{ "device": "cuda", "math_check": { "convex_interpolation_inside_bounds": true, "d_mean_alpha_d_beta_at_zero": 0.25, "initial_alpha_beta_minus4": 0.01798621006309986, "suppression_gradient_sign": true }, "results": { "correct": { "free": { "rmse": 0.13216720521450043, "free_branch_rmse": 0.13216720521450043, "geom_branch_rmse": 0.13216720521450043, "mean_alpha": 0.0, "gate_entropy": 1.7310190969510586e-06 }, "geom": { "rmse": 0.011932441964745522, "free_branch_rmse": 0.011932441964745522, "geom_branch_rmse": 0.011932441964745522, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "concat": { "rmse": 0.01182486955076456, "free_branch_rmse": 0.01182486955076456, "geom_branch_rmse": 0.01182486955076456, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "global": { "rmse": 0.029833368957042694, "free_branch_rmse": 0.3897355794906616, "geom_branch_rmse": 10.78345775604248, "mean_alpha": 0.03491225093603134, "gate_entropy": 0.15142330527305603 }, "mixer": { "rmse": 0.010266127064824104, "free_branch_rmse": 0.41977766156196594, "geom_branch_rmse": 0.27537474036216736, "mean_alpha": 0.6074919700622559, "gate_entropy": 0.6697317361831665 }, "global_lambda_0.02": { "rmse": 0.021190768107771873, "free_branch_rmse": 0.42418360710144043, "geom_branch_rmse": 12.315999984741211, "mean_alpha": 0.03323910012841225, "gate_entropy": 0.14582732319831848 } }, "wrong": { "free": { "rmse": 0.06733770668506622, "free_branch_rmse": 0.06733770668506622, "geom_branch_rmse": 0.06733770668506622, "mean_alpha": 0.0, "gate_entropy": 1.7310190969510586e-06 }, "geom": { "rmse": 0.06290043145418167, "free_branch_rmse": 0.06290043145418167, "geom_branch_rmse": 0.06290043145418167, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "concat": { "rmse": 0.07601242512464523, "free_branch_rmse": 0.07601242512464523, "geom_branch_rmse": 0.07601242512464523, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "global": { "rmse": 0.0752580538392067, "free_branch_rmse": 0.08121613413095474, "geom_branch_rmse": 0.6568082571029663, "mean_alpha": 0.044423263520002365, "gate_entropy": 0.18175531923770905 }, "mixer": { "rmse": 0.029876839369535446, "free_branch_rmse": 0.4139192998409271, "geom_branch_rmse": 1.8147188425064087, "mean_alpha": 0.1801556646823883, "gate_entropy": 0.41717299818992615 }, "global_lambda_0.02": { "rmse": 0.03819429129362106, "free_branch_rmse": 0.046958278864622116, "geom_branch_rmse": 1.1284319162368774, "mean_alpha": 0.02048567123711109, "gate_entropy": 0.09992331266403198 } }, "none": { "free": { "rmse": 0.06042604148387909, "free_branch_rmse": 0.06042604148387909, "geom_branch_rmse": 0.06042604148387909, "mean_alpha": 0.0, "gate_entropy": 1.7310190969510586e-06 }, "geom": { "rmse": 0.050840113312006, "free_branch_rmse": 0.050840113312006, "geom_branch_rmse": 0.050840113312006, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "concat": { "rmse": 0.06793663650751114, "free_branch_rmse": 0.06793663650751114, "geom_branch_rmse": 0.06793663650751114, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "global": { "rmse": 0.060362111777067184, "free_branch_rmse": 0.10155397653579712, "geom_branch_rmse": 2.0144524574279785, "mean_alpha": 0.03692173585295677, "gate_entropy": 0.15803472697734833 }, "mixer": { "rmse": 0.041153136640787125, "free_branch_rmse": 0.14096201956272125, "geom_branch_rmse": 0.7465592622756958, "mean_alpha": 0.10575458407402039, "gate_entropy": 0.32703617215156555 }, "global_lambda_0.02": { "rmse": 0.08374335616827011, "free_branch_rmse": 0.4030120074748993, "geom_branch_rmse": 9.209486961364746, "mean_alpha": 0.04062730446457863, "gate_entropy": 0.16993264853954315 } }, "half": { "free": { "rmse": 0.16700385510921478, "free_branch_rmse": 0.16700385510921478, "geom_branch_rmse": 0.16700385510921478, "mean_alpha": 0.0, "gate_entropy": 1.7310190969510586e-06 }, "geom": { "rmse": 0.12250637263059616, "free_branch_rmse": 0.12250637263059616, "geom_branch_rmse": 0.12250637263059616, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "concat": { "rmse": 0.13126066327095032, "free_branch_rmse": 0.13126066327095032, "geom_branch_rmse": 0.13126066327095032, "mean_alpha": 1.0, "gate_entropy": 1.7310190969510586e-06 }, "global": { "rmse": 0.15563292801380157, "free_branch_rmse": 0.559171199798584, "geom_branch_rmse": 7.377834796905518, "mean_alpha": 0.06807496398687363, "gate_entropy": 0.24863073229789734 }, "mixer": { "rmse": 0.12156328558921814, "free_branch_rmse": 1.4644699096679688, "geom_branch_rmse": 0.5877335071563721, "mean_alpha": 0.7179057002067566, "gate_entropy": 0.5880767703056335 }, "global_lambda_0.02": { "rmse": 0.14033490419387817, "free_branch_rmse": 0.5717989802360535, "geom_branch_rmse": 6.770478248596191, "mean_alpha": 0.07580511271953583, "gate_entropy": 0.2684025168418884 } } } }