 1.7713 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 17.95% Current Learning Rate: 1.000000e-04 --- Epoch 4 completed. Average Training Loss: 1.6633 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 19.42% Current Learning Rate: 1.000000e-04 --- Epoch 5 completed. Average Training Loss: 1.5895 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 20.97% Current Learning Rate: 1.000000e-04 --- Epoch 6 completed. Average Training Loss: 1.5427 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 22.40% Current Learning Rate: 1.000000e-04 --- Epoch 7 completed. Average Training Loss: 1.5022 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 23.50% Current Learning Rate: 1.000000e-04 --- Epoch 8 completed. Average Training Loss: 1.4664 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 24.32% Current Learning Rate: 1.000000e-04 --- Epoch 9 completed. Average Training Loss: 1.4350 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 25.14% Current Learning Rate: 1.000000e-04 --- Epoch 10 completed. Average Training Loss: 1.4023 --- Validation Exact Grid Accuracy: 0.00% Validation Per-Cell Accuracy: 25.93% Current Learning Rate: 1.000000e-04