{ "device": "cuda", "math_check": { "linear_profile_g_min": 0.9999999403953552, "linear_profile_g_max": 1.0, "linear_profile_hessian_min": 0.6939469575881958, "bad_profile_hessian_min": -11.702886581420898 }, "baseline_dot": { "final_loss": 0.01768084429204464, "mean_last20": 0.019677276071161032, "loss_std_last50": 0.004114333493313028, "max_grad": 0.9343541860580444, "final_negative_g_fraction": 0.0, "mean_attention_entropy": 0.7476468443870544 }, "unconstrained_metric": { "final_loss": 0.01860710233449936, "mean_last20": 0.02084847129881382, "loss_std_last50": 0.004466183089062249, "max_grad": 0.9493426084518433, "final_negative_g_fraction": 0.0, "mean_attention_entropy": 8.152075793077529e-06 }, "constrained_metric": { "final_loss": 0.018606653437018394, "mean_last20": 0.020852698013186455, "loss_std_last50": 0.004467031694072602, "max_grad": 0.9493432641029358, "final_negative_g_fraction": 0.0, "mean_attention_entropy": 2.02627506951103e-06 } }